eeueupiaveserubet aeweat xb. pubespededeenes oat
eit avaseesours dee Nae Bad aot doh on praseer a =vae bre romange gteuemmres to mtg e repret ene ea eTT ane e eae pemaiahncnnneencen pheaiagecs - asa onencnd rretearmeortiewsrtata ence sence ar -
ee ete Pare as eS ert wir a Smite Eke ee ee +: e<hathadetvasteeball~ eas’ ae = rete oe

DUA Gad
\

og 8 Vi W4eg9 eopmet
FT, |
Hibrary of the THuseum

OF

COMPARATIVE ZOOLOGY,
AT HARVARD COLLEGE, CAMBRIDGE, MASS.

Founded by private subscription, in 1861.

Deposited by ALEX. AGASSIZ.

A749 Lp
/ No. 4 J ol , \ i
| 4 / ay. ia Ab ; ;
SRL /4 $93 — APuL Ras ¥
{ /
vu

7
vf

wg il :
ae ny

ee mi

nap ;
bi

; ;
is Ba
we \ fy} Lae Pi
ve) in "|

Li Wy

a Pal: i) a x iA) (ie Mh
1 i " U

iu Vee,

RGN

nie ee
Wy al aie

4
le ‘etal
mi

r M a ' | mi i, pit i a t. i } i n i i
| yh AT , ' A ) ’ i |
Nt ales wh aA Ni ie th sade
At ae id ey an ily i oy
) ian ie loos aane pga ge) Oe
| Ls ia Ww a Avo ; “? | yay i"
ee cw) by ee ise. ml ly | i
7 I .

r 4

j

if
f ;
if

j i 1 ii ;

hi "iy fi 1° 7 i
i 4
ia i
£4 {
}

/4 49
/-21

QUARTERLY JOURNAL

OF

MICROSCOPICAL SCIENCE:

EDITED BY

HE. RAY LANKESTER, M.A., LL.D., F.R.S.,
Linacre Professor of Comparative Anatomy, Fellow of Merton College, and
Honorary Fellow of Exeter College, Oxford.

WITH THE CO-OPERATION OF

ADAM SEDGWICK, M.A., F.R.S.,
Fellow and Lecturer of Trinity College, Cambridge ;

A. MILNES MARSHALL, M.A., D.Sc., M.D., F.R.S.,
Late Fellow of St. John’s College, Cambridge ; Professor in the Victoria University ,
Beyer Professor of Zoology in the Owens College, Manchester ;

AND

W. F. R. WELDON, M.A., F.B.S.,
Jodrell Professor of Zoology and Comparative Anatomy in University College, London;
Fellow of St. John’s College, Cambridge.

VOLUME 35.—New Szrgizgs.

With Aithographic Plates and Engrabings on Wood.

LONDON:
J. & A. CHURCHILL, 11, NEW BURLINGTON STREET.
“1894.

f SP NAREER
DUP ON Si OOret ae
AW TATA TAS

— a

CONTENTS.

CONTENTS OF No. 137, N.S., JULY, 1898.
MEMOIRS:

On the Morphology and Physiology of the Brain and Sense Organs
of Limulus. By Writ1am Parren, Ph.D., University of North
Dakota, Grand Forks, U.S.A. (With Plates 1—5)

The Structure of the Pharyngeal Bars of Amphioxus. By W.
Buaxtanp Brenuam, D.Sc.Lond., Hon. M.A.Oxon., Aldrichian
Demonstrator of Comparative Anatomy in the University of
Oxford. (With Plates 6 and 7)

On the Perivisceral Cavity of Ciona. By A. H. L. Newsreap,
B.A., late Scholar of Christ’s College, Cambridge. (With
Plate 8) . ; : . ; ;

The Early Stages in the Development of Distichopora vio-
lacea, with a Short Essay on the Fragmentation of the Nucleus.
By Sypney J. Hickson, M.A. Cantab. et Oxon., D.Se.Lond.,
University Lecturer on the Morphology of Invertebrates; Fellow
of Downing College, Cambridge. (With Plate 9)

CONTENTS OF No. 138, N.S., SEPTEMBER, 1893.

MEMOIRS :

Studies on the Comparative Anatomy of Sponges. By Artuur
Denpy, D.Sc., Fellow of Queen’s College, and Demonstrator and
Assistant Lecturer in Biology in the University of Melbourne.
(With Plates 10—14)

b

PAGE

97

119

129

159

1V CONTENTS.

Some Points in the Origin of the Reproductive Elements in Apus
and Branchipus. By J. E.S. Moors, A.R.C.S., from the Huxley
Research Laboratory, Royal College of Science, London. (With
Plates 15 and 16) ‘ : - 3 P

Notes on the Peripatus of Dominica. By HE. C. Potarp,
B.Sc.Lond. (With Plate 17)

Studies on the Protochordata. By ArtHur Wiutey, B.Sc.Lond.,
Columbia College, New York. (With Plates 18—20)

CONTENTS OF No. 139, N.S., JANUARY, 1894.

MEMOIRS :
Observations on the Development of the Head in Gobius capito.
By H. B. Pottarp, Oxford. (With Plates 21 and 22) .

On the Head Kidney of Myxine. By J. W. Krrxatpy, Somerville
Hall, Oxford. (With Plate 23) 5 :

Report on a Collection of Amphioxus made by Professor A. C.
Haddon in Torres Straits, 1888-9. By ArrHur WILLEy,
B.Se.Lond.

The Orientation of the Frog’s Egg. By T. H. Moreay, Ph.D.,
Associate Professor of Biology, Bryn Mawr College; and Umé
Tsupa, Teacher in the Peeress’ School, Tokio, Japan. (With
Plates 24 and 25)

On the Fossil Mammalia from the Stonesfield Slate. By E.S.
GoopricH, F.L.S., Assistant to the Linacre Professor of Com-
parative Anatomy, Oxford. (With Plate 26) .

A Polynoid with Branchie (Eupolyodontes Cornishii). By
FLORENCE BucHanan, B.Sc. (With Plate 27) . :

On some Bipinnarie from the English Channel. By Water
Garstane, M.A., Fellow of Lincoln College, Oxford, Naturalist
to the Marine Biological Association. (With Plate 28)

Octineon Lindahli (W. B. Carpenter): an Undescribed Antho-
zoon of Novel Structure. By G. Herbert Fowrer, B.A.,
Ph.D., Demonstrator in Zoology, University College, London.
(With Plates 29 and 30). ; , 2 ‘ ;

PAGE

259

285

295

335

353

361

373

407

433

45]

461

CONTENTS. Vv

CONTENTS OF No. 140, N.S., MARCH, 1894.

MEMOIRS: PAGE
Studies in Mammalian Embryology. III.—The Placentation of
the Shrew (Sorex vulgaris, L.). By A. A. W. Husrecat,
LL.D., C.M.Z.S., Professor of Zoology in the University of
Utrecht. (With Plates 31 to 39) : : : . 48)

Some Further Contributions to our Knowledge of the Minute
Anatomy of Limnocodium. By R. T. Gintuer, B.A., Lecturer
of Magdalen College, Oxford: (With Plate 40) 4 . 539

Note on the Mesenteries of Actinians. By A. Fraser Dixon . 55]

tewa se
ee

On the Morphology and Physiology of the
Brain and Sense Organs of Limulus.

By

William Patten, Ph.D.,
University of North Dakota, Grand Forks, U.S.A.

With Plates 1, 2, 3, 4, 5.

ConTENTS.
INTRODUCTION . ‘ : ; : ‘ : 6 Mite

Part I—Sense Organs.

I. Gustatory ORGANS.

A. Experimental Demonstration of the Presence of Three
Sets of Gustatory Organs in the Mandibles, p. 7. Method of
stimulating the organs, p. 7. Description of the reflexes caused by
stimulation of the same, pp. 7—9. Location of the organs—(1) By local
stimulation, p. 10. (2) By amputation of mandibles, p. 10. (3) By shav-
ing off of spines, p. 10. Temporary suspension of reflexes, pp. 11, 12.
General remarks.

B. Structure of Gustatory Cells of the Mandibular Spines, p.
12. Chitinous tubules, p. 138. The spindle, the axial, and external nerves,
p. 14. Gustatory cells, p. 15. Gigantic ganglion-cells, p. 15. Double nature
of gustatory cells, p. 15. Cuticular canals, p. 16.

C. Experimental Demonstration of the Presence of Gustatory
and Temperature Organs in the Chele, p. 16. Stimulation with
clam, ammonia, and warm air; reflexes, p. 16. Experiments on amputated
chelea—description of the organs, p. 17.

II, OtFactory OrGans, p. 17.

A. Structure of the Olfactory Organ: external appearance, general
topography, cuticula, air in pore canals, spines, p. 17. Olfactory buds,

VOL. 35, PART 1.—NEW SER. A

2 WILLIAM PATTEN.

p. 18. Shape, structure; central ganglion-cells; non-glandular nature of
the buds; lumen of the buds; chitinous tubule ; cuticular canals; nerve-
supply, p. 22. Lateral olfactory nerves; origin, course, distribution,
structure ; lateral olfactory ganglion. Median olfactory nerve, p. 22. Struc-
ture, distribution, termination in buds. Problematical organs.

B. Physiology of the Olfactory Organ, p. 26. Absence of reflexes
on stimulation with food, ammonia, &c.; stimulation with electricity; nature
of reflex in the males; same in females; function of the organ; effect on
males on excising the organ.

C. The Development of the Olfactory Organ, p. 29. Primitive
olfactory thickenings; origin of the ganglion in the lateral olfactory nerve ;
origin of definitive olfactory organ, p. 30. Origin of median olfactory nerve ;
origin of olfactory lobes.

D. Relation of Primary Olfactory Organ to segmental sense
organs; to lateral eyes of scorpions; to frontal organ of Phyllopods; cause
of change of function, p. 32.

III. Gustatory Bups oF THE INNER MANDIBLEs, p. 33.

Position and structure of those in the inner mandibles; effect of various
reagents.

IV. TEMPERATURE SENSE.

Experimental demonstration of the same, p. 34.

A. Experiments showing course of temperature impulses; section of
tegumentary nerves; of spinal cord; longitudinal median section through
hind brain and vagus region; section of left crus; of both crura; tem-
perature centre in the fore-brain ; supplementary centres in legs, p. 36.

B. Position of Gustatory Centre, p. 39.

C. Structure of Temperature Organ, p. 39; probably much the
same as the gustatory buds.

D. Use of temperature sense, p. 39.

V. Tactite Senses, p. 40.
VI. Summary AND CoMPARISON.

Structure and nature of sense organs, p. 40; innervation; multiplication by
division; comparison with Gephyreans, Galeodes, Mutilla. Very wide dis-
tribution of the sense buds, their specialisation as isolated organs, and their
fusion to form complex aggregates.

Part II—The Morphology of the Arthropod Brain.

I. Tue Brain oF Insects anp Mynriapops, p. 43.

A. Structure of Cephalic Lobes: fore-brain; mid-brain: their dif-
ferent origins, p. 46.

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 3

B. The Stomodeal Nerves, p. 46. The labrum, p. 47. The an-
terior pons stomodei; the posterior pons stomodei; the anterior stomo-
deal ganglion; the lateral stomodzal ganglia; the lateral and the median
sympathetic nerves of the trunk.

C. The Convex Eyes and their Ganglia, p. 48. They do not belong
to the cephalic lobes, but to the mid-brain neuromeres; absence in scorpions ;
origin from the cheliceral segment in Limulus; is the organ of Tomésvary in
Myriapods the “anlage” of the convex eye? Conversion of the larval optic
ganglion in Acilius into that of the adult; relation of the frontal ocelli to
the larval ones.

If. Tue Brain oF ARACHNIDS.

A. The Cephalic Lobes, p.50. Semicircular lobes, p.50. Formation
of primitive cerebral vesicle, p. 51. Homology of the eyes of spiders and
scorpions, p. 51.

B. The Mid-brain, p. 53. Homology of fore- and mid-brair of scorpions
with those of insects, p. 53. Stomodzal nerves.

C. Development of the lateral stomodzal nerves, p. 54.
I). Comparison of cephalic lobes of Arthropods and those of Annelids, p. 55.

KH. Are the stomodzal nerves derived from the circumoral nerves of a
Celenterate? p. 56.

III. DevEeLopMeNntT oF THE Brain oF LIMuLvs, p. 57.

A. The Cephalic Lobes of Limulus, p. 57. Ganglionic invagina-
tions ; origin of the cerebral hemispheres ; fusion of the two invaginations of
the second segment to form the parietal eye-tube, or epiphysis.

B. The Mid-brain, p. 60. Stomodzal nerves.

C. Later Modifications of the Brain: fate of the invaginations ;
formation of medullary folds ; fate of the anterior neuropore, p. 61.

Development of the Cerebral Hemispheres: the posterior lobe ;
the anterior lateral lobe; the internal median, or corpus striatum. The
semicircular lobes, p. 66. ‘Their relation to nerves of parietal eye; the
optic ganglia; modification and origin of the various brain-vesicles.

D. Comparison with Vertebrates, p. 67. The semicircular Jobes
and the infundibulum of Vertebrates, p. 69. Changes in the position of the
parietal eye ; peduncles to the parietal eye of mammals. The third ventricle;
the lateral eyes; explanation of choroid fissure in Vertebrates, p. 72. The
“iter ;” the fifth ventricle; the fourth ventricle; summary, p. 73.

IV. Tue Parietat Eyes, p. 75.

Endo- and ecto-parietal eyes; origin; four nerves to the same; the epi-
physis; the stalk of parietal eye-nerve; the transverse tube; the blood-
vessel; comparison with Vertebrates; nature of the epiphysis ; white pigment

4, WILLIAM PATTEN.

of the endoparietal eye; comparison with white pigment in the eye of Pe-
tromyzon; Gaskell’s comparison of parietal eyes without sufficient data.
V. Otractory Oreay, p. 79.

Comparison with olfactory organ in Vertebrates ; its four nerves correspond
to the four olfactory nerves of Amphibians. Relation in Limulus and Ver-
tebrates of the olfactory nerves to the optic thalami; comparison of the
olfactory organs of Vertebrates and Limulus with the supra-branchial or
mandibular sense organs. The function, distribution, and multiplication of the
mandibular sense organs of Limulus are like those of the supra-branchial
organs of Vertebrates. The mandibles probably eer the rudimentary
endopodites of the thoracic appendages.

VI. ComPARISON OF OTHER BRAIN ReEGions oF LIMULUS WITH THOSE
oF VERTEBRATES.

Comparison of the fore-, mid-, hind, and accessory brains and their nerves
with the corresponding regions in Vertebrates, p. 83. Concluding remarks,
p. 85.

INTRODUCTION.

Some two years ago I published ashort paper in this Journal
calling attention to many striking resemblances between
Arachnids and Vertebrates. I maintained that the Verte-
brates are descended from a great group of Arthropods,
in which I included the Arachnids, Trilobites, and Mero-
stomata ; and that the remarkable palzozoic fishes Pterichthys,
Bothriolepis, and Cephalaspis resemble merostomatous Arthro-
pods like Pterygotus, Eurypterus, &c. Now, the internal
structure of the Merostomata and Trilobites cannot differ
greatly from that of Limulus, judging from their resemblance
in external characters; therefore, although Limulus itself is
not in the main line of Vertebrate descent, a study of its
structure will best enable us to understand that of the Mero-
stomata and of the primitive Vertebrates.

In my preliminary paper the lines were indicated along
which I had found evidence of relationship between Ver-
tebrates and Arachnids. It was shown—and this first drew
my attention to the subject—that the invaginations, which in
insects give rise to the optic ganglia, in scorpions and Limulus

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 5

become so extensive as to enfold not only the optic ganglia,
but the eyes and the fore-brain as well. A cerebral vesicle is
thus formed, from the floor of which arise the fore-brain and
the optic ganglia, and from the roof a tubular outgrowth, at
the end of which lie the inverted retinas of the parietal eye.
Such a condition is found only in Arachnids and Vertebrates;
and in my judgment this fact, when all its bearings are con-
sidered, affords as trustworthy evidence of relationship as the
presence of a notochord or of gill-slits.

It was further shown (1) that the lateral eyes of Limulus
could be compared with the lateral eyes of Vertebrates; (2)
that there isin Arachnids a cartilaginous endocranium similar
in position, shape, and development to the primordial cranium
of Vertebrates; (3) that there is in scorpions and in other
Arthropods a subneural rod similar to the notochord of Ver-
tebrates; (4) that in the Arachnids and in the Vertebrates
the brain contains approximately the same number of neuro-
meres; it is divided into the same number of regions, i.e.
fore-brain, mid-brain, hind brain, and accessory brain; and
while in each region there is a different number of neuro-
meres, i.e. 3, 1, 5, 2 to 4, the number in the corresponding
regions in Vertebrates and Arachnids is very nearly the same;
(5) the nerves of each brain region in both Vertebrates and
Arachnids show in a general way the same relation to sense
organs, the same ganglia, and the same distribution and fusion
with one another; (6) finally, there is a striking similarity
between the cephalo-thoracic shields of Arthropods and the
cephalic bucklers of the earliest fishes, such as Zenaspis, Both-
riolepis, Pteraspis, Auchenaspis, &c. The shape and micro-
scopic structure of the shields, and the arrangement of the
eyes upon them, are practically the same in both groups.
The three median and two lateral eyes of Cephalaspis
Campbelltonensis as figured by Whiteaves (‘ Trans. Roy.
Soc. Canada,’ vol. vi, 1888, pl. x) have exactly the same
arrangement as those of Limulus. I have carefully revised
the paleontological aspect of the subject, and I hope in a
separate paper to give it the careful consideration it deserves.

6 WILLIAM PATTEN.

Attention was also called to resemblances of a more general
character between Vertebrates and Arachnids, such as, for
example, in the structure of the muscles, the nerves, and the
liver; in the position and the net-like structure of the sexual
organs; in the origin of the ova and the spermatozoa; in the
origin of the germ-layers, and in the general formation of the
embryos.

To this formidable array of evidence I can now add the
following:—(1) It is possible to identify nearly all the important
lobes and cavities characteristic of the Vertebrate fore-brain in
the fore-brain of Limulus. (2) The coxal sense organs are shown
by conclusive experiments to be gustatory, and to correspond to
the supra-branchial sense organs of Vertebrates. (3) An
extraordinary organ has been discovered in Limulus, having
all the characteristic morphological features of the olfactory
organ in Vertebrates. It is united with olfactory lobes
that arise as outgrowths from the fore-brain. It consists of
an upright layer of epithelium containing olfactory buds similar
to those in the coxal (supra-branchial) sense organs. It is
supplied by four nerves, the median ones resembling, in histo-
logical structure, those of Vertebrates. One pair of these
nerves arises from the cerebral hemispheres, the other from the
optic thalami. I long ago looked for something answering to
the olfactory organ in Vertebrates, but finally gave up the
search because that part of the cephalic lobes where, as I
supposed, they ought to appear was invaginated, consequently
any sense organ derived from that region must also be invagi-
nated, and could not be homologous with the olfactory organ
of Vertebrates.

Most of the physiological results were obtained during the
summer of 1892 at the United States Fish Commission
Laboratory at Wood’s Holl, Mass., the facilities of which were
generously placed at my disposal by Commissioner MacDonald.

The descriptive part of this paper deals mainly with Limulus,
but incidental references are made to scorpions and other
Arthropods. As I aim to point out resemblances between
Vertebrates and Arachnids, I shall not enter into histological

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 7

details that might obscure the meaning of the broad facts I
wish to present. However, I shall describe the morphology
and physiology of the olfactory organ, and of the gustatory and
temperature organs on the appendages and elsewhere in detail.
In order to justify the comparisons instituted between these
sense organs and those of Vertebrates I have given a general
account of the structure and development of the brain and
median eyes.

I do not hesitate to say that I believe the results herewith
presented prove beyond any reasonable doubt that the Verte-
brates are descended from the Arachnids,

Part I.—Sense Organs.
I. Gustatory ORGANS.

a. Experiments on the Gustatory Organs of the
Mandibles.—If an adult horseshoe crab be placed on its back
with its abdomen hanging over the edge of the table, it makes
fruitless movements of the legs and abdomen to recover its natu-
ral position. The muscles, however, soon relax, and the animal
usually becomes perfectly quiet, except that after long intervals
the gills are raised and lowered a few times, and then held up
motionless a few seconds till every part, expanded to its full
extent, is thoroughly aérated; they then sink slowly back to
their original position. If, while in the quiescent condition, the
jaw-like spurs or mandibles (PI. 1, fig. 3,0. md.) at the base of the
legs are gently rubbed with some hard object, such as a piece
of wood, glass, or iron; or if water or air, the temperature of
the surrounding medium, be gently poured over them; or if
the animal be vigorously fanned, or loud noises be made near
it, only slight aimless movements of the legs or abdomen are
produced, usually none at all. But if a very small piece of
clam or other edible substance be rubbed ever so gently over
the stout spines that arm the mandibles, very characteristic
chewing movements are immediately produced. If all the
mandibles are touched in this way, or even moistened with a
few drops of water in which pieces of clam have been soaked,

8 WILLIAM PATTEN.

the chelicere snap and work alternately back and forth, as
though tucking something into the mouth; at the same time
the metastoma are moved forward and backward. But the
most constant feature is the following movement of the
second to the fifth pair of appendages; the second and
fourth pairs of mandibles move in unison inward toward the
median plane, and downward toward the mouth; then back
again in the reverse order. When they are farthest from the
mouth the corresponding legs (except the second pair in both
males and females) are quickly raised, flexed, and the tips
carried toward the mouth, where they remain an instant, and
then fall back on to the under side of the carapace ; the corre-
sponding jaw movement then begins again. The third and fifth
pairs of appendages and the corresponding jaws work in unison
in the same manner, but they alternate with those of the second
and fourth. At intervals these movements cease, the abdomen
is raised, and the stout crushing mandibles on the sixth pair
of appendages, which have heretofore remained motionless, are
slowly closed with great force, as though to crush some object
too large to be swallowed whole, or to kill some struggling
prey. These powerful jaws then slowly relax their convulsive
grasp, and the chewing movements are resumed. All these
movements go on with the greatest precision and regularity,
so that any food placed on the jaws is forced into the mouth,
and gradually disappears down the cesophagus. These chew-
ing movements are produced when drops of soluble food, or
almost any bit of animal matter, or wads of blotting-paper wet
in nutritive animal fluids, come in contact with the mandibles.
Drops of water from the interior of a clam will set the whole
complicated mechanism to working in exactly the same manner
as during the actual process ofeating. Again, chewing move-
ments of the mandibles are produced whenever ammonia
vapour, ether, or chloroform is blown over them with a medicine-
dropper, or when they are stimulated by a weak interrupted
current ; but in such cases these movements are rarely accom-
panied by the leg movement. Ifthe irritation with ammonia or
acids has been rather great the mandibles work apparently as

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 9

in eating, but the chelicere move rapidly back and forth,
making frantic snapping movements toward the mouth, as
though to pick away some disagreeable object. These move-
ments usually last a long time. If wads of blotting-paper
wet in ammonia or picric acid are used the chewing movements
are reversed, and the offensive object is sometimes snapped up
by the chelicerze and rejected.

Holding strong-smelling food as close as possible to the
mouth or to the jaws produces no effect, although chew-
ing movements are instantly produced when the jaws are
touched by it.

When a very small piece of clam or of pecten is touched to
the surface—say of the third mandible on the left side—care
being taken not to touch any other parts, that leg will be
promptly raised and the tip bent toward the mouth; it soon
falls back on to the cephalo-thorax again, and then its man-
dible moves back and forth, alternating with the leg movement,
as in eating; but all the other mandibles and appendages
remain quiet. One may start in this way one appendage after
the other (except the chelicerze, which have no mandibles),
until all of them, first on one side and then on the other, are
working in perfect rhythm.

If the mandibles of the post-oral appendages on one side
are stimulated, the chelicera of that side, although not stimu-
lated itself, soon has its tip extended straight backward as
far as it can reach, and may remain some time in this rigid,
unnatural position ; or else it begins those characteristic back
and forward movements, snapping its chela from time to time,
as though to seize something and lift it up, or else thrusting
the chela down the mouth as though forcing some piece of food
into the cesophagus. If the jaws on the opposite side are now
stimulated, the chelicera of that side begins to work also. It
is a curious fact that the chelicere rarely move when the
mandibles of the second or third appendages are stimulated ;
not till the last one or two pairs of mandibles are set in motion
do they begin their characteristic movements. It is evident
that the chewing movement produced in these various ways

10 WILLIAM PATTEN.

is a reflex act caused by the stimulation of gustatory organs
about the mouth.

The following experiments show that there are three kinds
of these gustatory organs, and that they are situated in the
inner and outer mandibles of the second to the fifth pair of
appendages. The organs of the first kind are located in the
mandibular spines, the second on the smooth concave margin
of the inner mandible (fig. 3, g. 6.), and the third are scattered
over the surface of the mandibles between the spines.

(1) If the outer surface of one of the mandibles—say the
second or third on the right side—be very gently touched with
a piece of clam, as small even as the head of a pin, the cha-
racteristic chewing movements of that mandible and the corre-
sponding leg are alone produced. With care all the appendages
on one side, or any number of them, can be made in this way
to go through the chewing movements, while the other legs
and mandibles remain perfectly quiet.

(2) If any one mandible, or any combination of them, be
amputated, then stimulation of the mouth region with food
produces chewing movement of the remaining mandibles and
of the corresponding legs, but those legs from which the man-
dibles have been removed remain perfectly quiet, even though
food be rubbed on the mouth, the rostrum, the soft skin
about the base of the legs, or on the proximal part of the
leg itself. This proves that the gustatory organs are located
in the mandibles alone, and not in or around the mouth, in the
rostrum, or in the base of the leg.

(3) If the stiff spines be shaved off of one or more mandibles
and a piece of clam rubbed over the outer anterior surface of
the shaved mandibles, either no effect at all is produced, or
else feeble movements of the mandibles only, without the leg
movement. But the least contact of the clam against the
unshaved mandibles produces at once the characteristic man-
dible and leg movements ; it makes no difference whether the
shaved mandibles are all on one side or not, or what combination
may be selected. This proves that a large proportion of the
gustatory organs are situated in the mandibular spines ; it also

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 11

shows that the reflex in each leg and mandible is due solely to
the stimulation of its own gustatory organs, and that it is
entirely independent of the reflex in the adjacent appendages,
either of the same or opposite side: this, however, does not
apply to the chelicere.

(4) If the shaved mandibles be rubbed on their outer ante-
rior surface, movements are rarely produced; but they are
fairly well marked whenever a piece of clam is touched against
the smooth concave surface of the inner mandible (fig. 3,g.0.). In
performing this experiment, the shaved or unshaved mandible,
it is immaterial which, is gently raised with a pair of forceps,
care being taken not to touch the animal with the fingers, and
a very small piece of clam rubbed on the smooth surface in
question. Immediately the inner mandible is retracted by the
muscle shown in fig. 3, m. 7. m., and then the whole mandible
begins its rhythmic movements. This proves that, besides
those organs in the spines, the smooth inner surface
of the inner mandible contains a second set of
gustatory organs, which when stimulated produce
reflex chewing movements.

(5) The results of the following experiments may be stated
in this way :—Destroying a certain number of gustatory organs
in one or more mandibles suspends the reflex in the mutilated
appendages. The reflex may be partially restored by destroy-
ing a corresponding number of sense organs in each of the
remaining mandibles.

The following experiments were performed successively on
the same individual :—(a) A healthy crab that was known to
be very sensitive to gustatory stimulation was deprived of a
part of its gustatory organs by shaving off all the gustatory
spines on the mandibles of the right side. Ten days after, on
equal stimulation of both sides, the shaved appendages
remained perfectly motionless, while the unshaved
ones began the normal chewing movements. But the
shaved mandibles could be made to act when vigorously stimu-
lated. (4) When the mandibles on the left side also were
shaved, the reflexes were impaired for some time. However,

12 WILLIAM PATTEN.

a week or ten days after, vigorous movements on both
sides were produced by rubbing pieces of clam well over
the mandibles. As might be supposed from experiment No. 4,
the movements were especially well marked when the clam was
rubbed over the under side of the inner mandibles, but they
could be produced when only the outer part of the mandibles
was touched. After the reflex had been once restored
it required but little more stimulation to start the
reflex in the shaved mandibles than it did before
in the unshaved ones. (c) Now if we cut off the inner
mandible of the right side the reflex on that side will again
be suspended, but it can be once more partially restored by
(d) cutting off the inner mandibles of the other side. The
restored powers are each time feebler than before, but nothing
short of amputating both inner and outer mandibles will
completely and permanently destroy the reflex. The feeble
movements caused by stimulation after the spines and the
inner mandibles have been removed are produced by scattered
gustatory buds distributed between the spines over the anterior
face of the mandibles. The reflex chewing movements in one
appendage, therefore, are not lost in proportion to the destruc-
tion of sense organs in it, but depend rather on the relation
between the number of sense organs retained in it and those in
the other appendages. Hence if the reflex in one appendage is
suspended by destruction of a certain number of sense organs,
it may be partially restored by reducing the number of sense
organs in the other appendages in a like degree. In other
words, the reflex impulse enters the widest door.

The reflex flows more readily along the most recently used
lines, as shown by the following experiment :—If mandible a
be slightly stimulated with food, whether enough to produce
reflex movements or not, subsequent (say five minutes after)
stimulation of all the mandibles to the same extent will pro-
duce reflex movement in mandible a first, and afterwards
feebler movements in all the others.

B. Structure of the Gustatory Organs of the Man-

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 13

dibular Spines.—lIt is not difficult to find the three sets of
gustatory organs, the existence of which is demonstrated above.
Examination under alow magnifying power shows that the man-
dibular spines are thickly covered on their sides nearest the
mouth with minute pores arranged in from eight to ten inter-
rupted vertical lines (fig. 2). Each line consists of several
subordinate groups composed of from two to twelve or more
pores. Longitudinal sections of one of the spines (fig. 1) show
that the cuticula is perforated by parallel canals, in each of
which is a delicate chitinous tubule (ch. ¢.); the latter
terminates at the outer opening of the canal flush with the
surface ; at the opposite end it bends nearly at right angles
towards the base of the spine, where it soon expands into a
clear, spindle-shaped body (sp.); beyond the spindle it is
continued as a very long slender process that constitutes the
body of the gustatory cell, the nucleated end of which unites
with other cells to form great ganglion-like masses (gsc.).

The spindle, when more highly magnified (fig. 5), seems
to be merely an inflation of the cell-wall, and contains, besides
a watery fluid, a number of fibrille, arranged in a single layer
along its inner wall, and continuous with fibrille in the stalk of
the cell. The proximal half of the spindle is usually stained a
little darker than the rest, and in it each fibril expands into a
fusiform thickening that stains deeply in acetic acid carmine.
The fibrille converge toward the distal half of the spindle to
form an axial bundle that can be followed nearly to the free
outer end of the chitinous tubule. From two to eight slender
bipolar cells surround each spindle, and send their fibrous pro-
cesses outward along the outer surface of the chitinous tubule
(figs. 5, 6,9. ¢.).

It is difficult to tell just where the tubule begins. A short
distance beyond the spindle (figs. 5 and 6) it appears to be
continuous with the cell-wall, which there becomes rather
distinct owing to its separation from the cell contents. It is
slightly indented in some places, and is apparently enclosed
by a second membrane, probably the outward prolongation
of the nerve-like cells surrounding the spindle, for when

14 WILLIAM PATTEN.

thoroughly macerated these cells fall off, and the outer mem-
brane is then absent. The sharp inner wall, the investing
membrane, and the axial nerve produce a picture very much
like that of a Vertebrate medullated nerve.

In macerated preparations the cell usually breaks just
beyond the spindle, as in figs. 5—7, but in some instances
nearly the whole tubule is isolated. When the tubule
breaks near the spindle, the axial nerve usually projects a
long distance from its open end, either as a single fibre
(figs. 5, 6) or as a delicate brush of fibrille (fig. 7). Beyond
this region it breaks with a clean fracture, as though made of
glass, and a protruding axial fibre is rarely seen. Towards
its outermost end the axial nerve cannot be seen under any
circumstances, but this is due to the fact, I believe, that the
canal in the tubule is so small that it is completely filled by
the nerve. The tubule is thickest where it enters the cuticular
canal (fig. 1, s. ch. ¢.); but it becomes smaller and smaller, as
well as the canal in which it lies, until, at the surface, the tubule
just fills the canal. In surface views (fig. 2) the black dot is
the pore of the canal, completely filled by the tubule; the clear
halo surrounding it is cuticula more transparent than the
rest.

The stalk of the gustatory cells, just below the
spindle, is very finely striated, and when macerated and
broken, as in fig. 5, the striation is seen to be due to the
presence of excessively fine fibrille, much more numerous
than those in the spindles. There are no nuclei to be seen on
the long nerve-like stalk of the cell, between the spindle and
the nucleus. On the distal side of the nucleated part of the
cell are usually a few yellowish-brown pigment granules, that
in some cases are large and very numerous, as in fig. 4.

The nucleated ends of the gustatory cells are arranged in
long clusters along roughly parallel lines, each band of cells
corresponding to a line of pores. Judging from the number
of cells it contains, the cluster at gs. c., fig. 1, is probably
connected with a row of pores running the whole length of the
spine, although this is difficult to determine with certainty.

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 15

There are a few very large ganglion-cells, with coarse,
anastomosing, plasmodia-like processes, that run at pretty
regular intervals one above the other around the inside of
the spines (fig. 1). One may often isolate great masses of this
finely fibrillate reticulum without finding a nucleus in it. It
usually overlies and unites large bundles of the slender stalks
of the gustatory cells, but the latter appear to pass through
the plasmodium without change.

The body of the gustatory cells resembles that of the
double retinal cells of Molluscs and Arthropods, described by
me in Arca, Pecten, Acilius, Lycosa, and others, in
that the large nucleus is nearly always excentric, and a
spiral partition separates the cell into two more or less distinct
portions (fig. 12). In one lies the main nucleus, n!; in the
other a small unstained body that I regard as the aborted
nucleus of the second cell, n?.. The proximal end of the cell
is sometimes forked, each branch being continued into a
nerve-fibre. The interesting fact is thus established that
double cells are not confined to the retina.

The cuticular canal of the gustatory organ can be readily
distinguished from other canals by the presence of a peculiar
swollen knee, or bend, near its outer third, the surrounding
cuticula of which stains more deeply in borax carmine than
elsewhere (fig. 1). Under favorable conditions one can see in
the knee”? a kind of spiral thread, caused by what in some
cases appears to be a spiral ridge on the wall of the canal, in
others by a spiral nerve-like fibril that seems to:surround more
or less loosely the chitinous tubule (fig. 13). A similar spiral
thread is sometimes seen on the isolated chitinous tubule
(fig. 6, a); when treated with dilute potash the tubule swells
a little, and a thin, finely granular layer is seen about it,
together with what appears to be an extremely delicate fibril,
wound spirally about the tubule and its sheath (fig. 14).

In some of the cuticular canals there are a few slender
nerve-like cells loosely surrounding the chitinous tubule
(fig. 1); in other cases these cells seem to be absent. That
the tubules of the gustatory organs and of the others described

16 WILLIAM PATTEN.

below are chitinous is shown by their resistance to caustic
potash, and by the fact that they are shed during ecdysis. On
examining cast-off shells one can see the perfectly preserved
tubule projecting out of the inner ends of the cuticular canals.

Each spine on the jaws is crowded with the organs just
described, and contains also a large blood-vessel. There can
be no doubt that they are the gustatory organs, which, when
stimulated, produce the reflex chewing movements described
above.

c. Experiments on the Gustatory and the Tem-
perature Organs of the Chele—Two varieties
of organs, having nearly the same histological
structure and arrangement as those on the mandi-
bular spines, are found on the last two joints, or
chele, of the first to sixth appendage. One kind
is a gustatory, the other, in all probability, a tempe-
rature organ.

The presence of the gustatory organs may be demonstrated
by the following experiments :—Place a crab on its back and
allow it to become quiet; then if the chele, which are
usually lightly closed, are rubbed with a small piece of clam,
they will open wide, and remaining so, move about rather
vaguely, as though trying to grasp something. They are
specially sensitive at the tips and along the cutting edge.
Ammonia vapour will produce the same result, but it cannot
be produced by any purely mechanical stimulus.

The temperature organs betray their presence by an entirely
different action, for if one breathes very gently, or blows warm
air on the chele, they suddenly close and open once, and will
repeat the action without variation as often as they are
stimulated in this way.

A very curious fact is the following:—If the chele are
amputated at the penultimate joint, they remain perfectly
quiet if left alone. But for five or ten minutes after the
operation they will snap once, i.e. close and immedi-
ately open again, every time they are gently

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 17

breathed upon. But stimulation with food or
ammonia produces no effect whatever. These facts
seem to indicate that there is a reflex centre in the chelz for
the temperature impulses, but none for the gustatory ones.
A rather hurried examination, however, failed to show the
presence of any centre there, unless the few scattered tripolar
ganglionic cells found everywhere in the subdermal nerve-
plexus can be regarded as such.

Description of the Organs.—As one might expect
after the above experiments, surface views of the chele show
the presence of two kinds of organs. Those of one kind
appear as small pores surrounded by a clear halo, and resem-
bling, in their arrangement in lines and in every other par-
ticular, the gustatory organs on the mandibular spines (figs.
10 and 11, g.o.). Seesection C. The others are less numerous
than the first, but larger, and over each canal there is a
saucer-shaped depression, from which projects a short blunt
spine. A chitinous tubule passes up the wide cuticular canal,
and terminates in the spine. The same kind of organ is also
found abont the bases of the larger mandibular spines. As the
first organs are just like those in the mandibular spines, and
as no other organs are found near the tips of the chelz, they
are without doubt the gustatory organs. The second kind
must, in all probability, be the temperature ones. Sections of
the chelez, blackened in osmic acid to show the distribution
of the gustatory canals (Pl. 3, fig. 44), show that they are
very abundant along the cutting edge of the chele, and even
more numerous at the flattened apex of the fingers—in other
words, just where they are most sensitive to taste, and where
they would be most likely to come in contact with foreign
bodies.

The pedal nerve in the propodite, or the next to the last
joint, divides into four branches which run along the anterior
and posterior margin respectively of each arm of the chele.
Whether this division of the nerve is due to a separation of
fibres into nerves, going some to temperature organs and others
to gustatory ones, could not be determined.

VOL. 35, PART 1.—NEW SER. B

18 WILLIAM PATTEN.

II. Ouractory Organs.

A. Structure of the Olfactory Organ.—The olfactory
organ is visible from the exterior as an irregular yellowish-
brown, wart-like thickening of the cuticula, from 5 to 8 mm.
broad, and situated about 25 or 80 mm. in front of the
mouth. In specimens from 2 to 4 inches long it is usually
raised into a beak-like projection directed backwards.

Directly beneath the ectoderm are a great many—at a
rough estimate from 1500 to 2000—clear, flask-shaped sense
buds, each of which is connected by a narrow neck with a
cuticular canal. The distribution of the olfactory buds, as I
shall call them, is fairly well shown by surface views of the
olfactory region (Pl. 2, fig. 19). In this preparation, which is
probably from an adult male, there are two unusually well-
defined median elevations containing many more canals than
elsewhere. The olfactory buds underlying this portion are sup-
plied by a large median nerve (fig. 18, m. ol. n.); and this fact,
together with the method of development, shows that it con-
stitutes a distinct part of the olfactory organ. The lateral
portions are clearer and smoother, and contain comparatively
few canals; this is specially the case on the posterior lateral
borders immediately over the bulb-like termination of the
lateral olfactory nerve (fig. 19). In young individuals, 2—4
inches long, the cuticula of this part is more transparent than
elsewhere, aud looks like a small lens. This fact, together
with the presence of pigment there in the early stages, was
what led me, in my paper on the ‘‘ Origin of Vertebrates,”’ to
regard this organ as a degenerate pair of eyes.

In the adult the cuticula over the olfactory organ often
appears a dirty silvery white in reflected light, and black by
transmitted light, owing to the inclusions of air in the ex-
tremely minute “pore canals.” These “ pore canals” are
found equally abundant elsewhere, but they do not contain air.
The cuticula in the median part of the olfactory organ often
contains irregular cavities (fig. 19), as though some animal

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 19

had eaten into it, or it may contain a network of membranous
canals, evidently the tubes of some Annelid; they are usually
most abundant in the median portion of the olfactory organ,
and either lie on the surface or are buried in the cuticula.

Scattered over the olfactory organ are many blunt back-
wardly curved spines. The olfactory cuticula can be easily
peeled off in successive layers, but it adheres strongly around
the pores leading to these spines. When it does come away
large tufts of pigment-cells are seen projecting from the spine
pores, and the outer surface of the inner layer of cuticle pro-
jects in a crater-like collar around the pore. There are similar
spines on the cuticle surrounding the olfactory organ, but they
do not act like the ones just described. For this reason I sup-
posed at first that the olfactory spines were perhaps the true
sense organs supplied by the olfactory nerves, but I can find
but little evidence in support of this view. The large pores
leading up into the spines over the olfactory organ are lined
with thin cells and crowded with pigmented tissue, and in some
cases contain a transparent, fragmented coagulum. A chi-
tinous tubule, similar to that of the gustatory cells, usually
runs the whole length of the canal, and becomes continuous
with a minute canal extending from base to summit of the
spine. The latter is pinnately striated in section, as though
its central canal were connected with the exterior by innu-
merable radiating canals. Under favorable conditions a rather
large nerve may be seen to enter the base of the spine canal.
The spines are suspended in sockets, and are moveable. They
are undoubtedly of a sensory nature, but they seem to play
only a very subordinate part in the olfactory region, and, con-
trary to what I at first surmised, cannot be compared with the
large gustatory spines on the mandible.

A section through the olfactory organ (fig. 21) shows the
larger branches of the nerve-plexus arising, in the main, from
the median olfactory nerve; also the densely pigmented layer
of ectoderm confined to the olfactory region, the clear olfac-
tory buds (ol. 6.), the small clusters of dark cells looking lke
ganglia, and numerous branching blood-vessels. One layer of

20 WILLIAM PATTEN.

cuticula has been peeled off, so that below where the tooth-like
spines should be there are large pores with projecting crater-
like summits (s. sp.), containing many pigmented cells.

The olfactory buds vary a good deal in size and form.
They are usually spherical or pear-shaped, and composed of a
varying number of large pyramidal cells, the apices of which
sometimes surround a perfectly clear spherical lumen. The
appearance of the buds varies greatly according to the method
of preparation, and other causes not clearly understood. In
some they appear perfectly empty, so that the organs look
like so many blank spaces in the tissue, with only a few cell
outlines visible ; or, in organs isolated by maceration in Bela
Haller’s fluid, a few cells may contain a very delicate spongy
reticulum (fig. 9), while others in the same organ may be
densely crowded with refractive globules, so that they resemble
certain gland-like cells that I have found associated with
sensory cells on the tentacles and mantle edge of Molluscs,
such as Arca, Pecten, and Lima (‘ Eyes of Molluscs and
Arthropods,’ p. 722). A dark multipolar cell with a large
nucleus can usually be seen in the interior of the organ; it
looks like a ganglion-cell with two or more fibrous ends, the
course of which cannot be followed very far in sections. It is
undoubtedly the same dark ganglion-like cell so conspicuous
in the young stages of these organs (fig. 23).

The clear lumen seen in the younger specimens, the chitin-
ous, duct-like tubule, and the whole appearance of these
remarkable structures point toward their glandular nature.
On the other hand, their extraordinarily rich nerve-supply,
and the unquestionable derivation of the whole group of organs
from a primitive segmental sense organ, seem to show equally
clearly that they are sense buds.

Not till I was able to demonstrate experimentally that
similar organs in the inner mandibles were gustatory organs
did I feel satisfied that the “olfactory buds” were of a sensory
nature. I then studied them anew by macerating fresh mate-
rial, paying special attention to the central ganglion-cell and
its relation to the chitinous tubule. I did not succeed in

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 21

getting such perfect isolation as with the gustatory cells, and
there still remain some points of importance unanswered. But
the cardinal point at issue, whether the buds are glandular or
sensory, was settled beyond doubt, for I was able to demon-
strate that there is rarely a lumen in the fully formed buds,
and that when it does occur it does not communicate with the
exterior. Moreover I proved that the chitinous tubule cannot
be a duct to the gland, since it is in reality a direct pro-
longation of the central ganglion-cell, and may be
compared with the tubule in the distal end of the
gustatory cells. ‘Thus every reason for regarding these
sense buds as glands disappears.

The lumen of the buds varies greatly in its appearance. It
is most commonly present in the newly formed buds, where it
is spherical and sharply circumscribed (fig. 23). In the adults
it seldom has this appearance, and may be entirely absent, or
it may be reduced to a small irregular space between the cells.
Although I have looked carefully for them I have never seen
any of the clear globules of the gland-like cells in the lumen
of the gustatory buds. The lumen appears to be something
like a much-retarded invagination cavity, although, as nearly
as I can make out, the organs arise by a solid ingrowth from
the ectoderm.

The tubule can be followed in sections from near the centre
of the bud, through the cuticular canals, to a point very near
the outer surface ; here it becomes very faint or disappears. It
may be either straight, very much coiled, broken at intervals as
though it were very brittle, or may have one or more spindle-
shaped swellings. The tubule is undoubtedly composed of
chitin, for, as with the gustatory tubules, they can still be
seen in the cast-off shells of immature specimens and in the
fresh shells cleaned with potash.

The cuticular canals of the olfactory buds are easily dis-
tinguished from the gustatory ones by their shape. Each
canal in the adult has a nearly uniform diameter, except that
near the outer surface it suddenly narrows and communicates
with the exterior by a transverse slit (figs. 9 and 12, C and D).

22 WILLIAM PATTEN.

In the younger specimens they are slightly expanded near the
top, and a flange-like rim is formed by the projection of the
cuticula into the outer end of the canal (fig. 28).

The canals contain, besides the chitinous tubule, a varying
number of fibres, with here and there a few minute nuclei
(fig. 9, n. c.). They can be traced a short distance only
toward the outer end of the canal; in the opposite direction
they seem to run either over the outer surface of the olfactory
bud, or apparently between its cells toward the interior. I
have never been able to trace them with certainty up to
nerve-fibres, although they appear to have such connections.

The tubules isolated by maceration in Haller’s fluid stain
deeply in methyl green and in acetic acid carmine, resembling
the chitinous tubules of the gustatory organs. They are
usually collapsed, and appear to be perfectly empty. Al-
though I have examined them in many ways, paying special
attention to their broken ends for protruding fibres, I have
never seen a trace of the axial fibres so conspicuous
in the other gustatory tubules. Isolated tubules from
the gustatory buds on the inner mandible sometimes show,
when treated with potash, a protoplasmic-like envelope with
spiral markings (fig. 14); others have two or more coarse
refractive fibres, often thrown into numerous irregular folds,
extending along their outer surfaces (fig. 16, a). The tubules
in the olfactory organ are very rarely convoluted, and they
never have the two refractive fibres just mentioned. When
the olfactory buds are isolated and examined whole, spindle-
shaped cells are often seen adhering to their outer surface,
also a few scattered nuclei that appear to belong to the
delicate membranous investment of each organ.

The nerves that supply the olfactory buds are small
strands arising from an extensive anastomosing plexus found
everywhere beneath and around the olfactory buds. The
plexus itself arises from three large nerves, that I shall call
the lateral and the median olfactory nerves (Pl. 2, fig. 18;
see also Pl. 4, figs. 48 and 49).

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 23

The lateral olfactory nerves arise apparently from the
anterior part of the brain, but in sections one can follow their
roots on to the ventral surface into the optic ganglia (fig. 49).
In the adult the proximal ends of the nerves consist of coarse,
transparent nerve-tubes, while their distal extremities contain
many large ganglion-cells, which form an elongated swelling
at the tip of the nerve: the latter terminates abruptly just
beneath the cuticle on the lateral edge of the olfactory organ
(fig.49). The lateral olfactory nerve is accompanied by a large
blood-vessel that divides into numerous brauches, supplying the
tissues in front of the olfactory organ (fig. 18, 0/. v.) ; small
nerve-filaments accompany some of these blood-vessels, and
probably supply the ectoderm of the same region.

Several larger nerve branches leave the median border of
the lateral nerves a little distance back of the olfactory organ,
and mingle with the plexus formed by the median nerve
(fig. 18). Some of these branches terminate in small, rounded,
ommatidia-like clusters of large cells, which contain irregular
refractive plates like those seen on the borders of the young
lateral eyes. The lateral olfactory ganglion contains a great
many of these clusters of cells. A fair idea of their appear-
ance may be had from fig. 20, which represents some of them
at the root of the lateral nerve in an immature specimen.
Some near the tip of the lateral nerve, in a still younger
specimen, are shown in fig. 22 (g. ol.n.). The lateral ganglion
and the isolated clusters of cells are the remnants of a
primitive sense organ! derived originally from the margin of
the brain, but which is now converted into these ganglion-like
cells. They are retinal cells that have lost their pigment,
and now have all the appearances of ganglion-cells, except
that they still retain their rods. In other words, we have
caught a sense organ in the act of being transformed into a
ganglion—the only case of this nature on record, so far as I
know.

It is an extraordinary fact that the lateral nerve terminates
abruptly in this great mass of ganglion-cells, which are

1 See pp. 29-30.

24, WILLIAM PATTEN.

apparently neither connected with surface sense organs nor
themselves in a position to receive stimuli from without.

The median olfactory nerve is of an entirely different
nature from the one just described. If differs greatly in size
and complexity in different individuals, and if the supposition
shortly to be advanced is right, it is better developed in the
males than in the females. It arises long after the lateral
nerves (after the third larval moult) as an outgrowth from the
anterior wall of the median eye tube (Pl. 3, fig. 43). In the
adult itis a solid nerve composed of two portions, a distal and a
proximal one. The latter is composed of a mixture of nerve-
fibres and ganglion-cells. The nerve-fibres are not the
apparently hollow nerve-tubes seen in the lateral olfactory
nerves and elsewhere, but appear to be more solid and refrac-
tive, with a yellowish cast, and nuclei here and there. At
intervals throughout the proximal portion of the nerve there
are spindle-shaped ganglia composed of small densely crowded
and deeply stained nuclei, exactly like those in the fore-brain.
They vary in number and size, and may extend directly into
the brain-tissue at one end of the nerve, or up to the olfactory
organ at the other (fig. 18).

The distal end of the median nerve divides into many
diverging branches, which can be followed by means of a hand
lens to the posterior edge of the olfactory organ; they there
begin to anastomose, and form a dense plexus underlying the
olfactory region, but a considerable number of fibres extend
beyond the olfactory region to the neighbouring ectoderm.
Here and there the larger strands of the plexus contain small
groups of the dark-coloured nuclei, similar to those in the
brain; or the smaller strands may contain a single large
tripolar ganglion-cell, with granular protoplasm and a large
clear nucleus (fig. 9, g. ¢.).

A large blood-vessel accompanies the median nerve ; under
the olfactory organ it breaks up into numerous branches, some
of which are crammed full of blood-corpuscles that stain
dark red, and under a low power might be mistaken for
ganglia (fig. 18, 0. v.).

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 20

Termination of the Nerves.—On the terminal joints
of the exopodites to the abdominal appendages sense buds
like those in the olfactory organ are very numerous ; and as
there is little surrounding tissue, one can peel off the cuticle
organs, together with the after maceration in nitric acid.
When such a preparation is placed in dilute glycerine and
examined from the inner surface, the nerve-plexus underlying
the organs can readily be seen. Branches like those shown in
fig. 9 are seen anastomosing with one another in a most
intricate manner. In the olfactory organs one cannot obtain
such instructive surface preparations, owing to the crowding
together of the sense buds and the abundance of connective
tissue and blood-vessels; but sections and isolated sense buds
show clearly that the plexus is much the same as in the ab-
dominal appendages, only a little denser. In the olfactory
region the larger branches of the plexus usually lie a little
below the organs, but they may lie directly on their inner
surfaces, or may be squeezed in between adjacent buds. In
some cases large branches seem to penetrate into the interior
of the organ, but such appearances may be deceptive owing to
the crowding of the organs.

The nerves actually connected with the organ are delicate,
transparent, nucleated filaments, easily overlooked; they spring
from the coarse strands and spread over the surface of the
buds, as shown in fig. 9. I could discover no uniformity in
number, or any of that drawing out of the cells often seen
where nerves unite with sense organs. They can be followed
a short distance as very faint, but rather wide fibrillated bands,
some of which appear to dip down between the cells toward
the central cavity. Occasionally one sees an irregular, poorly
defined body containing several nuclei, from which arise a few
nerve strands that spread out over the surface of the bud (fig.
9, g. c.). They sometimes contain a single large multipolar
ganglion-cell, with dark granular protoplasm and a clear
nucleus. The exact method in which the nerve strands ter-
minate is very difficult to determine.

About the neck of the buds are numerous fibrous strands

26 WILLIAM PATTEN.

which extend outwards over the surface of the buds into the
large pore canals (fig. 9, n. c.).

There are some remarkable organs scattered about in the
subdermal tissue of the olfactory region of the adult (fig. 21, .).
They are irregular, usually oblong, spindle-shaped, or branched
masses of small cells, whose nuclei stain deeply. They are
consequently very conspicuous, and I at first took them for
ganglia connected with the nerve-plexus; but I have found
them in other parts of the body where there seemed to be no
plexus, so their nature is doubtful. When examined under a
high power the nuclei appear to be surrounded by concentric
layers of protoplasm, giving the whole mass a very character-
istic appearance. In some places the cells are so loosely
packed that they fail to touch one another; they then lose
their concentric striations, and appear like masses of blood-
corpuscles. Usually a blood-vessel passes through the centre
or along the side of the bodies in question. In cross sections
the central blood-channel is seen to be either empty, filled
with a dark coagulum, or crammed with blood-corpuscles,
which in some cases are difficult to distinguish from the cells
composing the surrounding tissue. The nuclei of the cells
usually arrange themselves concentrically about the blood-
vessel. In other cases these problematical organs contain
a central canal, or irregular space, through which passes one
or more nerve-strands. The nerve-strands may run over or
through these organs, dividing into several branches on the
way, but emerging at the opposite end without any apparent
diminution in size, and without any intimate connection with
the organ,

B. Physiology of the Olfactory Organ.—The anato-
mical features of the olfactory organs are sufficiently remark-
able to make any physiological experiments as to their function
of great interest and importance. Their similarity to the
gustatory organs on the inner mandibles might lead one to
expect that stimulating them with food would produce reflex
chewing movements. But although I have made repeated

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 27

attempts to stimulate the olfactory organ with various kinds of
food, with acids and ammonia, I have never succeeded in pro-
ducing any characteristic reflex movements in that way. Even
drops of strong hydrochloric acid or ammonia seem to have no
more effect than when applied to other parts of the body ; they
cause a slight start, nothing more.

In order to remove all doubt as to its glandular nature, I
have excised the olfactory organ, wiped its outer surface dry,
and then stimulated with electricity the attached nerves, but
have never seen any traces of a secretion, such as might be
formed provided it were a gland.

Although these negative results are a little surprising, they
do not render the sensory nature of the olfactory organ any
less probable; for we cannot expect every sense organ to give
on stimulation such beautifully ‘‘ diagrammatic reflexes ” as
those on the mandibles. However, I finally discovered that
electrical stimulation of the olfactory region produced at once
very remarkable leg movements, such as are never seen under
any other circumstances.! The experiment is not always suc-
cessful, but when it is, the moment the electrodes are
applied to the olfactory organs of the male, rapid
chewing movements of the mandibles are produced,
accompanied by vigorous snapping of the chelicere,
which may finally become rigid and stretched out
backward at full length. At the same time the
second pair of legs, which during all our preceding
experiments on the gustatory organs have remained
motionless, are now quickly and repeatedly flexed,
as though trying to hug or grasp some object and
force it toward the mouth; all the other legs remain
motionless. Stimulation of the region about the olfactory organ,
or along the median line between the olfactory organ and the
brain, or above the brain, may produce the same effect. It isa
remarkable fact, for which I can give no explanation, that one
does not always get the same results on stimulating the

" T subsequently observed similar movements of the second pair of ap-
pendages when the olfactory organ of a male was excised.

28 WILLIAM PATTEN.

olfactory organ, although there can be no doubt about the
character of the response when it does occur. For example,
some males would never respond, although repeatedly stimu-
lated at different times. Others would respond immediately
and at almost every stimulation. There was one male upon
which stimulation had at first no effect, but which responded
beautifully a short time after its mandibles had been stimu-
lated into action by rubbing clam upon them. Another would
not respond at first, but did after it had lain on its back in
the air for two or three hours. Again, it might happen that
repeated stimulation produced at first no effect, when sud-
denly the characteristic movements of the second pair of legs
and the cheliceree would begin of themselves. Subsequent
stimulation, however, failed to reproduce this response, al-
though the animal would start whenever the electrodes were
applied, showing that the current had passed through the
cuticula into the underlying tissue.

It is also a curious fact that persistent stimulation of both
sides will sometimes produce the characteristic leg movements
on one side only; then suddenly both sides, or perhaps the
opposite side alone, respond. When the mandibles were
shaved or amputated on one side, the opposite side usually
responds first on stimulating the olfactory organ.

In one female, in which all the mandibles of the right side
were amputated, stimulation of the olfactory organs produced
sudden raising of the second to the fifth legs on the left side,
and forcing of their tips toward the mouth ; the movements
of the second leg were most marked. The legs on the right
side remained motionless. This is the nearest approach I
have seen in the female to the movements so characteristic
of the male under the same circumstances. In every other
case that came under my observation (at least
twenty) stimulation of the olfactory organ of the
female produced only slight starts of the legs and
abdomeu.

These experiments point toward a double function of the
olfactory organ. The reflex chewing movements indicate its

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 29

association either as a tasting or a smelling organ with the
process of eating. But it is very difficult to explain why
these movements are not produced by direct stimulation with
food. On the other hand, the extraordinary hugging and
grasping movements of the second pair of legs in the males
clearly show that they are in some way functionally asso-
elated with the olfactory organs. Now it is well known
that these legs in the male are specially modified for grasp-
ing the female during copulation; and, as I have shown
that they are the only legs not involved in the reflex
chewing movements caused by stimulating the mandi-
bles, I can conceive no other explanation for these facts than
that the olfactory organis used by the males in detect-
ing the females. This is, moreover, a function for which
its position is well suited. That an organ for this purpose
must be present seems certain, for the males during the breed-
ing season hunt out the females and attach themselves to them
with great precision. It is hardly probable that this could be
done by means of touch or vision. While I can find no im-
portant difference between the structure of the olfactory organ
in the adult males and females, in the young I find that the
median olfactory nerve has an enormous ganglionic enlarge-
ment in some individuals and a much smaller one iv others,
and I suspect that the former are males and the latter are
females. Again, when sound males and females are put in
the same aquaria, the males usually attach themselves to the
abdomen of the females, but I have never seen a male whose
olfactory organ has been cut out (and I have had, at different
times, half a dozen such specimens) attached to a female. This
experiment might be conclusive if it were performed on a
larger scale. Unfortunately I did not have sufficient material
to do this.

I have tried several times to arouse movements of the second
pair of legs in the male by rubbing the olfactory organ with
fresh ova, and with the secretions of the oviduct, but without
success. Renewed experiments are necessary here, for I have
not given this aspect of the question the attention it deserves.

30 WILLIAM PATTEN.

c. The Development of the Olfactory Organs.—The
olfactory organs first appear in surface views as a pair of oval
ectodermic thickenings, the “primary olfactory organs”’ on the
lateral margin of the brain, just in front of the optic ganglion
(figs. 24, 25, 46, 47, 49, and 61). Each organ soon separates
from the brain and grows forward, leaving behind long, thick
strands of ganglion-cells, which constitute the lateral olfactory
nerves (fig. 25). As soon as the primary olfactory thickenings
are separated from the brain they become filled with white
pigment ; at the same time branched cells filled with white
pigment leave the thickening, and, extending under the ecto-
derm in all directions, form a gradually widening pigmented
plexus; that in the adult may be several square inches in
extent. In the young larve the pigmented or choroid
plexus is attached to the anterior edge of the primary olfac-
tory thickening by a stout stalk (figs. 46 and 47, p. st.). Hach
plexus lies beneath the ectoderm and a little in front of the
brain, which even at this stage it more than equals in size. At
a little later stage the two plexi become completely united
(fig. 45). In fig. 22, w. p., a part of the plexus is shown in
section under a higher power.

The primitive olfactory thickenings soon unite in the median
line to form an apparently unpaired organ lying some dis-
tance in front of the brain. Before this takes place the cells
constituting the thickening arrange themselves in irregular
clusters, and their walls develop those peculiar cuticula-like
thickenings that look so much like groups of visual rods in the
ommatidia. Some of theSe cell-clusters soon leave the main
thickening, and lie scattered about under the ectoderm, but
connected with the distal portion of the lateral olfactory nerve
by branching nerve-bundles (fig. 18, x); others remain in the
distal end of the nerve to form the terminal ganglionic swell-
ing. In specimens about two inches long there is nothing
left of the primitive olfactory thickening but an irregular mass
of large cells, mostly pear-shaped, in which the lateral olfactory
nerves terminate. There is nothing in the final position,
shape, or structure of these cells to indicate that they now

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 381

communicate with the exterior, or could function as sensory
cells, although there is every reason to believe that they are
derived from what were once sensory cells, like those in the
eyes.

The definitive olfactory organ does not appear till the
changes just described are nearly finished, i. e. in the larve
from one half to one inch in length. During this period buds
begin to appear in the ectoderm between what is left of the
primitive olfactory thickenings, or between what now consti-
tute the swollen ends of the lateral olfactory nerves. At about
the same time the median olfactory nerve arises from
the median eye-tube, near where it joins the brainas
an outgrowth from its anterior wall. Now this part of
the tube may be considered morphologically as a part of the
anterior neural wall or roof of the brain, just as a correspond-
ing part of the stalk of the pineal eye in Vertebrates or in scor-
pions might be regarded as a part of the brain roof (compare
Pl. 3, figs. 41 and 42). Moreover, as in their earliest stages
the median olfactory nerves contain numerous small ganglion-
cells which soon develop into two large irregular botryoidal
lobes, having the identical histological structure as the cere-
bral hemispheres; and as the nerve soon unites directly with
the cerebral hemispheres, from which it then appears to be a
direct outgrowth, I think we are justified in regarding
the median olfactory nerves and olfactory lobes as
outgrowths from the anterior wall or roof of the cere-
bral hemispheres.

Of course the fully developed lobes, as seen in fig. 17, do not
grow out from the cerebral hemispheres, but as the few cells
from which they arise do so, it amounts to the same thing. It is
evident that if the growth of the lobes and their separation
from the brain took place at the same time the olfactory lobes
would, asin most Vertebrates, make their appearance as massive
outgrowths of the cerebral hemispheres.

The olfactory lobes vary greatly in size in different indi-
viduals. They are relatively largest in immature forms from
about 3 to 6 inches long. They may consist of two very

oe WILLIAM PATTEN.

distinct botryoidal masses composed of irregular lobes of
small, densely packed, and deeply stained nuclei, each sur-
rounding a central mass of medullary substance (fig. 17). The
whole appearance of the lobes is very much like that of the
convoluted parts of the cerebral hemispheres (fig. 52). In
some cases there is only one large lobe, evidently formed by
the more or less complete fusion of two like those in fig. 17.
The stalk of the lobes in this stage is a solid column of small
nuclei, and passes without any perceptible change into the
cerebral hemispheres. The fusion is so complete that imme-
diately after reaching the brain all distinction of olfactory
nerve and cerebral substance is lost, for there is no trace of
any medullary strands or strings of cells that can be regarded
as roots to the nerve. After the young crabs reach a length of
from 5 to 8 inches the lobes are less conspicuous, breaking up
into spindle-shaped masses scattered along the middle third of
the nerve. ‘The distal third breaks up into many fine strands,
that form a plexus beneath the olfactory buds, from which the
buds are supplied. The proximal third shows a more or less
clearly marked division into two main strands, corresponding
with the olfactory lobes, and each strand goes to its respective
cerebral hemisphere. Thus, although I have called this nerve
a median nerve, it is in reality a paired one.

p. Nature of Olfactory Organs.—The primary olfactory
thickenings are undoubtedly segmental sense organs serially
homologous with the eyes, as first statedin my paper “ On the
Origin of Vertebrates from Arachnids,” p. 337. They cor-
respond exactly in position with the lateral eyes of scorpions
(P1.5, figs. 58 and 59), and in their histological structure they
show traces of ommatidia and retinal rods, like those in the
lateral and median eyes of Limulus. The degeneration of this
pair of eyes in Limulus was due probably to their being retained
on the under side of a broad shield-like carapace, where they
could be of little or no use. Their gradual degeneration and
conversion into ganglion-cells, and their subsequent incor-
poration with a new set of sense organs with an entirely

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 38

different function, can be followed in great detail, and furnishes
a most remarkable instance of change of function and structure.

A fact, however, of great morphological significance is the
striking resemblance between the structure and mode of
development of the olfactory organ in Limulus and that of
the so-called ‘frontal Sinnesorgan” of Phyllopods, as de-
scribed by Leydig, Claus, and others. Claus’s description
of this organ in Branchipus will serve equally well for Limulus.
The peculiar “kolbenformige” cells described by him as
originating from the brain; their position beneath the ecto-
derm in ommatidia-like clusters, and containing the refractive
“zinkage” needles; their position in relation to the median
eye, as well as their relatively late appearance, are much the
same as in Limulus. It seems to me there can be but little
doubt that the frontal organ of Branchipus, with its ganglion-
cell masses arising from the median part of the brain, corre-
sponds to the median nerve and median olfactory region of
Limulus; while the “ Kolbenzellen” organ, with its lateral
ganglionic nerves, corresponds to the lateral nerve and primi-
tive sense organs of Limulus. Of course at first sight the
appearance of the organ in Limulus is different from that in
Branchipus, but its fundamental structure and relations to the
brain are the same. These facts point conclusively to a much
closer genetic relationship between Limulus and the Phyllo-
pods than has been recently supposed to exist, and this sup-
position has further support in the similarity in the develop-
ment of the trifold median eyes. If a more careful compara-
tive study of the frontal organ in other Phyllopods—Apus,
for example—should confirm the above comparison, it would
settle once for all the vexed question of the relation of
Limulus to the Crustacea, and would furnish very strong
evidence of the common ancestry of Crustacea and the Arach-
nids in the trilobites.

III. Toe Gustatory Bups or THE MANDIBLES.

It will be remembered that our physiological experiments
VOL. 35, PART 1,—NEW SER. Cc

34 WILLIAM PATTEN.

demonstrated the existence of special gustatory organs on the
under side of the inner mandibles (4, p. 11). On examining
this place (fig. 8, g. 6.) one can readily detect just beneath the
smooth cuticle a yellowish granular mass, 7 or 8 mm. long and
2 or3 mm. deep. Sections show that the cuticle is perforated
with an immense number of canals, something like those in
the olfactory region. The canals contain chitinous tubules
extending into the yellowish mass, which consists of innu-
merable gustatory buds, apparently exactly like those in the
olfactory organ, only much more numerous, and densely
packed together many rows deep. The chitinous tubules are
coarser than in the olfactory organ, and many of them are
thrown into complicated folds, as in fig. 16, 6. I have wiped
off the surface over these organs very carefully, and stimulated
the organs and their nerves with electricity, but have never
seen a trace of any secretion appear there.

When treated with chromo-acetic osmic acid, and stained in
hematoxylin, most of the organs are darkened around the
base of the chitinous tubule, assuming the colour and appear-
ance of sensory tissues when treated with this reagent, while
the periphery stains a bright blue. It is possible that the
peripheral gland-like cells secrete a substance having special
powers to absorb certain chemical substances in the surrounding
media, and in this way the stimulation of the centrally placed
ganglion-cell is increased. But how the stimulus can reach
the organ through these long tubules, which in some cases are
much coiled, is not easily understood. Some of the older
and larger organs seem to be quite empty and dead; others
stain a dense blue-black in hematoxylin; while still others,
apparently young ones, show very little of this blue colour,
but stain dark brown in osmic acid, like ordinary sense organs.

The same kind of buds, but isolated, are found thinly
scattered over the surface of the outer jaws, between the bases
of the spines. It is probably these organs which, after the
spines and the inner jaws have been removed, produce the
faint reflex chewing movements referred to in our description
of the physiological experiments (d, p. 12).

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 30

IV. TEMPERATURE SENSE.

The whole body of Limulus is very sensitive to changes
of temperature. This may be easily demonstrated in the
following manner:—If a crab be placed on its back and
allowed to become perfectly quiet, one may grasp the appen-
dages or mandibles with forceps and gently move them about
without arousing the animal; or one may touch the upper or
lower surface of the carapace, or the gills, with any object the
same temperature as the air, but the instant one touches any
of these parts with the fingers, or drops water on them warmer
or colder than the surrounding air, the animal at once becomes
more or less agitated, and moves the appendages and abdomen
about in vain efforts to regain its normal position. There is
no way in which we can make the quiescent animals start more
quickly or violently than by very gently breathing on the gills
and under surface of the body, although quite violeut fanning
may produce no effect at all.

If, holding the head within about a foot of a crab, one blows
little puffs of warm air on the parts about the mouth, the
chelicere will snap with every puff. If the puffs of warm air
are made a little stronger the chelaria are brought forward,
and the chele of the first and third pairs of appendages close at
every puff; tie chele of the second pair meantime, in both
males aud females, remain motionless.

Whenever the sides of the cephalo-thorax of a quiescent crab
are touched with the fingers prompt movements of the appen-
dages follow, the legs opposite to the point of contact, and on
the same side that was touched, beginning first.

The gentle warmth of the hands held within two or three
inches of the sides of the cephalo-thorax, or from the face or
body when watching closely the experiments, usually produces
uneasy movements in crabs that were before perfectly quiet.

The temperature sense is very acute on the lateral margins
of the thorax and abdomen, and on the tips of the legs, and of
the abdominal appendages, being apparently most acute in the

36 WILLIAM PATTEN.

last-mentioned organs. The flat triangular area on the ante-
rior margin of the under side of the cephalo-thorax is unusually
blunt to temperature changes.

It is a remarkable fact that the regions so sensitive to
slight temperature changes can be touched with small wires
hot enough to singe the cuticula without producing any move-
ment. Butif aratherlargeiron, about 2 or3 mm. in diameter,
be held for a quarter of a minute on an abdominal appendage,
movements are produced, but they are evidently due to irrita-
tion of organs situated more deeply than those stimulated by
gentle breathing.

a. Course of Temperature Impulses.—The following
experiments show the course of the temperature impulses to a
temperature centre, located somewhere in the fore-brain region.

Experiment! A.—When a shallow longitudinal cut is made
on the ventral side through the skin along the lateral margin
of the right row of appendages, the temperature sense of the
carapace lateral to this cut is destroyed. On applying the
hand to the right side of the cephalo-thorax, either on its
dorsal or ventral surface, no movements are produced, but
when the same is done to the left side the legs on both
sides are set in motion. The roots of the great tegumen-
tary nerves (fig. 48, 2—6, a. m. p.) lie close to the ventral sur-
face along the outer margin of the legs, and as they are the
only ones severed by this proceeding, the experiment shows
that the temperature impulses travel centripetally
along the anterior and posterior hemal nerves of the
thorax, not in all directions through the subdermal
plexus. Itshowsalsothat the temperature impulses
not only pass up and down the crura on the side
stimulated, but on to the opposite side as well. This
is in marked contrast with the gustatory impulses which give
rise to reflexes in those mandibles only that are stimulated.

1 In my later experiments a ligature was drawn under the skin and tied
tightly around the nerve, thus obviating the disadvantage of excessive
bleeding.

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 37

Exp. B.—Cutting across the ventral cord just back of
the chelaria causes regular raising and lowering of the abdo-
minal appendages about twenty times a minute. They finally
come to rest, and are then left in an unnatural position, with
the right and left appendages of the same pair crossed over the
median line. The mandibles are pressed firmly together in a
sort of tetanus, the line of the meeting being irregular and un-
symmetrical. Strangely enough, stimulation of the mandibles
with food produced, in this instance, no regular movements of
mastication. But a warm hand placed on either side of the
cephalo-thorax produced immediate and simultaneous move-
ments of the legsof both sides. The temperature reaction
of the cephalo-thorax, then, is apparently not in-
fluenced in the least by section of the ventral cord,
but the respiratory, and perhaps the gustatory re-
flexes are strangely affected.

Exp. C.—In this experiment the crab operated upon had
already had the mandibles belonging to the second right appen-
dage removed, but was otherwise in perfect condition. A deep
median longitudinal cut was made, severing (as shown by
post-mortem examination) the post-oral cross-commissures of
the crura, and cutting through the junction of the crura in the
vagus region (Pl. 4, fig. 48). The animal lived nearly two
months in apparently good condition. It was then killed to
make sure of the direction of the cut. During this period it
ate with the normal movements of the mandibles, except that
they worked perhaps a little more slowly than usual, and the
cheliceree were not brought into action. On placing the
warm hand on one side of the cephalo-thorax of the
quiescent animal, responsive movements of the legs
on both sides were at once produced. The tempera-
ture reaction was unaffected.

Exp. D.—In this specimen the crus of the left side was sec-
tioned just back of the second pair of legs. The results were
very clearly marked. After about five hours, when the crab
had become perfectly quiet, placing the hand anywhere on the
left side of the cephalo-thorax produced no movements what-

38 WILLIAM PATTEN.

ever of the appendages back of the section, and only feeble
movements of those on the same side in front of it. On plac-
ing the hand on the right side, however, all the legs on that
side were immediately set in motion, and continued to
move as long as the hand was held there, but they
ceased immediately the hand was removed. These
results could be produced repeatedly without any perceptible
variation. After eighteen hours, during which period the crab
had lain on its back on a table without attention, the reaction
was less vigorous, but still very prompt and decided. There
was no movement of the mandibles after this period in response
to stimulation of the gustatory spines. Unfortunately I did not
try this at an earlier stage; neither did I try, as I should have
done, heat stimulation of the appendages back of the section
in the left crus. These experiments are sufficiently definite,
however, as regards the course of the temperature impulses,
for they show that, starting in the cephalo-thorax, they
travel inward along the hemal nerves to the corre-
sponding crus, which they ascend to the fore-brain ;
from there they must descend along both crura to the
pedal nerves.

If, as seems probable from this experiment, the temperature
centre is somewhere in the fore-brain, we ought to be able to
destroy all temperature reflexes by cutting both crura close to
the brain. This is very nearly what takes place, as shown by—

Exp. E.—In this specimen both crura were cut completely
across, just back of the second pair of appendages. There was
much loss of blood, and all the reactions were feeble. The
only spontaneous movements were those of the chelicerz and
second pair of appendages. Five hours after the operation no
response could be produced by heat stimulation of the sides of
the cephalo-thorax, but feeble movements of the legs
could be produced by breathing on them directly.
These experiments indicate the existence of subordinate
temperature centres in each crus (see also experiment on
amputated chele, p. 16), and prove that the main tempera-
ture centre is located somewhere in the fore-brain.

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 39

B. There is some doubt in regard to the position of the
Gustatory Centre. I did not have this point so much in
mind during these experiments following section of the crura,
and I did not always test for gustatory reactions. The centre
for the mandibular gustatory organs probably lies in the fore-
brain. One would naturally suppose, however, from the way a
given leg can be made to chew when its gustatory organs are
stimulated, that there was in each leg an independent gustatory
centre, which I at first supposed might be located in the
ganglion of the pedal nerves. However, if this be so, it is
hard to understand why sectioning either the ventral cord, or
the crura back of the fore-brain, should stop the reflexes, while
a median longitudinal section across their union in the vagus
region should have no effect (see pp. 37-8). There is also a
difficulty in the fact that I have amputated the whole leg,
including a portion of the crura with the attached pedal nerve
and its ganglion, and on stimulating the gustatory spines with
food have failed to produce reflex contraction of the leg,
although movements could be easily produced by applying
the electrodes directly to the pedal ganglion or nerve.

c. Structure of Temperature Organs.—It is not so
easy to identify the temperature organs as the gustatory ones.
However, I have found close beneath the epidermis, in all the
parts examined, including various regions on the upper and
lower walls of the carapace, the gills, the legs, &c., a loose
subdermal plexus of nerve-fibres and ganglion-cells. The
cuticula of all these regions is perforated with canals under
which are buds, in nerve-supply and structure like those in
the mandibles and in the olfactory region. As I can find
nothing else there that looks like sense organs, it is very
probable that these buds are the organs that are specially
susceptible to changes of temperature.

p. Function.—The temperature sense seems to be remark-
ably acute in Limulus. I know of no other Invertebrate that
approaches it in this respect. As Limulus spends the fall and

40 WILLIAM PATTEN.

winter months in deep water, where there is comparatively
little variation in temperature, it is hard to imagine what this
sense can be used for if not to aid the animal in mi-

grating to warmer shallow water during the breeding
season.

V. TactiLeE SENSE.

There are some wart-like sense organs about 5 mm. in
diameter on the endopodites of the abdominal appendages, to
the structure and function of which I have not given special
attention. They have already been briefly described by
Gegenbaur. The underlying cuticula is richly perforated
with peculiarly shaped canals, and chitinous tubules extend
into them similar to those in the gustatory spines. They are
sensitive to ammonia vapour, tactile impressions, and tempera-
ture changes, but stimulation of them in this way does not
produce any definite reflex action. Ifa camel’s-hair brush be
drawn gently across the terminal joint of the expodites to the
abdominal appendages, or over the region about the olfactory
organs on the abdominal appendages, prompt movements of
the legs and abdomen follow. This is not the case when the
basal joints of the legs, or the soft flexible skin about the
joints of the legs, or about the anus, are brushed.

VI. Summary.

If we review the results of our observations described in the
preceding pages, we see that there are two distinct kinds of
sense organs in Limulus. There are the single-celled sense
organs, each with a long chitinous tubule, the best representa-
tions of which are found in the mandibular spines and in the
chele of all the walking appendages; they are pre-eminently
gustatory: a slightly modified kind in the chelz probably serve
as temperature organs.

The second kind are rounded, solid clusters of gland-like
cells, containing a large multipolar ganglion-cell provided with
a chitinous tubular prolongation, the distal end of which

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 41

terminates in canals near the outer surface of the cuticula.
These sensory buds are distributed over the whole body, but
they are much more abundant in some places than in others.
They are as a rule innervated by delicate branches from an
everywhere present subdermal nerve-plexus, which is itself
connected with branches of the tegumentary nerves; but special
aggregations of buds may be supplied by special nerves, such as
the olfactory buds, and the gustatory cells and buds of the
mandibles ; both sets of the latter organs being supplied by
the two branches of the mandibular nerve (coxal nerve of
scorpions, supra-branchial nerve of Vertebrates).

All these sense buds or sense cells in Limulus multiply
by division. This process is easily studied in the buds of
the inner mandibles, and in the gustatory cells of the man-
dibular spines. The division begins at the summit of a
cuticular canal, and gradually extends down into the organ,
the process being the same in both sense cells and sense buds.
The chitinous tubule itself does not divide; a new one is
formed alongside of the old one; a longitudinal constriction
then appears at the summit of the old cuticular canal, dividing
it into two diverging arms, the separation of which gradually
progresses toward either the sense buds or cells, as the case
may be. I do not know how the latter divide.

There are at least five varieties of these two types that can
be recognised by their termination at the surface. (1) The
olfactory buds are connected with nearly straight canals, which
contract near the top into a very narrow slit (fig. 12, e and f).
(2) The gustatory buds of the inner mandibles are connected
with straight canals, resembling the ones just described, except
that they open out by excessively small pores (fig. 12, a, b, and c).
Just before reaching the surface the tubule expands into an
irregular, spindle-shaped body; beyond this it becomes
extremely small, but still it can be followed with certainty to
the outer surface, where it terminates in a very shallow de-
pression (fig. 12, a). A dark coagulum of some kind usually
fills the cuticular canal, and completely surrounds the chitinous
tubule. (3) The buds scattered over the surface of the man-

42 WILLIAM PATTEN.

dibles terminate in canals opening by a rather wide slit (fig.
12, d). (4) The gustatory cells always lie in bent canals
that gradually taper to a very fine opening (fig. 1). (5)
Finally, there is a set of canals like those of the gustatory
buds, except that they are capped by short spines of
varying shape and size, into which extends a chitinous tubule
(fig. 10).

The sense buds of Limulus have a certain resemblance to
the organs which in the Gephyreans have been described by
recent authors either as glands or sense organs. Organs
similar to those in Limulus, but generally described as glands,
are widely distributed in Arthropods.

The only thing resembling the gustatory cells in the mandi-
bular spines are certain organs in the extremities of the palps
and in the first pair of legs of Galeodes, as recently described
by P. Gaubert. They resemble the chitinous tubules seen in
Limulus, but histological details concerning them are so
meagre that it is impossible to identify them with certainty.

I will also call attention to the “ olfactory cones” of Mutilla,
as described by Franz Ruland. They consist of spindle-shaped
clusters of cells from which a delicate hyaline tubule runs to
the perforated summit of the overlying cone. At the base of
the tube is a spindle-shaped swelling with longitudinal stria-
tions resembling those on the spindle of the gustatory cells of
Limulus. The organ also contains a single large nucleus that
may be compared with that in the sense buds of Limulus.

The young stages of the sensory buds are much alike,
whatever the subsequent modification. They are probably
derived from the multiplication of a few ectodermic cells; but
however that may be, the cells soon arrange themselves to
form a rounded body with a small central lumen (fig. 23).
The buds in this early stage stain more deeply in osmic acid,
and have a different appearance from that of the older buds
(fig. 22). The large ganglion-cell and the halo around the
central cavity are much more conspicuous at this time than
afterwards. Moreover the ends of the cells that converge
toward the central cavity are provided with refractive rod-like

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 48

thickenings, which are highly suggestive of the rods on the
ommatidial cells of the lateral eye.

Very similar figures are seen in the inner mandibles of
immature specimens, as in fig. 8. Here one can see a distinct
fibrous process passing outward from the ganglion-cell, which
in this instance lies well to one side of the central cavity.

These buds, which in the young are found in all parts
of the body, and at first are everywhere alike, finally
undergo various modifications. Some may degenerate, some
may become olfactory buds, others gustatory, and_ still
others temperature organs. The resemblance of these buds to
ommatidia is so striking that we must include them both in
the same category. Wecan therefore reduce the whole
system of sense organs either to isolated sense
cells or sense buds, or aggregations of the same.
This agrees with my conclusion concerning the sense organs
of insects (see ‘ Zool. Anz.,’ 1890, Nos. 13, 14), where I main-
tained that the ommatidia of the compound eye were nothing
more than specialised cell clusters, which in other parts of the
body were supplied with various forms of spines or hairs, and
served as tactile, auditory, gustatory, or olfactory organs.
Moreover, in my earlier observations on the “Eyes of
Molluscs and Arthropods,” I showed that the eyes even of
Molluscs were composed of circles of cells or ommatidia, which
were also widely distributed over the surface of the body. In
Vertebrates we have evidence of the same condition. Isolated
sensory cells are there widely distributed, and, while essentially
alike in structure and appearance, have very diverse functions.
The olfactory and gustatory organs are but aggregates of
sense buds like those widely distributed over the body. The
presence of similar sense buds in the eye is shown by the
circles of rod-cells surrounding the cones.

Part I1.—On the Morphology of the Arthropod Brain.
The Brain of Insects and Myriapods.

Among the most remarkable features of the brain of Limulus

44 WILLIAM PATTEN,

are its various cavities, its cerebral hemispheres, and its infun-
dibulum. Although some of these parts can be identified with
a fair degree of certainty in insects, Myriapods, and Arach-
nids, the whole appearance of the fore-brain region in the
adult Limulus is totally unlike that of any other known Ar-
thropod. On the other hand, it bears such a striking resem-
blance to a Vertebrate brain that I believe no competent
person need hesitate a moment in picking out the correspond-
ing parts.

‘ Before one can understand the structure and development of
the brain of Limulus one must first have a clear idea of its
structure in those Arthropods in which the cephalic lobes are
present in their simplest and most primitive condition. I
shall therefore first call attention to certain features of the
brain of Myriapods, Insects, and Arachnids, that are probably
common to the brain of all Arthropods. I have not paid
special attention to the Crustacea, and I do not, unless ex-
pressly stated, include them in my speculations on the Arthro-
pod brain.

By a comparative study of the cephalic lobes of Acilius,
Blatta, Vespa, Hydrophilus, Scorpions, several species of
Spiders, and Limulus, specially prepared to bring out surface
contours, I am able to demonstrate that the cephalic lobes
of a typical Arthropod are composed of three distinct seg-
ments, each containing a segment of the brain, optic ganglion,
and optic plate; between the two latter is an invagination by
means of which the optic ganglia are more or less impeded.
These results have been in part confirmed by Wheeler and
Heider. Viallanes, who has made a careful study of the anatomy
and physiology of the adult brain, has quite a different con-
ception ofits structure. Embryological studies, however, do not
support his views ; they show that each of his three segments,
protocerebron, deutocerebron, and tritocerebron, comprise very
heterogeneous centres, not at all arranged according to their
morphological affinities. St. Remy, in his valuable work on
the brain of Myriapods and Arachnids, has followed Viallanes.
For lack of space I cannot review the works of these two

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 45

authors as I would like. But in what follows I shall try to
show that their basis of classification of the brain-lobes is
founded on vital misconceptions. They fail to recognise
the difference between the organs derived from the cephalic
lobes and those derived from the ventral cord, as well as the
segmental nature of the primitive optic ganglia in such forms
as Acilius and Scorpions, and their relation to the compound
eye. Moreover, when Viallanes subsequently attempted to
confirm his views by embryological study, he not only selected
in Mantis a poor type, but mistook the well-known trachea-
like invaginations for the ganglionic ones described by me,
and consequently he is quite right in asserting that they do
not give rise to any part of the optic ganglion.

Those who have heretofore touched on the development of
the brain of Insects have failed to appreciate the far-reaching
morphological importance of the ganglionic invaginations.
Korschelt and Heider, in their text-book of embryology, did
not understand their relations in Insects, Scorpions, and
Limulus. As I consider these invaginations the key to the
morphology of the Arthropod and Vertebrate fore-brain, I
shall try to explain in more detail my interpretation of their
significance.

In studying the development of the convex eyes great con-
fusion and difficulty was at first occasioned by the failure to
recognise that two distinct invaginations are sometimes pre-
sent, one for the optic ganglion, the other, less commonly
present, for the eye. Reichenbach and Kingsley made this
mistake ; both supposed that the purely ganglionic invagina-
tion described by them gave rise either to the whole (Kingsley)
or a part (Reichenbach) of the ommateum.

I was the first to show in Vespa that the two invaginations
are absolutely distinct, one giving rise to the three lobes of the
optic ganglion, the other to the compound eye. The latter
invagination, although very deep, probably does not close up,
and its outer or middle wall produce the corneagen, as I at
first supposed. Strangely enough, the invagination soon
straightens out, and the corneagen, as I found out subse-

46 WILLIAM PAT'TEN.

quently, is formed by the union, over the crystalline cones, of
two cells derived from the sides of each ommatidium. The
same process undoubtedly occurs in Crustacea.

T studied the subject again in Acilius, and fortunately this
insect furnished the most primitive and least modified type of
cephalic lobes yet described, and apparently the one from
which all others have been derived. I do not by any means
believe that the Coleoptera are the most primitive Arthropods.
But since we find the Acilius type of fore-brain clearly
repeated with more or less modification in other insects, in
Spiders, Scorpions, Limulus, and probably with still greater
modifications in the Crustaceans (although our knowledge of
the last group is too imperfect as yet to speak definitely about
them), it must be the nearest in structure to the ancestral type.

A. The Cephalic Lobes of Insects.—Basing our con-
clusions mainly on Acilius, we find that in insects the true
fore-brainis derived from the cephalic lobes, which
are composed of three segments. In each segment one
can readily distinguish a pair of brain-lobes, a pair of optic
ganglia, and two ocelli opposite each ganglion (PI. 5, fig. 57).
Between the ocelli and the ganglia are three pairs of invagina-
tions, which decrease in depth and extent from the first to the
third. The ocelli, after the closure of the invagination, still
remain on the margin of the cephalic lobes in their original
upright position.

The antennary neuromere is usually regarded as a part of
the cephalic lobes. I believe this is a mistake, as can be very
readily shown by comparison with Scorpions and Limulus. One
finds there, as well as in Acilius, no evidence whatever that it
forms part of the true cephalic lobes; but it is not possible to
show this without the proper material properly prepared. I
regard the antennary neuromere as morphologically post-oral
—it is unquestionably derived from the ventral cord of the
trunk. It moves gradually forward till it occupies a position
beside, or in front of, the mouth, constituting what I shall call
the mid-brain.

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 47

B. The cephalic, sympathetic, or the stomodzal
nerves are very important landmarks in this region. The
anterior stomodzal, or “frontal ’’ ganglion and its nerve arise
as a solid linear outgrowth from the median anterior wall of
the stomodeum. Its outer or proximal end is united by two
cross-arms with the lateral stomodeal ganglia situated either
on the median border of the antennary neuromere, or on the
cesophageal commissures (Pl. 5, fig. 57). This cross-commis-
sure I shall call the anterior pons stomodei (a. p. st.).
In the later stages it is usually shaped like an elongated V,
owing to the fact that the frontal ganglion, which is situated
at the apex of the V, is carried a long distance inward by the
growth of the esophagus. From the lateral stomodeal ganglia
the lateral stomodzal nerves extend inwards, one on either
side of the esophagus (fig. 57,/.n.). Their exact point of union
with the brain I have not been able to determine, because in the
forms that I have studied these nerves are either absent or im-
perfectly developed. However, from all I can gather about them
from the works of others, they must arise very near this point;
moreover that is their point of origin in Scorpions and Limulus,

The lateral and median stomodzal nerves expand at inter-
vals into ganglia united with one another by commissural
strands. These ganglia and strands are specially well developed
in Iulus, as shown by Newport (fig. 62).

The labrum, which in Acilius I have shown to be un-
questionably a paired organ, arising from the very anterior
margin of the cephalic lobes, is innervated, strangely
enough, from the roots of the anterior pons stomo-
dei, a condition which seems to prevail throughout the
Myriapods, Insects, Arachnids, and probably the Crustacea.
Moreover there appears to be present in a sufficient number
of cases to make it typical for the whole group of Insects a
small unpaired nerve directed outward and backward from the
frontal ganglion toward the labrum (fig. 53). It is obvious
that the innervation of the labrum is different from that of
any other Arthropod appendage, and I believe it has special
significance, as I shall presently indicate.

48 WILLIAM PATTEN.

The lateral stomodzal ganglia are usually situated on the
posterior median margin of the antennal neuromere, or some-
times on the cesophageal commissures. They are always united
with one another by a large transverse commissure that con-
tains, unlike the other post-oral commissures, many ganglion-
cells. In the young stages of Acilius (see ‘ Eyes of Acilius,
fig. 44) I have figured a segment of the median sympathetic
nerve which seems to be united with this commissure to form
a rudimentary ganglion, and in the fourth larval stage of the
lobster I have found a very distinct ganglion in the middle of
this commissure. No doubt this commissure ts characteristic
of the whole group of Arthropods ; it is well known in Myria-
pods, Insects, and Crustacea, and I have found it in Limulus.
Its apparent absence in the Arachnids is probably due to the
crowding together of the neuromeres in the mouth region.
This commissure, I am conviuced, also belongs to
the cephalic sympathetic system; but it may per-
haps contain fibres representing the cross-com-
missures of the antennal neuromere. I shall call
it the posterior pons stomodei.

Besides these nerves, most insects, as shown in the diagram-
matic fig. 53, are provided with a system of trunk-sympathetics
consisting of a chain of lateral and median sympathetic
ganglia; the latter, as I have shown elsewhere, is derived from
the ‘“ Mittelstrang” of Hatschek, and probably terminates
anteriorly in the posterior pons stomodzi.

c. The Convex Eyes and their Ganglia.—The convex
eyes form such an important part of the brain of Insects that it
is of great importance to determine exactly where they belong.
They are apparently so intimately associated with the fore-
brain that there seems little reason to doubt their derivation
from the cephalic lobes. But there are strong reasons for
supposing that they belong, not to the cephalic lobes proper,
but to the trunk, probably to the mid-brain neuromere.

That they did not belong to the cephalic lobes originally is
indicated by the fact that in Acilius not a trace of them appears

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS, 49

till long after the cephalic lobes, as such, have disappeared
(i. e. beginning of the pupal stage), but it is then impossible
to determine exactly their relation to the cephalic lobes. In
the Scorpion there is nothing comparable to the convex eye. In
Limulus they certainly are not derived from the cephalic lobes,
although as in insects the optic ganglion does have this deriva-
tion. My recent observations have shown that in Limulus the
convex eye arises farther forward than I at first supposed. In its
earliest stages it lies about opposite the cheliceral segment to
which it in all probability belongs; its subsequent union with
the optic ganglion of the cephalic lobes is therefore a secondary
affair. The same thing apparently takes place in Acilius, for
the convex eyes when they appear at the beginning of the
pupal stage, instead of being provided with a special
ganglion of their own, become united with the
ganglia of the degenerative ocelli.

These facts suggest a very interesting comparison with
Myriapods. ‘There seems to be present in nearly all Myriapods
a remarkable nerve arising from the optic ganglion and supply-
ing a peculiar sense organ situated at the base of the antenne.
St. Remy calls them the nerve and sense organ of Tomosvary
(fig. 62). Now I strongly suspect that this sense organ is the
rudiment of the convex eye of the higher Arthropods, for it
agrees in two important particulars with the convex eye of
Limulus, and presumably with that of all other Arthropods,
namely, in its situation at the base of the antenne, and in the
attachment of its nerve to the ganglion of the larval ocelli.

In Acilius, the compound eyes appear at the beginning of
the pupal stage as a sickle-shaped band on the dorsal and
median margin of the ocelli. They finally break away from the
surface, and their degenerated remains become attached to the
under side of the optic ganglia (fig. 60). The latter persist,
and form the ganglion of the compound eyes. The conversion
of the larval optic ganglion into that of the imago is brought
about in the pupal stage by rapid growth along three very
clearly marked regions (fig. 60, 0. g. 1—8). In all three
bands, or centres, which probably represent the three segments

VOL. 35, PART 1.—NEW SER. D

50 WILLIAM PATTEN.,

of the larval ganglion, there is rapid multiplication of cells,
but it is the middle lobe which increases most rapidly, and
which forms the main part of the adult ganglion; traces of the
other two bands may be seen even in the adult (see my paper on
the “ Eyes of Vespa”’). Now it is a remarkable fact that we
find these three identical bands of the pupe of Acilius in the
embryos of Vespa, but they are there formed by a
direct invagination of the ectoderm of the cephalic
lobes. Either the same kind of an invagination or a solid in-
growth is found in a variety of other forms, as in the Hymen-
optera, Orthoptera, and Hemiptera—that is, in forms that do
not pass through a free ocellate larval stage. This
proves that the Acilius type of cephalic lobes is the most
primitive, and that in such forms as the Hymenoptera and
Orthoptera, &c., important embryological processes are omitted ;
it also shows that the optic ganglion of the convex eye of
insects is formed by the fusion of the three larval ganglia.

As to whether the three frontal ocelli found in many adult
insects are derived from the larval ocelli, or are new for-
mations, like the compound eye, is a question of great
morphological importance, but we have as yet no evidence upon
which to found an opinion concerning them. Their position
and number suggest their identity with the three-lobed median
eye of Limulus and Crustaceans. Careful investigation of
some forms, the Sialide, for example, which pass through
an ocellate larval stage and possess frontal ocelli in the adult,
would probably settle this question. It is certain that early
in the pupal stage of Acilius and Cecropia all the larval ocelli
break away from the ectoderm and take up their position on
the under side of the optic ganglion, where they seem to
undergo complete degeneration.

II. Tot Brain or ARACHNIDS.

A. The cephalic lobes of Arachnids have at first the
same shape and appearance as in insects. They soon divide
into three segments, which can be identified with even greater

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 51

ease than in Acilius, especially the three invaginations of the
optic ganglion.

In the first segment of Scorpions! the ocelli, as well as the
distinction between brain-lobe and optic ganglion, have dis-
appeared. The invaginations, which in Acilius are separate,
here unite to form a great transverse furrow with thick walls and
small dark nuclei (fig. 58, s./.); the whole lobe sinks below the
surface, and moving backward, lies underneath the second
segment, where it forms the semicircular lobes (l’organ
stratifié, l?organ pédunculé), identical in almost every
particular with the semicircular lobes of Limulus and Spiders.
The second pair of invaginations are at first exactly like those
in insects; but the fold on the Jateral margin of the invagina-
tion soon advances rapidly inwards and backwards, and uniting
in front with a similar fold in the first segment, and behind
with one in the third, gradually extends backwards, in a broad
amnion-like fold, over the whole of the cephalic lobes. During
this process the eyes on the second segment are gradually
infolded, so that they finally lie inverted on the middle
wall ofthe fold. They then unite with one another
over the median line, and growing forwards come to
lie in a common sac at the distal end of ashort tube
(fig. 42). The eyes of the third segment are not involved in
the ganglionic invagination, consequently they remain upright
on the surface ectoderm, as in Acilius. Thus almost the
whole of the cephalic lobes are invaginated, or, to speak more
accurately, they are enclosed by amnion-like folds to form a
primitive cerebral vesicle. The floor of the vesicle is
formed by the fore-brain or the whole of the cephalic lobes,
except that part containing the lateral eyes. The vesicle has
a thin roof or pallium, from which arises a median tubular
outgrowth, with its terminal pineal eye. There is nothing
approaching this in any other animals, except in the Verte-
brates, where this condition has long been regarded as typical
of the group. I am convinced that this fact alone, when

* See also my figures of the cephalic lobes of the Scorpion “On the Origin
of Vertebrates from Arachnids,”’ vol. xxxi, part iii, of this Journal.

52 WILLIAM PATTEN.

looked at without prejudice, is sufficient proof of the genetic
relation between the Arachnids and Vertebrates. It certainly
is not inferior to evidence based on the presence of a notochord
or gill-slits. . However, this fact only furnishes the key-note
to the argument that is to follow.

In adult Scorpions and Spiders the brain shows more
clearly its origin from three segments than is the case in
insects. In fig. 61 I give a diagrammatic view of the brain of
Arachnids based on my observations on Scorpions and Spiders.
None of the tegumentary nerves are represented, and, excepting
the position of the lateral ocelli, the diagram would do as
well for one as for the other. I was the first one to point
out the homology of the eyes of Spiders with those of
Scorpions and Insects in my paper on the ‘‘ Segmental Sense
Organs of Arthropods.” I stated, p. 601, “‘In Spiders the
structure of the cephalic lobes is the same as that of Scorpions.
The two anterior median eyes belong to the second segment,
and are homologous with the median eyes of Scorpions, the
development being the same in both cases. The three
remaining pairs belong to the third segment, and are homo-
logous with the lateral eyes of Scorpions. They are invaginated
to form optic cups in the same way as those of Acilius.”
Korschelt and Heider have overlooked this statement, for they
credit Kishinue and Purcell with having shown the similarity
in the development of the median eyes of Scorpions and the
anterior median eyes of Spiders, although their papers did not
appear until two or three years after mine.

If further confirmation of this interesting fact is necessary,
it is found in the structure of the adult brain. In Scorpions
the optic ganglia of the second and third segments remain
distinct through life. They are carried by the movements of
the eyes on to the anterior face of the fore-brain, where they
remain as two pairs of conical projections, the anterior pair
united with the nerves to the median eyes (fig. 61). In the
brain of some Spiders, notably Epeira, according to St. Remy,
exactly the same condition prevails.!_ But the anterior ganglia

1 It must be remembered that the terms superior, anterior, and posterior,

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 53

supply the median eyes, or “ Hauptaugen ; ” while the others,
irrespective of their position on the head of the adult, are
supplied by the posterior ganglion, or that on the third segment
of the cephalic lobes (fig. 61).

The brain-lobes of the second and third segments gradually
become indistinguishable. In their place, in the adult, one
sees a pair of deeply stained lobes (c.’., fig. 61), which are
probably homologous with the cerebral hemispheres of
Limulus.

B. The Mid-brain.—There can be no doubt whatever that
the whole of the cephalic lobes of the Scorpion are homo-
logous with the whole of the cephalic lobes of Acilius. The
next neuromere, that of the chelicerz, must be homologous
with the antennal neuromere of insects, because both bear
identical relations to the stomodzal nerves.

In Scorpions and Limulus there are four nerves connected
with the cesophagus. The most important are the lateral
stomodeal nerves, extending from about the middle of the
median margin of the cheliceral neuromere inwards along the
sides of the cesophagus (figs. 59—6], J. st. n.). What I
formerly regarded as the pre-oral cross-commissures of the
cheliceral neuromere (a. p. st.) I am now convinced is merely
the much shortened anterior pons stomodzi of insects. It is
composed at first of numerous ganglion-cells surrounding a
cortical ‘‘ punct” substance, and in some Spiders, according to
St. Remy, contains in the centre a ganglionic enlargement
that I regard as the remnant of a frontal ganglion. The same
commissure, called by St. Remy “ pons stomatogastique,”’
occurs in Myriapods (fig. 62, a. p. st.). It there strongly
resembles the anterior pons stomodei of Arachnids, and yet
shows, by the presence of a distinct median ganglion, its
undoubted homology with the cross-arms of the frontal ganglion
of Insects (compare figs. 58—62).
when applied to the embryos, have entirely different meanings from those
when applied to the adults. This is owing to the doubling of the cerebral

lobes on to the back of the adult, so that the anterior border of the cephalic
lobes becomes the posterior border.

54. WILLIAM PATTEN,

An unpaired rostral nerve extends outward and back-
ward along the anterior ventral wall of the esophagus into
the rostrum (figs. 59—61, r. n.). In Limulus, perhaps in
Scorpions, there are also two lateral rostral nerves arising
from the lateral stomodzal ganglion, and extending outwards
to the rostrum (/.7..). They agree in origin and distribu-
tion with the lateral labral nerves of Insects, with which they
are homologous.

In Limulus the first post-oral cross-commissure is much
longer than the rest, and occupies a different position on the
posterior under wall of the cesophagus (figs. 46, 47, c*.). It is
difficult to account for this commissure if we do not regard it
as homologous with the posterior pons stomodzi of the
cesophageal ring of Insects. I cannot identify it in Scorpions
on account of the excessive crowding of the neuromeres about
the mouth, but there is no reason to doubt that it is present.

c. Development of Lateral Stomodeal Nerves.—
The lateral stomodzal nerves in Scorpions and Limulus
arise in part from the walls of the esophagus. Their develop-
ment is best studied in scorpions. They first appear in surface
views of Stage E (‘Origin of Vertebrates from Arachnids,’
fig. 2, sé. n.) as a pair of invaginations on the median border
of the cheliceral neuromere. The invagination gives rise to a
string of cells which at its inner end is continuous with an
evagination of the lateral wall of the cesophagus. By the in-
ward growth of the csophagus the nerves are gradually drawn
out to their full extent; their inner ends are for a long time
continuous with the proliferating thickening of the stomo-
dzum, and their outer ends terminate in the lateral stomodzal
ganglia derived from the thick-walled invaginations on the
margin of the cheliceral neuromere.

Thus, if we combine our observations on scorpions and
Limulus, we are able to identify in the Arachnids every cha-
racteristic nerve and ganglion of the stomodzal system of
Insects except the unpaired stomodeal nerve. It is not im-
possible that further observation on other forms will show the

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 59

existence of this nerve in Arachnids. It is not easy to over-
estimate the importance of these facts. They prove be-
yond doubt that there has been no addition or sup-
pression of neuromeres about the mouth either of
Insects or of Arachnids.

It is also obvious that in Arachnids the fore-brain, the cheli-
ceral neuromere, and the principal stomodzal nerves are respec-
tively homologous with the fore-brain, the antennal neuro-
mere,and stomodeal nerves of Insects and Myriapods.
The “ pons stomatogastric” of Myriapods is homologous with
the frontal ganglion of Insects, and the cross-arms uniting it
with the ganglion of the cesophageal commissure, or what I
have called in Insects and Arachnids the anterior pons stomo-
dei. In spiders the same organ is called by St. Remy the
“rostral ganglion,” or “lobe rostral.” The lateral rostral
nerves and the labral nerves throughout Myriapods, Insects,
and Arachnids, including Limulus, are the same, for in all
these cases they arise either very near to or directly from the
lateral stomodeal ganglion. It is hard to understand how St.
Remy, after making his careful study of the brain of Myria-
pods and Arachnids, could overlook, as he seems to have done,
these obvious homologies. It was probably due to the false
a priori assumption that the antenne are absent in Arach-
nids, and that their chelicere correspond to the mandibles of
insects.

We are thus able to reduce the fore- and mid-brain of
Myriapods, Insects, and Arachnids to the same ground plan.
Such a comparison has been heretofore impossible, owing to
the absence of the necessary embryological data. This,
however, has been no serious obstacle to those who find it so
easy to account for all discrepancies by assuming that one or
more neuromeres have been omitted as the necessities of the
comparison demanded. But there is apparently no biological
law more constant than that head and anterior trunk segments,
once formed, are never entirely omitted, or new ones intercalated
between them ; and a careful study shows that the head of
Arthropods offers no exception to this law. On the other

56 WILLIAM PATTEN.

hand, to assume that there have been such changes raises an
insurmountable barrier to any satisfactory comparison between
the brain of Arachnids and other Arthropods. There is no
reason to doubt that the brain of Crustacea will fall in line
with the above comparisons.

p. Comparison with Annelids.—Still another advantage
to be derived from my interpretation is that it enables us to
compare the cephalic lobes and stomodezeal nerves of Arthropods
with those of Annelids. According to Lang, the paired
stomodzal nerves of Annelids arise from the cesophageal
collar, that is near the point of union of the ventral
cord with the brain, just as they do in Arthropods.
According to Kleinenberg, they originate in Lopadorhynchus
from the lateral margin of the first post-oral neuromere. The
origin of this nerve, therefore, in Annelids and in Arthropods
marks the point of union of the neuromeres of the head and
trunk. Moreover, since the stomodzum nerves arise from the
first post-oral neuromere of Annelids and from the first
neuromere of the ventral cord of Arthropods, i.e. from the
antennal neuromere of Insects and Myriapods, the cheliceral
of Arachnids, and the first antennal of Crustacea, these
neuromeres must be homologous.

gr. Nature of Stomodzal Nerves.—The remarkable
mode of development of the stomodeal nerves from the walls
of the stomodeum, their constant presence, and their
voluminous size in the early embryonic stages emphasises
their importance, and indicates that they belong to a system
quite apart from that of the fore-brain and ventral cord. If
so, what is their significance? The stomodzum undoubtedly
represents the invaginated ectoderm formerly surrounding a
primitive mouth leading directly into the mesenteron. That
primitive mouth now lies at the junction of the mesenteron
with the stomodeum. We have only to turn the stomodeum
back to its primitive condition as ectoderm surrounding the
mouth, in order to obtain a clearer idea of the original position

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 57

of the stomodeal nerves. I have made this change in the
typical insect nervous system shown in figs. 54, 55, and 56,
and the transformation is very suggestive. With only slight
modifications of the nerve-rings back of the cesophagus, the
pontes stomodei, with the frontal and the lateral stomodeal
ganglia, are seen to form a circumoral nerve-ring ; and by
continuing the nerves, uniting the lateral sympathetics, p. p.n.,
till they meet on the dorsal side between the body and the head
our ideal insect embryo appears like the larva of Lopadorhyn-
chus, or of an advanced trochosphere; the stomodeal nerves
forming on the sub-umbrella a system of circumoral nerve-
rings, such as we might expect to find in a Celenterate. The
labrum is carried forward by the evagination of the esophagus
to the umbrella, and its nerves now appear as umbrella nerves
originating from a circumoral nerve-ring.

This transformation renders the homology of the labrum
with the pre-oral antenne of Annelids, as suggested by
Korschelt and Heider, very plausible, and at the same time it
explains their remarkable innervation from the stomodzal
ganglia. The connection of the anterior ends of the sympathetic
nerves of the trunk is not known. If they are connected with
the system of stomodzal nerves, as indicated in the figures,
they might be regarded as sub-umbrella nerves, drawn out to
their present extent by the growth of the trunk. The
median sympathetic nerve would then be antemeric with the
unpaired stomodeeal nerve, as in fig. 54. It is obvious, on
inspection of the figures, that the invagination of the sub-
umbrella must have been greater in front than elsewhere,
for the lateral stomodzal ganglia are carried only to the edge
of the permanent mouth, while the frontal ganglion is carried
far into the esophagus, bringing the labrum up to the anterior
border of the mouth.

The advantages of this view are obvious. It affords a means
of identifying, approximately, the ‘ trochosphere” in the
cephalic lobes of Arthropods; explains, among other things, the
anomalous innervation of the labrum, and the limitation of the
stomodeal nerves to the ectodermal portion of the alimentary

58 WILLIAM PATTEN.

canal. J maintained, in my paper on “ Acilius,” that the
segments of the cephalic lobes of Insects were originally post-
oral, and comparable with those in the trunk. While I still
recognise the great similarity of trunk and cephalic segments,
I do not now believe that they are morphologically identical.
The constant presence of the larval ocelli, and the absence of
typical appendages, besides other considerations, indicate a
sharp demarcation between the segments of the cephalic lobes
and those of the body.

III. DEVELOPMENT OF THE BRAIN oF LIMULUS.

A. The cephalic lobes of Limulus, before they divide
into segments, resemble more closely in shape and general
appearance those of Insects and Scorpions than those of Crusta-
ceans. We find there the same three pairs of invaginations
seen in Acilius and Scorpions, but so strangely modified as to
be at first sight hardly recognisable.

The first change that takes place is the formation of a great
furrow along the whole of their anterior margin. As the
furrow deepens, the lobes become a little shorter, and assume
the shape shown in fig. 61; at the same time they come to be
differentiated into four distinct swellings. The lateral ones
constitute the optic ganglia (op. g.), or, as I shall sometimes
call them, the thalamencephalon, owing to their double
relation to the nerves of the lateral eyes and olfactory organ,
and to the way they finally become incorporated into the adult
brain. In front of the optic ganglia the furrow becomes
deeper and broader, and constitutes the invaginations of the
optic ganglia. Just in front of the latter, and united with
them by thick nerves, is a pair of oval thickenings, the primitive
olfactory organs (p. ol.0.). The compound eyes, with which this
optic ganglion is subsequently united, have not yet appeared.
They are first seen near the outer end, and a little back of, the
ganglion, as though they belonged to the cheliceral segment.

The optic ganglia of Limulus evidently correspond to the
optic ganglia of the third segment of Scorpions and Insects,

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 59

but the opening to the invagination is at right angles to the
longitudinal axis of the body instead of parallel with it, as in
the early stages of the Scorpion (fig. 59). But even in Scorpions
it is at right angles with the long axis of the body at a later
period. The sense organ on the anterior margin of this
invagination in Limulus, the primitive olfactory organ, must
be homologous with the lateral eyes of Scorpions. In the
anterior median part of the cephalic lobes another enlargement
of the furrow appears (s. /.) ; its walls are thicker and stain
deeper than elsewhere, forming two oblong, slightly thickened
lobes, very conspicuous in surface views, and having the same
appearance as the semicircular lobes of Scorpions and Spiders,
with which they are unquestionably homologous. This is
shown not only by their position and appearance in the early
stages, but by the fact that they develop into the same kind
of organs in the adult. Back of the semicircular lobes are
two poorly defined swellings (d7.? and 6r.°), that I regard
as the second and third segments of the fore-brain. From
the anterior lobes arise the cerebral hemispheres; the
posterior ones do not undergo any marked specialisation, they
are concealed by the subsequent backward growth of the
cerebral hemispheres, and seem to bear the same relation to
the rest of the fore-brain that the “‘ tween-brain ”’ does to the
cerebral hemisphere in Vertebrates.

About midway between the invagination of the optic ganglia
and that of the semicircular lobes appears a small pore, which
rapidly moves forward and inward till it lies in front of the
invagination of the cephalic lobes. The position of this pore
at first made me regard it as belonging to a segment in front
of the semicircular lobes, as I stated in my paper on the
“Origin of Vertebrates from Arachnids.” Further study,
however, convinced me that it originates between the invagi-
nation of the semicircular lobes and that of the optic ganglion,
and consequently it belongs to the second segment of the
cephalic lobes. The pore leads into a short tube formed by
an invagination of the ectoderm. The tube, although easily
seen in sections, is not very clear-cut, and the lumen extends

60 WILLIAM PATTEN.

a short distance only into its interior. The distal end of the
tube is not specialised, and shows no trace of the eye that is
subsequently developed fromit. These two tubes soon unite to
form a common tube opening by a large pore, situated some
distance in front of the brain (PI. 3, fig. 24). Two bands of ecto-
derm, in which there is a shallow groove, lead on either side to
the point where the original invaginations of the eye-tubes were
situated (fig. 24, c.m.e.t.). The distal end of the tube is now
slightly swollen, and constitutes the “anlage” of the median
eye. The paired invaginations of the median eye-
tubes in Limulus, therefore, are homologous with
the invaginations on the second segment of Scor-
pions, with which they agree in their relation to the other in-
vaginations. They are, however, much smaller than those of
Scorpions, and instead of advancing backward and inward over
the cephalic lobes, they advance forward and inward to a
point in front of the brain, where they unite to form a single
invagination extending right into the yolk, and apparently
having no connection with the cephalic lobes. A comparison
of figs. 24 and 25 and the sections shown in figs. 41 and 42
will show that the opening into the median eye-tube really has
much the same position in Limulus as in the Scorpion, so that
there can be no doubt that the median eyes in these two forms
are homologous. The strange position of the median eye-tube
in Limulus is caused partly by the forward migration of the
invaginations, and partly by the shortening of the cephalic
lobes, which brings all the invaginations into a nearly straight
transverse line, instead of being distributed along a broad
semicircular curve, as in Scorpions. Moreover the rapid in-
vagination of the semicircular lobes and their backward growth
under the rest of the brain have something to do with the
apparent separation of the median eye-tube from the brain
(figs. 41 and 42).

A narrow commissure, similar to that seen in Myriapods
and in the early stages of Acilius, unites the right and
left halves of the cephalic lobes (compare figs. 57, 59, 61,
G26).

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 61

B. The Mid-brain.—The first post-oral neuromere, that
of the cheliceree, becomes intimately united with the cephalic
lobes, as in Insects and Scorpions. Its stomodzal nerves, with
one exception, are identical with those of Scorpions. There
isa large “lateral stomodeal ganglion” on the median
margin of the cheliceral neuromere, from which two large
lateral stomodeal nerves extend inward the whole length
of the stomodzum (figs. 43, 46, 49,61, st. n.). There are also
a small median and two lateral nerves extending outward and
backward along the cesophagus to the rostrum, comparable
with the paired and unpaired labral nerves of Insects and
Myriapods (figs. 47 and 48). The first post-oral commissure
differs from the others in that it is isolated from the rest, and
forms a long loop around the posterior under side of the ceso-
phagus. The other commissures extend straight across, and
are bound together by firm connective tissue. I have not
traced the ends of this commissure up to the lateral stomodzal
ganglion, but nevertheless it seems to me very probable that it
represents the “posterior pons stomodei”’ of insects and
Myriapods. The commissure just in front of the mouth un-
questionably corresponds to the anterior pons stomodei of
insects and Myriapods.

I can find nothing in Limulus corresponding to the optic
ganglion of the second segment in Scorpions. This seems to
be due to the fact that the median eye-nerves in Limulus
have shifted their points of attachment from the neural surface
of the brain to the semicircular lobes on the opposite side.

c. Later Modifications of the Brain.—After the stage
shown in fig. 59 the cephalic lobes rapidly change in aspect,
and one finds less and less resemblance between them and the
brain of other Arthropods; while, on the other hand, their resem-
blance to the fore-brain of Vertebrates becomes more apparent.

Among the important changes that take place in the dis-
position of the parts is the shortening of the optic ganglia,
and the forward movement of their distal ends; at the same
time they are drawn inward and backward till they lie in

62 WILLIAM PATTEN.

about the middle of the hemal surface of the brain,
enclosed by a common envelope for the brain and optic
ganglia (fig. 49). I know of no other Arthropod in which
the optic ganglia have this position. They usually project
from the sides, or, as in most Arachnids, from the neural
surface of the brain. In their position on the hemal surface
the optic ganglia of Limulus resemble the thalamencephalon
of Vertebrates, with which I regard them homologous. The
other important changes that take place are in the cerebral
hemispheres. They arise from disc-like thickenings
of the cephalic lobes, and correspond in position with
the brain-lobes of the second segment of Scorpions
(compare figs. 24, 59, and 61, c..). They grow upward at first,
and then their summits spread out mushroom-like, till they
completely conceal the remainder of the cephalic lobes, in-
cluding even the greater part of the segment of the mid-
brain (compare figs. 24, 25, 43, 47—52).

By this method of growth a series of chambers are formed,
some disappearing early, others persisting in the adult, that
correspond to certain cavities of the Vertebrate brain, such as
the fifth ventricle, the lateral ventricle, the primary cerebral
vesicle, and the cavities leading from the third ventricle into
the lateral eye-tubes.

We shall now describe these changes in more detail. I
have carefully studied these stages, both in surface views and
in sections, and have constructed a partial wax-plate model of
an embryo in about this stage. In fig. 24 I have represented
the cephalic lobes as they would appear in surface views.
Some details have been omitted, and the clearness of some
parts has been exaggerated, and in order to save repetition of
figures strict attention was not paid to synchronism. It is
therefore not so much an accurate picture of one stage, asa
diagram of two or three closely joined stages. A series of
sections of this period, drawn as accurately as possible, will
serve to make the meaning of the drawing clearer, and will
also show to what extent it is diagrammatic. Such aseries of
longitudinal sections is shown in figs. 35—40. They are taken

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 63

from a little earlier stage than that in fig. 24, and illustrate
the continuity of the invagination cavities. There were four-
teen sections in the series, of which the fourteenth is shown
in fig. 35 ; the eleventh, ninth, fifth, and second, in the four
succeeding figures. The other sections of the series showed
nothing of interest, and so were not represented. The
position of the sections is shown by the dotted lines in fig. 24.
The invaginations do not persist very long, and the method of
disappearing varies somewhat in the different parts. In fig. 39
the ectoderm has already broken away from the anterior wall
of the invagination, and is now pushing its way backward
over the optic ganglion, in order to unite with the thin ecto-
derm behindit. The anterior wall (a. w.), which assumes a little
darker colour, finally fuses with and forms a part of the
definitive optic ganglion. A section through this region in a
little later stage (fig. 33) still shows traces of the anterior wall
of the fold (a. w.). The figure also shows the characteristic way
in which the thick layer of ectoderm, constituting the ganglion,
becomes tilted over, so that its inner surface, where the
medullary substance is just appearing, is turned outward.
Another section of the same stage, across the distal end of the
optic ganglion, and showing the first traces of the convex eye,
is shown in fig. 44.

On the median side of the optic ganglion, to go back to the
stage we started with, the invagination is already quite faint
(figs. 37 and 38,7.v.), and in the next stage disappears, both walls
forming brain tissue. The same is true of the semicircular
lobes (fig. 36, 2. v. sl.), although the invagination cavity and its
walls can be distinguished as such for a considerably longer
period than in the optic ganglion. In the median section (fig.
35) one sees the median eye-tube cut lengthwise; the relatively
small commissure uniting the halves of the brain (m. c.),
and, at the tip of the rostrum, a great mass of transitory
tissue (7. m. s.), the nature of which could not be determined.
In fig. 26 a cross-section of the anterior margin of the cephalic
lobes of the same stage shows the extent of the two invagina-
tions for the semicircular lobes.

64 WILLIAM PATTEN.

Shortly after this stage the cerebral hemispheres become
more conspicuous. An outline of a section showing their
appearance at this time is seen in fig. 27. The whole brain
here appears to be a very thick, proliferating layer of ectoderm,
with no differentiation into an overlying hypodermis. The
underlying semicircular lobes are seen with their cavities
nearly obliterated (ce. s/.).

Soon after this the ectoderm begins to advance in an
obscure fold over the cerebral hemispheres, as in fig. 28, which
represents a section just back of the posterior margin of the
semicircular lobes. In some cases one can see a double-
walled fold like that on the lower left side of the figure, but
usually there is only a single layer, the edge of which creeps
over the hemispheres, hugging closely to their outer surfaces.
Where this layer has passed, one sees a layer of thin cells (0. s.),
that probably develop into a part at least of the brain envelopes.
Whether this layer comes from the brain itself, the mesoderm,
or the middle wall of the ectodermic fold, could not be deter-
mined.

Similar but more distinct medullary folds are seen advancing
over the margin of the mid-brain and ventral cords. Figs.
29—382 are selected from a series of cross-sections to illustrate
these folds. In fig. 29 the section passes just back of the
cerebral hemispheres in fig. 30, where the folds show best,
through the mid-brain, just in front of the chelicere ; in
fig. 31 back of the second post-oral appendage; and in fig. 32
through about the middle of the third post-oral neuromere.
These sections show that the nervous system is not separated
from the surface by a process of delamination like that which
occurs in nearly all other Arthropods, but by an infolding
from the lateral margins, very similar to that which takes
place in Vertebrates, especially of such forms as the sturgeon,
as described by Salensky. The folds in the post-oral region
never advance far enough to meet over the median line. What
becomes of the middle layer of the fold when it does occur
could not be determined. Over the middle of the ventral
cords, mainly over the “ Mittelstrang,” there is formed by

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 65

delamination a thin layer of superficial ectoderm, which pro-
bably unites with the medullary folds to form the continuous
layer that finally covers the ventral cords.

The margin of the ectodermic fold advances forward and
medianly over the cerebral hemispheres, as shown by the line
m.c. f., fig. 24. At the same time the narrow bridge of ecto-
derm leading to the pineal eye-tube is drawn backward, dimin-
ishing the clear triangular area behind it. Its anterior lip
then unites with the advancing fold on the cerebral hemi-
spheres, so that the surface of the hemisphere, not yet covered
by the ectoderm, has a contour something like that shown in fig.
25,¢.h. The cross-bar of the ectoderm is finally converted, by
the union of its anterior and posterior lips, into a tube opening
by a small round pore immediately over the proximal end of
the median eye-tube. This pore, which I shall call the anterior
neuropore (n. p.), leads directly into the median eye-tube and
through the cross-tube into all the cavities of the fore-brain.
The lips of this pore also represent the last point of
attachment of the fore-brain with the surface ecto-
derm.

Meantime the cerebral hemispheres have increased
rapidly in size. From the very start each hemisphere shows a
distinct separation into two lobes—a posterior one, drawn
out into long sharp points, and an anterior lateral lobe.
Both lobes are separated from each other by a deep fissure,
in the angle of which hes the slender fibrous peduncle on
which the hemisphere is supported (fig. 25, p. ch.). The
shape of the hemispheres in the second larval stage is accu-
rately shown in fig. 47. The fore-brain region is drawn from
a wax-plate model, the rest from dissected specimens. The
posterior lobes are not so slender and pointed as at an earlier
stage, and both lobes show the first traces of the convolutions
so characteristic of the brain at a later period. The hemi-
spheres have united with each other along the median line,
and the anterior lateral lobe is beginning to grow around on
to the under side of the brain. Two cross-sections of the brain in
this stage are shown, one through about the middle of the hemi-

VOL. 30, PART 1.—NEW SER. E

66 WILLIAM PATTEN.

spheres (fig. 51), another, farther back (fig. 50), showing the
growth of the lobes over the rest ofthe brain. A third cerebral
lobe, the median internal lobe, or “ corpus striatum,” is
now visible on the anterior inner face of each hemisphere
(fig. 51, c. s.). It is an oblong lobe, extending about half the
length of the median face of the hemispheres. Even at a
much later period its thick cortical layer of cells is never
convoluted like the rest of the hemispheres, and it contains a
great medullary core, terminating blindly in front, and behind
in a great bundle of medullary substance that forms a part of
the cerebral peduncles (fig. 43, p. c. s.). By comparing figs.
47—49 a fair idea of the enormous increase in size of the
cerebral hemispheres may be obtained. These are all young
specimens, however; the hemispheres are still larger in the
adult, as may be seen in fig. 52, a camera drawing of a section
through about the middle of the brain. The extraordinary
convolution of the cortical substance, formed of very small,
deeply stained nuclei, stands out in sharp contrast with the un-
stained medullary portions. One should observe the slender
peduncles of the hemisphere, the large space or vesicle (v.°),
sometimes filled with a mass of loose connective tissue, and
the now much-reduced semicircular lobes (s. é.).

The semicircular lobes undergo strange modifications.
The two curved lobes seen in fig. 59 grow inward and back-
ward beneath the braiu till they cover its whole median hzmal
surface. They are seen in cross-section in fig. 27, and their
extent is dimly shown through the rest of the brain in figs. 24
and 25. They reach their greatest relative extent in the
second larval stage, as shown in fig. 46 s. /. Some time before
this there are split off from the median eye-tube two clear
nerve-strands, which now terminate in conical swellings on the
anterior ends of the semicircular lobes (fig. 46). The latter
are now joined with each other at their posterior ends, and no
longer show any trace of their dual origin. The lateral and
posterior margins of the semicircular lobes contain small,
densely crowded, and deeply stained nuclei, which are appa-
rently multiplying rapidly, and when the brain is stained and

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 67

seen whole as a transparent object these cells appear as a
very conspicuous narrow band, showing clearly the peripheral
boundaries of the semicircular lobes. This narrow dark band
is best seen in larve about four or five inches long, but may
be distinguished even in the adults. After the stage shown in
fig. 46 the semicircular lobes become narrower and relatively
m aller (fig. 49). Its cortical layer is always smooth, and
contains many large ganglion-cells. Within the lobes is a
narrow stratified band of fine dense medullary substance,
something like a horseshoe in shape. Its arms are directed
forward and outward, and terminate blindly in the apex of
the lobes: the curved posterior part constitutes the anterior
commissure (fig. 43, a. com.).

The optic ganglia, soon after their first appearance, divide
into four medullary cores or centres, each covered with a
cortical layer of ganglion-cells. Three of these lobes are seen
in fig. 49; the fourth, which differs in some respects from the
rest, is situated at the base of the ganglion, concealed by the
semicircular lobes.

p. Comparison with Vertebrata.—It cannot be justly
denied that the brain of Limulus has a strong superficial
resemblance to a Vertebrate brain, and it is equally certain
that there is nothing else like its adult condition among other
Arthropods. Is this resemblance real—that is, is it pro-
found and varied, indicating genetic relationship, or merely a
superficial resemblance, which on closer inspection is seen to
be due to some deceptive quirk, but being in other respects
totally and irreconcilably different? It must be either one
or the other. Hither Limulus is closely related to the Verte-
brates, and will show this by a fundamental similarity
of structure, or else it is a world apart, and will show this
also by an equally profound difference in structure. The
resemblances already pointed out can hardly be called super-
ficial, involving as they do the similarity in the number of
neuromeres, nerves, and sense organs. I shall not discuss
these points now, but shall consider the various vesicles and

68 WILLIAM PATTEN.

invagination cavities, which are so strikingly like the cavities
of a Vertebrate brain that the resemblance cannot be due to a
mere coincidence, or to similarity of function. Before studying
the development carefully, it was a great puzzle to explain how
the adult brain of Limulus could resemble so closely the brain
of Vertebrates, and yet one be solid and the other hollow. How
would it be possible to convert the solid brain of Limulus into
the hollow one of Vertebrates, without at the same time
destroying its resemblance to the Vertebrate type? A careful
consideration indicates that we shall not have to deal with this
difficulty, for the brain of Limulus already contains, either
potentially or actually, all the important cavities of the
Vertebrate brain. In order to show this, let us go back and
consider the manner in which the brain and nerve-cord are
folded off from the surface. The imperfect folds of ectoderm
on the lateral margins of the ventral cord, and the single layer
of cells advancing over the cerebral hemispheres, are, I believe,
morphologically the same as typical double-layered folds, such
as the medullary folds of Vertebrates. No one doubts, for
instance, that the formation of the brain and spinal cord of
Teleosts by solid ingrowth is anything more than a modifica-
tion of the same process which in other Vertebrates is ex-
pressed in continuous folds. Either may be derived from
the other. In order to prove the identity of these folds in
Limulus with those in Vertebrates, it is necessary to show that
they advance over the brain, not as they do now in a typical
Vertebrate embryo, but as they did in the ancestral Vertebrate ;
or it will be sufficient if we can show that these im-
perfect folds in Limulus advance in such a manner
that, if they were perfect duplicatures, the cavities
they enclosed would have the same relation to one
another and to the brain-lobes as those in Verte-
brates. These requirements, that appear so formidable on
any other theory, can be fairly met in Limulus. It is neces-
sary to remember, however, that in Limulus the advancing
layers of cells, representing folds of ectoderm, do not enclose
spacious cavities between themselves and the brain; and that

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 69

neither these cavities nor the invagination cavities persist
till the adult stage is reached.

Lateral Ventricles.—Referring now to figs. 24 and 25,
it is evident the fold (m.c.f.), growing over the cerebral
hemispheres, encloses a very broad but flat cavity, correspond-
ing to the lateral ventricles of Vertebrates. The cavities are
as extensive as the whole upper surface of the cerebral hemi-
spheres, and they increase in extent as the latter increase in
size. They communicate with each other in front (fig. 25),
as in many fishes, and below with the cavity of the semicircular
lobes or infundibulum. The roof consists of a thin, non-
nervous membrane or pallium, and the floor of the thick pos-
terior, anterior lateral, and the median internal, lobes of the
cerebral hemispheres.

The Infundibulum.—On comparing figs. 24, 25, 41,
and 43, it is seen that the anterior wall of the cerebral hemi-
spheres passes downward and backward into the cavities of the
semicircular lobes (s. /.). These two cavities, which are at first
separate (fig. 59, s. /.), disappear about the time the lobes in
which they lie unite. Had they persisted a little longer they
would communicate with each other (as they do in Scorpions),
and this broad, backwardly directed cavity thus formed would be
similar to that in the infundibulum of Vertebrates. In
my first paper on the “ Origin of Vertebrates ” I advocated the
old view that the infundibulum represented a primitive ceso-
phagus. I have now abandoned that position for what I con-
sider a much stronger one, and we are thus left to account for
the ancestral cesophagus in some other way. It is possible that
the primitive Arthropod cesophagus broke through the narrow
band of nerve-tissue in front of it, and moving forward, was
converted into the one we now see in Vertebrates. However
that may be, there are excellent reasons for regarding the
semicircular lobes of Arachnids as homologous with the
infundibulum of Vertebrates: (1) they agree in a general way
in shape, and in their position on the hemal surface of the
brain; (2) they develop from invaginations in such a way as
to bend the anterior end of the brain-tube downward and back-

70 WILLIAM PATTEN.

wards on to the hemal surface, and form cavities communi-
cating directly with the third and the lateral ventricles ;
(3) their relations to the median or parietal eye are apparently
the same in both Limulus and Vertebrates.

Concerning the last fact, it seemed at first impossible to
satisfactorily account for what is apparently a great difference
in the position and attachment to the brain of the parietal
eye in Scorpions, Limulus, and Vertebrates. But these differ-
ences are very easily explained, and when thoroughly understood,
help to strengthen the comparisons already made. For instance,
in Vertebrates the parietal eye arises from the middle of the
roof to the thalamencephalon, back of the cerebral hemispheres.
In Limulus it seems to be about as far as possible from its
apparent position in Vertebrates, for it lies in front of the
cerebral hemispheres, and has its roots inserted into what, in
Vertebrates, would be the floor of the infundibulum
(figs. 43—49). But on looking at the spider’s brain (fig. 61),
which for the point under consideration will serve as well as
that of the scorpion, we see that the parietal eye les
back of what corresponds to the cerebral hemispheres in
Limulus (ec. 2.); and this position, as shown by the whole course
of development, seems to be the primitive one for Arachnids.
But by elongating its divergent nerve-roots the
median eye can be moved forward in front of the
cerebral hemispheres, so that its roots, instead of
describing a half-circle around the posterior neural
surface of the fore-brain, lie like an inverted Y on
its anterior hemal surface; the points of attach-
ment, however, will be the same in either case. In
Limulus this very change has taken place, owing to the move-
ment of the eye farther and farther forward on to the hemal
surface. It is a very significant fact that in mammals there
are two bands, the peduncles of the parietal eye, that extend
around the “‘tween brain” and terminate in the posterior
wall of the infundibulum, in the “corpora albicans.” I
have seen something similar to these bands in the brain of
Petromyzon, but have not yet had time to work them out

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 71

carefully. I am not aware that their existence has been
described in fishes, but it is hoped they will soon receive the
careful attention they deserve. However, if we can regard the
course and termination of these peduncles in mammals as
typical for the Vertebrates, it is obvious that they correspond
to the divergent roots of the parietal eye nerve of Limulus.
On turning to fig. 43, it is evident that the parietal eye could
be moved back to the position it occupies in mammals under
the backwardly projecting lobes of the hemispheres, for the
diverging roots (/.”.m.e.) could be slipped over the hemi-
spheres, and when shortened would form nerve-bands or
peduncles encircling the “‘tween brain” and terminating in the
infundibulum. By turning to figs. 24, 25, and 59, it is seen
that the olfactory organ would not stand in the way of this
change provided it were made early enough, since the olfactory
organs do not unite in the median line, and the median olfactory
nerve is not formed, till long after the parietal eye has assumed
its final position. The difference, then, between the position
of the parietal eye in Limulus and Vertebrates is more apparent
than real, and is probably due to the fact that Limulus moves
about on its neural surface, and the parietal eye, to be of any
use, must be on the opposite side; the extent to which it has
wandered from its original position on the cephalic lobes is
shown approximately by the great length of its nerve in the
adult. In Vertebrates the neural surface is turned upward,
hence the parietal eye can remain nearer its original position
than in Limulus.

The, at first sight, apparently inexplicable position of the
parietal eye in Limulus, and its connection with the infundi-
bulum, are seen, therefore, to be in harmony with Vertebrate
anatomy as soon as we recognise that the true roots of the
nerve to the parietal eye in Vertebrates do not terminate in
the roof of the ‘“‘ tween brain,” but in the infundibulum.

The Third Ventricle.—It is evident that in Limulus the
space between the lateral ventricles and the cavity of the in-
fundibulum corresponds to the third ventricle of Verte-
brates (compare figs. 25 and 43). The resemblance appears

72 WILLIAM PATTEN.

greater when we recollect that the invagination cavity of the
optic ganglion leads from the sides into this space at a level
midway between the infundibulum and the cerebral hemi-
spheres, just as the tubular lateral eye nerve of Vertebrates
leads into the sides of the third ventricle. The principal
difference between the optic cavities in the two cases is that in
Vertebrates the canal extends the whole length of the nerve to
the lateral eye, while in Limulus it only reaches to the root
of the nerve. It must have extended further along the lateral
eye nerve in forms more closely related to Vertebrates, other-
wise the lateral eye could not have been inverted, as it is now
in Vertebrates.

If this view is correct, then extending the ganglionic inva-
gination to the lateral eyes in Limulus ought to give rise to
conditions similar to those in the lateral eyes of Vertebrates.
This is indeed the case. It is a well-known fact that the
lateral eyes of Limulus, Trilobites, and the Merostomata
are kidney-shaped, a configuration they must necessarily
have in order to distribute the ommatidia economically over
the convex surface of the cephalo-thorax. The concave
edge of the eye is always directed hemally, as I have
shown, in a purely diagrammatic way, in fig. 24. Now
if the invagination of the optic ganglion had progressed
a little farther along the nerve the whole eye would have
been invaginated, and the ommatzum would then form
at the end of a long tube a kidney-shaped retina, but with its
concave edge turned in the opposite direction from before.
As a kidney-shaped retina would be no longer of any advan-
tage it would tend to assume a circular outline; but owing to
the peculiar distribution of the nerve-fibres and the prede-
termined method of growth, this could be most economi-
cally and advantageously accomplished by bringing the halves
of the concave edge together, thus producing a choroid
fissure, the position and direction of which would be
like that in Vertebrates. In other words, the kidney-
shaped retina of Vertebrates is due to the fact that the retinal
cells multiply faster on the convex margin than elsewhere,

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 73

This method of growth forces the concave margins together
and produces a choroid fissure. The shape and method of
growth of the vertebrate retina are inexplicable, unless we
assume it to be inherited from ancestral forms whose eyes
must of necessity have had the shape and position that the
eyes of Vertebrates would have if carried back to the exterior
where they originally belonged. The only Invertebrates having
such eyes are the marine Arachnids, Trilobites, Limulus, and
Mirostomata, &c.

If we turn again to fig. 24 we see that the space over the
mid-brain uncovered by the marginal folds (m. f. cr.) repre-
sents the imperfectly roofed-over cavity of the Sylvian aque-
duct, and that it extends forward around the posterior median
margins of the hemispheres into the uncovered lateral ven-
tricles as a shallow groove (F. M.) which corresponds in
position to the foramina of Monroe. Before a complete roof
can be formed to the mid-brain (We.'), and to the third fore-
brain segment or the “tween brain” (WNe.°), by the union
of their medullary folds, the cerebral hemispheres grow back-
ward, as they do in Vertebrates, and, uniting in the median
line, completely cover them (fig. 25). Therefore the cavity in
Limulus corresponding to the “iter” must communicate freely
above with the fifth ventricle, or the space between the inner
median faces of the hemispheres (fig. 52, v.). But if the
marginal folds of the embryo persisted they would occupy
about the position indicated by the dotted line in fig. 52, and
if they united completely two distinct chambers (v.° and i.)
would be found corresponding exactly to the “iter” and to
the fifth ventricle of Vertebrates.

If the marginal folds of the ventral cord united (fig. 32)
they would produce a medullary canal, bounded at the bottom
by the Mittelstrang, on the side by the ventral cords, and on
the roof by the united medullary folds.

Owing to the divergence of the cords in the hind-brain
region (fig. 48) a great rhomboidal cavity with a thin roof
would be formed that would correspond to the fourth ven-
tricle.

74 WILLIAM PATTEN,

To sum up what has preceded, we find that in the brain of
Limulus there is an actual or potential agreement throughout
with the Vertebrate brain. There is (fig. 43) the folding of
the anterior end of the cerebral vesicle downwards and back-
wards to form the infundibulum; the upward growth of the
co-ordinating centres, or the cerebral hemispheres, and their
expansion in all directions at the summit to cover the rest of
the brain. There are the lateral ventricles communicating in
front with each other, with the cavity in the epiphysis, or
median eye tube, and with that of the olfactory lobes, below
with the infundibulum, and backwards and downwards with
the “iter” and the third ventricle. Its roof is a thin mem-
brane or pallium ; its floor the three great lobes of the cerebral
hemisphere—the posterior, the anterior lateral, and the median
internal or corpus striatum.

There is the great cavity in the centre of the brain corre-
sponding to the imperfectly separated third and fifth ventricles
and the “‘iter.”” It is bounded in front by a thin membrane,
‘Jamina terminalis” (fig. 43, 1. ¢.), above by the corpora
striata (c.s.), and on the sides by the cerebral peduncles and
optic thalami, and on the floor by the middle and inferior
commissures. It communicates anteriorly and above with
the lateral ventricles, behind with the unroofed “iter,” and on
the sides with a cavity leading to the roots of the lateral eye
nerves.

There are three commissures to the brain—one the inferior
commissure in the roof of the infundibulum (a. com.), the very
large middle commissure in the floor of the third ventricle (m.
com.), and the posterior commissure in what would be the roof
of the mid-brain, just behind the posterior end of the cerebral
hemispheres (p. com.).

There is the fourth ventricle, a large rhomboidal space
imperfectly covered by the medullary folds (fig. 48). Its floor
is composed of the united cross-commissures of the thoracic
neuromeres in the position of the future ‘ pons,” and its sides
are formed by the great diverging masses of nerve-substance, or
crura, leading up to the fore-brain.

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 75

It is evident that the principal differences between the fore-
brain of Limulus and that of Vertebrates is a difference in
degree and not inkind. In Limulus the infoldings are obscure,
in not always showing the duplication of layers ; individualized,
in that in each important brain-lobe they progress indepen-
dently of the others ; and they are incomplete, in that they do
not always unite so as to entirely enclose the infolded parts.
Moreover some important cavities, such as the lateral ventricles
and the cavities in the optic ganglia and infundibulum, disap-
pear at an early stage, so that their relations to the permanent
cavities are not very obvious. In Vertebrates, on the contrary,
the infoldings are very simple because they have lost their in-
dividuality by fusing into one continuous fold, which appears
very early, and is usually completed before the brain-lobes are
specialized. In Limulus the phylogenetic processes from the
very first stages dominate over the pure mechanics of ontogeny.
In Vertebrates it is almost the reverse ; purely embryological
processes prevail at first, afterward phylogenetic ones are
manifest.

There is no greater difficulty in identifying all the charac-
teristic features of the Vertebrate brain in Limulus than there
is in comparing a fish brain with that of mammals. No other
Invertebrate will permit any approach to such comparison.
On the other hand, it is impossible that the comparisons, such
as I have instituted, could be carried so far without breaking
down, unless they rested on a sure foundation of actual facts
and correct premises.

It remains now to say a few words concerning the parietal
eye and the olfactory organ. If the relation of the parietal
eye tothe primary cerebral vesicle and to the infundibulum are
mere coincidences, having nothing to do with the genetic re-
lationship of Limulus with Vertebrates, then in all probability
the resemblance will cease there; it certainly cannot extend
any farther, say to the structural details of its nerves and its
terminal portions. But the resemblance does go a great deal
farther, as we shall now see.

76 WILLIAM PATTEN.

TV. Tue Parietat Eye.

After the median eye-tubes in fig. 59, p.e., unite, a single
tube is formed like that in figs. 24 and 25. The solid distal
end (fig. 35, m.e.) soon divides into three lobes; the two
outer ones just before hatching become filled with black pig-
ment, and form the retinas to the ectoparietal eyes (m. e.);
the inner, unpaired, one forms the endoparietal eye (m.e'.).
There is no distinct cavity in these lobes, but as they are
formed from the walls of an invaginated tube, and as the two
outer ones are homologous with the median eyes of scorpions,
where such cavities are present, we must regard them morpho-
logically as sac-like diverticula from the end of a tube. Their
similarity to the common vesicle of the median eyes of scor-
pions and their relation to the cerebral vesicle is easily seen
in figs. 41 and 42. During the first larval stage the endo-
parietal eye is a solid cylindrical mass of cells with rod-like
thickenings in their walls, like those in true retinal cells, and
filled with dense white pigment (fig. 63). In sections the
crumpled refractive plates on these degenerate retinal cells
look like coiled or zigzag fibres something like those in
fig. 20. As the animal grows older this mass of aborted retinal
cells buries itself deeply in the underlying tissue, away from all
connection with the exterior (fig. 74). In the adult it usually
lies below a conical tubercle, situated a little behind the two
lenses to the median eyes. In many old specimens this
tubercle is replaced by a clear, transparent spot, which is no
doubt the remnant of a lens. As the endoparietal eye
must itself have been formed by the fusion of paired retinas,
it is evident that the distal end of the parietal eye-
tube contains the retinas of four originally distinct
eyes. These facts are all the more interesting since Claus
has shown that the median eye of crustacea, which in some
cases is composed of three distinct eyes, is formed by invagi-
nation just as in scorpions and Limulus.

The distal portion of the primitive eye-tube is converted
bodily into the median eye-nerve. Just before hatching its

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 77

distal end splits up into four branches, two of which plunge
directly into the median diverticulum or endoparietal eye (fig.
63, n. en. p. e.), and the other two pass to the paired retinas of
the ectoparietal eyes (figs. 63—65, ec. p.e.). Two delicate
nerve-strands split off at a very early period from the proximal
end of the primitive eye-tube and its diverging arms (figs. 35,
l.n.m.e. and fig. 36, 7. m.e.n.). Compare also surface view
(figs. 46 and 47,7. m.e.n.). These two nerve-roots subse-
quently divide (fig. 49), so that there are at least four main
roots to the nerve, not including the epiphysis. Two of these
roots terminate in conspicuous medullary nuclei attached to the
horseshoe-shaped medullary core in the interior of the semi-
circular lobes (p.e.c.). From each nucleus a delicate strand
passes laterally to the peduncle of the lateral optic ganglia
(s222-p:):.

The epiphysis, or what is left of the median eye-tube after
splitting off the roots to the parietal eye-nerve, remains for
some time unchanged, as in fig. 43, m.e.t., also figs. 46—49
and 68—70. Finally its lumen disappears, and several
swellings appear in it, composed of small nuclei like those
in the cerebral hemispheres. In specimens four or
five inches long these swellings may be seen em-
beddedinthe brain evelope, and extending downward
from the root of the median olfactory nerve toward
the parietal eye-nerve (fig. 49, m.e.¢.). After this stage
the epiphysis loses altogether its connection with the root of
the parietal eye-nerve, and then disappears completely.

We have in the parietal eye seven distinct struc-
tures, which should be constantly borne in mind in
making any comparison with the parietal eye of
Vertebrates. (1) The primary brain diverticulum or
primitive parietal eye-tube; (2) the paired, and (3) the
unpaired diverticula at its distal end; (4) the solid
nerve-stalk to these diverticula; (5) the epiphysis, or
remnant of the primitive parietal eye-tube after
splitting off the (6) roots of the median eye-nerve, and
finally (7) the large blood-vessel accompanying the

78 WILLIAM PATTEN.

nerve. The primary parietal eye-tube of Limulus corresponds
to the epiphysial outgrowth from the brain-roof in Vertebrates.
The ectoparietal eyes containing the two separate retinas, and
the ventral diverticulum or endoparietal eye filled with white
pigment, correspond to the two distinct terminal organs found
in some Vertebrates. I venture to suggest that in Hatteria
the “ capsular-like structure,” described by Spencer, at the base
of the eye where the nerve enters, corresponds to the endo-
parietal eye of Limulus. Again, the extraordinary presence of
dense pigment in the middle of the outer wall of the eye in
Varanus giganteus, as described by the same author, may be
regarded as the remnant of the pigmented epithelium origi-
nally separating the paired retinas of the ectoparietal eyes of
Limulus and scorpions ; and finally it is possible that the two
separate pineal eyes that occur in certain reptiles, such as
Anguis and Lacerta, may be due to shortening of the primary
evagination, so that the ecto- and endo-parietal eyes arise
as two independent outgrowths from the roof of the thalamen-
cephalon.

The distal end of the primitive eye-tube in both Limulus
aud Vertebrates is converted bodily into the pineal eye-stalk.
The proximal end in Limulus, after giving rise to the nerve-
roots, persists as an inverted T-tube; the upright arm unques-
tionably corresponds to the epiphysis in Vertebrates, and the
cross-bar, perhaps, to the ganglion habenule. The nerve-
roots correspond to the “ peduncles” of the pineal eye as seen
inmammals. In both Limulus and Vertebrates a large blood-
vessel accompanies the nerve to the pineal eye. I formerly
regarded the epiphysis as one of the nerves to the parietal eyes,
but it now seems to me to be nothing but the ectoderm along
which the parietal nerves formerly extended. The nerves
separate from it just as the peripheral nerves separate
from the overlying surface ectoderm. It may be compared to
the epithelium lining the interior of the hollow lateral eye-
nerves.

A remarkable feature of the parietal eyes of Limulus and
of Vertebrates is the presence in them of great quantities of

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 79

dense white pigment. In Petromyzon the solubility of this
pigment in weak acids and its intensely white glistening
appearance in reflected, and deep black appearance in trans-
mitted light has been the cause of contradictory statements
concerning it, some claiming it is black, others white, and
others that it is absent, according to the method of prepara-
tion or the light (transmitted or reflected) by which it has
been examined. I have examined the white pigment in the
parietal eye of Limulus side by side with that in the parietal eye
of Petromyzon, and have found them to be identical in
appearance. In Limulus, as in Petromyzon, the white pig-
ment is much more soluble in dilute nitric acid than the black.
White pigment is comparatively rare in Invertebrates, and, so
far as I know, is confined to the Arthropods. It is present
about the base of the ommatidia in the compound eyes of the
crustaceans Pingus and Galatea (Patten), and in Limulus is
found in the infolded margin of the young lateral eyes and
in the choroid plexus lying in front of the brain. But no-
where in Arthropods is it more abundant than in the endo-
parietal eye of Limulus. This eye is apparently nothing but
a solid mass of the pigment, and when ruptured with needles
it flows out in a dense white chalky stream. Ahlborn claims
that the white granules in Petromyzon are composed of calcium
phosphate, like the “ brain-sand ” found in the parietal eye of
the higher Vertebrates. It would be very interesting to learn
its composition in Limulus.

Gaskell has compared the parietal eye of Vertebrates and
Crustaceans, but he failed to see the real points of resemblance
between them, for he knew nothing about their development
in Arthropods. There is no more resemblance between the
parietal eye of Vertebrates and the ocellus of Acilius, as he
puts it, than there is between the lateral eyes of Vertebrates
and those of a Cephalopod. If we fail to take into account
the development of the parietal eye in Arthropods there can
be no trustworthy grounds of comparison between it and the
parietal eye of Vertebrates.

While there remains a good deal of doubt concerning the

80 WILLIAM PATTEN.

significance of certain parts of the parietal eye in Limulus and
Vertebrates, that they are homologous with each other as a
whole is, I believe, beyond question. "

V. Tue Oxractory ORGANS.

We have in the olfactory organ of Limulus a structure pre-
senting the most striking and unusual features. It is as
different from the other cerebral sense organs as the olfactory
organ of Vertebrates is from the pineal eye. The features
that in Limulus distinguish it from all other sense organs are
the very ones characteristic of the olfactory organ in Verte-
brates. The organ itself consists of an upright epithelium
containing a large number of bud-like sense organs comparable
with the sense organs described by Blaue in the olfactory
organ of fishes. The whole organ, as it exists in the adult
Limulus, may be compared to the unpaired olfactory organ
found in the Cyclostomata. It is composed of three distinct
parts—a pair of primitive sense organs growing forward from
the brain, anda subsequently formed part lying between them.

It is probable that the dormant individuality of these
parts has reasserted itself in the higher Vertebrates, and
caused the separation of the olfactory organ into its con-
stituent parts, or the olfactory organ proper, and the organ of
Jacobson. The distribution of the nerves supports this view.
In Limulus there are three distinct nerves: the paired lateral
and hemal ones, having their roots in one of the centres of
the optic ganglia, or in the optic thalami; and an unpaired,
median, neural nerve terminating in the cerebral hemispheres.
The median nerve, which in its early stages shows evidence of
its paired origin, arises at a late period as an outgrowth from
the cerebral hemisphere. It is a new formation, in histological
structure and origin utterly unlike any other Arthropod nerve.
All four of these nerves are represented in Vertebrates, for it is
now known that each olfactory nerve of the higher Vertebrates
is represented in Amphibia by two distinct nerves, which have
been likened to the dorsal and ventral roots of a spinal nerve.

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 81

But if this were so they would differ from all other spinal
nerves in that both dorsal and ventral branches supply sense
organs. Moreover, on any supposition they are entirely
different from those belonging to the other sense organs of the
fore-brain, such as the lateral and parietal eye, and the auditory
organ. This condition is quite inexplicable on any theory
founded on Vertebrate anatomy. But this very thing occurs
in the olfactory organ of Limulus, although the meaning of it
cannot be explained any more there than in Vertebrates.

Now in Pipa dorsigera and Epicrium glutinosum,
according to Wiedersheim, there are four distinct olfactory
nerves—a neural and hemal pair. The Saracins in their
Ceylon work, and more recently Burckhardt (‘ Z. f. w. Zool.,’
1891), have further shown that in Ichthyophis and Triton the
heemal pair innervate the organ of Jacobson, and the neural
pair the rest of the olfactory organ. Exactly the same rela-
tions prevail in Limulus; the hemal pair are also lateral, and
they supply a distinct lateral depression in the olfactory organ.
The neural pair in Limulus innervates the greater part, if not
the whole, of the definitive olfactory organ; they are formed
by an outgrowth from the cerebral hemispheres ; they contain
the olfactory bulbs or lobes, and they are composed of pale
refractive nerve-fibres with a structureless nucleated sheath,
differing strikingly from the thick-walled apparently empty
nerve-tubes seen in the lateral olfactory and other nerves ;
moreover, as they supply by far the greater part of the olfac-
tory buds with nerves, they may be regarded as the olfactory
nerves proper. They therefore agree in these important mor-
phological and histological features with the dorsal pair in
Amphibians. In the higher Vertebrates all four nerves seem
to have united, and the features of the median pair impressed
on both.

Iam not aware that the course of the roots to the olfactory
nerves is accurately known in the lower Vertebrates ; but inthe
Mammalia there are four large medullary nuclei in the thala-
mencephalon, in one of which terminate the lateral roots to the
olfactory nerve. It certainly is not without significance that

VOL. 35, PART 1.—NEW SER. R

82 WILLIAM PATTEN.

the optic ganglion of Limulus, which for other reasons may
be regarded as homologous with the thalamencephalon, also
contains four nuclei, and from one of them (fig. 49) arises the
lateral olfactory nerve.

But we have not yet reached the end of the similarity
between the olfactory organs of Limulus and Vertebrates. In
Vertebrates the olfactory organs are supposed to belong toa
system of supra-branchial sense organs, from which arise the
scattered sense organs of the head and lateral line. This supposi-
tion is based on the similarity between the method of develop-
ment of the organs in both cases, and by the fact that the olfac-
tory organ in many fishes is at first composed of a collection of
sense buds similar to those in, or originating from the supra-
branchial sense organs. Now, although I do not attribute much
weight to these arguments, it is important to observe that we
have in Limulus the same relation between the olfactory and
other sense organs that we have in Vertebrates. We have, for
instance, in the thorax of Limulus a series of sense organs and
ganglia supplied by a purely sensory nerve, which in position
and connection is much like the supra-branchial nerve of Ver-
tebrates. Before I had paid much attention to these organs
in Limulus I was led, from a knowledge of these relations in
scorpions, to regard the epidermic thickening that gave rise to
the “coxal ganglion” of Scorpions as homologous with a
supra-branchial sense organ of Vertebrates. My recent ob-
servations on Limulus confirm this comparison, and broaden
its significance in a most unexpected manner. In scorpions
the great ganglionic thickening on the median edge of the
coxal joint is a transitory structure, and there is no perma-
nent or definite sense organ found there.’ In Limulus, how-

1 It seems to me now very probable that the mandibles on the walking
appendages of Limulus and the coxal spur of Scorpions, with its rudimentary
sense organ, represent much-shortened endopodites of the thoracic appendages.
In Scorpions the sense orgar comes in about the same position as the rudimentary
appendages described by Jaworowski (‘Zool. Anz.,’ xiv Jr., 363) in Trochosa
singoriensis, and regarded by him—rightly, I believe—as the endopodites

of the thoracic appendages. But he has not, in my judgment, produced satis-
factory evidence that the structure described by him as corresponding to the

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 83

ever, we find exactly the same nerves and the same ganglia,
but they are connected with permanent organs of taste.
Moreover some of these sense organs are single cells, strangely
modified, and recalling the isolated gustatory, olfactory, and so-
called ‘‘ tactile” cells so widely distributed in Vertebrates ;
others are sensory buds, practically identical with those scat-
tered so uniformly over the body, as well as with those in the
olfactory organ.

There certainly can be no doubt that the great masses of
sense organs in the jaws of Limulus have disappeared in Scor-
pions, leaving nothing but an enormous ganglion and nerve to
indicate their former existence. This fact is of special in-
terest in this connection, for it is exactly what Beard sup-
posed had taken place in the supra-branchial sense organs of
Vertebrates.

In the free swimming merostomata-like ancestors of the Ver-
tebrates, we may assume that a change like that in the Scorpion
has taken place. With the transformation of the appendages
into gill arches the endopodites or the gustatory spurs of the
walking appendages were probably reduced to mere sensory
patches, appearing in the ontogeny of Vertebrates as transitory
gangliogenic thickenings of the ectoderm, or supra-branchial
sense organs. I have already pointed out that in Scorpions
(“ Origin of Vertebrates from Arachnids”’) the coxal sense
organs, ganglia, and nerve, were like the supra-branchial sense
organs of Vertebrates in their development and distribution and
relations to the rest of the nervous system. Now we can
show, in addition to this, that in Limulus the mandibles may be
regarded as segmental gustatory organs. They are centres
composed of aggregations of gustatory buds and cells, and
from them, or near them, may arise diffusely distributed sense
buds. The segmental aggregates and the diffuse organs in
Limulus are largely gustatory, and the same is in all pro-
bability true of these organs in Vertebrates, various theories
to the contrary notwithstanding.

antenne of Insects is really an appendage ; it may be either a sense organ, or
the endopodite or exopodite of the cheliceral segment.

84. WILLIAM PATTEN.

Moreover the gustatory cells in Limulus, and the various
sense buds arising from the supra-branchial sense organs in
Vertebrates, are arranged in straight lines, a condition, so
far as I know, found in Arachnids and Vertebrates only. Now
whether this arrangement in lines is due to the method of
growth or to a physiological necessity is doubtful. Neverthe-
less this condition may be taken as evidence of genetic
relationship between the forms in which it occurs. And
finally, this being what I regard as of most importance, we see
that in Limulus the olfactory organ, besides its other resem-
blances to the Vertebrate organ, contains the same kind of
buds as those in the segmental gustatary organs. As is well
known, the same thing occurs in Vertebrates. Now we
might suppose, as has been done in Vertebrates, that the
olfactory organ is serially homologous with the mandibular
gustatory organs. But the whole development of the organ
and its nerves shows that in Limulus this view cannot be
entertained for a moment. The olfactory organ in Limulus
is obviously a new growth on an old foundation. The latter,
it is true, is a segmental sense organ, but it belongs to the
series including the ocelli and compound eyes, not to that of
the thoracic appendages.

The same interpretation applies, I believe, to the olfactory
organ of Vertebrates, as Limulus appears to be less intelligent
than animals like Lobsters, Scorpions, and Spiders, in which
there are no cerebral hemispheres or any organ that may be
regarded as a psychic centre. It is very surprising that it
should possess gigantic cerebral hemispheres, and in shape,
structure, and development like those of Vertebrates. This
may be a coincidence. But can it be a coincidence that in
both Limulus and Vertebrates these cerebral hemispheres are
connected with only one pair of nerves? Is it a coincidence
that these nerves arise from the same brain region, have the same
lobes, and supply sense organs that show striking similarities
in position, structure, development, and function? It cannot
be. Not one of these characters obtains elsewhere, and,in my
judgment, it is impossible that all of them should occur in

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 85

Limulus and Vertebrates unless they are genetically re-
lated.

VI. Comparison oF OTHER Brain Recions or LImMvutus
WITH THOSE OF VERTEBRATES.

In any near relative of Limulus there might be from five to
ten pairs of supra-branchial nerves, according as the cheli-
cere and the vagus segments retained their sense organs. In
Limulus there are five pairs, those belonging to the post-oral
thoracic appendages (figs. 47, 48, md. n. 2—6). This number
corresponds approximately to the number of supra-branchial
sense organs in Vertebrates, and affords us a satisfactory ex-
planation of their relation to the gill-arches and their absence
from the trunk in the early stages of development.

I have already pointed out in my paper on the “ Origin of
Vertebrates”’ how the cephalo-thoracic neuromeres of Scorpions
and Limulus are comparable with the entire brain of Ver-
tebrates. This is clearly shown by the similarity in the grouping
of the neuromeres, and by the modifications these groups
have undergone.

The modifications and homologies of the three segments
constituting the fore-brain of Limulus and Vertebrates have
already been considered. The mid-brain of Vertebrates is
generally conceded to consist of a single neuromere, distinct
from those in front and behind, and characterised by the fact
that it is provided with the only pair of cranial nerves arising
from the neural surface of the brain. We find the same isolated
neuromere in Arthropods, namely, the one belonging to the
antennal segment of Insects and Myriapods, and the cheliceral
segment of Arachnids. In the Arachnids it is easily dis-
tinguished from the true fore-brain on the one side and from
the post-oral neuromeres on the other. Owing to the position
of the chelicere, close together in front of the mouth, their
nerves invariably arise from the neural surface of the neuro-
mere, while all the others, except, perhaps, in the vagus
region, arise from the sides.

86 WILLIAM PATTEN.

The hind brain of Limulus (fig. 48, H. B.) is composed of
five typical thoracic neuromeres, each having the following
four pairs of nerves :—(1) The mandibular nerves (md. n., 2—6)
are purely sensory, and consist of two branches which terminate
in the gustatory organs inthe mandibles. They arise from the
neural surface of the ganglion at the root of the pedal nerve,
and usually possess well-defined swellings that contain a few
small ganglion cells scattered through a mass of interwoven
fibrous substance. Sections through the appendages of half-
developed embryos (fig. 72) show two sensory thickenings,
one corresponding to the inner mandible (¢. md.) and the
other to the outer (0. md.). In figs. 71 and 73 are seen
ganglion cells separating from the sense organs to form the
ganglia of the mandibular nerves in the adult (fig. 3, md. n.9.).
The mandibular nerves correspond to the supra-branchial
nerves and ganglia of Vertebrates. (2) The pedal nerve is a
mixed nerve arising from a large neural ganglion common
to it and the coxal nerve; it supplies the sense organs and
muscles of the walking appendages. In Scorpions it is at an
early stage probably connected with the lateral segmental
sense organs, but the existence of such a connection could not
be demonstrated in Limulus, owing probably to the transitory
nature and imperfection of the segmental sense organs. This
nerve corresponds to the ventral or branchial nerve of Ver-
tebrates, with which it agrees in innervating the only im-
portant muscles of the head, namely, those of the visceral
arches. All the other head muscles of Arachnids and Ver-
tebrates, owing to the complete anchylosis of segments, have
disappeared.

The neural and lateral ganglia of Beard probably correspond
respectively to the pedal ganglion and the ganglia to the
segmental sense organs of Limulus and Scorpions. The
anterior and posterior hzmal nerves correspond to the pre- and
post-trematic of Vertebrates.

All the nerves of the hind brain of neuromeres of Limulus
agree with those of Vertebrates in being separate, while the
same nerves in the trunk of both Vertebrates and Arachnids

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 87

are more or less united. This is especially true of the Scorpion,
where I have shown that the abdominal nerves are built
on the type of true spinal nerves, for their proximal ends
remain separate to form two roots, the dorsal or neural one
being ganglionated, the hzemal one being non-ganglionated.

The accessory brain forms the fourth and last brain
region of Vertebrates and Arachnids. It is formed of from
two to four neuromeres derived from the trunk and added at
a comparatively late period to the head. In spite of their late
origin, they form the most specialized and completely fused
neuromeres of the head. In both cases almost every trace
of the metameres to which they belong has disappeared,
consequently their nerves have wandered backward to new
territory. The nerves to these four neuromeres in Scorpions
are very unlike all others in front or back of them, for the
four ganglionated neural nerves are fused into one group, and
the eight hemal ones into another. They are,on the other
hand, strikingly like the vagus nerves of Vertebrates in their
direction, distribution, and general appearance (see ‘‘ Origin of
Vertebrates from Arachnids,” fig. 1). In Limulus there are
only one or at most two neuromeres in this region, and there-
fore it has not such striking vagus characters as in Scorpions.

The presence of these vagus segments is not confined to
Scorpions and Limulus, but is of very wide occurrence in
Arthropods. The last thoracic and first two or three abdominal
segments in Insects, and the four abdominal appendages bearing
segments in Spiders, probably belong here. Their presence in
trilobites is shown by a well-marked “ cervical suture” like
that which in the adult Limulus marks the presence of these
segments.

After we have torn off the deceptive Arthropod mask that
disguises Limulus we discover that the nervous system, with
all its complex and intricate modifications, shows as a whole a
profound structural similarity to that of Vertebrates. This
similarity extends to important aggregations of parts, as well
as to many minute details, all of which could not on any
reasonable assumption have occurred accidentally or be due to

88 WILLIAM PATTEN.

similarity of function or condition. The mere enumeration
of the resemblances which my as yet superficial study enables
me to point out ought to convince the most sceptical that
we are working in the right direction. Is there any group of
Invertebrates besides the Arachnids which by any reason-
able assumption can be made to resemble Vertebrates in
the way that Limulus does? How far do the comparisons
of either Ascidians, Balanoglossus, Nemertians, and Annelids
with Vertebrates carry us? They explain nothing either in
Vertebrate anatomy or in phylogeny, and only serve to render
the whole problem of the ancestry of the Vertebrates hope-
lessly perplexing and obscure. If we accept the Arachnid
theory all this is changed, for we can see, dimly perhaps, some
way out of the intricate problems bound up in the morphology
of the Vertebrate head. It is true many great problems will
be left unsolved, but their discussion will be shifted to the
vast field of Arthropod morphology, where there is hope of
their ultimate solution. And the problems involved are great
ones, notwithstanding a dawning tendency to subordinate
phylogenetic questions to profound studies on the mechanics
of cell division and kindred topics. If the Arachnid theory is
true, the Ascidians, Amphioxus, and perhaps Balanoglossus
and the Echinoderms become degenerate phyla, bearing some-
what the same relation to the Vertebrates that the parasitic
Copepods do to the Arthropods. It will furnish us with a new
basis for embryological interpretation, especially regarding the
problems connected with the formation of germ layers. If
the Arachnid theory is true, many current views on phylogeny,
ontogeny, and important problems in comparative anatomy are
based on false conceptions, and must be revised. That this is
no exaggeration is obvious from the objections most frequently
urged against the Arachnid theory, ones I had not anticipated
would have any weight. Instead of considering the question
on its own merits, it has been objected that “ it was contrary
to all our preconceived ideas,” and ‘it is quite impossible to
conceive highly specialized animals like Arachnids giving rise
to such highly specialized animals as Vertebrates.” I do not

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 89

repeat these objections to give them serious consideration, but
only to emphasize the fact that the solution of the Vertebrate
problem is not merely the filling in of a gap in our system of
classification, but this solution will necessitate the recon-
struction of a vast deal of that preconceived morphological
philosophy which at present forms the most serious obstacle
to the perception of the genetic relation between Vertebrates
and Arachnids.

Granp Forks, N. D.; Nov. 29th, 1892.

EXPLANATION OF PLATES 1 to 5.

Illustrating Mr. William Patten’s paper “On the Morpho-
logy and Physiology of the Brain and Sense Organs
of Limulus.”’

EXPLANATION OF PLATES 1 and 2.

Reference Letters to Plates.

ax. n. Axial nerve. 4/7. v. Blood-vessels. c.c. Cuticular canal. ch. ¢.
Chitinous tubule. cf. ¢’. Proximal end of chitinous tubule. g. c. Ganglion-
cells. g.o. Gustatory organs. g.p. ”. Ganglion to the pedal nerve. gs. c.
Gustatory cells. g. and ¢. or. Gustatory and temperature organs. 7. md.
Inner mandible. 7.0/.2. Lateral olfactory nerve. m. Muscle. md. 2. Man-
dibular nerve. m. 7. m. Flexor muscles to inner mandibles. m. o/. x. Median
olfactory nerve. 2. Nerve. 2. c. Slender nerve-like cells surrounding the
base of the chitinous tubule. 2. co. Nerve collar. . and x’. Primary and
secondary nuclei of the gustatory cells. 1. /6/. Cone of nerve-fibrille inside
the spindle. o. md. Outer mandible. p. gy. Yellowish-brown pigment
granules. p.z. Pedal nerve. s.c.c. Swelling in cuticular canal. s. ch. ¢.
Swelling on chitinous tubule. sp. Spindle on gustatory cells. sp. f. Spiral
fibre. s. sp. Canal containing pigment and fibrous cells leading to spines.
t.o. Temperature organs. v. Deeply stained varicosities on the nerve-
fibrille within the spindle. w.p. White pigment-cells.

90 WILLIAM PATTEN.

PLATE 1.

Fig. 1.—Longitudinal section through the anterior wall of one of the stout
gustatory spines of the mandibles, showing the cuticular canals, each con-
taining a chitinous tubule.

Fie. 2.—A gustatory spine from the mandibles, seen from its anterior sur-
face, and showing the lines of pores into each of which runs a chitinous
tubule. x 30.

Fic. 3.—Third appendage on right side of an adult female, seen from in
front, with the anterior wall removed, and showing the nerve to the appen-
dage with its mandibular branches. Natural size.

Fic. 4.—Nucleated end of one of the gustatory cells of the mandibular
spines, showing the peculiar yellowish-brown granules that are sometimes
found in these cells. Isolated by Bela Haller’s fluid.

Fie. 5.—Distal end of a gustatory cell from the mandibular spines, show-
ing fibrillar structure of the cell body and the spindle with its internal cone
of fibrillee, each fibrilla having an enlargement at the base of the spindle, and
all converging toward the apex, where they unite to form the axial nerve-
bundle that runs outwards through the chitinous tubule. x about 2000.
Macerated in Haller’s fluid and stained in acetic acid carmine.

Kies. 6 and 7.—Spindles from gustatory cells of mandibular spine, showing
the partly detached nerve-cells, and the projecting axial nerve-bundle, which
in Fig. 7 is broken up into its constituent fibrille. Haller’s fluid and methyl
green.

Fic. 8.—Very young gustatory bud from the inner mandible, showing the
circle of refractive rod-like plates and the single ganglion-cell with its fibrous
prolongation.

Fic. 9.—Two sense buds from the olfactory organ, showing the chitinous
tubule, central ganglion-cell, and the nerve-plexus. Slightly diagrammatic.

Fie. 10.—Three cuticular canals from the base of a mandibular spine.
a. Larger canals with short spine in a saucer-shaped depression at the
summit. May be one of the temperature organs. 3B. Ordinary gustatory
canal. c. Same, with a few bead-like globules in the chitinous tubule.

Fre. 11.—Surface view of the cuticle from the chele of the walking appen-
dages of a young Limulus about 8 inches long, showing the two kinds of
sensory pores, the gustatory pores at g.o., and the temperature organs at
t. 0.

Fic. 12.—The nucleated portions of two gustatory cells from the mandi-
bular spines, showing that they are formed of two imperfectly fused cells
similar to the double ones described by the author in ‘ The Eyes of Molluscs
and Arthropods.’

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 91

Fie. 12 a.—Outer end of a cuticular canal of the inner mandibular gusta-
tory buds. a. In longitudinal section. &. Seen from above. ¢. Gustatory
canal (outer end) from the chele.

Fie. 12 p.—Surface view of the outer end of a sensory bud of canal, from
the thorax.

Fic. 12 5.—Same view of olfactory bud of canal from olfactory organ of
female.

Fic. 12 Fr.—Same of male. All of Figs. 12 a—p drawn to same scale.

Fic. 13.—One of the node-like swellings in the cuticular canals seen in
Fig. 1 much enlarged, and showing the chitinous tubule within and the spiral
fibre.

PLATE 2.

Fic. 14.—TIsolated chitinous tubule from the gustatory buds of the mandi-
bles, showing the thin, granular, protoplasmic investment, with the spiral
lines. Treated with caustic potash.

Fic. 15.—Isolated ganglion-cell from the centre of a gustatory bud. Hert-
wig’s mac. fluid, 5 days.

Fic. 16.—Chitinous tubules from the mandibular gustatory buds. In a
the tubule is stained dark by the acetic acid carmine, and is surrounded by
two sinuous, refractive and colourless fibres. 3B shows the coiled tubules that
are frequently found in the mandibles.

Fic. 17.—Longitudinal horizontal section of the median olfactory nerves
and olfactory lobes of a young Limulus about 4 inches long. The lateral
olfactory nerves, /. o/. 2., which bend outwards beneath the olfactory organ,
are cut transversely. x 90.

Fic. 18.—Outline of the fore-brain, olfactory organ, and nerves, of an adult
female Limulus. x 4.

Fic. 19.—Surface view of the olfactory organ of an adult female, showing
the distribution of the cuticular canals leading to the olfactory buds. x 15.

Fie. 20.—Longitudinal section through the base of the lateral olfactory
nerve, showing the ommatidia-like clusters of cells with their refractive, rod-
like thickenings. x 200.

Fic. 21.—Cross-section of the adult olfactory organ. The outer layers of
the cuticula have been removed. x 46.

Fic. 22.—Cross-section of the olfactory organ in a young Limulus about 7
inches long. Flemming’s fluid (strong). x 90.

Fic. 23.—Portion of the preceding figure still further enlarged. The
thickness of the cuticula is not increased in the same proportion. x 550.

92 WILLIAM PATTEN.

EXPLANATION OF PLATES 3 and 4.
Reference Letters to Plates 3 and 4.

a.2—6. Anterior hemal nerves of the second to the sixth neuromere.
A. B. Accessory brain. a.e. Anterior edge of dactylopodite. a. w. Ante-
rior wall of the optic ganglion invagination. a. w.c. Anterior wall of cerebral
vesicle. 4. e. Last point of union of cerebral hemispheres with surface ecto-
derm. c. Posterior fore-brain commissure. c*. First post-oral commissure.
c. h. Cerebral hemispheres. c. 4’. Uncovered part of cerebral hemispheres.
c. m. Canal between the cerebral hemispheres and the thin median ectoderm.
c. m.e.¢t. Canal leading to the median eye-tube. com’. Posterior cerebral
commissure. c¢. 7. crura cerebri. c. s. Internal cerebral lobe = corpus
striatum. c.s/. Invagination cavity of the semicircular lobes. ez. Cuticula.
c. v. Cerebral vesicle. D.S. Dorsal surface. #. B. Fore-brain. g. l. ol. x.
Ganglion to the lateral olfactory nerve. g.2.1—5. Nerves to gills. g. p. x.
Ganglion to pedal nerve = “neural ganglion” of a Vertebrate cranial nerve.
g. st.. Ganglion to stomodeal nerve. H. B. Hind brain. &. 2. 2—9.
Hemal nerves or peripheral tegumentary nerves. #. Unroofed space corre-
sponding to ‘‘iter.” 7¢. v. Continuation of the invagination of the optic
ganglion with that of semicircular lobes. 7. v. op. g. Invagination cavity
in optic ganglion. 7. v. sl. Cavity of semicircular lobes. J. e. Lateral
eye. /.e.%. Lateral eye-nerve. J. ol, x. Lateral olfactory nerve. J, ¢.
Lamina terminalis. m.2—6. Median hemal nerve of the 2—6 neuro-
meres. M.B. Mid-brain. m.c. Middle commissure of the brain. m. ec.
1—2. Two cortical masses of ganglion-cells in front of cheliceral nerves.
m. cbr. l. Median cerebral lobe. m.c.f. Margin of fold growing over cere-
bral hemispheres. md.”.1—5. Mandibular nerve, or nerve to the endo-
podite. m.e. Median eye. m.e'. Diverticulum from median eye. m.e. ¢.
Median eye-tube. m./. Furrow leading from uncovered portion of cerebral
hemispheres to the uncovered portion of the crura. m.fc.7. Margin of the
ectodermic fold growing over the crura. wm. ol. ~. Median olfactory nerve-
m.p.com. Medulla of posterior commissure. m. s¢. Mittelstrang. mié. 2.
nerve to chelaria. ze. 1—4. Neuromeres. x. p. Anterior neuropore ; leads
into brain cavity and into median eye-tube. o/. 6. Olfactory buds. o/. /. Ol-
factory lobes. o/.0. Olfactory organ. op.g. Optic ganglion. o. g. 1—3.
Lobes to optic ganglion. op. z'. Nerve to operculum. yp. 2—6. Posterior
hemal nerve of 2—6 neuromeres. p.c. Posterior commissure. p.c.s. Ped-
uncle to internal cerebral lobe (corpus striatum). .e. Posterior edge.
p.e.c. Medullary nucleus at root of parietal eye-nerve. p.g. Pigment
granules. p.2.1—6. Pedal nerves. p. op. g. Peduncle to optic ganglion.
p.ol.o. Primary olfactory organ. op. s¢. Stalk to the pigmented plexus
arising from the primary olfactory organ. 7. Rostrum. 7. m.e.2. Roots

MORPHOLUGY OF BRAIN AND SENSE ORGANS OF LIMULUS. 93

to median eye-nerve. 7. m.e.¢. Remnants to median eye-tube. 7. ms.
Rostral mesoderm. ss. /. Semicircular lobes = infundibulum = l’organ stratifié.
st. g. Stomodeal ganglion. s¢. ~. Stomodeal nerves. s. ¢. p. Strand con-
necting the nucleus to parietal eye-roots with the peduncle of optic ganglion.
t.e.c. Thin ectoderm between cerebral hemispheres. 74. 7. Thickened ring
surrounding the cephalo-thorax. v. 5th ventricle. va. x. Vagus neuromeres
and nerves. v.s. Ventral surface.

PLATE 3.

Fic. 24.—Slightly diagrammatic view of the brain of an embryo in a stage
when the legs and from two to three pairs of gills are developed. The draw-
ing is constructed from surface views and sections, and is intended to show the
relations of the various invaginations and the lines along which the imperfect
ectodermic folds (m. c.f, m. /f. c. 7.) advance to enclose the brain.

Fie. 25.—Similar view of an older embryo, showing the greatly increased
cerebral hemispheres and the areas still uncovered by the advancing folds,
m. c.f. and m.f. er.

Fic. 26.—Camera drawing of a cross-section through the brain in a stage
like that shown in Fig. 19. The section passes through the end of the line,
p. inf.

Fic. 27.—Cross-section of a brain in a later stage than that shown in
Fig. 24, and in a plane that would pass through about the end of the linec. m.
of Fig. 24.

Fies. 28—32.—Cross-section of the brain and nerve-cord in about the same
stage as that shown in Fig 24.
In Fig. 28 the section plane passes through about the middle of the cere-
bral hemispheres (compare Fig. 24).
In Fig. 29, just back of the cerebral hemispheres.
In Fig. 30, just in front of the first pair of appendages.

In Fig. 31, through one of the lateral nerve-cords, just posterior to the
second pair of appendages.

In Fig. 32, through the ventral cords, a little back of the middle of the
third post-oral neuromeie. x 200.

Fic. 38.—Longitudinal section of a later stage than that of the preceding
series, passing through the base of the optic ganglion. It shows the ante-
rior wall of the invagination, and first differentiation of the primary olfactory
organ. xX 200.

Fic. 34.—Part of a longitudinal section in a little older stage than that of
the preceding series. The section shows the early stages of the lateral eye,
while it is in its primitive position close to the optic ganglion.

94, WILLIAM PATTEN.

Fies. 35—40.—A series of longitudinal sections through the brain in
about the stage shown in Fig. 24. There were 28 sections in the series, sec-
tion 14, Fig. 25, passing through the sagittal plane. The position of the
sections is shown on Fig. 24 by the numbers 35—40. Only those sections are
represented that show some variation in the invaginations.

Fie. 41.—Diagrammatic longitudinal section through an early stage of the
brain of Limulus to show the relation of the median eye-tube to the anterior
wall of the brain.

Fic. 42.—Same of the scorpion.

Fic. 43.—Right half of the brain of a young Limulus about 3 inches long,
viewed from its cut surface. The outlines and proportions of all the parts
are drawn from a wax plate model, same as that in Fig. 49, and is diagram-
matic only in so far as the invagination cavities of the infundibulum, optic
ganglion, median eye, and olfactory organ and neuropore are represented as
persisting up to this period instead of disappearing soon after their formation.

Fic. 44.—Cross-section through the dactylopodite of a Limulus about 8
inches long, showing general distribution of the cuticular canals for the gusta-
tory and temperature organs. The gustatory canals are most abundant along
the anterior cutting edge of the joint. The cuticula about the canals is in
some places stained black by the osmic acid used for that purpose.

Fie. 45.—Under surface of a young Limulus 61 mm. long, showing the
position of the olfactory organs in reference to the mouth, also its radiating
choroid plexus of white pigment-cells. x 2.

EXPLANATION OF PLATE 4.

Fic. 46.—Nervous system of Limulus in its second larval stage seen from
the dorsal or hemal surface. x 66.

Fie. 47.—Cephalo-thoracic nervous system of the same seen from the
ventral or neural surface. x 66.

Fie. 48.—Nervous system of Limulus about 24 inches long, seen from the
neural surface. Constructed from sections and dissections. The sympathetic
system is not represented in Figs 36—88. x 15.

Fig. 49.—Fore-brain region of a young Limulus about 3 inches long, seen
from the hemal surface. The drawing is made from a wax plate model,

enlarged about 100 diameters. The drawing is reduced to about 30
diameters.

Figs. 50 and 51.—Cross-sections of the brain of Limulus in the second
larval stage (see Figs. 46 and 47).

In Fig. 50 the section plane lies some distance back of the peduncles, and
shows the overlapping of the cerebral hemispheres.

MORPHOLOGY OF BRAIN AND SENSE ORGANS OF LIMULUS, 95

In Fig. 51 it passes through the middle commissure, the peduncles of the
cerebral hemispheres, and the last connection of the hemispheres with
the ectoderm. xX 275.

Fic. 52.—Cross-sections of the brain of an adult Limulus, passing through
the slender peduncles on which the enormous cerebral hemispheres are sup-
ported. ‘The plane of section cuts the anterior border of the middle commis-
sure. Its position is shown on a younger brain by the line 21, Fig. 43.
x 15.

EXPLANATION OF PLATE 5.
Reference Letters to Plate 5.

a. p. st. Anterior pons stomodei. at. x. Nerve to antenne. at. neu.
Antennary neuromere. a. st. g. Anterior stomodeal ganglion. a. st. x.
Anterior stomodeal nerve. dr. 1—3. Brain lobes. ec. Brain commissure.
c.e. Convex eye. ch.g. Ganglion to cheliceral nerves. c. /. Cerebral lobes.
ch, m. Cheliceral nerve. ec. p. e. Ectoparietal eye. ez. p. e. Endoparietal
eye. ep. Epiphysis. /. B. Fore-brain. g. ab. Transverse tube of nervous
substance. g. v. 1—3. Invagination of optic ganglia. 7. md. Inner mandible.
t. md. m. Inner mandibular nerve. 7. Labrum. J/.e. x”. Lateral eye-nerve.
1. st. g. Lateral stomodeal ganglion, 7. sy. Lateral sympathetic nerve. M. B.
Mid-brain = antennal or cheliceral neuromere. m, sy. Median sympa-
thetic nerve. 2. ec. p.e. Nerve to ectoparietal eye. x. ex. p.e. Nerve to
endoparietal eye. . 7. Labral nerve. x. oc. Nerve to ocelli. 2.7. p. e.
Nerve-roots to parietal eye. o. Zo. Nerve of Tomosvary. oc. Ocelli. 0. e.
(sophagus. o. gy. 1—3. Optic ganglia. o. md. ”. Outer mandibular nerve.
o. md. Outer mandible. p. e. Parietal eye. p. gy. Pedal ganglion. p. m.
Primitive mouth. p. 2. Pedal nerve. p.p. st. Posterior pons stomodei.
s. J. Semicircular lobes, v. c. Ventral cords.

Fig. 53.—Diagram of nervous system of an insect, showing relations of
stomodeal and trunk sympathetics to the cephalic lobes and ventral cords.

Fic, 54.—The same, with the stomodeum evaginated, to illustrate what
was probably the original distribution of the stomodzal nerves.

Fic. 55.—Hypothetical embryo seen from the side, showing distribu-
tion of stomodeal nerves.

Fig. 56.—The same, with the stomodeum evaginated.

Fie. 57.—Diagram of the cephalic lobes and stomodeal nerves of an insect
embryo (Holometabolic).

Fig, 58.—Same of a scorpion.

Fic. 59.—Same of Limulus.

Fic. 60.—-Brain of Acilius at beginning of pupal stage, slightly diagram.
matic.

96 WILLIAM PATTEN.

Fic. 61.—Diagram of a spider’s brain, showing relation of the ocelli to the
brain-segments.

Fie. 62.—Brain of a Myriapod (Iulus).

Fig. 63.—Cross-section of parietal eye, just after hatching (early trilobite
stage); depigmented. xX 265.

Fi1es. 64—70.—Series of cross-sections of the perietal eye, after first larval
moult (trilobite stage) ; not depigmented.

Fig. 64. Section through about the middle of the endo- and ecto-parietal
eyes. X 265.

Fig. 65. Section through the root of the eye, showing the four nerves.
x 570.

Fig. 66. Section of parietal eye-stalk just below the eye. x 570.

Fig. 67. Section shows the primitive roots to the parietal eye-nerve
separating from the epiphysis. x 570.

Fig. 68. Section of the epiphysis just below the preceding section.
x 570.

Fig. 69. Section through the base of same. x 570.

Fig. 70. Section through the base of the epiphysis, showing its union
with the transverse tube, and the communication of the cavities in the
same. X 570.

Fig. 71.—Cross-section through the base of one of the legs, showing the
developing sense organs, nerves, and ganglia in the inner and outer man-
dibles. Trilobite stage. x 265.

Fig. 72.—Section of same in an embryo just after the shedding of the
chorion. X 265.

Fic. 73.—Same just before hatching. x 87.

Fie. 74.—Parietal eyes of a Limulus about 3 inches long. The eyes have

been dissected away from the cuticle and surrounding connective tissue,
mounted in glycerine and viewed from above. xX 38.

Fic. 75.—Three sensory hairs from the cephalo-thoracic shield of Limulus
just after the trilobite stage. Their position and relation to the adult sensory
hairs were not determined.

Fic. 76.—Sensory bud from thorax of same stage.

STRUCTURE OF THE PHARYNGEAL BARS OF AMPHIOXUS. 97

The Structure of the Pharyngeal Bars of
Amphioxus.

By

W. Blaxland Benham, D.Sc.Lond., Hon. M.A.Oxon.,
Aldrichian Demonstrator of Comparative Anatomy in the University of
Oxford.

With Plates 6 and 7.

Iv may be thought that the structure of the pharyngeal bars
of Amphioxus is sufficiently known, after the description
by Lankester, and more recently by Spengel; but there
still remains a certain amount of doubt as to some points in
the structure of the tongue or secondary bar, although recent
authors are in agreement as to the general structure of the
primary bar. It is to the tongue bar, therefore, that I have
more particularly directed my attention.

Spengel contradicted Professor Lankester on several points,
both with regard to matters of observation, and more espe-
cially with regard to certain interpretations, in a very dog-
matic and, indeed, discourteous manner. I was surprised to
find that Professor Spengel himself is by no means correct
in sundry matters of mere observation.

It will be remembered that Lankester, in addition to his
account of the structure of the gill bars—which was, like his
figures, in great advance of the work of previous writers on
the subject—made certain statements with regard to the spaces
within these bars; he attempted to distinguish, not only in the

VOL. 35, PART 1.—NEW SER. G

98 W. BLAXLAND BENHAM.

bars, but in Amphioxus as a whole, coelomic spaces from
blood-vessels. It is chiefly with regard to the interpretation
of the spaces within the tongue bars that Spengel joins issue
with him, and it was to this part of the subject that I directed
my attention last summer in the endeavour more especially to
decide whether the cavity traversing the “chitinoid” rod of the
tongue bar be celom (Lankester) or blood-vessel (Spengel).

I hoped to decide the question by the examination of speci-
mens which had been fed with carmine; and for this purpose
Professor Lankester requested Mr. A. Willey, who was then at
Naples, to feed some animals and preserve them for me. My
thanks are due to Mr. Willey for so doing. Unfortunately,
however, these carmine-fed specimens have not been of much
value for my purpose, and I was compelled to fall back upon
carefully preserved specimens (in picro-sulphuric acid) stained
in various media. This was the method followed both by
Lankester and by Spengel—careful tracing of spaces from
section to section in order to ascertain their communication
with other cavities about whose nature there is no doubt.

Before I had completed my work, Boveri’s most interesting
paper on the nephridia was published. Boveri found, as I had,
that Spengel, both in incomplete observation and in certain
interpretations, had fallen into errors similar to those which
with an accompaniment of gratuitous insolence he had attri-
buted to Lankester.

This short note may conveniently be divided as follows:

a. Description of the tongue bar, according to my own ob-
servations, and comparison of it with the primary bar.

B. Interpretations of the parts of the tongue bar.

c. The observations of recent observers.

p. Certain abnormalities in the pharyngeal bars.

The authors who have described and figured the gill bars of
Amphioxus, to whom I shall have occasion to refer, are—

1. Strepa.—“ Studien iib. d. Amphioxus lanceolatus,” ‘Mem. Ac. Sci.
Pétersbourg,’ xix, 1878.

STRUCTURE OF THE PHARYNGEAL BARS OF AMPHIOXUS. 99

2. Lanceruans.— Zur Anat. d. Amphioxus lanceolatus,” ‘ Arch. f.
mikr. Anat.,’ xii, 1876.

3. ScuneipER.—‘ Beitrige zur Vergleich. Anat. und Entwickel. der Wir-
belthiere,’ 1879.

4, LANKESTER.—‘‘ Contributions to the Knowledge of Amphioxus lanceo-
latus,” ‘ Quart. Journ. Micr. Sci.,’ xxix.

5. Spencet.— Beit. z. Kenntniss d. Kiemen des Amphioxus,” ‘ Zool.
Jahrbuch’ (Anat.), iv, 1891.

6. Bovert.—‘‘ Die Nierencanalchen des Amphioxus,” ‘ Zool. Jahrbuch’
(Anat.), v, 1892.

In addition to these, a bibliography relating to Amphioxus

will be found in Lankester’s paper.

a. The Tongue Bar.

The tongue (or secondary) bar is usually distinguished from
the primary bar (a) in being supported by a tubular skeletal
rod in place of the double rod of the primary bar, (4) and in
being without any ceelom between the rod and the atrial epi-
_ thelium. It is unnecessary to describe the relations of the
bars to one another or to neighbouring parts of the animal, as
these matters have been fully described and illustrated by re-
cent writers on the subject.

The structure of the bar is most readily seen in its transverse
sections, but such sections—accurately transverse to the bar
—are not so easily obtained ; in sections transverse to the long
axis of Amphioxus, only one or two bars on each side of the
pharynx will be cut transversely, though in the pre-hepatic
region more bars are so cut, and still more are very nearly
transversely cut, than is the case posteriorly. But by varying
the obliquity of the plane of section to that of the long axis of
the body, I was able, ultimately, to obtain sections which cut
nearly the whole series of bars in any section almost accu-
rately at right angles to their length.

It appears to me that this is most important, for the discre-
pancies in various descriptions and figures of the bar are doubt-
less due to the more or less obliquity of the sections.

Another matter which must be taken into serious account is
the mode of preparation and the character of the stain; for I

100 W. BLAXLAND BENHAM.

have noted in my series—treated differently in both these
respects—various differences due to shrinkage and such effects,
which suggest the cause of certain omissions by some authors,
and of wrong interpretations and other errors.

I have found that Kleinenberg’s fluid (picro-sulphuric acid)
is the preservative which produces less distortion than other
reagents. Cochineal stains, especially Mayer’s alcoholic cochi-
neal, serves best for the demonstration of blood in the vessels
and for the examination of the skeletal tissues.

Hematoxylin is, of course, useful for the nuclei, but cochi-
neal gives better results in the case of cell-bodies.

Some of the animals were stained in bulk; in other cases
the sections were stained on the slide.

I proceed now to a description of a transverse section of a
tongue bar, as elucidated by the examination of sections
treated in different ways, as well as of isolated portions of
pharyngeal wall.

Such a section is a narrow bar (about three times as long as
it is wide), presenting two ends and two sides: one end, the
Outer end, projects into the atrium, and is covered by atrial
epithelium ; the opposite end, or Inner end, projects into the
cavity of the pharynx, and is lined by hypoblast. The sides
are directed (roughly) anteriorly and posteriorly.

The bar consists almost entirely of columnar epithelial
cells, the inner ends of which abut upon a membrane, the cutis
(I follow here Professor Lankester), throughout the greater part
of the bar; whilst at the outer end the cells rest upon the
skeletal rod, which I regard as merely a specialised part of
this cutis.

The character of the epithelium, however, differs at the two
ends and at the sides. At the Inner or pharyngeal end the
nuclei are arranged in three groups, a central group and a
lateral group on each side, as Lankester was the first to point
out. The nuclei of the central group are arranged in two
more or less curved rows (as in fig. 1), but this arrangement
is due not to the existence of two layers of cells, but to the
fact that alternately the nuclei are situated nearer to or

STRUCTURE OF THE PHARYNGEAL BARS OF AMPHIOXUS. 101

further from the base of the cell. The nuclei are elongated
and oval ; they are placed much closer to one another than my
figure represents, and as they stain deeply are very conspi-
cuous. In thick sections it is difficult to determine whether
one has to do with a single row of very elongated cells (as
Lankester believed), or with several rows of them; this diffi-
culty is emphasised when the sections are not accurately
transverse. But in thin sections, successfully cut, it is easily
seen that the nuclei are arranged as I describe them.

The cells to which these nuclei belong are, therefore, as
long as the epithelium is thick; they are very narrow peri-
pherally and swollen at the nuclear level, producing the flask-
shaped appearance of the whole group. These cells carry
cilia which are considerably longer than those carried by the
lateral groups of cells at this inner end ; and it is curious that
of previous observers, only Spengel and Boveri have noted
this special bundle of cilia. The free ends of the cells are pro-
vided with a very finely striated border, which comes out well
in cochineal preparations, but which is not differentiated by
other stains used.

Each lateral group of nuclei consists of a single row curving
downwards from the central group towards each side, some-
what in the way represented by Lankester; but I find that
these nuclei, which are long and narrow, are not arranged
quite in the fan-shaped manner represented by him.

The cells containing these nuclei carry quite short cilia—
entirely overlooked by Spengel and Boveri,—and their free
ends are not provided with a striated border. The shape of
this inner end has been very variously represented, as the
copies of previous figures on Pl. 6 will show.

The side of the bar presents some four or five rows of small
nuclei forming a broad band, about two thirds the whole width
of the epithelium. These nuclei are not circular, as most ob-
servers have represented them (owing to the obliquity of the
section, as I know from experience), but are oval, with the
long axis directed vertically to the plane of the surface.
Langerhans appears to have noted their oval shape.

102 W. BLAXLAND BENHAM.

Here, again, these rows of nuclei do not represent as many
layers of cells, for there is but a single layer of very long and
very narrow cells, with the nuclei at different levels in neigh-
bouring cells. In some of my preparations these cells are more
or less macerated, so that I was able to confirm the description
of them given by Langerhans (see fig. 19). Each cell, or at
any rate most cells—for it is impossible to be absolutely certain
that all the cells reach the surface, though I believe this to be
the case—carries a single very long cilium, and presents a
finely striated border. The nuclei of the lowermost row ap-
pear to be slightly longer than those of other rows, and are
close to the septal membrane or cutis, along which they form
a very well-marked series. Towards the extremity of the side,
both outwards and inwards, i. e. towards the atrial end and the
pharyngeal end, the lowest row of nuclei curve upwards to-
wards the surface, so that the epithelial cells become shorter
and shorter, and the number of rows less and less, till finally
there is but a single nucleus contained in a cell not much longer
than itself (see fig. 1). At the inner end of the bar the thin
membrane-like cutis curves outwards towards the surface, and
naturally the row of nuclei take on this curvature. At the
opposite end of the bar the row of nuclei follow the curve of
the chitinoid rod; but at this end, for a short space, the epi-
thelium is overlapped, as it were, by a row of three or four
cubical cells containing pigment, and not carrying cilia. These
cells are part of the atrial epithelium, the invaginated epiblast
of the pharyngeal slits. These pigmented cells differ from
those hitherto mentioned in containing circular (spherical)
nuclei, and herein agree with the cells constituting the epithe-
lium of the Outer end of the bar.

This atrial epithelium is characterised by the larger size of
its cells, the absence of cilia, and the presence of relatively
small round nuclei, which are placed at different levels in the
cells (fig. 1). These cells are highest in the middle of this end
—i.e. at the end of the long axis of the section through the bar
—and decrease in length at each side of this point till they
graduate into the cubical pigmented cells which overlap the

STRUCTURE OF THE PHARYNGEAL BARS OF AMPHIOXUS. 103

cells of the sides of the bar. In the tongue bar these cells are
all of one kind (fig. 18), whereas in the primary bar two kinds
of cell (fig. 17, a, 6) are present, one (a) being vacuolated, the
other (5) more granular. Langerhans noted this fact, though
he exaggerated the difference. So much for the epithelium of
the bar. I shall point out later how far these statements of
fact differ from those of my predecessors.}

Turning now to the cutis of the bar, i.e. the septal mem-
brane and the chitinoid rod: the septal membrane forms
a very thin sheet of tissue traversing the greater part of the
long axis of the section, and separating the epithelium of the
two sides. At the base of the epithelium of the Inner or
pharyngeal end of the bar the septal membrane splits
into two, and each of the two branches curves outwards to-
wards the surface; this forking of the membrane leaves a
V-shaped space, which is converted into a triangle by a mem-
brane (cutis) at the base of the pharyngeal epithelium.

In this triangular space is a blood-vessel, as Spengel and
Boveri have already described (figs. 7,8). It may be called the
internal or Visceral vessel. At the opposite end the septal mem-
brane similarly divides into two, each half of which appears to
be continuous with the corresponding side of the rod (fig. 1).
This space, which differs somewhat in shape according to the
shape of the rod, but which is, on the whole, triangular, also
contains a blood-vessel—the external, or somatic vessel. This
vessel was observed by Lankester (fig. 6), but overlooked by
Spengel (fig. 7), although he represents the space here, whilst
Boveri described it as existing only in the primary bar. From
the fact that, at each extremity, this septal membrane forks,
and from theoretical considerations, I believed this membrane
to be in reality double, as indeed it is represented by Stieda’s
figure (Pl. 6, fig. 3). But I was for a long time unable to
assure myself of this fact. I was unable to satisfy myself as to
the presence of two membranes here, for, owing to the refran-
gibility of the structure, it is difficult to make certain whether

* [have not observed the ‘ muscle-cells” in the bar described by Rohon
and by Langerhans, J, Miiller and Schneider.

104 W. BLAXLAND BENHAM.

one sees a double outline to the single membrane (such as
Boveri represents, Pl. 6, fig. 8) or two closely apposed mem-
branes.

Spengel vehemently animadverts on Lankester’s interpre-
tation of this membrane as a mesoblastic cutis, and insists on
its being a “ basement membrane,” i.e. epiblastic. Lankester
was, it seems to me, in error in referring the origin of this
membrane to the deepest layer of nuclei in the sides of the
bar; at the same time, if we consider the relations of this
membrane to the rod and to the vessels at each end, we cannot
doubt that the rod and the membrane have the same origin.
The rod would scarcely, I presume, be referred by Spengel
to the epiblast. I believe, however, that I have decided that
the rod is mesoblastic by the discovery of the flattened nuclei
pressed against its inner surface (Pl. 7, fig. 13); and we may
conclude that the rods in both bars are produced by the
flattened epithelium which, as Hatschek has pointed out, forms
the “‘ connective tissue” throughout the body of Amphioxus.
If the rod is mesoblastic, then a priori we may believe the
eptal membrane, which is absolutely continuous with it, to
be also mesoblastic ; but further, I have detected flattened
nucleiin this axis of the bar, i. e. between the two halves
of the closely apposed membranes. I searched long and care-
fully for any nuclei in relation to the septal membrane, and
ultimately, in my series of accurately transverse sections, I
was able, with the aid of Zeiss’s apochromatic, to observe some
structures which I believed to be nuclei. However, I was not
absolutely certain of their existence, owing to the refrangibility
of the membrane and the denseness with which the epithelial
nuclei are packed; but in a series of sections intended to be
horizontal, and stained in hematoxylin, some of the bars were
cut longitudinally for a considerable distance—some six or
seven times the length of an ordinary transverse section of a bar,
—and here I saw distinctly elongated and much compressed
nuclei (fig. 21) of fair size, lying between the two membranes.
These, like the nuclei surrounding ccelomic spaces, are not
rich in chromatin, and do not take the stain easily. We may,

STRUCTURE OF THE PHARYNGEAL BARS OF AMPHIOXUS. 105

therefore, conclude—as, from a priori reasoning, we should
be led to believe—that this septal membrane is double,
and is mesoblastic, and not a basement membrane.

The chitinoid rod of the tongue bar is distinguished, as is
well known, from that in the primary bar by the presence of a
canal—it is a perforated rod.

The shape of the section of the rod varies considerably, both
in different bars and even in different parts of the same bar,
but that represented in this figure may be regarded as the
most general shape. Not only the general outline, but the
shape and extent of the contained space are included in the
above statement as to the variability of the rod. Professor
Lankester has already figured several such variations (loc. cit.,
pl. xxxvi, B), and others will be found on the plate illus-
trating this note (Pl. 7, figs. 13—16). But most gene-
rally the rod presents a somewhat triangular section, with
rounded angle at the base, and a notch—more or less _pro-
found—at the apex. This notch forms a part of the internal
triangular cavity partially bounded by the septal membrane
and occupied by the somatic blood-vessel.

The chief cavity—the canal—of the rod is similarly more or
less triangular in section, rarely round, as Lankester represents
it, though this shape does occur. The thickness of the outer wall
presents very interesting variations: more usually it is nearly
as thick as the sides, but it may be very much thinner (as in
fig. 14); it may be represented merely by a thin membrane little
thicker than the septal membrane, and much thinner than the
cutis (basement membrane—Spengel) below the atrial epithe-
lium of the primary bar. Further, this rod may be represented
by acouple of curved pieces, which do not quite meet externally,
so that the curtained cavity is without an outer wall (fig. 15).
This is always the case at the points where a synapticulum
joins the tongue bar (fig. 31), but it also occasionally occurs
elsewhere.

This cavity frequently appears quite empty, invariably so in
my hematoxylin preparations; but in sections stained with
cochineal, granular matter is very frequently present in greater

106 W. BLAXLAND BENHAM.

or less abundance ;! and I have sometimes noticed an apparent
division of this cavity by a transverse partition, the granula-
tions having a slightly different appearance in the two sides of
this partition (fig. 14). Moreover in this same series of
sections, as also in sections stained with Weigert’s picro-
carmine, the blood in undoubted blood-vessels, such as hepatic
vessels, dorsal and subendostylar vessels, takes on a character-
istic deep red colour—deeper than the tint taken by the
skeletal tissues,—so that I am able most definitely to state that
there are three blood-vessels traversing the tongue
bar, not two, only as Spengel and Boveri believed (c.f. fig. 1
with figs. 7, 8).

Of these three vessels, two occur always at opposite ex-
tremities of the septal membrane, in the triangular spaces
already mentioned ; the third lies inside the rod, and may
be called the skeletal vessel. Usually it has the position
represented in fig. 1—in the chief canal of the rod, but it does
not appear to fill this canal; I can generally detect a
slight space around it. This may, of course, be due to shrinkage
of the clotted blood; but the apparent existence in some Cases
of a partition (see fig. 14 and the explanation of it) favours my
view, as also does the condition of things represented in fig. 13,
where the vessel is passing out of the cavity, that this cavity
of the rod is celom, which contains a blood-vessel.
This view is further strengthened by the fact that, both in my
hematoxylin preparations (where the blood-vessels are not
evident) and in my cochineal sections, I have detected flattened
nuclei pressed against the inner surface of the rod, as repre-
sented diagrammatically in fig. 1, and accurately drawn from
the object in fig. 13. This, I may say, has not been an occa-
sional occurrence, but can be seen in many accurately trans-
verse sections of the bar.

In some of the variations from the normal the rod presents
a small cavity about midway between the main canal and the
apical notch (figs. 18, 14), and this cavity is usually connected

1 Spengel mentions the presence of finely granular material in the canal of
the hollow rod (loc. cit., p. 278).

STRUCTURE OF THE PHARYNGEAL BARS OF AMPHIOXUS. 107

with the former by a narrow channel; in this accessory cavity
I have sometimes seen a vessel in addition to that in the chief
canal, and sometimes I have not detected the latter. I take it
that there may be occasional anastomosis between the ‘‘somatic”’
vessel in the notch and the skeletal vessel in the canal.

Comparison of the Tongue Bar and Primary Bar.

I wish now to compare such a transverse section of a tongue
bar with that of a primary bar, so far as regards the Outer
end. Compare my figures 1 and 2 and that copied from Boveri
(fig. 12). In the primary bar Boveri describes, as I myself
find, three vessels—(a) the inner, or visceral, and (0) outer, or
somatic, as is the case in the tongue bar, and (c) the third or
skeletal (first observed by Spengel) outside the ccelom of the
primary bar, between the atrial epiblast and the ‘‘cutis” (base-
ment membrane of Spengel), This last vessel may project
through the cutis into the ccelom, and at the base of the bar,
where it springs from the endostyle, and where the rod forks,
this blood-vessel lies in the angle of the fork, i.e. in the celom
itself (fig. 30, a), as Spengel has figured. In the tongue bar the
first two of these vessels are identical with those of the primary,
and one can scareely resist the idea that the third, or skeletal
vessel inside the rod, may correspond with the subepidermal
vessel of the primary bar; it differs from it, however, in one
very important point, namely, in the absence of any connection
with the subendostylar vessel. But now turn to the rod itself.
This is, in the primary bar, made up of two pieces, more or less
fused according to the region of the bar (see fig. 30, a, 3, c),
so as to form a triangle, usually with a more or less pro-
nounced notch, or linear channel, arising from near the base
(see fig. 9); or again, as Schneider figured and as I have fre-
quently observed, a triradiate split in its centre! (fig. 30, c),
Outside the rod comes the ceelom—every one is agreed about
that—lined by flattened cells; and outside them the cutis,

1 This usually is due to the softer nature of the central part of this rod,

and is not really a cavity: the rod presents irregularly concentric markings,
as if shrunk, and is firmer externally.

108 W. BLAXLAND BENHAM.

which varies considerably in thickness, and can be traced, as
Boveri’s figure shows, into the rod at each extremity (PI. 6,
fig. 12). This cutis stains exactly like the rod in cochineal,
and also like the rod is unstained in hematoxylin. Sometimes
the cleft in the rod is more pronounced, and gives rise to a more
definite channel (see Lankester, xxxvi B, fig. 3, f) open to the
celom. Such a condition of things is represented in my PI.
7, fig. 31, which passes through a primary rod (P. 1) at the
level of a synapticulum,! and should be compared with certain
variations in the rod of the tongue bar in the same figure and in
fig. 15, and one is struck with the resemblance between the two.

I would suggest that the distinction between the two rods,
viz. that of the primary bar and that of the tongue bar, is not
so profound as one would be led to think from the use of the
terms solid (or bifid) rod and hollow rod. We have seen that
not infrequently the rod of the tongue bar is formed of two
pieces (fig. 15), whilst, on the other hand, the rod of the
primary bar may enclose a cavity. But I think the real dis-
tinction between them is that in the tongue rod the subepi-
thelial portion is usually and typically as thick as the sides,
and distinctly continuous therewith; whereas in the primary
rod the extra-ceelomic piece (subepithelial) is thinner, and,
owing to the greater development of the ccelom here, is more
widely separated from the rest of the rod. This suggestion
occurred to me forcibly in examining the connections of the
synapticula with the two bars.

In a lucky series of sections, cutting the bars very accu-
rately transversely, one often gets the whole synapticulum
in sections, passing from one primary bar to the next, and
showing the connection of the rod in the transverse bar with
those of the main bars (fig. 31). Starting with the tongue
bar (7.), the rod forks, so that its contained cavity is no longer
bounded by a chitinoid wall; one branch of the fork passes
towards each of the adjacent primary bars, and is con-
tinuous, not with the main part of the rod itself, but with

1 Spengel gives a figure very similar to this one in pl. xvii, fig. 13, illus-
trating his paper.

STRUCTURE OF THE PHARYNGEAL BARS OF AMPHIOxuS. 109

the cutis (basement membrane of Spengel), or extra-ccelomic
portion of the rod, as I would regard it. But this apparent
forking of the tongue rod is due merely to the passage of the
contained blood-vessel out of the rod to the vessels of the
adjacent primary bars; so that in reality, as many sections
show, the rod of the connecting bar, i. e. the synapticulum, is
connected on the one hand with the extra-ccelomic portion of
the tongue rod, and on the other with the extra-celomic
‘cutis ” of the primary bar.

A second difference between the rods is that the skeletal
vessel is inside the coelom in the tongue bar, but outside
it in the primary bar over the greater part of its extent,
though, in the lower part of the latter bar, it is intra-coelomic,
as in the tongue bar (fig. 30, a). I have sometimes seen a
small cavity outside the rod in the tongue bar, but I have
not been able to satisfy myself as to its nature; it may be
merely artifact.

B. Summary of my Observations and Interpre-
tations.

1. The epithelium of the bar is everywhere only one cell
in thickness.

2. The arrangement of the cells at the pharyngeal end of
the bar, both in the primary and in the tongue, is much more
definite than previous observers, except Lankester, have
figured and described; the central group, contrary to
Lankester’s opinion, presents two rows of nuclei, and carries
a bundle of very long cilia; the lateral groups carry quite
short cilia.’

3. The nuclei at the sides of the bar are oval and not
round, and the lowest row has nothing to do with the septal
membrane.

4. There are three blood-vessels in the tongue, as in the

1 This differentiation of the cilia round the bar may be compared with that
occurring in the gill filaments of Lamellibranchs, and in the cirri of Bra-
chiopods, where there are similarly bundles of long cilia, situated at the sides
or angles, and shorter cilia elsewhere. The existence of a skeletal tissue in
these cases is a further analogous resemblance.

110 W. BLAXLAND BENHAM.

primary, bar; (a) the visceral vessel at the pharyngeal end of
the bar; (4) the somatic vessel in the apical notch of the rod;
and (c) the skeletal vessel inside the rod: the last two anasto-
mose here and there.

5. The cavity of the rod is celom, is lined by flat cells,
and contains this skeletal blood-vessel.

6. The outer wall of this celom is homologous with the
extra-ceelomic cutis (Spengel’s basement membrane) of the
primary bar.

7. The septal membrane and the rod are mesoblastic, and
the nuclei of the cells forming them are here recorded for the
first time.

c. The Observations of Recent Observers.

I will now pass on to a brief survey of the various descrip-
tions and figures of the tongue bar of previous writers, copies
of whose figures are represented on Pl. 6, and in the expla-
nation of these, the references whence the figures are taken.
Stieda, Langerhans, and Schneider saw no modifications of
the epithelium at the pharyngeal end of the bar. Lankester
was the first to observe this, and his figures represent more
accurately than do those of his successors, Spengel and Boveri,
the actual arrangement. As I have already pointed out, how-
ever, he missed the fact that the middle group presents two
rows of nuclei. This is equally overlooked by Boveri, while
Spengel’s drawing is not quite clear on this point, and neither
of these latter indicate the much greater length of the nuclei
in this part of the bar. So far as the general shape of the
pharyngeal end of the bar is concerned, Spengel’s and Boveri’s
drawings show it fairly. Fig. 12 from Boveri is taken from
quite the upper end of the bar, hence the peculiar shape of
this end. Lankester is alone in regarding—as I believe
wrongly—the epithelium of the sides of the bar as in several
layers. Spengel points this out; but he, like all my prede-
cessors, with the doubtful exception of Schneider and Langer-
hans, represents small round instead of oval nuclei here.

The authors, as will be seen from the reproductions of

STRUCTURE OF THE PHARYNGEAL BARS OF AMPHIOxvuS. I111

their figures, represent the characters of the atrial epithelium at
the upper end of the bar correctly, with the exception of Lan-
gerhans, who describes each cell here as bearing a flagellum.

Stieda and Langerhans did not recognise the differences in
the length of the cilia round the bar. Schneider is the first to
figure the long cilia at the sides, and the remaining authors
follow him. Spengel is the first to note this tuft of long
cilia at the pharyngeal end; but both he and Boveri! over-
looked the short cilia carried by the lateral groups of cells at
this end.

With regard to the contained vessels in the primary bar,
Langerhans entirely missed the vessel at each end of the
septal membrane. Schneider and Lankester saw only one
vessel—that at the outer end of the bar at the apex of the rod
—the somatic vessel. Spengel most unaccountably missed
this vessel, although he states that he looked carefully for one
here, and his figure of the tongue bar (PI. 6, fig. 7) shows a
very narrow cleft in its position, whilst he describes the inner
or visceral vessel (vn.), which is usually less noticeable than the
other. In this respect his figure shows no advance upon that
of Stieda, who also saw, but did not describe, the space at the
inner end of the bar.

Boveri, however, observed this vessel (and figures it) in the
primary bar (fig. 12, ve. 1), in the position already given to it
by Lankester, but overlooked it in the tongue bar (fig. 8). As
a matter of fact, as I have stated above, there are three
vessels in each bar. Schneider and Lankester found one, the
“outer or somatic” vessel; Spengel found one, the “ inner or
visceral vessel,’ but denied the somatic vessel. Boveri found
both these in the primary bar, but overlooked the “ outer
vessel”’ in the tongue bar.

As to the nature of the rod cavity, Stieda and Langerhans
leave it aside.” Schneider represents (at A, fig. 5) “a blood-

1 It should be borne in mind, however, that Boveri did not pretend to
describe the bars except in so far as they are related to the nephridia.

? Stieda regards the granular substance in the centre of the rod (fig. 2, a)
as an axial part of the rod itself.

Lt W. BLAXLAND BENHAM.

vessel communicating with the branchial artery.” Lankester,
from its relation to the subendostylar coelom, regarded this
canal in the rod as ‘‘celom;” whereas Spengel and Boveri
look upon it as a blood-vessel, and deny its coelomic character.
Lankester gives half a dozen figures (loc. cit., pl. xxxvi B, figs.
5—9) of as many consecutive sections representing the canal
communicating with the subendostylar celom. Spengel denies
this altogether, and states that the rod becomes solid before
it reaches the endostyle, and there becomes continuous with
the subendostylar skeleton. He further denies any communi-
cation of this rod cavity with the dorso-pharyngeal coelom.

Now the tracing of these cavities is an extremely difficult
matter, owing to the difficulty of making sections in the right
plane, and I searched through section after section before I
could satisfy myself as to the mode of termination of these
rods. Time after time it seemed to me that Spengel was right,
so far as the non-communication of the rod cavity was con-
cerned ; but in certain lucky sections I found that the plane
was convenient for this purpose, and I represent five consecu-
tive sections which show, as I believe, the continuation of the
subendostylar ccelom into the cavity of the rod (Pl. 7, figs.
22—26). The series shows most certainly no continuity be-
tween the rod and the subendostylar skeleton on which Spengel
insists.

I was not successful in tracing the rod cavity into the dorso-
pharyngeal ceelom; but I attribute this to the difficulty of ob-
servation, and am by no means inclined to conclude that such
a connection does not exist.

The “ skeletal vessel ” ceases some little way before the rod
does, being connected with the vessels in the neighbouring
primary bars at the lowest synapticulum.

If the hinder gill-slits be examined in a preparation of the
pharynx, flattened out on a slide, and the mode of development
of the “ tongue ” observed, it will be seen that the rod in the
tongue is double at its upper end; the two pieces diverge
and constitute the arcade connecting the series of rods. The
tongue bar, as is known, is a downgrowth from the upper

STRUCTURE OF THE PHARYNGEAL BARS OF AMPHIOXUS. 113

boundary of a primary slit ; the celom is here in the form of a
canal giving off shoots to each bar, and fro them appearance
presented in such a preparation, it is not an inconsistent inter-
pretation to regard the ccelom as sharing in the downgrowth
of its ventral wall. The appearance presented by the rod in
the hinder gill-slits is thus: M.

p. Certain Abnormal Bars.

In one series of sections the pharynx presented a few bars
containing sometimes three and sometimes two rods. In all
cases these abnormal bars are “ primary” bars, i.e. the rods
contained within them are cleft rods, similar to those found in
normal primary bars; and on each side of such abnormal bars
there lies a normal tongue bar, separated by a pharyngeal slit
from the abnormal one. On Plate 7, fig. 27, I represent
one such bar cut at about the middle of its length, containing
three rods, which are in all respects similar to one another.
The bar itself is otherwise normal, and presents the usual
characters of the epithelium in its various parts. Naturally
the bar is much wider than an ordinary one, especially at its
atrial end. The septal membrane (s.m.) is very short, but the
two usual blood-vessels (viv.v. and som.v.) are present. There is
not a vessel at the apex of each rod, as one might expect; the
ceelom is very extensive, and dips in between the rods as the
nuclei show. The usual subepidermic blood-vessel (shel. v.) is
also present. But this bar, if traced upwards towards the
epibranchial groove and downwards towards the endostyle,
presents variations of this condition in different regions.

At its origin dorsally the three rods are fused together, but
this occurs for only a very short space, and we soon find three
rods. Towards its lower end two of these (Nos. 1 and 8) are
fused, so that the bar has but two rods; and still further a
junction between these two is effected (fig. 28), and there
appear to he but one rod, rather larger than usual, but evidently
consisting of two united rods. Soon, however, these separate
again, and I take this union as representing a synapticulum.

But now the relative position of the two rods changes ;

VOL. 35, PART 1—NEW SER. H

114 W. BLAXLAND BENHAM.

hitherto they have been side by side, or rather anterior and
posterior ; but now (fig. 29) one (No. 2) lies outside the other
(Nos. 1 and 3), which occupies the normal position. I was
unable to track the rods quite to their junction with the floor
of the pharynx owing to the imperfection of some of the sec-
tions in this series.

Most of the other abnormal bars contained two rods, but I
did not trace them all out from top to bottom, so that I am
unable to state whether they always contain, at some part of
their length, three rods. But I am inclined to answer this in
the negative ; so frequently are there only two rods that I
think this is the more “‘usual”’ abnormality. There is another
triple-rod bar on the opposite side of the pharynx, at about the
same level as the one described, and usually, as far as I could
observe, the abnormal bars are opposite.

A bar with a double rod might conceivably be produced by
closure of a primary slit between two primary bars; but there
is no evidence of any formation of a slit and subsequent closure
and fusion of the bounding bars. One would expect if this
had happened that the ceelom and subepidermic vessel would
be doubled, and that the pharyngeal end of the bar would
exhibit some peculiarity ; this I did not find to be the case. A
triple-rod bar might be explained by an extension of this sug-
gestion, namely, that two slits had remained imperforate,
whilst the rods had been formed nevertheless.

EXPLANATION OF PLATES 6 and 7,

Illustrating Dr. W. Blaxland Benham’s paper on “ The
Structure of the Pharyngeal Bars of Amphioxus.”

In most of the figures the blood-vessels are represented black, but in fig. 14
I have drawn the actual appearance presented by the sections.

Fics. 1 and 2.—Transverse sections of a tongue and a primary bar, so far
diagrammatic in that each figure is a combination of several drawings of

STRUCTURE OF THE PHARYNGEAL BARS OF AMPHIOXUS. 115

different sections, which have been treated in different ways. The two figures
are drawn to scale, and therefore represent the true relative sizes of the bars.

Fig. 1. A transverse section of a tongue bar of the pharynx of Amphi-
oxus (Branchiostoma lubricum, Costa). It represents, accu-
rately as I believe, a typical section of the bar. Most of the parts
are fully named on the plate. The rod contains a cavity—the celom,
lined by an epithelium, whose nuclei are shown (see also Fig. 13),
Within this ccelom lies a blood-vessel—the skeletal vessel. Two other
blood-vessels are present in the bar, the ‘‘visceral” (Visc. Bl. vessel)
and the ‘‘somatic” blood-vessel ; these lie at either end of the septal
membrane, between the two layers of which are shown three nuclei
(see also Figs. 20, 21). The atrial epithelium consists of only one
kind of cell. The grouping of the epithelial nuclei at the pharyngeal
end of the bar may be noted.

Fig. 2. A transverse section of a primary bar (see explanation of Fig. 1).
Here the ccelom is more extensive, and the “cutis”—or outer wall of
the ccelom—is thinner than in the tongue bar. The rod itself is made
up of two pieces, meeting along the middle line of the bar, and a
third piece wedged between. ‘The skeletal blood-vessel is outside the
celom, but is in reality surrounded by the rod, i.e. “cutis.” The
atrial epithelium consists of two kinds of cells (see Fig. 17), viz.
a, the vacuolated, and 4, the granular cells.

Fies. 3 to 12 are tracings of figures published by previous observers; the
only alterations that have been made are (1) colouring the rod yellow, and (2)
filling in with black the spaces regarded as blood-vessels by the authors.

Fig. 3. Copied from Stieda’s fig. 6, pl. i. ‘Transverse section of a
branchial plate. a. The rod. 6. Deep layer of the epithelium. cc.
Superficial layer with cilia.”

Fig. 4. Copied from Langerhans, fig. 24, pl. xiii. ‘“ Transverse section
through a gill bar. m. Epithelium. 4. Nuclei of branchial epithelium
p.e. Pigmented epithelium. 4. The atrial epithelium. s. Hollowgill rod.”

Fig. 5. Copied from Schneider’s fig. 4, pl. xiv. ‘Transverse section
of a thin (that is, tongue) gill bar. a@. Triangular space, a blood-
vessel. 4. A blood-vessel in communication with the branchial artery.
k!, The rod. p.p. Peritoneal (i.e. atrial) plate.”

Fig. 6. Copied from Lankester, pl. xxxvi B, fig. 2.‘ Transverse
section of a tongue bar. a/. Left inner epithelial band. ar. Right
inner epithelial band. am. Median inner epithelial band. cod. Columnar
lateral cells, with long cilia. x. Superficial nuclei. 71. Deeper nuclei.
sept. Clear septal tissue. B/. v. Supposed blood-vessel, ending blindly
at the ventral extremity. pig. Lateral groups of pigment in the atrial
epithelium. at. ep. Atrial epithelium (epidermic).”

Fig. 7. Copied from Spengel, fig. 19, pl. xviii. “Transverse section of
a tongue bar. vz. Chief vessel of the bar. vz, Accessory vessel,”

116 W. BLAXLAND BENHAM,

Fig. 8. Copied from Boveri’s fig. 14, pl. xxxiii, “Transverse section
of tongue bar at the level of the nephridium. eph. Position of nephri-
dium. vi!!, Inner axial vessel. ve!!. Outer axial vessel.”

Fig. 9. Copied from Schneider, pl. xiv, fig. 3. ‘‘ Transverse section of
a thick (i.e. primary) gill bar. 4!, 4?. Blood-vessels which are in
communication with the branchial artery. a@. Triangular space, blood-
vessel. p. p. Peritoneal (i.e. atrial) plate. V. Vein in communication
with the subvertebral vein, and the venous space around the branchial
artery. 4. The thick rod.”

Fig. 10. Copied from Lankester, pl. xxxvi n, fig. 1. “ Transverse
section of a primary bar. B.v. Blood-vessel connected with the lateral
branches of the median endostylar vessel. fiss, Fissure due to the
bilateral origin of the rod. ca. ep. Celomic epithelium. pg. Pigmented
atrial epithelium.”

Fig. 11. Copied from Spengel’s fig. 12, pl. xvii. “Outer part of a
transverse section through a primary bar. 4. m. Basal membrane.
cok. Coelomic canal of the bar. sp. Skeletal rod. v#. Chief vessel in
the primary bar.”

Fig. 12. Copied from Boveri, fig. 6, pl. xxxii. ‘‘ Transverse section of
a primary gill bar, vé!. Inner axial vessel. ve. Outer axial vessel.
ve. Celomic vessel. ce. Colom.”

Fic. 13.—A tongue rod in section, to show the skeletal blood-vessel (sh. 6.
v.) running in a special canal; it is probably about to anastomose with the
somatic vessel (som. 6. v.). Around the ccelom (ce/.) three nuclei (x. ¢. ep.)
are present, embedded in granulations (? protoplasm). In this section the
two blood-vessels, as is frequently the case in cochineal preparations, are
stained deep red. The ccelom is quite clear. cz. cutis.

Fic. 14.—A rod from a tongue bar, exhibiting granulations in the cavity,
which are of two kinds: a coarser, less deeply stained, in the outer part, which
is probably the protoplasm of ccelomic epithelium—in that region marked
ce.—ccelom; and a finer and more deeply staining mass (shel. b. v.), the
skeletal blood-vessel. The two are apparently separated by a partition.
n. is the nucleus of the ccelomic epithelium. som. 4. v. The outer or somatic
blood-vessel. cz. The “cutis” or outer wall of celom. From a preparation
stained with Mayer’s alcoholic cochineal. Drawn under Zeiss’s homogeneous
immersion.

Fic. 15.—A modification in the shape of the section of the rod, which is
not unusual, especially near the synapticula. The rod is here composed of
two pieces. The outer wall of the ccelom is free of skeletal tissue (cf. Fig.
31), and the blood-vessel (shel. 5. v.) is passing out of the ccelom (c@/.), and
lies just below the atrial epithelium (a?. ep.).

Fic. 16.—Another modification in the tongue bar, which occurs near the
upper end of a bar,

STRUCTURE OF THE PHARYNGEAL BARS OF AMPHIOXUS. 117

Fic. 17.—Cells from the atrial epithelium of a primary bar, from a partially
macerated section (cochineal). «a. Vacuolated cells. 6. More granular,
narrower cells, the nucleus being near the free end. c. A portion of a cell,
which is probably similar to d, with a narrow peripheral portion, and dilated
nuclear region below.

Fic. 18.—Three cells of the atrial epithelium of a tongue bar, from a
partially macerated section.

Fig. 19.—Cells from the lateral epithelium of a gill bar, from a partially
macerated section. The nuclei of the cells are seen to be at different levels.
Each cell carries a single long cilium. The small cell to the left may be a portion
of a larger cell, or may be really a basal interstitial cell.

Fie. 20.—A small piece of the septal membrane, with a nucleus (z.) (from
a transverse section of a bar). The membrane appears to be double, and the
nucleus lies between the two sheets.

Fig. 21.—A portion of a longitudinal section of a bar, showing the septal
membrane and three elongated, compressed nuclei (#.). Only a small por-
tion of the epithelium of the bar is filled in, in order to show the relative
size of the nuclei of the septal membrane and those of epithelium (ep.).
(Hematoxylin. Under Zeiss’s apochromatic 4 mm., aperture ‘95, with com-
pensating ocular 8.)

Figs. 22—26.—Five consecutive sections to show the passage of a tongue
bar into the endostyle, and the communication of the rod cavity with the sub-
endostylar ccelom.

Fig. 22. Transverse section of the endostyle and a neighbouring tongue
bar. The details as to arrangement of nuclei are only approximately
accurate. It is merely desired to exhibit the general topographical
relations of the structures. ezd. Endostyle. end. sk. Subendostylar
skeleton. end. cel. Subendostylar celom. ca. ep. Nuclei of the
celomic epithelium. a¢. ep. Atrial epithelium. at. Subendostylar
artery. P. The lowermost part of a primary bar fused with the endo-

. style. «. The peculiar tissue of elongated vacuolated cells (not
reticular tissue). Z. 7. Tongue bars. vod. The contained hollow rod ;
on the left the bar has already reached and fused with the subendo-
stylar tissue, on the right it is still free.

Fig. 23. The next section. The rod of the tongue bar on the left appears
to have a narrow cleft in its wall, so that its cavity communicates with
the subendostylar celom; but this is not so distinct as in the case of
the right rodin Fig. 26. (The lithographer has exaggerated this cleft.)

Fig. 24. The right portion only of the next and following sections is
drawn. Here the rod of the tongue bar (7.) is distinctly smaller, and
approaches the corner of the subendostylar ceelom,

118 W. BLAXLAND BENHAM.

Fig. 25. The rod has passed further into the subendostylar tissue, and
lies above the corner of the ccelom.

Fig. 26. The cavity of the rod communicates with the subendostylar
cclom most distinctly.

Fics. 27—29.—Transverse sections of an abnormal primary bar at different
levels.

Fig. 27. Section about the middle of the length of the bar. The epi-
thelium and general structure of the bar is normal, but contains three
separate primary rods, 1, 2, 8, each partially surrounded by a layer of
ceelomic epithelium. vise. v. The visceral blood-vessel. som. v. The
somatic blood-vessel. s.m. The very short septal membrane. ce. Ceelom.
skel. v. Extra-celomic skeletal blood-vessel. :

Fig. 28. A section of the same bar lower down. ‘There are here only two
rods, partially fused; the rod to the right is formed by the fusion of
rods 1 and 3 (in Fig. 27).

Fig. 29. The same rod lower down, nearer the endostyle. The two rods
have separated again, but have shifted their relative positions.

Fic. 30.—Three sections of a normal primary rod at different levels. (a)
Close to the endostyle, where it consists of two independent pieces. (4) Slightly
higher up, where a third piece (sometimes isolated) plugs the gap between
the two pieces, with which it is more or less fused. (c) The general condition,
in which the fusion of the three pieces is complete, except for a narrow axial
cleft (? or less refractive substance). &. The rod, in the usual sense. cz.
“ Cutis,” or outer wall of ccelom. ca. Colom. shel. b. ». Skeletal blood-vessel.
som. v. Somatic vessel. (Cf. Figs. 13—16.)

Tic. 81a.—The connection of synapticulum with the primary and tongue
bars. The section is accurately transverse, and includes three bars and the
synapticulum. P!, P?. The two primary bars. 7. The intervening tongue
bar. sy. To the left, half a synapticulum cut along its whole length, and
continuous, as is seen, with one half of the tongue rod and the extra-ccelomic
“cutis” (ew.) of the primary rod. syz!, syx*. The two portions of the other half
of the synapticulum (the intervening portion is in a section not figured). . 4. '.
The subepidermic skeletal blood-vessels of the primary bar. 4. v?. The skeletal
vessel of the tongue bar, which is seen to divide into a right and a left branch,
which fall into J. o!. of the primary bars. ca. Celom of the primary bar.

Fre. 314.—The next section of the primary bar (P?) showing the passage
of syn?. into the cutis (cw.). Other letters as in Fig. 31a.

ON THE PERIVISCERAL OAVITY OF OIONA. 119

On the Perivisceral Cavity of Ciona.
By
A. H. L. Newstead, B.A.,

Late Scholar of Christ’s College, Cambridge.

With Plate 8.

THe genus Ciona differs from other simple Ascidians in
having the alimentary canal (except the rectum) and other
organs contained in a definite cavity, which is situated in the
region of the body posterior to the pharynx—the so-called
‘**body-cavity” of Ciona. From its position round the viscera
this cavity may be called the perivisceral cavity, which implies
nothing as to its nature, while the term body-cavity might be
taken to assume its coelomic nature as definitely known.

This investigation had for its object the determination of
the exact nature of this cavity, and was carried on partly at
the Zoological Station of Naples during the summer of 1891,
and since then at Cambridge.

A. Previous Researches.

The existence of this cavity was first made known by
Kupffer,' who described it as a coelom or body-cavity. It is
separated from the atrial cavity by a septum which is per-
forated by the cesophagus, rectum, and genital ducts. In this

1 I take the account of Kupffer’s work from the paper of van Beneden and
Julin (2, p. 430). I regret that I have been unable to refer to Kupffer’s
original paper.

120 A. H. L. NEWSTEAD.

septum Kupffer described an orifice leading from the atrial
cavity to the perivisceral cavity, by means of which water could
penetrate the latter. He also described two orifices at the
bottom of the pharynx, one on each side of the posterior
groove which extends between the end of the endostyle and
the cesophageal opening ; these orifices he described as opening
into the atrial cavity.

Roule (1), in his monograph on Ciona, gives a detailed
account of the relations of this cavity, which he calls “la
cavité générale du corps,” but does not mention any openings
into the atrial cavity or pharynx; on the contrary, he denies
the existence of any possible communication between it and
the atrial cavity: ‘vers chacune de ces ouvertures [i. e. the
Openings in the septum for the passage of the cesophagus,
rectum, &c.], la lame péritonéale, insérée sur tous les organes
qui la traversent, envoie entre eux de petits prolongements, de
telle sorte qu’il ne peut exister aucune communication, si
minime qu’elle soit, entre la cavité péribranchiale et la cavité
générale” (p. 107). He considers the perivisceral cavity as
the remains of the primary blastocele of the larva, which has
become reduced and, as it were, pushed back to the posterior
end of the body by the development of the atrial cavity.

Van Beneden and Julin (2) discuss the question as to the
nature of this cavity ; they consider that it cannot be part of
the original blastoccele, since, according to the description of
the vascular system given both by Kupffer and Roule, it does
not communicate at all with that system.

Two other possibilities remain :

(a) It may be derived from the atrial cavity ; or

(b) it may be derived from the pharynx.

The latter view is considered by van Beneden and Julin to
be the more probable one; they suppose the perivisceral cavity
to be homologous with the epicardium described by them in
Clavellina. In “ Vopinion la plus probable, elle constitue une
dilatation de l’épicarde, auquel cas elle devrait communiquer
non pas avec les cavités péribranchiales, mais bien avec le sac
branchial... . . Si espace périviscéral répond 4 la cavité épi-

ON THE PERIVISCERAL OAVITY OF CIONA. 121

cardique des autres Ascidiens, les communications avec le sac
branchial doivent exister & droite et 4 gauche de ce sillon, a
supposer toutefois que ces orifices persistent pendant toute la
durée de la vie chez ces Ascidiens” (p. 432). Van Beneden
and Julin state that they were able to confirm the existence of
the two orifices at the posterior end of the pharynx mentioned
by Kupffer, but not his statement that they opened into the
atrial cavity : “Le seul point que nous n’ayons pas pu con-
firmer, c’est que ces orifices déboucheraient dans les cavités
péribranchiales” (loc. cit., p. 432). They consider that these
two orifiees most probably open into the perivisceral cavity
and not into the atrial cavity, in which case water could enter
into the perivisceral cavity, as described by Kupffer.

B. Perivisceral Cavity of the Adult.

A detailed account of the anatomy and relations of the
perivisceral cavity of Ciona has been given by Roule (1, p. 105),
so that it will be unnecessary to give along description. The
perivisceral cavity is separated from the atrial cavity by a
septum or diaphragm attached all round to the body-wall ex-
ternally, and internally to the posterior end of the pharynx on
either side of the posterior groove which extends between the
end of the endostyle and the mouth of the cesophagus ; the
atrial or outer aspect of this septum is lined by the atrial
epithelium, and internally it is lined by the lining epithelium
of the perivisceral cavity. This internal layer extends on to
the cesophagus and rectum, which pierce the septum, and forms
numerous folds or mesenteries passing to the various organs
contained in the cavity; the chief of these mesenteries passes
to the pericardium, completely surrounding it, and attaching
it to the end of the pharynx underneath the posterior groove,
the body-wall, and the stomach. The heart, V-shaped, lies in
the pericardium, and the vessels going to and from it lie in
the mesentery around the pericardium.

Two points require elucidation :

(i) Do the openings described by Kupffer at the end of the
pharynx open into the atrial cavity, as he described, or do they

122 A. H. L. NEWSTEAD.

open into the perivisceral cavity, as supposed by van Beneden
and Julin ?

(ii) Is there any communication between the atrial cavity
and perivisceral cavity, as described by Kupffer, and denied by
Roule?

These points I hnve tried to work out by means of sections
of small individuals, about 1—2 cm. in length; such indi-
viduals have completely undergone metamorphosis and acquired
the adult form.

A transverse section of a small Ciona through the posterior
end of the pharynx between the end of the endostyle and the
cesophageal opening—i. e. through the posterior groove—is
drawn in fig. 1, and a portion of the same section, more
enlarged, in fig. 2.

The perivisceral cavity (pv. c.) is separated from the atrial
cavity (at.) by the septum (s.), in which is seen a blood-vessel
on one side. Inside the perivisceral cavity is seen the peri-
cardium (pe.), attached by its mesentery (m, fig. 2) on the one
hand to the base of the posterior groove (p. g.), and on the
other hand to the stomach (s¢.); at its attachment to the pos-
terior groove is a large vessel (v., fig. 2) running beneath the
groove. Inside the pericardium is the heart (h.), which is cut
twice owing to its peculiar V-shape; each part is seen to be
attached to the wall of the pericardium. On each side of the
posterior groove of the pharynx is an opening (0.), by which
the perivisceral cavity communicates with the pharynx. These
two openings are situated, one on each side, between the pos-
terior groove and the edge of the septum which elsewhere is
attached to the base of the groove; they may be traced only
through a few sections (12—15), and the left is distinctly
larger than the right.

On working through a series of sections no communica-
tions between the perivisceral cavity and the atrial cavity can
be seen.

The above section, shown in fig. 1, then, shows that definite
communications between the pharynx and perivisceral cavity
do exist ; and, since these openings occur in the same position

ON THE PERIVISOERAL CAVITY OF CIONA. oe

as the orifices described by Kupffer, it is extremely probable
that the supposition of van Beneden and Julin is correct, and
that the orifices observed by Kupffer open into the peri-
visceral cavity, and not, as he described, into the atrial cavity.

Since no openings can be found in a whole series of sections
between the atrial cavity and the perivisceral cavity, and since
Roule expressly denies their possible existence, it appears cer-
tain that such communications between the two cavities do not
exist.

c. Development in the Larva.

The early stages of development of Ciona are passed through
very quickly; the tailed larve are formed in the first twenty-
four hours of development, and fixation takes place on the
second day, the changes at this time being passed through
very rapidly, so that the development is very difficult to follow
through all its stages.

I have unfortunately not been able to follow out completely
the development of the perivisceral cavity in the early stages,
but the comparison of the stages observed with the develop-
ment of the epicardium of Clavellina, as observed by van
Beneden and Julin (1), shows the process of development to be
very similar in the two forms, and supports the hypothesis put
forward by these authors that the perivisceral cavity of Ciona
is homologous with the epicardiac tubes of Clavellina.

The process of development in Clavellina is shortly as fol-
lows:—Two tubes are first formed as outgrowths of the pharynx,
called by van Beneden and Julin the “ epicardiac tubes ;”
these later become fused with one another posteriorly, and the
posterior fused portion then becomes separated from the epi-
cardiac tubes to form the pericardium. The dorsal wall of the
pericardium invaginates to form the heart, so that the heart is
at first completely open along its whole length to the primi-
tive blastoceele of the larva; later the pericardium becomes
closed, while the heart remains attached to its dorsal wall by
a kind of mesentery, and open only at its ends to the general
blastoceelic space, which develops into the vascular system of

124 A. H. L. NEWSTBAD.

the adult. The epicardiac tubes become connected with the
process of budding,

Van Beneden and Julin suggest that the perivisceral cavity
of Ciona represents the epicardiac tubes of Clavellina, which
have become enlarged, growing completely round the viscera
and fused together.

The fact shown above that the perivisceral cavity communi-
cates with the posterior end of the pharynx by a pair of open-
ings affords a strong support to this view, which, however, can
only be completely accepted if it be confirmed by a study of
the embryological development of Ciona. .

The first stage of development which I have been able to
obtain is that of a larva which has recently become fixed. Figs.
3—5 are transverse sections of a larva at this stage. Fig. 3 is
the most anterior, and passes through the lower end of the
pharynx (ph.) below the opening of the cesophagus ; below the
pharynx are seen the stomach (s¢.) on one side, and the intes-
testine (i.) on the other, the whole being surrounded by the
general blastoceele space (4.), containing numerous scattered
cells.

A section a little further back (fig. 4) shows the endostyle
(end.) completely separated from the more dorsal portion of
the pharynx, which is now divided completely into two parts
(ep.) by a septum ; in a section still further back (fig. 5) we
find these two portions fused together, and much less in size.
From their mode of origin from the posterior end of the pharynx,
their fusion at their posterior ends, and their similar relations,
these two parts are evidently homologous with the epicardiac
tubes of Clavellina, so that the early stages of development of
Ciona are very similar to those observed in Clavellina.

The first formation of the pericardium from these epicardiac
tubes I have not been able to follow; the next stage of which
I have been able to get satisfactory preparations is one much
later, when the pericardium is fully formed and separated from
the epicardiac tubes which have grown round the viscera, and
become completely fused dorsally, though still separated ven-
trally by the pericardium. This stage is one during the

ON THE PERIVISCERAL CAVITY OF CIONA. 125

middle of metamorphosis, while the larva is still fixed by a
long stalk ; it shows the formation of the heart by the invagi-
nation of the dorsal wall of the pericardium just as it has been
shown to be formed in Clavellina.

Fig. 6 is a transverse section through a larva at this stage,
passing through the heart (4.), which is seen to be formed by
the invagination of the dorsal wall of the pericardium (pe.),
and to be still open to the general blastoccele space (d.), which
is now very much reduced, the chief portion of it being con-
fined to a small space round the alimentary canal, into which
the heart opens in the neighbourhood of the stomach (st.).
The whole of the viscera are now surrounded by a large peri-
visceral space (ep.), which may be distinguished from the
general blastoccele space by its containing no free cells. This
space evidently corresponds to the epicardiac tubes of the
preceding stage, now very much enlarged and fused together
completely, except at the region of the pericardium between
the pharynx and the stomach. A comparison of this stage
with the preceding, and with the development of Clavellina as
described by van Beneden and Julin, can leave no doubt that
the pericardium in Ciona is separated off from the epicardiac
tubes just as it is in Clavellina, and also that the perivisceral
cavity of Ciona corresponds to the epicardium of Ciona.

The later stages of development, as regards the perivisceral
cavity and associated organs, are very simple; the heart
becomes completely closed except at each end, where it still
opens to the original blastoceele space, which becomes reduced
to the blood-vessels ; the two halves of the perivisceral cavity
approach each other between the pericardium and stomach till,
on the closure of the heart and the reduction of the blood-
space round the stomach, they only become separated by their
thin walls, which unite to form the mesentery attaching the
pericardium to the stomach, a similar process also taking place
on the ventral side of the pericardium between it and the
pharynx.

The derivation of the heart from the dorsal wall of the
pericardium is still indicated in the adult, where the heart,

126 A. H. L. NEWSTEAD.

although it has become twisted, still remains attached to the
wall of the pericardium along one edge, so that a study of the
development of the perivisceral cavity shows it to be very
different from the original blastoccele, and corroborates the
hypothesis put forward by van Beneden and Julin that it is
homologous with the epicardium of Clavellina.

One small point may here be noticed, which, though of
little value by itself, yet when taken in conjunction with the
other evidence affords another point of similarity between the
perivisceral cavity of Ciona and the epicardiac tubes of Clavel-
lina. In Clavellina van Beneden and Julin state that the left
epicardiac tube is always larger than the right, and the same
fact may be noticed in the openings from the perivisceral cavity
into the pharynx in Ciona, the left opening in this case being
similarly larger than the right.

vp. Conclusions.

Two morphological conclusions may be drawn from the
above facts.

(1) The primary condition of the epicardium is undoubtedly
that found in Clavellina, where it has the function of a budding
organ. The condition in Ciona is that of an organ which has
become very much modified while losing its original function
as an organ of budding. Since prolongations of the epicar-
dium or perivisceral cavity extend into the stolons of Ciona,
this supports the view which is adopted by Herdman (3, 4,
p. 139), that the stolons of Ciona are modified budding organs
which have lost their original function; the opposite view,
that the stolons of Ciona may be regarded as nascent organs,
which have not yet acquired their function of budding, is nega-
tived by the above facts, since the perivisceral cavity of Ciona,
if it be a modified epicardium as shown above, cannot be
regarded as a primitive condition.

(2) This leads us to the second conclusion to be drawn from
the above facts. Roule (1) looks upon the perivisceral cavity
of Ciona as a primitive condition, corresponding to the general
blastoceele space which we find in the larve, as well as in

ON THE PERIVISCERAL CAVITY OF CIONA. 127

Appendicularia. He considers that the further development
of the atrial cavity has reduced and obliterated this cavity in
the other simple Ascidians. This view is not supported by
the above facts, which lead us to look upon the perivisceral
cavity of Ciona as a specially modified epicardium which has
become greatly enlarged. The perivisceral cavity is certainly
not homologous with the general blastocele space of Appen-
dicularia, and we have no reasons for believing that the other
simple Ascidians pass in development through a stage in which
the epicardium is modified and enlarged, as in Ciona, to be
afterwards reduced ; so that we cannot look upon the perivis-
ceral cavity of Ciona as a primitive condition: in this respect
Ciona is the most modified of the simple Ascidians.

In conclusion, it is my pleasant duty to express my sincere
thanks to Mr. Harmer for his kind help and assistance, to the
various members of the staff of the Zoological Station at
Naples for the great kindness I received at their hands, and
to the Master and Fellows of Christ’s College for the grant of
a scholarship during my residence at Naples.

REFERENCES.

1, L. Rovutz.—“ Recherches sur les Ascidies simples des cdtes de Provence,”
‘Annales du Musée d’Histoire Naturelle de Marseille,’ vol. ii, 1884-5.

2. E. van BENnEDEN and C. Juxin.—‘‘ Recherches sur la morphologie des
Tuniciers,” ‘Archives de Biologie,’ vol. vi, 1887.

3. W. A. Herpman.— On the Evolution of the Blood-vessels of the Test in
the Tunicata,” ‘ Nature,’ vol. xxxi, 1885, p. 247.

4, W. A. Herpman.— Report on the Tunicata,” part iii, ‘ Challenger
Reports,’ “ Zoology,” vol. xxvii.

128 A. H. L. NEWSTEAD.

EXPLANATION OF PLATE 8,

Illustrating Mr. A. H. L. Newstead’s paper “On the Peri-
visceral Cavity of Ciona.”

Explanation of Figures.

at. Atrial cavity. 4. Blastoccele. end. Endostyle. ep. Epicardiac tube.
g. Generative organ. 4. Heart. 7. Intestine. m. Mesentery attaching the
pericardium to the posterior end of the pharynx. o. Openings from the peri-
visceral cavity into the pharynx. pc. Pericardium. p.g. Posterior groove
of pharynx. pd. Pharynx. pv. e. Perivisceral cavity. s. Septum between
the perivisceral cavity and the atrial cavity. s¢. Stomach. v. Blood-vessels.
x. Artificial space (in Fig. 6).

Fic. 1.—Transverse section of a small adult individual of Ciona through
the posterior end of the pharynx (1 in. obj.).

Fic. 2.—Portion of the same section further magnified (3 in. obj.).

Figs. 3—5.—Three transverse sections of a recently fixed larva, Fig. 3
being the most anterior (4 in. obj.).

Fic. 6.—Transverse section of a larva at a more advanced stage of meta-
morphosis, showing the formation of the heart (4 in. obj).

DEVELOPMENT OF DISTICHOPORA VIOLACEA. 129

The Early Stages in the Development of Disti-
chopora violacea, with a Short Essay on
the Fragmentation of the Nucleus.

By

Sydney J. Hickson, M.A. Cantab. et Oxon., D.Sc.Lond.,
University Lecturer on the Morphology of Invertebrates ; Fellow of
Downing College, Cambridge.

With Plate 9.

Tue material upon which I have made my investigations was
in part collected by me in N. Celebes, and in part by Professor
A. C. Haddon in Torres Straits. Some of the specimens were
treated with strong alcohol alone, others with corrosive sub-
limate followed by alcohol. For decalcification I have entirely
used nitric acid.

I have tried a great many different stains and combinations
of stains. Borax carmine, Biondi’s fluid, methyl green, and
hematoxylin all give fairly good results; but I find that the
best treatment is to place the sections, when fastened to the
slide, in a strong solution of eosin in 90 per cent. spirit for an
hour, then to wash in 90 per cent. spirit and stain in weak
hematoxylin for twenty minutes. This treatment gives a
beautiful double stain which shows the nuclei and the chromatin
granules better than I have seen them in any preparations
treated with carmine.

My researches were entirely carried on in the morphological
laboratory at Cambridge.

VOL. 30, PART 1.—NEW SER. I

130 SYDNEY J. HICKSON.

I. The Early Stages in the Development of
Distichopora.

The ovum of Distichopora, like that of Allopora and other
Stylasterids, is provided with a large amount of yolk, and lies
in a cup-shaped trophodisc.

In young immature ova the germinal vesicle is situated in
the middle of the egg, is spherical in shape, is provided with a
well-defined membrana limitans, a germinal spot, and a fine net-
work of protoplasmic fibrils with thickened nodes (PI. 9, fig. 1).

When examined with a high power the germinal spot may
be seen to contain a few clear vacuoles (fig. 2).

In some ova with a full complement of yolk-spheres the
germinal vesicle is irregular in shape, and provided with
processes resembling the pseudopodia of ameba. The outlines
of these processes are usually difficult to observe, the mem-
brana limitans being apparently wanting, and the intra-nuclear
and extra-nuclear protoplasm perfectly continuous (fig. 8).

In these cases there may be seen a few large rod-shaped
granules (the chromosomes), which stain deeply with carmine
and other stains.

These amceboid germinal vesicles are without doubt passing
from the centre of the ovum towards the periphery. In those
that are near the periphery the chromosomes are more
numerous and very much smaller than they are in those nearer
the centre of the ovum. In one case I have observed these
bodies arranged in a row parallel to the surface of the ovum,
and dividing the nucleus into two unequal halves (fig. 5).

When and in what manner the polar bodies are formed I
cannot say, but it is probable that in some cases the nuclei of
the polar bodies are formed before the germinal vesicle reaches
the periphery, and are absorbed in the substance of the ovum.
The germinal vesicle finally reaches the periphery of the ovum,
and when it is in that position the fertilisation most probably
occurs.

It is clear that the germinal vesicle must remain at the

DEVELOPMENT OF DISTICHOPORA VIOLACEA. 131

periphery for a very considerable time, for of the numerous
unfertilised ova that I have examined a large majority have
their germinal vesicles in that position.

In the next stage the membrana limitans of the inner half
of the vesicle disappears, the network and the germinal spot
break down into numerous very minute scattered granules
(fig. 7).

Then the membrana limitans entirely disappears, and lastly,
the substance of the vesicle, or, as it should now be called, the
oosperm nucleus, becomes scattered through the substance
of the ovum.

Fig. 8 is a careful drawing of a stage in which the mem-
brana limitans has just disappeared, and I have three or four
complete series of sections through ova in which no trace of
nuclear structure can be found nor any area, such as that
shown in this figure, which represents the vanished nucleus.

As these two stages are of the greatest importance in the
consideration of what follows, it is necessary to say that not-
withstanding very careful search with high powers, no trace
of karyokinetic figures could be observed.

The ova of these stages are not sufficiently numerous, nor
are the methods of preservation sufficiently perfect to enable
one to assert that such figures do not occur. Corrosive sub-
limate followed by alcohol, although giving excellent general
histological results, does not always bring out the full details
of nuclear division; and it will be necessary to confirm these
purely negative results as regards karyokinesis by observa-
tions made upon specimens preserved in Flemming’s solution
and other reagents before any general statements regarding
fragmentation of the oosperm nucleus of Distichopora can be
accepted.

Nevertheless it is my belief that we have here an instance
of nuclear fragmentation, for reasons which I propose to dis-
cuss in the third section of this paper.

In the next stage that I have observed, a few small islands of
protoplasm may be seen in the yolk (fig. 9), and the exa-
mination of broken sections, in which part of the yolk has

132 SYDNEY J. HICKSON.

been washed away, shows that these islands are connected
together by a very coarse mesh-work of fine protoplasmic
strands.

In a later stage the islands are seen to be more numerous,
and the protoplasmic mesh-work somewhat finer. A complete
nucleus may be seen in some of these islands, but in others all
that can be made out are a few deeply staining granules
(figs. 10 and 12).

In a later stage the nuclei have increased in number in the
midst of the yolk, and a few make their appearance in the
protoplasmic sheath that surrounds the ovum.

In these last three stages I have described a process which
can only be compared with the so-called free nuclear forma
tion in early insect embryos. Nuclei make their appearance
in places which were previously apparently devoid of any
nucleus or nuclear structure. Moreover nuclei of various
sizes and shapes may be seen in the embryo at the same time.

It is not reasonable, however, to assume on the insufficient
evidence before us that ‘free nuclear formation” does actually
occur. It seems to me to be much more probable that minute
fragments of nuclear substance scattered through the proto-
plasmic mesh-work collect together in places, and form by
their fusion true recognisable nuclei. In other words, the pro-
cess we have under observation is rather one of ‘‘nuclear re-
generation’’ than one of “ free nuclear formation.”’

I have often noticed in ova of these stages an aggregation
of the yolk into spherical, polygonal, or irregular lumps, sug-
gesting that the egg has undergone some form of complete
segmentation (fig. 13). This is not a true process of seg-
mentation, however, since the distribution of the nuclei in the
spaces between the aggregations and not in their centres shows
that it affects the yolk only. It is remarkably similar in
appearance to the so-called yolk segmentation of Arthropods,
the appearance of the embryo at this stage being very much
like that of such a form as Peripatus nove-zealandiza, as
described by Miss Sheldon (58). This segmentation of the yolk
seems to be only temporary, for in embryos in which the

DEVELOPMENT OF DISTICHOPORA VIOLACEA. 133

ectoderm has commenced to be differentiated it cannot be
observed. In later stages of the development the ectoderm is
gradually formed. Nuclei appear in the peripheral sheath of
protoplasm, and the protoplasm accumulates in the form of
cellular blocks around each nucleus, as in Allopora.

I have carefully examined the endoderm in these stages in
the hope of finding out the manner in which the nuclei divide,
and although I have found a few dumb-bell-shaped forms, and
no satisfactory evidence of karyokinesis, I do not feel justified
in asserting that the nuclei always divide amitotically.

As far as the ectoderm is concerned, I can assert most
positively that indirect nuclear division does occur.

Numerous dumb-bell-shaped nuclei and nuclei connected
together in pairs may be seen in the developing ectoderm, and
in these faint achromatic lines may be seen connecting the
chromatin rodlets. The nuclei are too small to enable me to
make out all the details of the process, but there can be no
doubt that there is a true process of karyokinesis in the
divisions of these nuclei (fig. 18, a, b,c, d, and e). Ihave not
been able to decipher anything like the “spheres of at-
traction.”

One very remarkable and important point in the develop-
ment of all the Hydrocoralline, so far as they have at present
been investigated, is the fact that thereis no segmentation of
the ovum, either complete or partial, nor is there any for-
mation of cells with a definite outline until a very late stage.

At the time when (in Allopora and Distichopora) there are
ten or fifteen nuclei, the young embryo is a simple multinu-
cleated plasmodium, loaded with yolk. In the later stages the
nuclei have increased in numbers, and a certain number of
them are arranged in a row at the periphery of the embryo.

The yolk in the immediate neighbourhood of these peripheral
nuclei disappears, probably by absorption, and thus they are
situated in a clear peripheral sheath or envelope of protoplasm.
In a later stage this peripheral sheath of nuclei breaks up into
blocks, each block containing one nucleus, and thus the
ectoderm is formed.

134 SYDNEY J. HICKSON.

The ectoderm is, then, a differentiation of the periphery of a
multinucleated plasmodium.

What becomes of the inner part of the plasmodium ?

We have no answer to this question so far as the Hydro-
corallines are concerned ; but, judging from the other Celente-
rates, there can be little doubt, I think, that it becomes the
endoderm.

In the development of Aglaophenia (Tichomiroff, 55) we
find a stage that is almost precisely similar to the solid planula
of Distichopora and Allopora, and in a later stage this central
yolk-laden plasmodium breaks up into blocks, which become
the endoderm-cells of the adult.

The difficulty that we have now to face is, how can these
facts concerning the origin of the germ layers be brought into
line with those of other Ceelenterates ?

We find in the Stylasteride no segmentation, no process of
invagination to form the endoderm, and no process that can be
compared with ordinary primary delamination; but still it is
probable that this method of the formation of the germ layers,
if it is not itself the primitive one, has been derived from those
of other Celenterates, and I shall endeavour to show in the
next section how the transition has taken place.

II. On the Formation of the Germinal Layers in the
Celenterata.

During the last ten years our knowledge of the early stages
of the development of the Ceelenterata has very considerably
widened, but still we seem to be no nearer to the solution of
many interesting phylogenetic questions than we were before.
The various theories that have been put forward, based upon
the study of a few forms, have in no instance received the un-
qualified approval of the principal authorities on the group, and
we find ourselves in a maze of conflicting theories, none of
which seem to conform entirely to our knowledge of facts.

This unfortunate state of affairs is due to the fact that in the
group of the Celenterata we find many very different types of

DEVELOPMENT OF DISTICHOPORA VIOLACEA. 135

development, and no one of them seems to be particularly
predominant.

The development of a gastrula by invagination probably
occurs only in the group of Scyphomeduse.

The formation of a planula by delamination (i. e. the primary
delamination of Metschnikoff) occurs only in the group of the
Geryonide.

The formation of a sterrula by secondary delamination
occurs in most of the Anthozoa (McMurrich) and in many of
the Hydroids.

The formation of a sterrula by hypotropic invagination
occurs in many Sertularide and Campanularidz.

The formation of a planula or sterrula by polypolar immi-
gration of cells into a hollow blastula occurs in a few forms.

Lastly, the formation of a multinucleated plasmodium
without segmentation, which is followed by the differentiation
of epiblast-cells at the periphery of a solid plasmodium (the
endoderm), occurs in the Hydrocoralline and in some Al-
cyonarians.

Between these various types of development many inter-
mediate forms have been found, so that we have as it were a
complete series of developmental histories, with the typical in-
vaginate gastrula at one end and the multinucleated plasmo-
dium at the other.

We may represent this series by the following plan :

A. Gastrula formed by invagination. Large segmentation
cavity.
Examples: Cotylorhiza (Claus, 8), Pelagia nocti-
luca, and Nausithoé (Metschnikoff, 42).

a. Intermediate forms between type 4 and B are found
in Aurelia flavidula (Smith, 52), in which the
clump of cells that are invaginated is at first solid,
and in Cyanea capillata (McMurrich, 41), in
which this clump of cells remains solid longer
than in A. flavidula.

B. A solid planula (sterrula) formed by hypotropous im-
migration of cells into a large segmentation cavity.

136 SYDNEY J. HICKSON.

Examples: Clytia, Tiara, Rathkea, Obelia, Tima,
ARquorea (Metschnikoff, 42), and Cyanza arctica
(McMurrich, 41).

6. Intermediate forms, in which the migration takes
place mainly at the hind end, occur in Mitrocoma
(Metschnikoff, 42).

C. A sterrula is formed by polypolar immigration of cells
into a large segmentation cavity, these cells being
formed by the radial fission of the cells of the ccelo-
blastula.

Example: Aginopsis (Metschnikoff, 42).

c. Intermediate form, in which the cells that immigrate
are formed partly by radial and partly by tangential
division.

Example: Hydra (Brauer, 5).

D. A planula is formed by primary delamination, the
endoderm-cells formed by tangential division only.
The segmentation cavity is large.

Example: Geryonia (Metschnikoff, 42).

d. Numerous intermediate forms in which the segmen-
tation cavity is small.

Examples: Tubularia (Brauer, 5a), Bougainvillea
(Gerd, 14).

E. A sterrula is formed by precocious delamination
(secondary delamination of Metschnikoff). No seg-
mentation cavity formed.

Examples: Aglaura (Metschnikoff, 42), Rhopalo-
nema (Metschnikoff), Eudendrium and Sertularella
(Tichomiroff, 55).

e. Intermediate forms in which the segmentation is at
first incomplete.

Examples: Renilla (Wilson, 63), Gorgonia (von
Koch, 36), and probably other Alcyonarians.

F. A multinucleated plasmodium is formed. There is no
segmentation and no segmentation cavity.

Examples: Algaphenia (Tichomiroff, 55), Mille-

DEVELOPMENT OF DISTICHOPORA VIOLACEA. 137

pora (Hickson, 17), and the Stylasteride (Hickson,
18 and 19).

It is not my purpose to discuss fully the various views that
have been put forward concerning the origin of the Metazoa
from the Protozoa. The gastrula theory, the planula theory,
the plakula theory, and the phagocytella theory have each
received in their turn the consideration of naturalists, and
nothing would be gained were an attempt made in these pages
to reopen the discussions that they gave rise to.

But I cannot pass on without expressing my opinion that
the developmental history of the Hydrocoralline lends some
support to the so-called “ plasmodium ” theory. Many years
ago, Jehring (27) and Saville Kent (32) put forward the view
that the Metazoa are derived from a multinucleated Protozoan
like Opalina. Sedgwick (57) has supported this view, as a
result of his important work on the development of Peripatus,
and considers that the ancestral Metazoan was probably of
“the nature of a multinucleated Infusorian, with a mouth
leading into a central vacuolated mass of protoplasm.”

In discussing Saville Kent’s views Metschnikoff (42) says
that there is no evidence of the formation of such a multi-
nucleated cell in the lowest Metazoa.

Now I have already pointed out that in the earliest stages
of the Stylasteride and of Millepora the embryo is nothing
more nor less than a multinucleated cell; that is to say, it is a
single undivided mass of protoplasm, containing numerous
nuclei. It might be urged that it is a syncytium, a number
of cells fused together; but there is no more evidence for
such a view than for the view that it isa single multinucleated
cell.

Similarly it may be urged that Tubularia (Brauer, 5a),
Aglaophenia (Tichomiroff, 55), Aleyonium (Kowalewsky, 37),
Gorgonia (von Koch, 86), and Renilla (Wilson, 63) all pass
through a stage in their development in which the embryo is
simply a multinucleated cell.

The fact that such a condition as this occurs in many
different groups of the animal kingdom widely separated from

138 SYDNEY J. HICKSON.

one other also lends support to the view that it may have
some important phylogenetic significance.

Instances of the occurrence of an unsegmented multi-
nucleated plasmodium are found not only in the Coelenterata
above mentioned, but in Peripatus, Myriapods, Spiders (Kishi-
nouye, 34, and Morin, 44), Insects, Crustacea, Elasmobranchs,
and probably many other forms with large eggs.

It might be urged as an argument against the plasmodium
theory that the multinucleated plasmodium occurs principally
in the development of those forms whose ova contain a large
amount of food-yolk, that the segmentation is modified by the
presence of this yolk, and that consequently the phylogeny is
obscured.

But it does seem to me that in the ovum that is perfectly
clear and homogeneous we have a cell that is any nearer to the
ancestral Protozoan than the ovum that contains a moderate
amount of yolk.

It is almost certain that the ancestral Protozoan normally
contained some food-vacuoles, and it is quite as probable as
not that it had some coutractile or simple water-vacuoles for
floatation purposes as well.

It is quite as reasonable to suppose that the Metazoa are
derived from an Actinospherium-like ancestor with vacuoles in
the outer regions as well as in the inner mass, as it is to derive
the Metazoa from a “ multinucleated Infusorian with a mouth
leading into a central vacuolated mass of protoplasm.”

If this is the case, then we can no longer consider the yolk-
bearing eggs to be secondarily modified, and the small
transparent eggs to be the primitive types from which all the
others are derived; but we may expect to find in the develop-
ment of eggs with a moderate amount of yolk just as much or
even more evidence of ancestral history as in eggs that are
practically yolkless.

It must not be forgotten, moreover, that the occurrence of a
multinucleated plasmodium is not confined to those cases in
which the ovum contains a large amount of yolk.

In the ovum of Millepora there is no yolk, and yet the

DEVELOPMENT OF DISTICHOPORA VIOLACEA. 139

oosperm nucleus fragments without any segmentation occur-
ring, giving rise to a simple multinucleated plasmodium.

The eggs of Aphis (Will, 62) and some other insects contain
very little yolk, and do not segment until a large number of
nuclei are formed.

The segmentation of the ovum, then, and the subsequent
formation of a morula mass of cells, are phenomena not
entirely dependent upon the absence of yolk. Many, compara-
tively speaking, large eggs, such as that of Rana, segment,
whilst others, such as that of Aleyonium, do not.

We cannot, consequently, assert that when an ovum segments
it is simply repeating an ancestral phase, and that when it does
not segment it is prevented from doing so by the physical
obstruction of the yolk.

The reverse of this is more probably true. The recent
brilliant researches of Driesch (9) prove that the segmentation
of the ovum is due to physical or mechanical laws, and we
cannot or should not derive any phylogenetic conclusion from
the phenomena of segmentation.

We may even go further than this, and say that the develop-
ing ovum would not segment, but would naturally pass
through the stage of a multinucleated plasmodium, were it not
for the action of certain purely mechanical forces, with which
we are not at present fully acquainted. When these forces
cannot act upon the egg, or are in some way counteracted, the
ovum does not segment, whether it is laden with yolk
(Stylasteride, many Insects, Elasmobranchs, &c.) or not
(Millepora).

III.—On the Fragmentation of the Oosperm Nucleus.

It is the belief of many eminent histologists that any process
of division of the nucleus other than that by karyokinesis or
mitotis is a sign of the degeneration of the nucleus, and the
approaching end of the life of the cells.

Flemming says, ‘‘ Fragmentation of the nucleus, with and
without subsequent division of the cell, is universally a

140 SYDNEY J. HICKSON.

process in the tissues of Vertebrates which does not lead to the
physiological multiplication and reproduction of cells, but, on
the contrary, represents where it occurs a degeneration or
aberration, or perhaps, in many cases, is subservient to the
metabolism of the cell by increasing the periphery of the
nucleus.”

Ziegler (65), who quotes the above passage from Flemming’s
work, discusses in detail some of the many instances of
amitotic nuclear division, and comes to similar conclusions.
He says that amitotic division of the nucleus always indicates
the end of the series of divisions, and considers it hardly
probable that nuclei which have arisen by amitotic division
will ever again divide by mitosis.

If Flemming, Ziegler, and those who agree with them are
right, then it is clear that the oosperm nucleus does not and
cannot fragment. It must divide regularly by karyokinesis.
But Ziegler’s views are, it seems to me, altogether unten-
able. By simply denying, or passing over in silence, many
instances of fragmentation of the nucleus, which do not
support his views, he has given undue weight to mitosis, and
leaves an unsatisfactory gap in the list of cases which support
his theory.

Verson (56), Frenzel (12), and Lowit (40) have, since the
publication of Ziegler’s paper, called attention to cases of
amitotic division of the nucleus which are most certainly not
followed either by nuclear degeneration or by a cessation of
cell multiplication.

A review of the recent literature of cell division shows that
the cases given by these authors may be supplemented by many
others, and, indeed, leads one to a conclusion quite different
from that of Ziegler and Flemming, namely, that indirect nuclear
division rarely occurs unless it is preceded by or accompanied
by some partial or complete segmentation or division of the
surrounding cell substance.

It is undoubtedly true that in many cases amitotic frag-
mentation of the nucleus is followed by its degeneration and
the death of the cell. The numerous examples quoted by

DEVELOPMENT OF DISTICHOPORA VIOLACEA, 141

Ziegler prove that this is the case. But I shall endeavour to
show that we are by no means justified in assuming that
amitotic fragmentation is a sign of degeneration.

In the first place, it can be shown that there is considerable
evidence for believing that the oosperm nucleus of some ova
does not divide by normal karyokinesis, but does split up
amitotically into a large number of minute fragments.

-I have already described (17 and 18) such a process of
fragmentation in the case of Millepora, Allopora, and Disticho-
pora, but the following considerations prove that the same is
probably true of many other eggs.

Celenterata.—In the development of Alcyonium the
germinal vesicle entirely disappears, and no traces of the
karyokinetic division of the oosperm nucleus can be found.
Kowalewsky! (87) gives a figure of the ovum without any
nucleus, but my own observations show that at a stage cor-
responding to the one he figures the nucleus is in the form of
a number of minute fragments scattered through the substance
of the ovum.

The failure to find karyokinetic division of the oosperm
nucleus cannot be attributed to imperfect methods of pre-
servation or staining, because young embryos, preserved and
stained in precisely the same way as the fertilised ova, exhibit
beautiful and typical karyokinetic figures.

The early stages in the development of Gorgonia cevolini,
described by G. von Koch (86), seem to be precisely similar to
those of Aleyonium. In the unfertilised ovum there is a large
germinal vesicle containing an exceutrically placed germinal
spot, but in the eggs that he believed to be fertilised there was
no nucleus. “Thre Structur weicht von der des unbefruchteten
Eies wesentlich ab. Es fehlt nimlich vor allem der Kern, von
dem ich keine Spur mehr auffinden konnte.”? The fact that
von Koch, after carefully examining over a hundred series of
sections through fertilised ova, could find neither traces of
segmentation nor the division of the oosperm nucleus, suggests

" As Kowalewsky’s paper is written in the Russian language I am unable
to read it.

142 SYDNEY J. HICKSON.

very forcibly that the ovum of Gorgonia does not segment at
first, and that the oosperm nucleus fragments as it does in
Alcyonium.

Arthropoda.—In the development of Peripatus capen-
sis, Sedgwick (51) has described the division of the ovum
into two blastomeres, and the large and easily seen karyo-
kinetic figures which mark the first division of the oosperm
nucleus. The fertilised ovum of Peripatus nove-zealandiz,
however, does not segment, and Miss Sheldon (53) was unable
to find any karyokinetic figures in the divisions of its
nucleus.

It is a very striking fact in support of my views that in
two species of the same genus we should find such a well-
marked difference in this respect, the ovum that does seg-
ment showing clear and unmistakable nuclear mitosis, and
the ovum that does not segment showing no signs of karyo-
kinesis.

But this is not the only example of the relation between the
segmentation and the division of the nucleus.

In a recent paper on the “‘ Embryology of the Macroura”
Herrick (6) states that it is a rule with the decapod
Crustacea that the nuclei of the segmenting eggs divide with
karyokinesis. There is an exception to this rule, however,
in the case of Alpheus minus. “ The fertile egg of A.
minus is pervaded with a remarkably fine reticulum which
encloses spherules of minute and uniform size. The nucleus is
central or nearly so, and consists of an ill-defined mass of
protoplasm, in which a fine chromatin network is suspended.
In the next phase the nucleus is elongated and about to
divide. Division appears to be direct andirregular. Ata
somewhat later stage the phenomena of the most interest
occur. Each product of the first nucleus has developed a
swarm of nuclear bodies which seem to arise by fragmentation.
These bodies take the form of spherical nuclei in clear masses
of protoplasm. ... In the last stage obtained the whole
egg is filled with several hundred very large elements, which
are descended more or less directly from some of the nuclear

-

DEVELOPMENT OF DISTICHOPORA VIOLACEKA. 1438

bodies just considered, but the intermediate stages have not
been considered.”

In the species Alpheus Saulcyi and Alpheus hetero-
chelis (two varieties) the segmentation is normal and regular,
of the centrolecithal type, and the division of the nuclei
indirect. In Alpheus minus alone is the segmentation
extremely irregular and the nuclear division direct.

Among Mpyriapoda we find that the ovum of Julus
terrestris is very similar in many respects to that of the
Stylasteride. There are no signs of segmentation, and there
is no formation of cells until the time when the epiblast is
formed. Heathcote (20), who carefully studied the early
stages in the development of this species, could not find any
signs of karyokinesis in the first divisions of the oosperm
nucleus.

There is, according to Kingsley (33), a disappearance of
the germinal vesicle of the American Limulus, and it is a
suggestive fact that Kishinouye (35), in his careful paper on
the development of Limulus longispina, does not refer to
the first nuclear divisions.

It is possible that there may be a fragmentation of
the oosperm nucleus in the ova of some other Arachnida.

In the development of many Insecta there are many
facts that point to the conclusion that the oosperm nucleus
fragments.

It is noteworthy in the first place that, notwithstanding the
fact that several excellent embryologists have carefully studied
the development of the common blow-fly, not one of them has
been able to give a satisfactory account of the first division of
the oosperm nucleus.

Blochman (8), who figures the spindles of the nuclear
divisions in the formation of the polar bodies, and also the
spindles of the nuclear divisions of the later stages of embryonic
development, did not apparently observe the first division of
the oosperm nucleus. He says, “Als erste Theilung des
Eikernes kann man die Bilder wohl nicht auffassen, weil, wie
ein Blick auf die spateren Figuren zeigt, bei Theilungen die

144. SYDNEY J. HICKSON.

Tochter kernplatten stets so fort weit aus einander riicken.”
Henking (21), too, was unable to find the first division of the
oosperm nucleus of Musca.

Now, in Musca, and in many other eae in which the early
divisions of the oosperm nucleus have not been made out, the
occurrence ‘‘ of free nuclear formation” has been described in
the young embryo. Whence come these free nuclei? It can
hardly be believed that they are actually formed in the cell
substance from something that is not directly derived from a
pre-existing nucleus. All the evidence of modern histology
tends to prove that nuclei are derived from nuclei, and nuclei
only, and it is only reasonable to suppose that the so-called
“free nuclei” of insect embryos are formed by the growth or
fusion of fragments of the oosperm nucleus.

The evidence in support of this hypothesis is not the purely
negative evidence of the absence of any direct proof of mitotic
division of the first nuclei, but the fusion of minute chromatin
bodies to form larger ones has actually been observed by
Henking (23) in the embryos of Pieris, Pyrrochoris, and Lasius.

But it is extremely probable that fragmentation of the
oosperm nucleus is of very frequent occurrence in the eggs of
insects. In many cases, both in large yolk-laden eggs and in
small yolk-free eggs, the fertilisation is followed by the
appearance of numerous nuclei in the substance of the egg.

In Neophalax concinnus, one of the Phryganids, the
division of the oosperm nucleus was not observed by Patten
(46), and the following is his account of the early stages :—
“Within ten or twelve hours after oviposition—the time
varying with the temperature—a clear space makes its ap-
pearance at the surface of the egg, and gradually increases
until it has attained the breadth of the future blastoderm.
In this layer, which has been called the ‘blastema,’ the pro-
toplasm has, under ordinary conditions, a very homogeneous
appearance, with occasionally lighter, less refractive spots,
which appear like vacuoles, but in which, when observed more
closely and under slight pressure of a cover-glass, or especially
when treated with a very little acetic acid, faintly marked

DEVELOPMENT OF DISTICHOPORA VIOLACHA. 145

nuclei make their appearance in greater or less numbers ac-
cording to the more or less advanced stage of the blastema.”
It is extremely improbable, if the minute nuclei in the
blastema could be observed by the simple method of treat-
ment with acetic acid, that the karyokinetic divisions of the
large oosperm nucleus, if they really occur, would have been
overlooked.

Many other instances could be given from the writings of
naturalists during the last twenty years of the failure to trace
the divisions of the oosperm nucleus in insect eggs, and of the
occurrence of “free nuclear formation” in the eggs after
fertilisation ; but in many of these instances it might be urged
that sufficient patience was not exercised, or that the methods
of preservation and staining were imperfect.

An important paper has, however, been recently published
by Henking (23) containing an extremely elaborate account
of his investigations upon many different species of insects
carried on with the aid of the best modern methods of re-
search. It would take me far beyond the limits of this paper
to give even an outline sketch of Henking’s important results,
but a brief reference to some of the points bearing upon the
subject of this essay must be made.

In Pyrrochoris, one of the Hemiptera, Henking finds that
in the formation of the polar bodies the nucleus divides by a
process of karyokinesis, the chromatin bodies being of con-
siderable size and definite in number.

After fertilisation a new spindle is formed with the chromo-
somes arranged in an equatorial plate, but before the division
is completed the chromosomes disappear. Later on the
chromosomes reappear in the form of extremely minute and
numerous granules, which fuse together into threads, and
arrange themselves in the equatorial plate of a new spindle.

Similarly, in Agelastica alni, a Coleopteran, the chroma-
tin entirely disappears after the division of the segmentation
nucleus.

In the Hymenopteran Lasius the chromatin of the first
two segmentation nuclei completely disappears, and when the

VOL. 30, pARLT 1.—NEW SER. K

146 SYDNEY J. HICKSON.

nuclei are about to divide again reappears in the form of
extremely minute granules, which fuse together to form the
chromosomes of the next division.

A similar disappearance has been described in the unfer-
tilised egg of Rhodites, and in this form there is no mem-
brane surrounding the nuclei.

These researches prove, then, that in some insects there is
a “disappearance ’”’ of the chromatin substance of the nucleus
after its first division.

To what is this disappearance due? MHenking thinks that
it is due to some chemical change in the chromatin substance,
as in some cases the outline of the chromosomes may be ob-
served after the disappearance of the colouring matter. Never-
theless it is a fact that commonly the chromosomes lose their
compact form during the colourless stage, and become very
finely divided. We can attribute the disappearance, then,
partly to the change in the chemical character of the chro-
matin, and partly to the very minute and scattered condition
of its elements.

Further, in some cases (Rhodites) not only does the chro-
matin disappear, but also the membrane surrounding the
nuclear area, so that we have (as in Distichopora, &c.) a
condition in which the nucleus is practically indistinguishable
from the surrounding protoplasm.

It is during this condition that some of the nuclear fragments
may be distributed through the substance of the ovum, and
give use to the nuclei of the so-called “free nuclear formation”
by subsequent fusion.

It must be obvious to anyone who carefully studies
Henking’s figures that in many insects the spindle of the first
division of the oosperm nucleus is very irregular, that the
chromosomes are not always arranged with the same mathe-
matical precision that they are in typical karyokinetic figures,
and further, that in consequence of the disappearance and
extremely fine division of the chromatin substances there are
still some steps in the nuclear divisions at the commencement
of development which have not been satisfactorily traced.

DEVELOPMENT OF DISTICHOPORA VIOLACEA. 147

We may go further than this, though, and say that some of
Henking’s figures, such as figs. 335, 336, and 337 of Lasius,
can only be interpreted on the supposition that the nucleus has
fragmented. The little clusters of chromatin granules, of very
irregular size and indefinite arrangement, that are here
figured scattered through the substance of the ovum, cannot be
considered to be the product of regular mitosis.

It seems to be extremely probable that in the group of
insects we have a series of stages intermediate in condition
between regular mitotic division of the oosperm nucleus or its
immediate successors and irregular fragmentation.

In Aphis (Will, 62) we may have regular karyokinesis at all
stages of the segmentation, the chromosomes being divided
into two equal halves at each division of the nucleus; but in
Musca, in Lasius, and perhaps in several others in which the
earliest stages are passed through with great rapidity, the
nuclei fragment with greater or less irregularity.

That the occurrence of karyokinesis is in some way dependent
upon forces manifesting themselves in the cell substance of
the ovum and acting upon the nuclei is rendered probable (1) by
the fact that in Aphis, where the nuclei divide by karyokinesis
in all stages, there is, as Will points out, a distinct aggregation
of protoplasm round the nuclei, and (2) by the fact that in
nearly all insects the karyokinetic figures of the nuclear
divisions that take place in the formation of the polar bodies
are much more regular and constant than they are in the early
stages of development.

But I shall discuss this point and general significance of
mitosis in greater detail later on.

That a similar process of fragmentation of the oosperm
nucleus may also occur in some Vertebrata seems to be probable
from the recent researches of Kastschenko (31) upon Elasmo-
branchs. It must be remembered that the early stages of the
development of Elasmobranchs and birds have been carefully
studied by numerous observers for the last twenty years, and
although the karyokinetic spindles in the developing blasto-
derm and its surrounding yolk have been described by nearly

148 SYDNEY J. HICKSON.

all of them, we have not received any account of the first
division of the oosperm nucleus.

It is quite unreasonable to suppose that all these observers
would have overlooked a nuclear division—which we might
expect, if it exists at all, to be the largest and most conspicuous
of the whole series. Nor can we suppose that the methods of
preservation or staining was so consistently bad at the first
stage as to prevent the observation of the figure, and so fre-
quently good in the later stages as to show the whole process of
karyokinesis clearly and distinctly.

Now Kastschenko (31) shows that in Elasmobranchs a
number of nuclei appear in the blastoderm and the surround-
ing yolk before the formation of the segmentation furrows,
which appear not in regular sequence, but simultaneously and
irregularly. ‘Die bekannte regelmiassige Reihenfolge des
Erscheinens des Segmentationsfurchen existiert bei Selachiern
fast gar nicht. Nur in seltenen Fillen bemerkt man das
urspriingliche Erscheinen einer Segmentationsfurche, welcher
dann gleichzeitig mehrere andere unregelmissig sich kreuzende
folgen. In den meisten Fallen aber erscheinen schon vom
Anfang an mehrere Segmentationsfurchen gleichzeitig und
somit zerfiallt die Keimscheibe direct in mehrere verschieden
grosse Segmentationskugeln, welche sich dann weiter aber nicht
gleichzeitig theilen.”’

We have, then, at the commencement of the development of
the Elasmobranch a multinucleated plasmodium, and Kast-
schenko is of opinion that all the nuclei of this plasmodium are
formed by repeated divisions of the first segmentation nucleus-
But, like all his predecessors, Kastschenko was apparently
unable to observe these repeated divisions of the first nucleus,
and it seems extremely probable that in Elasmobranchs, as in
insects, Hydrocorallines, and others, we have at this stage a
true process of nuclear fragmentation.

I have already called attention to the fact that in all of these
cases in which the fragmentation of the oosperm nucleus
probably occurs the ovum does not segment immediately after
fertilisation ; that there is, in fact, for a time in the early em-

DEVELOPMENT OF DISTICHOPORA VIOLACEA,. 149

bryonic development a multinucleated plasmodium without
any definite cell walls or cell areas.

There can be little doubt, I think, that in all holoblastic eggs,
such as those of Echinoderms, worms, Amphioxus, &c., the first
segmentation is accompanied by typical karyokinetic division
of the nucleus.

We may go further than this, and say that in many mero-
blastic eggs the first divison of the oosperm nucleus is also an
indirect one. Vialleton (57) and Watase (59) have observed
this division in the egg of Cephalopods, and Oppel (45) has
observed it in the egg of the lizard, Anguis fragilis. But
in both these cases the segmentation furrows occur regularly
and in sequence from the commencement of development, and
we have, consequently, evidence that the same forces are at
work in the protoplasm as those which produce the more or
less complete blastomeres of holoblastic eggs. Even in those
eggs of insects in which the nuclei are known to divide by
karyokinesis there is evidence of the drawing together of the
protoplasm along certain lines of force in the “ plasmatische
Strahlungen” of Henking, which surround the nuclei.

But if there is any truth in the view that I have here put
forward, that karyokinesis is primarily due to the forces which
bring about cell division, and that in those cases in which cells
or cell areas are not formed the nucleus may fragment or
divide directly in some other way, then we should expect to
find some further evidence of fragmentation of the nucleus in
other tissues. There is ample evidence of this in other tissues.

In the formation of the spores in Protozoa the nucleus of
the parent cell often divides long before there is any division
of the cell protoplasm, and in nearly all such cases division of
the nucleus is direct. In some cases the nucleus disappears,
and it is probable, as in the case of the oosperm nucleus
quoted above, that this may be due to the extremely fine
division of the chromosomes and fragmentation.

I will give just a few examples to illustrate these points.

Wolters (64), in describing the conjugation of Monocystis
magna and agilis, says, “ Kurz nach erfolgter Encystirung

150 SYDNEY J. HICKSON.

soll der Kern, respective die Kerne der beiden Copulanten
sehr undeutlich werden. Sie entziehen sich zuletzt dem beo-
bachtenden Auge ganz undsind im Inhalte der ausgequetschten
Cyste nicht mehr zu finden.’ The author figures, it is true, an
achromatic spindle in the encysted forms after the extrusion of
the polar bodies, but the chromatin bodies are very minuteand
irregularly scattered through the substance of the protoplasm.
“Rs gelangzwar nicht,” he says, “eine zusammenhingende
Reihe von Bildern fiir die Constatirung der mitotischen Theil-
ung an den Sporogonien zusammen zustellen, doch liess sich
mit Sicherheit constatiren, dass die Kernmembran an manchen
Kernen der ungetheilten Sporogonie geschwunden war und die
farbbare Substanz in zwei, durch einen grosseren Zwischen-
raum getrennte Reihen angeordnet war.”

But the evidence in favour of a process of fragmentation of
the nucleus seems to be much more conclusive in the case of
Clepsidrina blattarum (Wolters, |. c.). In this form
nuclei are found “in denen unziahlige kleine chromatische
Korner lagen, wie es schien regellos, ohne besondere Anord-
nung vertheilt. Allen bisheran geschilderten Kernformen
war dagegen eine scharf contourirte Kermembran gemeinsam.
Im Gegensatz da zu stehen Formen, die ebenfalls haufig beo-
bachtet wurden, welche einer solchen Membran entbehrten.
Der Kern breitet sich sternformig mit seinen Fortsatzen in
die Leibessubstanz des Thieres aus und steht mit dem proto-
plasmatischen Gefiige derselben in directen unterbrochenen
Zusammenhange.”

This account of the fragmentation of the nucleus of Clep-
sidrina blattarum is confirmed in all its essential details by
the more recent work of Marshall (40a), who was unable to find
at any time any traces of karyokinesis. A very similar account
is given by Schneider (49) of the division of the nucleus of
Klossia.

It may be that in some forms, such as the Gregarina
irregularis of Holothuria nigra (Minchin, 48), a regular
form of division with mitosis does occur, but this does not
detract from the importance of the fact that in many

DEVELOPMENT OF DISTICHOPORA VIOLACEA. 151

Gregarines which form during encystment a vast number of
spores, no karyokinetic figures can be observed.

Many years ago Hertwig (24) described a curious method of
the fragmentation of the nucleus without karyokinesis in the
spore formation of Thalassicola, and more recently Brandt (4)
was unable to find karyokinesis in the divisions of the nucleus
to form the nuclei of the spores of the Sphzrozooids.

Gruber (16) has described several instances among the
ciliate Infusoria in which the nucleus apparently fragments
into extremely minute granules, which become scattered
through the protoplasm of the body and collect again into
lumps.

Jickeli (29) has described fragmentation of the nucleus of
Stylonychia, Parameecium, and other Ciliata.

Quite recently, too, Lister (39), in his researches upon
Orbitolites, has not been able to discover any signs of karyo-
kinesis in the division of the nuclei.

There is probably, too, a method of fragmentation in the
spermatogenesis of many animals. I have myself carefully
examined the earliest divisions of the nucleus of the sperm
mother-cell of Millepora, Allopora, Distichopora, and Alcy-
onium, and I can find no trace of karyokinesis. It is, in fact,
only in a few exceptional cases, such as Ascaris (Hertwig, 25),
where the cell outlines of the spermatocytes are very early
delineated, that karyokinesis has been observed in the division
of the nuclei of the sperm mother-cells.

Verson (56) shows that in Bombyx mori the primordial cells
have at first a giant nucleus, which divides amitotically to form
numerous secondary nuclei, and these divide mitotically to
form the nuclei of the Spermatids,

Bolles Lee (38) found amitotic division of the nuclei of the
spermatogonia of Chetognatha and Nemertines and regular
karyokinesis in the division of the nuclei of the spermatocytes.

Dostojewski found the same thing in the spermatogenesis
of Amphibia (see Waldeyer, 58, p. 39), and other examples
could be quoted from the writings of La Valette St. George,
Gilson, Sabatier, and others (see Waldeyer, 58, p. 39).

152 SYDNEY J. HICKSON.

In some Annelid worms the nucleus of the spermatogonium
disappears, and there is no evidence at present that the nuclei
of the spermatocytes are derived by repeated mitotic divisions
of this nucleus (Jensen, 28, and others). In the recent work
on spermatogenesis, by Pictet (47) no mention is made of the
manner in which the nuclei of the spermatogonia divide in
Polychetes.! |

A study of the literature of spermatogenesis shows that
when there is a distinct division of the protoplasm to form the
spermatocytes or spermatogonia, distinct karyokinesis of the
nuclei may generally be seen; but when, on the contrary,
~ multinucleated cells are formed, which eventually give rise to
the spermatocytes, the nucleus of the spermatogonia either
disappears or divides amitotically.

It is not necessary for me to discuss in detail the numerous
cases of indirect fragmentation of the nucleus that have been
described by Arnold in his numerous papers in ‘ Virchow’s
Archiv’ and the ‘Archiv fiir mikroscopische Anatomie,’ by
Werner (61), Schottlaender (50), Hess (26), Geelmuyden (18),
Beltzow (2), Strobe (54), Goppert (15), and others. Many of
these cases are those of the nuclear division of giant-cells, and I
believe I am quite correct in saying that in all of them the
fragmentation of the nucleus is not immediately followed by
cell division.

The general conclusions to be drawn from the evidence
before us are—1. That fragmentation of the nucleus is a
normal method of nuclear division, and is not always a sign
of pathological change. 2. That in many of the instances in
which the nucleus is supposed to disappear there is, as a
matter of fact, minute fragmentation. 3. That fragmentation
only occurs where there is no cell division; and 4. That
karyokinetic division of the nuclei is caused by the forces in the
cell protoplasm which bring about the division of the cytoplasm.

That there may be many forms of nuclear division inter-

1 It is a noteworthy point that O. von Rath (48), who believes that the
nuclei of the spermatogonia and spermatocytes always divide mitotically, does
not refer at all in his paper to the spermatogenesis of Polychetes.

DEVELOPMENT OF DISTICHOPORA VIOLACHA,. 153

mediate in character between fragmentation and bipolar
karyokinesis seems to be probable from the discovery of
pluripolar mitosis in the inflamed cornea by Schottlaender
(50), and other atypical nuclear divisions in the spleen of the
white mouse by Arnold (1), &c.

We have, then, a series of phenomena in the division of
nuclei, with typical karyokinesis at one end and direct frag-
mentation at the other. The occurrence of any one kind or
the other is, in my opinion, determined by the forces which act
simultaneously upon nucleus and cell plasm. If these forces
are of such a kind as to drag the cell plasm into two equal
halves, the nucleus is also dragged into two equal halves with
mitosis; if, on the other hand, the forces are irregular and act
from many centres at the same time, the nucleus fragments
irregularly.

These views seem to me to be supported by the statement of
Flemming (11) quoted by Sedgwick, that “ the first change
observable in a cell whose nucleus is about to divide is in the
extra-nuclear protoplasm,” and by Biirger’s (7) recent con-
clusions concerning the meaning of the spheres of attraction.

List oF REFERENCES.

e

ArnoLtp, J.—“ Weitere Mittheilungen iiber Kern- und Zelltheilung in
der Milz,” ‘ Archiv f. mikr. Anat.,’ xxxi, p. 541.

2. Bextzow, A.—‘‘ Zur Regeneration des Epithels der Harnblase,” ‘ Vir-
chow’s Archiv,’ 97, 1884.

8. Buocumann, F.—‘* Ueber die Richtungskorper bei Insecteneiern,”
‘Morph. Jb.,’ xu, 1887.

4, Branpt, K.— Die Koloniebildenden Radiolarien,” ‘Fauna und Flora
des Golfes von Neapel,’ xiii, 1885.

5. Braver, A.—‘ Ueber die Entwicklung von Hydra,” ‘ Z. wiss. Zool.,’
lii, Hft. 2, 1891.

5a. Braver, A.—‘‘ Ueber die Entstehung der Geschlectsprodukte und
die Entwicklung von Tubularia mesembryanthemum,” ‘Z, wiss.
Zool.,’ lii, Hft. 4, 1891.

6. Brooks, W. K., and Herrick, ¥. H.—‘‘The Embryology and Meta-

morphosis of the Macroura,” ‘U.S. National Acad. of Sciences,’ vol.

v, Fourth Memoir.

VOL. 85, PART 1.—NEW SER. L

154 SYDNEY J. HICKSON.

a.

8.

16.

a6

18.

19.

20.

21.

22.

23.

24.
25.

26.

Bireur, O.—“ Was sind die Attraktionsspharen und ihre Centralkérper,”
‘Anat. Anz.,’ vii, 1892, p. 222.
Craus, C.—‘ Ueber die Entwicklung des Scyphostoma von Cotylorhiza,

Aurelia, und Chrysaora,” &¢., ‘Arb. a. d. Zool. Inst. Wien,’ ix, p. 85,
1891.

. Driscu, H.—‘‘ Entwickelungmechanische Studien,” ‘Z. f. wiss. Zool.,’ 55.
. Fiemmine, W.—“ Ueber Zelltheilung,” ‘ Verh. Anat. Ges.,’ 1891, pp.

125—144.

. Fiemmine, W.—‘ Zellsubstanz, Kern- und Zell-theilung,’ Leipzig, 1882.
. Frenze1, J.—<Zur Bedeutung der amitotischen (direkten) Kerntheilung,”

‘Biol. Centralbl.,’ xi, pp. 558—565.

. GEELMUYDEN.—“ Das Verhalter des Knochenmarks in Krankheiten,”

* Virchow’s Archiv,’ Bd. evi.

. Gerp, W.—“‘ Zur Frage iiber die Kernblatterbildung bei den Hydro-

medusen,” ‘ Zool. Anz.,’ xv, p. 312, 1892.

. Gorrert, H.— Kerntheilung durch indirekte Fragmentirung,” &c.,

‘ Archiv f. mikr. Anat.,’ Bd. xxxvii.

Gruper, A.—‘‘ Ueber Kern und Kerntheilung bei den Protozoen,”’ ‘Z.
f. wiss. Zool.,’ 40, p. 121.

Herrick, F. H.—See Brooks, W. K.

Hickson, 8. J.—‘‘ The Sexual Cells and Early Stages in the Develop-
ment of Millepora plicata,” ‘ Phil. Trans.,’ clxxix, 1888.

Hicxsoy, 8. J—‘‘On the Maturation of the Ovum and Early Stages
in the Development of Allopora,” ‘Quart. Journ. Micr. Sci.,’ xxx,
1890.

Hickson, 8. J.—‘‘ The Medusew of Millepora Murrayi and the Gono-
phores of Allopora and Distichopora,” ‘Quart. Journ. Micr. Sci.,’
xxxii, 1891.

Hearucotg, F. G.—‘‘ The Early Development of Julus terrestris,”
‘Quart. Journ. Mier. Sci.,’ xxvi, p. 449.

Henxine, H.— Die ersten Entwicklungsvorgange im Fliegenei und
freie Kernbildung,” ‘Z. f. wiss. Zoologie,’ xlvi, 1888.

Heyxine, H.—“ Untersuchungen iiber die Entwicklung der Phalangi-
den,” ‘Z. f. wiss. Zool.,’ xlv, 1887.

Henxine, H.—< Untersuchungen iiber die ersten Entwickelungsvor-
gange in den Hiern der Insekten,” ‘ Z. f. wiss. Zool.,’ 54, 1.

Hertwic, R.—‘ Zur Histologie des Radiolarien,’ Leipzig, 1876.

Hertwic, O.—“ Vergleich der Ei- und Samenbildung bei Nematoden,”

. Archiv mikr. Anat.,’ xxxvi, 1890.
HeEss.— Ueber Vermehrungs- und Zerfallsvorgange an den grossen

27.

28.

29.

30.

31.

32.
33.
34.

35.

36.

37.

38.

39.

40.

DEVELOPMENT OF DISTICHOPORA VIOLACEA. 155

Zellen in der acut hyperplastischen Milz der Weissen Maus,” ‘ Beitr.
zur pathol. Anat. herausg. von Ziegler,’ Bd. viii.

JHERING, H. von.—‘ Vergleichende Anatomie des Nervensystemes und
Phylogenie der Mollusken,” Leipzig, 1877.

JENSEN, O. 8.—“ Recherches sur la Spermatogénése,” ‘ Archiv. de Biol.,’
iv, p. 1.

Jickeui, C. F.—‘‘ Ueber die Kernverhaltnisse der Infusorien,” ‘ Zool.
Anzeiger,’ 1884.

KastscHenko, N.—Ueber den Reifungsprocess der Selachiereies,” ‘ Z.
f. wiss. Zool.,’ 50, p. 428.

Kastscurenxo, N.—‘‘Zur Eutwickelungsgeschichte des Selachierembryos,”’
‘Anat. Anzeiger,’ ii, 1888, p. 448.

Kent, W. Savitie.—‘ Manual of the Infusoria,’ vol. ii.

Krnestey, J. S.—“ The Embryology of Limulus,” ‘J. Morph.,’ vii, p. 92.

KisuinovyE, K.—‘Qn the Development of the Araneina,” ‘Journ.
Coll. Sci. Japan,’ iv.

KisuinovyrE, K.—‘‘ On the Development of Limulus longispina,’’
‘Journ. Coll. Sci. Japan,’ v.

Kocu, G. von.— Die Gorgoniden des Golfes von Neapel,” ‘Fauna und
Flora von Golfes von Neapel,’ Monographie xv.

Kowatewsky, A.—‘ Researches on the Development of the Ccelenterata’
(Russian), Moscow, 1873.

Lez, A. Bouttes.—<‘ La Spermatogénése chez les Chétognathes,” ‘La
Cellule,’ tom. iv, 1887, p. 107.

Lister, J. J—‘‘ On the Reproduction of Orbitolites,” ‘Camb. Phil.
Proc.,’ 1893.

Lowit, M.—* Ueber Amitotische Kernteilung,” ‘ Biol. Centralbl.,’ xi,
pp. 518—516.

40a. Marsnatt, W. 8.— Beitrage zur Kenntniss der Gregarinen,” ‘ Archiv

41.

42.
43.

fiir Naturg.,’ lix, p. 25.

McMovraicg, J. P.—‘ Contributions on the Morphology of the Actinozoa.
II: On the Development of the Hexactinie,” “J. Morph.,’ vol. iv,
1891, p. 303.

Mertscunikorr, E.—‘ Embryologische Studien an Medusen,’ Wien, 1886.

Mincuin, EH. A.—“ Observations on the Gregarines of Holothurians,”
‘Quart. Journ. Micr. Sci.,’ xxxiv, p. 279.

44, Morin, I.—“ Zur Entwicklungsgeschichte der Spinnen,” ‘ Biol. Cen-

45.

tralbl.,’ vi, p. 658.

Opret, A.—‘Die Befruchtung des Reptilieneies,” ‘ Archiv f. mik.
Anat.,’ 39, p. 215.

156

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.
64.

SYDNEY J. HICKSON.

Parten, W.— The Development of Phryganids,” ‘Quart. Journ. Mier.
Sci.,’ xxiv, p. 549.

Pictet, C.—‘ Recherches sur la Spermatogénése,” ‘ Mitth. Zool. Stat.
Neap.,’ x, p. 75.

Ratu, O. von.— Ueber die Bedeutung der amitotischen Kerntheilung
im Hoden,” ‘ Zool. Anz.,’ 1891, p. 381.

Scuneiper, A.—‘‘Sporulation du Klossia octopiana,” ‘Arch. de
Zool. exp. et génér.,’ 1883, p. 77.

ScHoTTLAENDER, J.—‘ Ueber Kern- und Zelltheilungsvorgange in dem
Endothel der entziindeten Hornhaut,’”’ ‘ Archiv f, mikr. Anat.,’? Bd.
Xxxi, p. 426.

Sepewick, A.—‘ Development of the Cape Species of Peripatus,”
‘Quart. Journ. Micr. Sci.,’ xxvi, p. 175.

Smitu, F.—“ The Gastrulation of Aurelia flavidula,” ‘ Bull. Mus.
Comp. Zool.,’ xxi, 1891.

Surtpon, L.—“ On the Development of Peripatus nove-zealandiz,”
‘Quart. Journ. Micr. Sci.,’ xxvili, p. 495.

Strope.—‘ Ueber Kerntheilung und Riesenzellenbildung in Geschwiils-
ten und im Knochenmark,” ‘ Beitrg. z. pathol. Anat. herausg. von
Ziegler,’ Bd. vii.

TicHomrrorr.—‘ The Embryology of the Hydromeduse’ (Russian),
Moscow, 1887.

Verson, E.—< Zur Beurteilung der amitotischen Kernteilung,” ‘ Biol.
Centralbl.,’ xi, pp. 556—558.

Viatteton, M. L.—“ Recherches sur les premieres phases du développe-
ment de la Seiche,” ‘ Ann. Sci. Nat.,’ vi, 1888.

Wa.pEyEerR, W.—“ Ueber Karyokinese und ihre Beziehungen zu den
Befruchtungsvorgangen,” ‘ Arch. f. mikr. Anat.,’ xxxil.

Watasz, S.— Studies on Cephalopoda. I: Cleavage of Ovum,” ‘J.
Morph.,’ iv, pp. 247—3802.

Weismann, A.—“ Beitrage zur Kenntniss der ersten Entwicklungs-
vorgange im Insectenei. In Beitrage zur Anatomie und Embryologie,”
&e., ‘ Festschrift fir J. Henle,’ Bonn, 1882.

WerRNER, W.—‘“ Ueber Theilungsvorginge in den Riesenzellen des
Knochenmarkes,” ‘ Virchow’s Archiv,’ Bd. cvi.

Wu, L.— Entwicklungsgeschichte der viviparen Aphiden,” ‘ Zool.
Jahrbicher,’ ii, 1888.

Witson, E. B.—* The Development of Renilla,” ‘ Phil. Trans.,’ 1883.

Wotters, M.—* Die Conjugation und Sporenbildung bei Gregarinen,”
‘ Arch. mikr. Anat.,? Bd. xxxvii, 1891.

DEVELOPMENT OF DISTICHOPORA VIOLACEA. 157

65. Zirciter, H. E.— Der biologische Bedeutung der amitotische Kern-
theilung im Thierreich,” ‘ Biol. Centralbl.,’ xi, 12 and 13, pp. 372—
389. Transl. in ‘Ann. N. H.,’ viii, p. 362.

DESCRIPTION OF PLATE 9,

Illustrating Dr. Sydney J. Hickson’s paper on the
“* Development of Distichopora.”

Fic. 1.—Young ovum of Distichopora, situated in the cup-shaped tropho-
disc (¢7.). The germinal vesicle, G.v., is situated near the centre of the
ovum.

Fic. 2.—Germinal vesicle of the same stage, showing the vacuoles in the
nucleolus.

Fic, 3.—Germinal vesicle of the ovum of Distichopora, migrating from the
centre to the periphery. The membrana limitans becomes obscure over the
pseudopodial processes.

Fig. 4.—A peculiar condition of the germinal vesicle, observed in only one
preparation.

Fic. 5.—A germinal vesicle, with chromatin granules arranged in a row.

Fic. 6.—Germinal vesicle, containing numerous minute chromatin granules
situated at the periphery of the ovum.

Fic. 7.—A stage showing the disappearance of the inner part of the mem-
brana limitans.

Fic. 8.—A stage showing the complete disappearance of the membrana
limitans.

Fic. 9.—Section of a young embryo which shows only two large nodes of
protoplasm, each of them containing a few deeply staining granules. The
yolk is omitted from the lower part of the section in order to show the loose
protoplasmic mesh-work which pervades the embryo.

Fic. 10.—A later stage in the development, showing several nodes of
various sizes, some with nuclei, some without.

Fic. 11.—A stage in the development corresponding to that of Fig. 10, to
show the relation of the nuclei to the yolk. Hach nucleus is situated in
a small protoplasmic area or node, and the yolk granules close to it are ex-
tremely small.

Fic. 12.—The same stage as Fig. 10, showing the different phases in the
formation of the nuclei. The details of the yolk are omitted.

VOL. 35, PART 1.—NEW SER. M

158 SYDNEY J. HICKSON.

Fic. 13.—A section of a young embryo, which shows yolk segmentation.

Fic. 14.—A section through an older embryo, the yolk being omitted,
showing the first stages in the formation of the ectoderm (ect.). ec. Ecto-
derm. ez. Endoderm of the gonophore.

Fig. 15.—A section through a still older embryo, showing a later stage in
the formation of the ectoderm and its connection with the central endoderm
plasmodium.

Fic. 16.—A solid planula of Distichopora, just before it escapes from the
gonophores.

Fic. 17.—A small portion of the endodermic plasmodium more highly
magnified. Some of the smaller yolk granules are omitted.

Fie. 18.—a, 6, c, d, e. Six stages in the division of the nuclei of the

ectoderm.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 159

Studies on the Comparative Anatomy of
Sponges.

V. Observations on the Structure and Classification of the
Calearea Heterocela,

By

Arthur Dendy, D.Sc.,
Fellow of Queen’s College, and Demonstrator and Assistant Lecturer on
Biology in the University of Melbourne.

With Plates 10—14.

I. PREeFAce.

Tue object of the present paper is to give a general account
of the anatomy, histology, and classification of the Calcarea
Heteroccela, from the point of view of one who has for some
time past been engaged in an independent study of the group.
The exceptionally fine collection of Calcarea Heterocela
which I am fortunate enough to have at my disposal must be
my justification for making this attempt.

Concerning the anatomy and histology of the group, I have
perhaps not much to say that is new, but I hope that by
bringing together our information on the subject in a collected
form I may be of some use to students of spongology. Serious
errors have from time to time crept into our information as to
the anatomy of the Heteroceela, and I can only hope that I
have not reproduced any of them in this paper. As, however,
by far the greater part of my information has been derived
from personal observation, I feel tolerably safe in this respect.

With regard to classification, I have been obliged to depart

VOL. 35, PART 2,—NEW SER. N

160 ARTHUR DENDY.

widely from the lines laid down by previous writers. The
necessity for doing so was forced upon me when preparing my
“ Synopsis of the Australian Calcarea Heterocela” (4). In
that paper I proposed a classification of the group based upon
a personal examination of forty-seven species, and a careful
consideration of the published accounts of species which I had
not seen. In the Synopsis, however, I had not space to justify
the classification proposed, merely giving it as a skeleton upon
which to arrange my descriptions of species. The task of
justification I reserved for the present occasion, and I have
endeavoured to fulfil it rather by a critical re-examination
of the anatomy of the group than by a detailed criticism of the
systems of other spongologists. A great deal depends upon
whether one regards the canal system or the skeleton as
affording the most reliable guide to the systematist, for the
characters of the two certainly appear to contradict one
another. Here, as in similar cases, I believe that a com-
promise is the only satisfactory way out of the difficulty,
as neither set of characters is solely reliable. The skeleton
evidently follows the canal system up to a certain stage of
organisation, and then begins to vary independently. Up to
this stage I believe the canal system to be most important ;
after it I think the skeleton has prior claims, while the canal
system becomes of secondary value. According to this idea,
which will be found elaborated later on, I regard the Leuconoid
type of canal system as of polyphyletic origin, as also the
« Sylleibid ”’ type, and I abandon the old families Syconide
(Sycones) and Leuconide (Leucones) of Haeckel, as well
as the more recent Sylleibidz of von Lendenfeld. I have
previously shown (9) that the very generally accepted family
Teichonide must be abandoned, and I am glad to see that
von Lendenfeld follows my lead in this (10), although he
declines to acknowledge any indebtedness to my writings.

As I have naturally adopted my own classification as a
basis for the arrangement of the subject-matter of this paper,
it may be convenient to the reader if I give an outline
thereof at once, reserving further discussion as well as the

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 161

necessary diagnoses till later on. The following, then, is the
classification proposed :

Families. Genera.

. Leucascus.
. Sycetta.
. Sycon.
. Sycantha.
. Grantia.
Sub-genus Grantiopsis.
6. Ute.
Sub-genus Synute.
. Utella.
. Anamixilla.
. Sycyssa.
. Leucandra.

“4
1. Lelapia.
12 Leucyssa.
Ms ;
11
Ul

1. Leucascidee .

2. Sycettide

Oo HE O 2

3. Grantide

Oo oaos

3. Grantessa.
4. Heteropia.
5. Vosmaeropsis.

4, Heteropide.

6. Heteropegma.
7. Amphoriscus.
8. Syculmis.
9. Leucilla.

5. Amphoriscide .

Before concluding these introductory remarks it is my
pleasant duty to express my sincere thanks to various friends,
without whose assistance this paper could not have been
prepared. To Mr. J. Bracebridge Wilson, M.A., I owe, as
usual, the greater portion of my material, and I am also
indebted to Professor W. Baldwin Spencer, Mr. T. Whitelegge,
and the authorities of the Adelaide Museum for a number of
very valuable Australian sponges; while numerous fragments of
type specimens from the British Museum, most generously
forwarded to me by Dr. Giinther, have been of the greatest
service. Lastly, I must again thank Professor G. B. Howes
for most kindly undertaking the correction of the proof in my
absence from England.

The numbers in parentheses, in the text, refer to the list of
literature at the end. The technical descriptions of a large

162 ARTHUR DENDY.

number of the species referred to, or references to places where
they are to be found, are given in my “Synopsis of the
Australian Calcarea Heterocela” (4).

MELBOURNE ;
January, 1893.

Il. Tur Cana SysteM oF THE CaLcAREA HETEROCG@LA.

The simplest type of canal system in the group is found in
the genus Sycetta. This is, unfortunately, a genus which |
have never had the opportunity of personally examining, and I
am indebted for my information concerning it to Haeckel’s
great work on the Calcarea (5).

A Sycetta individual consists, in the first place, of a central
tube, which bears at its summit a single osculum leading from
the cavity of the tube (gastral cavity) to the exterior. This
central tube gives off all around and throughout its length
numerous short, hollow, conical diverticula (the radial flagel-
lated chambers). Each radial chamber communicates by a
single exhalant opening at its proximal end with the central
gastral cavity, and each has its wall perforated by numerous
much smaller apertures (the prosopyles), through which the
water passes from the exterior into the cavity of the chamber.
It is important to notice that the radial chambers, of which
there may be a large number, all stand perfectly free from one
another and do not touch at any point, so that the water is
free to circulate between them without obstruction of any kind,
and to penetrate right in to the outer surface of the wall of the
central gastral cavity. Hence there is no true inhalant canal
system, further than the small prosopyles by which the water
gains access to the interior of each chamber. The wall of the
central gastral cavity is very thin, so that the chambers open
almost directly into the latter, the wide “ exhalant canals ”
which conduct the water through the wall being comparatively
short and inconspicuous.

The collared cells are, of course, confined to the interior of
the radial chambers, while the central gastral cavity and the

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 163

outer surfaces of the radial chambers and of the central tube
are doubtless caused by a layer of pavement epithelium, as in
other Heteroceela.

Such, then, is the simplest type of canal system met with
amongst the Heterocela; and, as it appears to me, all the
higher types met with in the group may be derived from
some such simple radiate one by modification in one or more
of the following ways:

(1) By the outer surfaces of adjacent radial chambers
coming in contact with one another and fusing together.
This fusion at first takes place irregularly and partially, and
then more completely and throughout the length of the
chambers, so as to divide the water-containing space which
surrounds the chambers into a series of more or less well-
defined inhalant canals, sometimes called “ intercanals ”
(compare figs. 2—5).

(2) By the closing in of the inhalant canals at their distal
ends by the outgrowth of a thin, pore-bearing, dermal mem-
brane from the walls of the radial chambers at or near their
distal extremities. In this way true ‘‘ dermal pores” are
formed, through which the water gains access to the inhalant
canals (compare figs. 3, 6—8).

(3) By the increase in thickness of the dermal membrane
aud the development in it of a speciai skeleton, so as to form
a thick cortex, which not only stretches between but also
covers over the ends of the radial chambers. The formation of
such a thick cortex necessitates the development of a more or
less highly specialised “ cortical inhalant canal system,” which
places the dermal pores in communication with the deeper
parts of the inhalant canal system (compare figs. 9—12, 23).

(4) By the branching of the radial chambers, and conse-
quently of the inhalant canals which lie between them
(compare figs. 7, 9, 10, 18, 20).

(5) By increase in thickness of the wall of the central
tube (gastral cortex), and consequently also in the length of
the exhalant canals (compare figs. 4, 9, 10).

(6) By retreat of the collared cells towards the distal

164 ARTHUR DENDY.

extremities of the chambers, their place being taken by
pavement epithelium. This results in elongation and branching
of the exhalant canals, and in corresponding shortening of
the chambers, which may be thus converted from the elon-
gated radial chambers of the Syconoid type to the short,
rounded, and irregularly arranged chambers of the Leuconoid
type. Branching of the primitively straight chambers is of
course also necessary in order to effect this change (compare
figs: MO WLOM21s 7 aiG)ee

(7) By evagination or folding of the wall of the central
gastral cavity, which also results in elongation of the exhalant
canals, and possibly also in branching of the same (compare
fig. 7). As pointed out by Sollas (6), it is often extremely
difficult, if not impossible, to decide which of the two sets
of causes indicated in this and the preceding paragraph
respectively have operated in the production of a particular
type of canal system, though in some cases the arrangement
of the skeleton (fig. 7) may furnish a clue to the problem.
Though I do not doubt that in some cases a certain amount
of folding of the wall of the gastral cavity has taken place,
I do not believe that this cause has operated to any great
extent, in the production of the Leuconoid from the Syconoid
type of canal system.

(8) By the fusion of different Syconoid or Leuconoid
individuals! of a branching colony to form a compact whole, in
which the individuality of the different members is more or
less completely obliterated (compare fig. 15).

Other minor causes have also doubtless aided in the produc-
tion of various modifications in the canal system; such are the
enormous dilatation of the gastral cavity and osculum in
Grantia labyrinthica (9), and the strong development of
the mesoderm in the walls of the flagellated chambers which
takes place in some forms. The causes (or sets of causes)

1 By the terms Syconoid and Leuconoid individuals 1 mean simply in-
dividuals or “ persons” consisting each of a single, central osculum-bearing
tube enclosing the gastral cavity, and surrounded by flagellated chambers and

canals arranged according to the ‘‘Sycon”’ or “ Leucon” type, as the case
may be.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 165

given above are, I believe, the principal ones to which we
must look for an explanation of the peculiarities in the canal
system about to be described in different genera of Heterocela.

The various genera (with the possible exception of Leucas-
cus) might, as regards the canal system, be arranged in a
gradually ascending series, commencing with Sycetta and
ending with those genera, such as Leucandra, in which
the Leuconoid type reaches its maximum development. Such
a series would not, however, as I believe, represent a natural
arrangement of the group, for there is, as I hope to show later
on, strong reason for concluding that the most highly modified
Leuconoid type of canal system has been independently arrived
at in several distinct genera. At present, however, we need
not concern ourselves with this question, but pass on to
consider to what extent the various causes indicated in the
above scheme have operated in modifying the canal system
in each separate genus. I propose to deal with the genera in
the order in which they occur in my system of classification.
In this manner, having first of all disposed of Leucascus, we
shall be able to start with Sycetta, and trace the gradual
evolution of the canal system from its simplest Syconoid form
in that genus, to its most complex Leuconoid one as exhibited
in Leucandra. We shall then, in each of the two remaining
families (Heteropide and Amphoriscide) which are dis-
tinguished by skeletal peculiarities, be able to start fresh with
a Syconoid type of canal system, and work up again to the
Leuconoid.

Leucascus (fig. 1).

In this genus the sponge is more or less massive or lobate,
and it is not possible to distinguish a single central gastral
cavity. There may be several oscula, which perhaps indicates
that the whole sponge is then to be regarded as a colony com-
posed of several fused individuals. The arrangement of the
canal system in L. simplex, as seen in a vertical section
through the osculum, is shown in fig. 1. The flagellated
chambers are very long, and copiously branched. Their blind
distal extremities lie beneath the dermal surface, towards

166 ARTHUR DENDY.

which they are directed more or less at right angles, so that
the chambers certainly exhibit a more or less radial arrange-
ment. At their proximal ends the chambers open into wide
and long exhalant canals, which converge towards the osculum.
The distal ends of the chambers are covered over by a thin mem-
brane, strengthened by spicules and perforated by numerous
inhalant dermal pores. ‘These pores lead into a series of quite
irregular spaces lying between the branching chambers. These
spaces represent the inhalant canals, and convey the water to
the prosopyles in the thin walls of the chambers. In all speci-
mens which I have seen of L. simplex and in one of L. cla-
vatus (4) the mesoderm is feebly developed, so that the dermal
membrane, the walls of the chambers, and the walls of the ex-
halant canals are all very thin, and the entire sponge has
consequently a soft and delicate texture. In one of the speci-
mens which I have referred to, L. clavatus, the mesoderm
is very strongly developed; the sponge thereby acquires a
dense and solid texture, and the canal system is correspondingly
reduced in dimensions. This strong development of the meso-
derm is perhaps to be associated with the fact that the speci-
men contains very numerous embryos; for the normal condition
of the genus appears to be one with all the canals and chambers
thin-walled.

This genus, as already indicated, does not appear to come
into what may be called the typical series of Heteroccela; and its
relationships and systematic position will be discussed later on.
Though I do not suppose that it has ever passed through a
Sycetta stage in its history, it is easy to see how its canal
system may have been derived from a radiate ancestral type, by
modification along the lines suggested above.

Sycetta.

The simple form of canal system met with in this genus has
already been described. If we exclude from the genus
Haeckel’s S. strobulus, 8S. cupula, and S.stauridia, which
are all corticate species, and include, as von Lendenfeld (10)
has rightly done, Sycetta (Sycaltis—H.) conifera, we are

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 167

left with three species, all of which exhibit the same type of
canal system, in which the radial chambers are short, straight,
and unbranched, and project quite separately and indepen-
dently from the wall of the central gastral cavity.

According to Haeckel (5) there is in Sycetta primitiva,
and also in certain more highly organised species of Syconoid
Heteroceela, a single larger aperture at the distal end of each
radial chamber, which he terms a “ dermal ostium.” Poléjaeff,
however, has (8) thrown considerable doubt on the existence
of “dermal ostia” in any case, and I cannot help agreeing
with him in attributing the supposed presence of these
structures to an error of observation on the part of their
describer.

Sycon (figs. 2—7).

In this genus we meet with a considerable amount of varia-
tion in the canal system. The simplest form is found in such
species as the Huropean 8. ciliatum (Bauerbank’s Grantia
ciliata) and S. raphanus, and in the Australian 8S. Carteri
and S. minutum (4). For a detailed account of 8S. raphanus
I may refer the student to Schulze’s well-known and admi-
rable memoir (7).

These species mark but a slight step upwards from the
Sycetta type. In Sycon Carteri, for example (fig. 2), the
radial chambers are rather short, and more or less thimble-
shaped; they touch one another in some places, and there fuse
together by their outer surfaces. ‘Their distal ends, however,
project freely, and thus torm well-marked ‘distal cones ”
on the surface of the sponge. By the fusion of adjacent
chambers at the points of contact, the originally continuous
water-containing space which surrounds all the chambers
becomes broken up into more or less definite inhalant canals
(‘‘inter-canals”’?), whose exact form depends, of course, upon the
shape of the chambers and the extent to which they fuse to-
gether, and is ofnogreat importance. In the simplest cases, such
as Sycon Carteri (fig. 2) these canals appear, in sections taken
along the length of the chambers, as straight narrow gaps

168 ARTHUR DENDY.

between adjacent chambers. In the case of Sycon raphanus,
Schuize (7) has shown that the different inhalant canals often
inter-communicate through gaps left by the incomplete fusion
of adjacent chambers, and this condition probably occurs to a
greater or less extent in a great many Cases.

Probably in most species of the genus Sycon, and certainly
in the four with which we are now more immediately con-
cerned, the distal ends of the inhalant canals (“ inter-canals ”’)
are in no way closed in; in other words, the spaces between
the distal cones of the radial chambers remain widely open.
Thus the water still has direct access to the prosopyles, with-
out having to pass through the pores of a dermal membrane.
In Sycon ensiferum (4), a species closely related to S.
raphanus, the basal rays of many of the most distally
situated tuber triradiates are very strongly bent outwards from
the walls of the chambers, so as to curve over and protect the
entrances to the inhalant canals.

In the common Australian 8. gelatinosum (figs. 3—6)
we meet with a slight but very interesting advance in organi-
sation. The radial chambers, though of considerable length,
are still straight, and usually unbranched, subcylindrical
tubes. The mesoderm of the chamber walls is strongly de-
veloped, and the inhalant canals between the chambers are
very well defined and squarish in transverse section (fig. 8).
The gastral cortex is rather thick, and consequently the
exhalant canals take the form of distinct, though short
and wide, tubes, sharply marked off from their respective
chambers by well-developed chamber diaphragms, which
appear to be almost universally present in calcareous sponges
(fig. 4). A tangential section of the dermal surface (fig. 6)
shows that the distal ends of the chambers, each crowned with
its tuft of oxeote spicules, do not touch one another, but are
separated by rather narrow gaps or spaces, across which
stretches a delicate pore-bearing membrane (seen in section in
fig. 3). This membrane is doubtless formed as an outgrowth
of the ectoderm and mesoderm of the walls of the chambers.
In Sycon gelatinosum it contains no spicules. The fusion

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 169

of adjacent chambers does not commence until a short distance
within this pore-bearing membrane, so that the dermal pores
lead at first into a continuous subdermal space, from which
the true “ inter-canals” penetrate between the chambers.

In another Australian species, to which I have given the
name Sycon boomerang (4), we meet with yet further com-
plication. The principal features in the anatomy of this
species are shown in figs. 7 and 8. The radial chambers are
very long, thin-walled, and very much branched, especially
towards their distal ends.! The irregularity in the branching
causes the tufts of oxea at the distal ends of the chambers to
form an irregular series of prominences on the surface of the
sponge (fig. 8). Owing to the branching of the chambers the
inhalant canal system also becomes very irregular. The wall
of the gastral cavity is rather thick, and the length of the
exhalant canals appears to be further increased by an irregular
and not very extensive folding of the same (fig. 7). As in S.
gelatinosum, a thin pore-bearing membrane extends be-
tween the ends of the radial chambers, but in S. boomerang
a few spicules are found in this membrane (fig. 8).

In Sycon giganteum (4), a very large species from the
Gulf of St. Vincent, which closely resembles 8S. gelatinosum
in structure, the radial chambers are narrow and greatly
elongated ; they branch repeatedly, and the branches run
parallel with one another to the dermal surface. They come
municate with the gastral cavity through rather long exhalant
canals, which commence at some distance beneath the gastral
cortex. These canals appear to have been formed by modifi-
cation of the proximal portions of the radial chambers, from
which they are separated as usual by diaphragms. They may
unite together before opening on the gastral surface. The
inhalant canals are irregular and very narrow, opening on the
dermal surface through narrow, irregular chinks between the
tufts of oxeote spicules which cram the distal ends of the
chambers. Ihave not been able to detect any pore-bearing

* It appears from Schulze’s researches that branching of the chambers
may take place even in such a simple form as S. raphanus (7).

170 ARTHUR DENDY.

dermal membrane stretched across the openings, which are so
narrow that such a structure could hardly be required.

In Sycon Ramsayi, a species common in Port Jackson
and described by von Lendenfeld (11), the exhalant canals,
though confined within the limits of the thick gastral cortex,
are extraordinarily well-defined cylindrical tubes, and it is
interesting to notice that although the long radial chambers
are themselves unbranched, and there is no indication of the
exhalant canals having been formed by modification of their
proximal ends, yet these canals may unite together before
opening on to the gastral surface, and thus present a branching
structure. In other respects Sycon Ramsayi agrees in canal
system with the simpler species of the genus.

Sycantha.

We owe our information concerning this sponge entirely to
von Lendenfeld, who has recently described it from the Gulf
of Trieste, and from his work (10) the following details as
to the canal system are taken.

Sycantha tenella (the only species) is a large, tubular
sponge, with a single terminal osculum. The gastral cortex is
thin, and forms a cylindrical tube. From the outer surface of
the gastral cortex project tufts, which are partially united
together by membranes, and attain a length of 2—4 mm.
Each tuft runs out into a number of flexible points. These
points are the free distal ends of the flagellated chambers of
which the tufts are composed. Ten to twenty chambers,
united at their bases, form a tuft, each of which is thus a
group of flagellated chambers. The chambers themselves are
long and narrow. At the base they are irregularly prismatic,
generally quadrangular in cross-section. The free, always
unbranched, terminal portion is circular in cross-section, and
runs out into a conical point. In their basal portions the
chambers touch one another, and here there are no inhalant
canals (inter-canals) between them, so that collared cells are
found on both sides of all membranes which occur in the
interior of the basal portions of the tufts. In these membranes,

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 171

which separate the different chambers from one another,
roundish openings occur, which place the cavities of all the
chambers of a group in direct communication with one
another. One or several of the chambers of each group are
placed in connection with the central gastral cavity through
openings in the gastral cortex.! Small circular inhalant pores
(prosopyles) occur in large numbers in the free distal portions
of the chamber walls. Sycantha tenella is said to be
distinguished from all other Sycons by the fact that the
chambers do not all communicate directly with the central
gastral cavity, and also by that of their non-communica-
tion with one another in groups, the stream of water passing
partly from chamber to chamber before it reaches the gastral
cavity.

I am inclined to think that the apparent intercommunication
of the chambers may be due to an error of observation. It is
by no means the first time that such a condition has been
described. Poléjaeff observes (8) that ‘there is no doubt that
what Haeckel declares to be ‘dermal ostia’ and ‘dermal pores’
in the individuals of his ‘Syconusa type’ were merely the pores
of the intercanals; and what he calls ‘ conjunctive pores,’
these latter uniting, according to him, the cavities of the radial
tubes, were nothing but the common pores on the side walls of
the radial tubes connecting these latter with the intercanals.
To anyone who will notice Professor Haeckel’s remark that
these ‘conjunctive pores’ are best to be observed in sections of
dry Sycones, the error into which he fell will be easily
comprehended.” It is a noteworthy fact in this connection
that von Lendenfeld himself observes that his specimens of
Sycantha tenella had been preserved for some time in not
very strong spirit, so that he was unable to make any observa-
tions on the finer histological characters. He also refers to
the extraordinary delicacy of the mesoderm in all parts.

I may also remind the reader that Carter (12) bases an

1 Later on in the work referred to (p. 192) we are told that “jede

Kammergruppe ist durch eine einzige grossere Oeffuung in der Gastralmem-
bran mit dem centralen Oscularrohr in Verbindung.”

172 ARTHUR DENDY.

entire genus (Hypograntia) on the supposed presence of such
“large holes of inter-communication between the chambers.”

On the whole it appears to me that Sycantha tenella
probably presents a slight modification of the simpler Sycon
type of canal system, having the radial chambers united in
groups.

Grantia (figs. 9, 10).

The canal system of this genus may be regarded as derived
from the Sycon type by the conversion of the thin, pore-
bearing, dermal membrane which we first met with in Sycon
gelatinosum (figs. 3 and 6) and then, with the addition of a
few spicules, in S. boomerang (figs. 7 and 8), into a more or
less strongly developed spicule-bearing cortex, which not only
extends between the ends of the radial chambers, but also
covers them over, so that we no longer find in this genus
distal cones projecting from the dermal surface.

A thoroughly typical example of the genus is afforded by
the Australian Grantia extusarticulata (4), the anatomy
of which is represented in fig. 9. The radial chambers are
almost straight, cylindrical, and only slightly branched;
between them lies the irregular and more or less lacunar
inhalant canal system. The dermal or inhalant pores are
irregularly scattered through the dermal cortex, which is well
developed and about 0:07 mm. thick. The gastral cortex is of
about the same thickness, and is perforated by the short, wide
exhalant canals; one coming from each chamber, and separated
from it by a constricting diaphragm.

Grantia Vosmaeri (4), another Australian species, whose
anatomy is represented in fig. 10, illustrates in an extremely
interesting manner the gradual shortening of the radial
chambers and the corresponding elongation of the exhalant
canals, accompanied by a strong development of the mesoderm.
which surrounds the latter. Here, again, the junction between
exhalant canals and flagellated chambers is clearly marked by
well-developed diaphragms. The dermal cortex is very thick,
and the dermal pores communicate with the deeper parts of

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES, 178

the inhalant canal system through distinct, though not very
regular, cortical canals.

I have already, in an earlier paper (9), given a detailed
account of the anatomy of the somewhat aberrant and very
remarkable Grantia labyriuthica. In its essential features
the canal system agrees with that of G. extusarticulata,
and there are only two points to which I need refer again in
this connection. The first concerns the inhalant pores, which,
in this species, tend to be collected together in groups or pore
areas, where the cortex forms only a thin dermal membrane,
and overlies wide lacunar spaces perhaps comparable to the
subdermal cavities of some higher sponges, though not at all
sharply separated from the deeper parts of the inhalant canal
system. The second concerns the central gastral cavity and
osculum, which, though in most calcareous sponges so uniform
in structure as to require no special notice, are here enormously
enlarged, so that the entire sponge takes the form of a funnel
whose thin wall is thrown into deep folds, and the margin of
the osculum is no longer evenly curved, but extremely sinuous.
Since my memoir on the species was written Mr. Bracebridge
Wilson has dredged some very large specimens, one of which
measures no less than five inches across the top, so that the
actual circumference of the osculum is something enormous.

Grantia labyrinthica also illustrates very clearly the
branching of the radial chambers, which is of very general
occurrence in Syconoid sponges higher than the Sycetta

type.
Grantiopsis (fig. 11).

The only species of this sub-genus, Grantiopsis cylin-
drica (4), though evidently derived from the ordinary Grantia
type by no very great amount of modification, is in many ways
a very remarkable sponge. The entire sponge forms long
cylindrical tubes, which may branch and which are provided
with single terminal oscula. The largest tube which I have
seen is unbranched and slightly crooked, 57 mm. long, and
with a nearly uniform diameter of 5 mm. The wall of the

174 ARTHUR DENDY.

tube surrounding the central gastral cavity is about 1 mm.
thick, and is divided into two sharply defined concentric layers
of about equal thickuess. The outer of these layers forms a
firm cortex, with a very strongly developed skeleton. The
inner layer is soft and spongy, consisting almost entirely of
the thin-walled radial chambers.

Fig. 11 represents a portion of a transverse section of the
sponge. It will be seen from this that the radial chambers
are arranged side by side with great regularity. Each is a
straight, wide, unbranched (or very slightly branched), thin-
walled tube, extending completely through the chamber layer.
In cross-section the chambers vary from nearly square to
nearly circular. Hach opens directly and separately into the
gastral cavity, the gastral cortex being so thin that no special
exhalant canals are required. Each is provided at its proximal
end with a membranous diaphragm, which, in spirit specimens,
almost closes the exhalant opening. There is, of necessity, a
well-developed cortical inhalant canal system. The inhalant
pores, scattered over the dermal surface, lead into sharply
defined cortical canals, which unite into larger trunks, which
conduct the water to the ordinary “intercanals”’ between the
radial chambers. Hence it appears that, as regards the canal
system, the points to be specially noticed in Grantiopsis are
the strong development of the cortical canal system and the
great elongation of the gastral cavity.

Ute (figs. 12—14).

The canal system of a typical Ute agrees essentially with
that of a typical Grantia. In Ute syconoides (4) the canal
system is unusually regular, and the radial chambers have a
remarkably definite and constant form. The anatomy of this
species, which in general form closely resembles the European
U. glabra, is represented in figs. 12—14. The radial
chambers are straight, and circular in transverse section for
the greater part of their length (fig. 13). Towards their
distal extremities, however, they widen out in a direction at
right angles to the long axis of the gastral cavity of the

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 175

sponge, and at the same time become hour-glass shaped in
transverse section (fig. 14), while each has a slight depression
at its extreme end (fig. 12). The extraordinary regularity
with which these chambers are arranged, and the very definite
relation which they bear to the lines of huge oxeote spi-
cules in the dermal cortex, will be sufficiently evident from
the figures referred to. The inhalant pores are thickly scattered
over the dermal surface between the parallel lines of oxeote
spicules. They lead into spaces which are at first somewhat
lacunar, but soon give place to sharply defined “ inter-canals ”
as they pass in between the radial chambers (vide figs.). The
gastral cortex is very thin, and the exhalant canals are con-
sequently very short.

The canal system of Ute argentea, another Australian
species, has been figured by Poléjaeff (8). It appears to
differ from that of U. syconoides principally in the relatively
greater thickness of the dermal cortex, and consequent
shortness of the radial chambers, and in the less regular
arrangement of the inhalant canal system.

In a colonial species from Port Jackson, which I have named
Ute spiculosa (4), we find the canal system much more
irregular, in accordance with an unusually strong develop-
ment of the mesoderm and its contained skeleton. The gastral
cavity is narrow and cylindrical, occupying about one third of
the total diameter of the sponge. The flagellated chambers
are long and narrow, and more or less radially arranged. They
do not extend nearly through the entire thickness of the
sponge wall, and they communicate with the gastral cavity
through long, sometimes branched exhalant canals. The
inhalant canal system consists of scattered pores on the dermal
surface, leading into elongated canals which lead down between
the chambers, but the typical radial arrangement of the canal
system is greatly obscured by the strong development of the
mesoderm, and the dense, irregular skeleton. There is a very
thick dense cortex on both dermal and gastral surfaces.

In Ute Spenceri (4), from the same locality, we find
another species with a very strongly developed mesoderm and

VOL, 35, PART 2,—NEW SER, 0)

176 ARTHUR DENDY.

skeleton. This sponge is solitary, and is remarkable for its
globular or subspherical shape, with correspondingly situated
gastral cavity and narrow osculum. The dermal cortex is
very strongly developed, occupying more than one third of the
entire thickness of the sponge wall. The inhalant pores,
scattered over the surface of the sponge, lead into wide, irregular
subdermal cavities lying in the cortex, from which inhalant
canals lead down between the radial chambers. The chambers
themselves are arranged parallel to one another with consider-
able regularity. They are long and narrow, and at their distal
ends they branch in a curiously irregular manner, the branches
sometimes penetrating for some little distance into the dermal
cortex. The proximal ends of the chambers are all situate at
about the same level, which is some little distance from the
gastral cavity, and even from the gastral cortex, which latter
is very much thinner than the dermal cortex. Hence we find
a number of rather short, cylindrical, radially arranged ex-
halant canals, which look exactly like continuations of the
radial chambers without the collared cells, and which may
unite together in groups before opening on the gastral surface.
The points of junction of these exhalant canals with the radial
chambers are marked as usual by diaphragms. The “ inter-
canals ” between the chambers are narrow and irregular.

Ute spiculosa and U. Spenceri occupy a position in the
genus Ute very similar to that occupied by Grantia Vos-
maeri (fig. 10) inthe genus Grantia. In both cases the typical
radial arrangement of the canal system is more or less dis-
turbed in accordance with the strong development of the
mesoderm and skeleton.

Synute (fig. 15).

In this interesting sub-genus we meet with a very unusual,
if not unparalleled condition, in the complete fusion of a large
number of Syconoid individuals to form a compact, solid
sponge invested in a common cortex. The anatomy of Synute
pulchella, as seen in horizontal section, is represented in
fig. 15. As I have already pointed out (18), the canal system,

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 177

apart from the fusion of the individuals, closely resembles
that of Ute argentea, as figured by Poléjaeff (8). A horizontal
section of the colony shows a number of circular spaces
scattered at intervals, generally, but not always, in a single
row. These are the gastral cavities of the Syconoid individuals
cut across. Each cavity is surrounded by its own radial
chambers, arranged in a perfectly normal manner, except for
the fact that on the adjacent sides of any two neighbouring
gastral cavities the chambers are much shortened, and some-
times appear bent outwards as though to avoid one another.
The inhalant pores are scattered over the dermal surface, and
lead into irregular canals, which pierce the thick cortex to
reach the radial chambers. At their lower ends the gastral
cavities of the fused Syconoid individuals all communicate
with one another, indicating that this peculiar form of colony
has arisen from fusion of adjoining individuals of a branching
colony. The radial chambers are approximately octagonal in
transverse section, while the much smaller “ intercanals ”’
between them are square. The exhalant openings of the
chambers are protected by very well-developed diaphragms,
and each gastral cavity has also a single large, well-developed
diaphragm situate just within the osculum.

Probably Synute pulchella is the only known example of
a calcisponge with a Syconoid canal system, which has become
at all strikingly modified in accordance with the principle set
forth in section 8 of the scheme given above.

Utella.

This genus, founded entirely on skeletal characters for the
reception of Haeckel’s Sycandra hystrix, and perhaps also
Schmidt’s Ute utriculus (5), presents no features of special
interest in its canal system, which appears to conform to the
ordinary corticate Syconoid type, like that of Grantia.

Anamixilla.

This genus is also founded entirely on skeletal characters,
and the only known species, A. torresi, Poléjaeff (8), appears

178 ARTHUR DENDY.

to possess the typical corticate Syconoid canal system, well
shown in Poléjaeff’s figure of the anatomy of the sponge.

Sycyssa.
Here, again, the same remarks apply as in the case of the last

two genera. The anatomy of Sycyssa Huxleyi is illustrated
by Haeckel in his great monograph (5).

Leucandra (figs. 16, 17, 25).

Tn this genus we meet for the first time with the well-known
Leuconoid type of canal system, the essential features of which
are shown in fig. 16, representing a portion of a transverse
section through the wall of Leucandra phillipensis (4).
Tt will be seen from this figure that the flagellated chambers
are small and more or less rounded, and scattered quite irregu-
larly but abundantly through the thickness of the sponge wall.
Each communicates by several small prosopyles with the
irregular inhalant canal system, and by a single larger
opening with the exhalant canal system. The gelatinous
ground substance of the mesoderm is sparingly developed,
except in the region of the dermal and gastral cortex, and
the inhalant and exhalant canals are very wide and irre-
gular. Numerous small inhalant pores, scattered over the
dermal surface, lead first into small canals in the dermal
cortex; these unite into much larger ones, which lead inward,
and break up in the thickness of the sponge wall. The very
wide exhalant canals open into a perfectly well-defined. gastral
cavity from which the water passes out through a single ter-
minal osculum.

Such a sponge as Leucandra phillipensis forms a
thoroughly typical Leuconoid individual, and we find the type
of canal system above described repeated in a large number of
species with singularly little variation; the flagellated chambers
being irregularly scattered, ovoid or subspherical in shape
(fig. 25), and about 0°] mm. in diameter.

In a few species of Leucandra, however, we find the flagel-
lated chambers much larger, more or less elongated in form,

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 179

aud arranged radially around the exhalant canals, features
which characterise von Lendenfeld’s family Sylleibide (10).

The only instance of this Syllectoid condition which I have
myself observed in the genus Leucandra is in the case of L.
australiensis (4), the anatomy of which is represented in
fig. 17. This is a rather large, solitary species, of the typical
sac-shaped form. The flagellated chambers are irregularly
sac-shaped, and average about 0°3 by 0:1 mm. in size, though
very variable. Their radial arrangement around the exhalant
canals is not nearly so well marked as in some other species to
be mentioned directly ; but it is obvious that the chambers of
any sponge must be arranged more or less radially around these
canals if, as in Leucandra, a large number open directly
into them.

In Leucandra aspera, according to Vosmaer (14), we find
much the same condition, except that the radial arrangement
of the somewhat elongated, sac-shaped chambers around the
exhalant canals is more regular, a condition which is very
closely paralleled by Leucilla uter (vide infra and fig. 21).

Von Lendenfeld’s Polejna telum and Vosmaeria corti-
cata (10), both of which belong to the genus Leucandra as
here understood, likewise afford, according to this author’s
figures, good illustrations of the more regular Sylleibid type
of canal system, with the flagellated chambers somewhat elon-
gated, and arranged radially around the exhalant canals.

This Sylleibid condition, first described by Poléjaeff (8),
appears to be intermediate between the typical Syconoid and
the typical Leuconoid conditions, but approaches more nearly
to the latter than totheformer. Indeed, it would be extremely
difficult to say where the Sylleibid condition ends and the
Leuconoid begins, for even in such a typically Leuconoid form
as L. phillipensis (fig. 16) some of the chambers, especially
towards the outside of the sponge, may be markedly elongated,
and thus approach the Syconoid type; and the same variation
in the form of the chambers is found to an even greater extent
in some species belonging to other genera, as will be seen later
on. This fact, and the fact that the Sylleibid type of canal

180 ARTHUR DENDY.

system is met with in species with very distinct types of
skeletal arrangement (as will be seen subsequently), cause me
to believe that it is of no value for systematic purposes, although
it is of great interest, as probably representing a late stage in
the evolution of the Leuconoid from the Syconoid type of
canal system by branching of the radial chambers and restric-
tion of the collared cells to the branches. (It is also conceiv-
able that the Sylleibid type may have originated from the
Syconoid by folding of the wall of the gastral cavity, but there
is very little evidence in favour of this view; while, in some
species of Sycon and Grantia, we have a certain amount of
evidence, in the arrangement of the skeleton, for believing in
the direct conversion of flagellated chambers into exhalant
canals. That both processes may have taken place together I
do not for a moment deny.)

In some species of Leucandra we find, as in the case of
Synute pulchella, a very strong tendency towards the forma-
tion of massive, irregular colonies, by the more or less com-
plete fusion of a number of Leuconoid individuals. Hence we
find species which, instead of having a single well-defined
central gastral cavity, with a single terminal osculum and a
correspondingly definite sac-shaped external form, as in Leu-
candra phillipensis, exhibit an irregular massive form with
a larger or smaller number of oscula scattered over the surface,
each osculum being the outlet of a wide exhalant canal, which
probably corresponds to the gastral cavity of a single Leu-
conoid individual. Synute pulchella is of great interest
as showing us how such forms have probably arisen. As far
as my experience goes, all such massive colonial species have
in other respects a typical Leuconoid canal system with small
rounded chambers, as in the Australian L. gladiator (4).

Lelapia.

The canal system of this remarkable genus is unfortunately
unknown, but from the fact that Mr. Carter (12), who has
personally studied it, places it amongst the ‘ Leucones,” it
seems probable that it conforms to the Leuconoid type.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 181

Leucyssa.
This genus also appears, from the very scanty information
which we possess with regard to the canal system (5), to con-
form to the ordinary Leuconoid type.

Grantessa (fig. 18).

In this genus we find ourselves returning to a canal system
which is emphatically Syconoid, and agrees closely with that
which we have already described in the genus Grantia. This
will be rendered obvious by a comparison of fig. 18, represent-
ing a portion of a transverse section of Grantessa intusar-
ticulata with fig 9, representing a similar section of Grantia
extusarticulata. In Grantessa sacca, one of the most
beautiful of our Australian sponges, the chambers are very
long and copiously branched (4). All the known species are
solitary Syconoid individuals or branching colonies, never
completely fused into solid, massive forms. In Grantessa
erinaceus we meet with a very striking peculiarity, in the
presence of ingrowths of mesoderm from the gastral cortex,
covered by a siugle layer of flattened epithelium, into the gastral
cavity. These ingrowths form aseries of irregular “‘endogastric
septa” without any spicules. They are present in both the speci-
mens in my possession, and, as Mr. Carter also mentions them
(12), they would seem to be constant in the species. In G.
erinaceus also, the flagellated chambers are very irregular and
much branched, and they communicate with the gastral cavity by
unusually long exhalant canals, which unite together in groups.

The anatomy of Grantessa (Amphoriscus) poculum
has been figured by Poléjaeff (8) and re-investigated by myself.
The canal system of this species agrees closely with that of G.
intusarticulata (fig. 18).

Heteropia.

The only species which I propose to retain in this genus is
Carter’s Aphroceras ramosa (15), which appears to have a
typical radiate Syconoid canal system, complicated only by
the development of astrong dermal cortex resembling that of

the genus Ute.

182 ARTHUR DENDY.

Vosmaeropsis (fig. 19).

In this genus we find (4) at present only three species—V.
macera, V. depressa, and V. Wilsoni,—all of which are in-
teresting with regard to their canal system. ‘The canal system
is never truly radial and Syconoid, but the shape and size of
the flagellated chambers, and in V. macera (fig. 19) the
arrangement also, clearly indicate a condition intermediate be-
tween the typical Syconoid and the typical Leuconoid plan ; in
other words, a Sylleibid condition. In V. macera, the anatomy
of which is represented in fig. 19, the chambers are thimble-
shaped, and mostly widely separated from the central gastral
cavity ; they communicate with this by wide exhalant canals,
into each of which a number of chambers discharge their con-
tents. Each chamber has, as usual, whether it be large or
small, a number of small inhalant prosopyles, and a single
much larger exhalant aperture, guarded by a well-developed
diaphragm (fig. 31). Those chambers which lie next to the
dermal surface still exhibit a more or less radial arrangement
in relation to the central gastral cavity.

Vosmaeropsis depressa is, unfortunately, known only
from a single specimen, so that it is impossible to tell how far
the peculiarities of its canal system are constant. There is
no single wide gastral cavity, but several large branching
exhalant canals converge to a single small osculum situate
near the middle of the upper surface of the cushion-shaped
sponge. The inhalant canal system is quite irregular, com-
mencing in wide lacunar spaces beneath the thin pore-bearing
dermal cortex. ‘The flagellated chambers are irregularly but
thickly scattered through the thickness of the sponge, with no
trace of radial arrangement around a central gastral cavity.
They are, however, more or less sac-shaped or thimble-shaped,
measuring about 0°2 by 0°09 mm. This sponge, from the fact
that it possesses but a single osculum, probably corresponds
to a single individual, but it is interesting to note how the
central gastral cavity has become indistinguishably merged
into the branching system of wide exhalant canals.

Vosmaeropsis Wilsoniisa large species found abundantly

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 183

in the neighbourhood of Port Phillip. The sponge is colonial,
and consists of short, thick, subcylindrical or truncatedly conical
individuals, united together basally in larger or smaller agglo-
merations. Hach individual has a single well-defined wide
gastral cavity, with a single terminal osculum, protected by a
remarkably distinct membranous diaphragm situated a short
distance within its margin. The wall of the sponge surrounding
the gastral cavity is thick, and there is a dense thick cortex on
both gastral and dermal surfaces. Between the dermal and the
gastral cortex lie the flagellated chambers, thickly but irregu-
larly scattered. These chambers vary to a remarkable extent
both in shape and size, from approximately spherical ones of
about 0°072 mm. in diameter, to elongatedly sac-shaped ones
of as much as 0°37 by 0:13 mm. It is right to state that these
measurements were taken from different specimens, but the
species is so well characterised that it would be difficult to
make a mistake in identification, and we also find considerable
variation in the chambers, even in the same section. Most
remarkable, however, is the inhalant cortical canal system, a
portion of which, from an osmic acid preparation, is represented
in fig. 23. The inhalant pores, thickly scattered over the
surface of the sponge, lead each into a separate narrow canal
lined by a flattened epithelium. These canals unite together
into larger and larger canals as they penetrate the dermal
cortex, and also form frequent anastomoses by cross-branches.
This canal system conducts the water to the chamber-bearing
layer of the sponge wall and distributes it to the chambers;
from these it is collected again by the exhalant canals, which,
uniting into tolerably large trunks, penetrate the gastral cortex
and open into the central gastral cavity. I have never seen
the inhalant cortical canal system so well illustrated as in this
sponge, doubtless because I took the precaution to treat a
piece of a living specimen with osmic acid. It finds a close
parallel in the corresponding cortical canal system of the
Syconoid Grantiopsis cylindrica, and may probably be
taken as representing the typical condition for Heterocela
having a very strongly developed dermal cortex.

184 ARTHUR DENDY.

Heteropegma (fig. 20).

In this genus we again go back to a Syconoid arrangement
of the canal system, though not a very typical one. In H.
nodus-gordii, the anatomy of which has been admirably
illustrated by Poléjaeff (8), and of which I also venture to give
a drawing (fig. 20) based upon personal examination, the
sponge forms irregular agglomerations of small Syconoid
individuals, each with a single osculum and a central gastral
cavity. The flagellated chambers have extremely thin and
delicate walls, and branch in an extraordinarily copious and
irregular manner (fig. 20). The gastral cortex is also extremely
thin, but the dermal cortex is very strongly developed, and pene-
trated by irregular canals which lead from the inhalant pores .
on the surface of the sponge into the quite irregular lacunar
system of spaces (‘‘inter-canals”) between the chambers.

In Heteropegma latitubulata, which is found off
the south coast of the continent of Australia (H. nodus-
gordii being found off the north), we meet with an iden-
tical canal system and external form (4). The peculiarities
of the canal system in these, the only known species of the
genus, are brought about by the copious and irregular branching
of the chambers and the extremely slight development of the
mesoderm everywhere except in the dermal cortex.

Amphoriscus.

In this genus the canal system is more typically Syconoid,
such as we find in any of the Syconoid genera, like Grantia,
with a dermal cortex. As I have never myself met with the
genus, I may refer the reader to the account of Amphoriscus
(Sycilla) cyathiscus and A. (Sycilla) cylindrus given by
Haeckel (5).

Syculmis.

Here, again, the canal system appears to differ in no par-
ticular from the typical corticate Syconoid plan, like that of
Grantia. I must again refer the reader to Haeckel’s great
work (5) for an account of the only known species (Syculmis
syuapta).

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 185

Leucilla (figs. 21, 22).

This genus includes species which never possess a truly
radiate Syconoid type of canal system, but either a thoroughly
typical Leuconoid arrangement with small, more or less
rounded chambers, as in Leucilla australiensis (fig. 22),
L. saccharata, and L. prolifera (4), or an arrangement like
that of Leucilla uter (fig. 21), which indicates by the
elongated form of the chambers and their radial arrangement
around the exhalant canals a Sylleiboid condition, intermediate
between the Syconoid and Leuconoid types.

The genus Leucilla, in short, occupies a position in the
family Amphoriscide exactly analogous to that occupied by
Leucandra amongst the Grantide, aud we find in both
genera the same variations in canal system. Thus the canal
system of Leucilla uter! appears to be almost exactly paral-
leled by that of Leucandra aspera (14), while that of
Leucilla australiensis and the other species with small,
rounded, and irregularly arranged chambers, is paralleled by
the similar arrangement found in numerous species of Leu-
candra, as will be evident on comparing figures 16 and 22.

Most remarkable is the canal system of Poléjaeff’s Leucetta
vera (8), which appears to belong to the genus Leucilla as
now constituted. In this species, according to Poléjaeff (8),
the chambers of the inner half of the sponge wall are rounded
and irregularly scattered, while those of the outer half are elon-
gated and radially arranged, and it thus affords a noteworthy
commentary on the value of the canal system for purposes of
classification.

A similar variation in the form of the chambers, only on a
much smaller scale, is to be found in Leucilla australiensis
(fig. 22).

Another interesting variation in the canal system is afforded
by Leucilla cucumis (Haeckel’s Leucandra cucumis, 5),

1 For further information as to this species vide Poléjaeff (8). My draw-

ing is made from my own preparations of a portion of the type specimen from
the British Museum.

186 ARTHUR DENDY.

In this species we find a series of very distinct subdermal
cavities supported by a special skeleton, for further informa-
tion with regard to which the reader is referred to the sections
of the present paper dealing with the skeleton and classifica-
tion of the Heteroceela.

We find in some species of this genus also the same tendency
to form colonies, by complete fusion of Leuconoid individuals,
as we met with in Leucandra. An admirable example of
this is afforded by Leucilla (Teichonella) prolifera, of
which I gave an illustration in a previous memoir (9), showing
the oscula arranged side by side in rows. .

SUMMARY.

Thus we find that the canal system varies considerably in what
appear to be closely related genera of Heteroceela Calcarea; and
that, if my view of the relationships of the genera be correct, the
Leuconoid type has been independently evolved from the Syco-
noid type, along the lines indicated above, no less than three
times. The Leuconoid type cannot, however, be produced until
the corticate Syconoid type has been arrived at; and when this
condition has been reached the conversion of the originally
long and radially disposed chambers into short, rounded, and
irregularly arranged ones seems such a simple matter that it
may well have taken place again and again. The variation in
shape and size of the chambers, even in the same species, may,
as I have already shown, be very great. The branching of the
radial chambers in Syconoid forms is of such common occur-
rence in the most diverse genera as to excite no surprise wher-
ever we meet with it; and the shortening of the chambers and
corresponding elongation of the exhalant canals, due simply
to shifting of the limits of the lining of collared cells, has been
repeatedly observed in various genera.

We may sum up our observations on the canal system of the
Heteroccela by indicating again the various stages which appear
to have been passed through in the gradual evolution of the
most complex Leuconoid from the most simple Syconoid type.

Stace A (Sycetta stage).—The flagellated chambers are

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 187

perfectly straight, unbranched, and radially arranged. They
do not touch one another at all, and there is no trace of a
dermal cortex, hence there is no enclosed system of inhalant
canals, but the water circulates freely between the chambers
without any interruption. Sycetta appears to be the only
genus in which this most simple condition is retained.

Stace B (Sycon stage).—This stage differs from the fore-
going only in the more or less complete fusion of the walls of
adjacent radial chambers wherever they come in contact. This
results in the formation of more or less well-defined inhalant
<‘inter-canals.”” The cllaambers may also branch. In this
stage we find most species of the genus Sycon, and we may
perhaps also include Sycantha. Those species of Sycon in
which a thin pore-bearing membrane covers over the ends of
the “inter-canals,” as described above, are intermediate be-
tween stages B and C,

Stace C (Grantia stage).—The chambers are still elon-
gated and radial, but their distal ends and the ends of the
“‘inter-canals ” between them are covered over by a dermal
cortex, which contains true inhalant pores, and sometimes a
complicated cortical inhalant canal system. In this stage we
find Grantia, Grantiopsis, Ute, Synute, Utella, Ana-
mixilla, Sycyssa, Grantessa, Heteropia, Heteropegma,
Amphoriscus, and Syculmis.

Stace D (Sylleibid stage).—The chambers are no longer
arranged radially around the central gastral cavity, but are
still more or less elongated, aud arranged radially around the
usually radial exhalant canals. It was this condition, first
described by Poléjaeff (8), which gave rise to von Lendenfeld’s
Sylleibide (10,11). It is not, however, characteristic of any
particular genus, much less of any particular family, but is
found in a few isolated species, such as Leucilla uter,
Leucandra aspera, and Vosmaeropsis macera.

Stace E (Leucandra stage).—The chambers are small,
more or less spherical, and irregularly scattered through the
sponge wall; and the inhalant and exhalant canal systems are
correspondingly developed. We find this condition in most

188 ARTHUR DENDY.

species of Leucandra and Leucilla, and presumably also in
Lelapia and Leucyssa.

No one of these five stages is very sharply marked off from the
stage below it, and the five appear to me to indicate a process
of evolution which has actually taken place. Von Lendenfeld
(10) attributes to the Leuconoid type of canal system an inde-
pendent origin from the Homoceele type, through his very pro-
blematical “ Leucopside,” without passing through a radiate
Syconoid stage at all, although he admits the derivation of
the Sylleibid type from the latter. With this view I cannot
at all agree. Considering the canal system alone, we have
cogent reasons for opposing it, and when we come to discuss
the skeleton we shall find others.

III. Tue SKELETON oF THE CALCAREA HETEROCG@LA.

The Spicules.—So much has been written about the
spicules of calcareous sponges, and their variations in minute
details of shape are of so little interest from a morphological
point of view, that I propose to say very little about them in
this place, and only to recapitulate those facts which it is
necessary to consider in discussing the arrangement of the
skeleton.

In calcareous sponges, whether Homocele or Heterocele,
three principal types of spicules are met with:

(a) Triradiate,
with three rays diverging from a common centre and lying
typically in one plane, though frequently curved more or less
out of that plane.

(6) Quadriradiate,

resembling the triradiate but with an additional ray, known as
the apical ray, coming off from the ceutre in a plane at right
angles to the plane of the other three (facial) rays.

In both triradiate and quadriradiate spicules we can dis-
tinguish three chief varieties :—(1) Regular, the three facial

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 189

rays being all alike, with the angles between them equal. (2)
Sagittal, when two of the facial rays or two of the angles form
a pair differing in some respects from the remaining ray or
angle. In this case the paired rays are termed oral, the odd
ray is termed basal, and the odd angle between the oral rays
is also termedoral. (3) Irregular, when the three facial rays
conform to neither of the above plans.

(c) Oxeote,

or uniaxial spicules (oxea), in which there is only a single
axis. These spicules vary greatly in details of shape, from
symmetrically fusiform, with two sharply pointed ends, to nail-
shaped, with one end biunt and swollen.

As in the Homocela, so in the Heteroccela, the mere
presence or absence of one or other of these three types of
spicule is of very slight—in my opinion, of not more than
specific—value for purposes of classification. With the arrange-
ment of the spicules I believe the case to be totally different,
and I find, in the strncture of the skeleton as a whole, charac-
ters of family value to the systematist.

The Arrangement of the Spicules.—Here, as in the
case of the canal system, we shall find it most convenient to
take the various genera one by one and deal with them
separately. We shall find that, up to the Grantia stage, the
arrangement of the skeleton follows, and appears to be to a
large extent controlled by, that of the canal system. But on
reaching the Grantia stage the development of a strong
dermal cortex introduces new possibilities with regard to the
skeleton, which commences to vary independently of the canal
system, and branches off along three lines corresponding to the
families Grantide, Heteropide, and Amphoriscide. The genus
Leucascus appears to occupy an isolated position.

Leucascus (fig. 1).
In this genus (fig. 1) the arrangement of the skeleton is
extremely simple, and exhibits no trace of that radial sym-
metry which is so characteristic of the Heteroccela. It

190 ARTHUR DENDY.

- resembles, on the other hand, the arrangement of the skeleton
in such simple Homocele sponges as Leucosolenia proto-
genes, excepting that there is a true dermal skeleton in the
pore-bearing dermal membrane. In both species of Leu-
cascus (4) the skeleton consists of small, regular triradiates,
irregularly scattered in the walls of the elongated chambers
and exhalant canals, and in the dermal membrane. In Leu-
cascus simplex these are the only spicules present, but in
L. clavatus we find in addition some large, club-shaped,
uniaxial or oxeote spicules, partly projecting from the dermal
surface.

Sycetta.

Here, in accordance with the arrangement of the canal
system, we can distinguish between two main parts of the
skeleton,—(1) the gastral skeleton, supporting the wall of the
gastral cavity; and (2) the tubar skeleton, supporting the walls
ofthe radial chambers. The gastral skeleton consists of tri-
radiate or quadriradiate spicules, whose three facial rays lie in
the thickness of the wall of the gastral cavity, while the apical
ray, if one happens to be developed, projects into the cavity.
These spicules may be sagittal, asin Sycetta (Sycaltis) coni-
fera (5), and then the basal ray is found to be directed down-
wards, away from the osculum and towards the base of the
sponge, a position which is so constant in this and other genera
as to have given rise to the term “basal ray.” The tubar skeleton
consists exclusively of triradiate spicules, which lie in the thick-
ness of the chamber walls, and which always have the basal ray
directed along the length of the chamber and away from the
gastral cavity. Hence the oral or paired rays are spread out in
a direction at right angles to the length of the chamber, and as
several spicules generally lie at the same level, the tubar
skeleton forms a series of more or less definite joints or rings,
and is hence said to be articulate. This articulate arrange-
ment, however, which prevails in most genera with a Syconoid
type of canal system, is usually very irregular.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 191

Sycon (figs. 2—8).

Here, again, we meet with distinct gastral and tubar skeletons,
as shown in fig. 2, and the general plan of the skeleton is the
same as in Sycetta. Quadriradiates are usually, if not invari-
ably, present in the gastral skeleton, the apical ray projecting
freely into the gastral cavity (figs. 2, 4, 7), and doubtless
serving as a protection against the ingress of parasites. The
gastral skeleton may be very strongly developed so as to form
a thick cortex, which may be continued inwards along the
exhalant canals, as in Sycon boomerang (fig. 7). Where
this is the case—but it is very rare—it seems to me to indicate a
folding of the wall of the gastral cavity. The tubar skeleton is
always “articulate,” and the number of “joints” which it
exhibits depends upon the length of the chambers. Usually
the spicules composing it are more or less strongly sagittal and
triradiate, but occasionally apical rays may be developed, as in
S. boomerang (fig. 7). The position of the spicules is always
the same as in Sycetta, with the basal ray pointing away from
the gastral cavity ; where an apical ray is developed it projects
freely into the cavity of the chamber. Usually a special set
of spicules, known as “ subgastral sagittal triradiates” are
developed, as is well shown inSycon Carteri (fig. 2). These
have the oral rays extended very widely in the outer part of
the wall of the gastral cavity, and thus forming a part of the
gastral cortex, while the basal ray is usually very long, and
points towards the distal end of-the chamber in whose wall it
les. These spicules, or at any rate their basal rays, form the
first “joint” of the articulate tubar skeleton, which is com-
monly a good deal longer than any of the succeeding joints
(fig. 2). The subgastral sagittal triradiates may, however, be
almost, if not quite, indistinguishable from the ordinary tubar
triradiates of the sponge. Usually the three rays of the tubar
triradiates do not lie all in one plane, but the oral rays are
curved or inclined towards one another, so as to partially
embrace the chamber in the thickness of whose wall they lie
(fig. 5).

VOL, 35, PART 2.——-NEW SER, P

192 ARTHUR DENDY.

In this genus we always find, at the end of each radial
chamber, a more or less dense tuft of oxeote spicules (figs. 2,
8, 6—8). The shape of these spicules varies greatly, and
influences in a remarkable degree the character of the surface
of the sponge. They may be enormously elongated and pro-
ject far beyond the surface, which thus becomes covered with
a coating of long silky hair, as in Sycon Ramsayi; or their
outer ends may be swollen and nail-shaped and project but
very slightly, so as to form a dense crust, as in Some specimens
of Sycon gelatinosum (fig.6). All conditions intermediate
between these two may also be met with. These tufts of uni-
axial spicules, which are extremely characteristic of the genus
Sycon, doubtless serve to protect the surface of the sponge
generally, and also to filter the water before it passes into the
“inter-canals.” The entrances to the “ inter-canals,” between
the tufts of oxeotes, may also be protected by the basal rays
of the more distal tubar triradiates, which may curve outwards
from the chamber wall, as in Sycon ensiferum (4).

In Sycon boomerang, as I have already had occasion to
point out, we meet with the first indication of a true dermal
skeleton distinct from that of the chambers. It consists of a
few scattered triradiate and oxeote spicules lying in the thin,
pore-bearing membrane, which stretches across and protects
the entrances to the intercanals (fig. 8).

In this genus we frequently meet with a more or less specially
developed “ oscular skeleton,” which when present, either in
this or any other geuus, always consists of a fringe of oxeote
spicules, projecting more or less markedly around the osculum.
Usually the fringe is vertical, or inclined only at a slight angle
to the long axis of the gastral cavity ; but occasionally a second,
almost horizontal, fringe is developed, which projects almost
at right angles all around the base of the first one. A beau-
tiful illustration of such a horizontal fringe is given by
Haeckel in the case of his Sycarium elegans (8, pl. lviii),
and I have also met with it in a variety of Sycon gela-
tinosum from Port Jackson. The oscular skeleton is,
however, a very variable structure, and of very little import-

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 193

ance from the point of view either of the morphologist or the
systematist.

Sycantha.

The skeleton of this genus appears to conform exactly to the
normal Sycon type, which appears to favour my view that
the canal system is but a slight modification of the Sycon

type.
Grantia (figs. 9, 10).

In Grantia we find the skeleton built upon the same
essential plan as in Sycon, but there is, in addition, a well-
developed dermal skeleton lying in the dermal cortex, which
covers over the ends of the radial chambers and inter-canals,
while we no longer find each chamber surmounted at its distal
extremity by a tuft of uniaxial oxea. The gastral skeleton
and the articulate tubar skeleton are precisely similar to what
we found in Sycon, as will be at once evident on referring
to fig. 9, representing the anatomy of Grantia extus-
articulata.

The skeleton of the dermal cortex in this genus consists
principally of triradiate spicules lying parallel to the dermal
surface. These spicules may be sagittal, with the basal ray
directed, as in the gastral cortex, towards the base of the
sponge. A good example of this arrangement is seen in
Grantia labyrinthica (9).

In addition to the triradiate spicules we also find in the
dermal cortex of Grantia, in most if not all species, a number
of oxeote spicules placed at right angles to the dermal surface.
These may either be very small and numerous, as in Grantia
extusarticulata (fig. 9), or they may be large and fewer, as
in G. Vosmaeri (fig. 10). In the former case they form a
kind of crust, and are spoken of by some writers as ‘‘ mortar-
spicules:” their inner ends only penetrate the dermal cortex
for a short distance. In the latter case their inner ends may
penetrate through nearly the entire thickness of the sponge
wall, In no case in this genus do they form tufts at the ends

194 ARTHUR DENDY. °

of the radial chambers, but they are always arranged entirely
without regard to these.

In Grantia labyrinthica (9) we find minute surface oxea
in the gastral as well as in the dermal cortex. This is a very
unusual occurrence, but is paralleled in Ute Spenceri (4).

In Grantia Vosmaeri (fig. 10), where we have a marked
shortening of the radial chambers and a corresponding elonga-
tion of the exhalant canals, together with a strong develop-
ment of the mesoderm surrounding them, we find the skeleton
in the inner portion of the sponge wall losing its regular
radially symmetrical character, and becoming diffuse and
scattered. This is an illustration of a general law, that the
skeleton of the chambers varies with the chambers themselves,
which will be found to hold good throughout the group, with
few exceptions. As the chambers lose their radial symmetry
and regularity of form and arrangement, so the skeleton loses
its radially symmetrical articulate character, and becomes
diffuse and scattered. This will be seen to follow as a natural
consequence, if we remember that the position of the spicules
is to a large extent determined by the position of the layer of
mesoderm (in the chamber wall) in which they are developed.
Hence it is that irregularity in the skeleton generally accom-
panies thickening of the mesoderm, for when this is effected
the spicules are free to take up a variety of positions which,
while embedded in a very thin layer of mesoderm, they are
unable to assume.

Grantiopsis (fig. 11).

In Grantiopsis cylindrica (4) we meet with a slight
modification of the Grantia type of skeleton, as shown in
fig. 11. The dermal cortex is enormously developed, but its
skeleton still consists of triradiate and oxeote spicules arranged
asin Grantia. In the thin gastral cortex quadriradiates are
present as usual, but of somewhat peculiar shape, with small
facial and enormous apical rays. The subgastral sagittal
triradiates of the ordinary Sycon and Grantia types are here
replaced by subgastral sagittal quadriradiates. The position of

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 195

these spicules agrees with that of the subgastral sagittal
triradiates of Sycon and Grantia, but an apical ray is
developed which projects into the gastral cavity, almost in a
line with the centrifugally directed basal ray. Hence these
spicules are not homologous with the subgastral quadriradiates
of some Amphoriscide, their position being different.

The articulate tubar skeleton of Grantiopsis cylindrica
is very remarkable, owing to the peculiar form of the spicules
of which it is composed. These spicules are the most extreme
modifications of the sagittal triradiate type which I have ever
seen or heard of. They consist almost entirely of the strangely
developed centrifugally directed basal ray, which is straight,
fusiform, gradually sharp-pointed at the distal end, and at the
proximal end provided with a pair of minute, widely divergent,
conical teeth, which represent the extremely reduced oral rays.
The basal ray measures about 0°3 by 0:008 mm., while the oral
rays are only about 0:003 mm. long. The entire tubar
skeleton is made up of these spicules and of the basal rays of
the subgastral quadriradiates, arranged usually in single series,
but with overlapping ends, each series comprising only about
three spicules (fig. 11).

Ute (figs. 12—14).

In this genus we find the skeleton arranged as in Grantia,
only with the addition of a number of very large oxeote
spicules disposed longitudinally in the dermal cortex. The
gastral skeleton is exactly like that of Grantia and Sycon, :
and so also is the tubar skeleton in most species. In Ute
argentea, however, there is, as Poléjaeff points out (8), a
tubar skeleton composed only of a single joint, i.e. of the rays
of the subgastral sagittal triradiates. To this tubar skeleton
the term “inarticulate” has been applied, but this appears to
me to be a mistake, for it is very different from the so-called
“inarticulate” tubar skeleton of the Heteropide and Am-
phoriscidee. As I shall show later on, however, the distinction
between “articulate” and “inarticulate” tubar skeletons is
not altogether satisfactory.

196 ARTHUR DENDY.

The development of the dermal skeleton varies much in
different species as regards its extent, but always follows the
same pattern. The large longitudinally arranged oxea always
form the most conspicuous feature, and frequently give to the
surface of the sponge a very characteristic striated appearance.
Nearly always, however, triradiates and minute surface oxea
are also present, arranged asin Grantia. In Ute spiculosa
and U. Spenceri (4) the triradiates are very numerous
indeed, while in U. glabra (5), U. argentea, and U.
syconoides (4) they are very scarce.

The presence of the longitudinally disposed oxeote spicules
in the dermal cortex, although often giving the sponge a very
characteristic appearance, is by no means peculiar to the genus
Ute, for the same character is found in Leucandra alci-
cornis (5) and in two species of Homocela Simplicia, viz.
Leucosolenia asconoides (4, 12) and L. uteoides (16);
and it is also characteristic of the genus Heteropia (4).

In Ute spiculosa (4), where the mesoderm is very strongly
developed, we find, in accordance with the rule laid down
above, that the skeleton of the chamber layer is dense and
irregular, though showing traces of the normal articulate
arrangement in the usually centrifugal direction of the basal
rays of the triradiates.

In Ute Spenceri (4) we find one or two slight pecu-
liarities in the skeleton. Thus we find, as in Grantia
labyrinthica, a number of minute oxeote spicules on the
. gastral as well as on the dermal surface ; and we find also that
the exhalant canals are protected by quadriradiates of special
form. The tubar skeleton is distinctly articulate, though
becoming confused as it approaches the dermal cortex.

Synute (fig. 15).

In this sub-genus the dermal cortex and its contained
spicules are enormously developed, so as to bind together all
the individuals of a branching colony into one continuous
whole (fig. 15). Here, as in other cases where the dermal
cortex is very strongly developed, the cortical triradiates lose

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 197

their parallelism with the dermal surface and become irregu-
larly scattered. In other respects the arrangement of the
skeleton follows the ordinary Ute plan.

Utella.

In this genus, of which the type species is Haeckel’s
Sycandra hystrix (5), we find a very remarkable modification
of the Grantia type of skeleton, for a layer of longitudinally
disposed oxeote spicules is developed in the gastral cortex,
instead of, as in Ute, in the dermal one. For a descrip-
tion and illustrations of Utella hystrix I must refer the
reader to Haeckel’s great work (5).

Anamixilla,

Here, again, the skeleton is very peculiar, as will be seen at
once on referring to Poléjaeff’s drawing thereof (8). The
flagellated chambers are elongated and radially arranged. The
gastral and dermal cortex, each with its skeleton, is developed
as in Grantia. There is, however, no articulate tubar
skeleton, although it is important to notice that subgastral
sagittal triradiates are still present, as in Ute argentea.
The most remarkable feature of this sponge is the presence, in
the chamber layer of the wall, of a number of large triradiate
spicules, out of all proportion to the size of the chambers and
arranged without regard to the direction of these, but generally
lying across them and more or less parallel to the dermal
surface. It looks as if the chamber layer had first lost the
greater part of its own skeleton, all except the subgastral
sagittal triradiates, and had then been invaded by large spicules
from the dermal cortex.

Sycyssa.

In this very remarkable genus, for our knowledge of which we
are entirely indebted to Haeckel (5), the canal system appears
to be arranged upon the regular Grantia plan, but we are
told that the skeleton is made up solely of oxeote spicules,
there being no radiate spicules present of any kind. The
skeleton of the gastral cortex consists of (1) a gastral layer of

198 ARTHUR DENDY.

irregularly arranged bundles of very slender oxea, and (2) a
subgastral layer composed of regularly arranged, parallel,
longitudinal bundles of large, stout oxea: this is comparable
to the similar layerin Utella. The tubar or chamber skeleton
is composed of the inner ends of enormous radially arranged
oxeote spicules, comparable to those found in Grantia
V osmaeri, but larger and apparently more regularly arranged.
The outer ends of these spicules project far beyond the dermal
surface. The skeleton of the dermal cortex is composed of
(1) a layer of very slender oxea, arranged irregularly but
parallel to the dermal surface, and (2) a layer of very slender
oxea, arranged radially and forming a kind of pile over the
outer surface, as in so many Heteroceele sponges.

This genus, of which the only known species is Sycyssa
Huxleyi, from the Adriatic, must be regarded as possessing a
very aberrant type of skeleton, and occupying a very isolated
position in the group.

Leucandra (figs. 16, 17).

To understand the structure of the skeleton in this genus
we must return to that of Grantia, of which it is evidently a
modification. We saw how in Grantia Vosmaeri (fig. 10)—
where the exhalant canals have begun to elongate and the
chambers to shorten, and where the mesoderm is very strongly
developed in the inner half of the sponge wall—the skeleton
has begun to lose its regular articulate character. The same
thing is carried to a much greater extent in Leucandra,
where the canal system has lost its symmetry and the
skeleton has followed suit (figs. 16, 17). The skeleton of the
gastral and dermal cortex, however, retains the same structure
as in Grantia; it is only the skeleton of the chamber layer
which is affected by the changes in the canal system. Hence
we find that in a typical Leucandra (fig. 16) the skeleton of
the gastral cortex is made up principally of quadriradiate
spicules whose apical rays project into the gastral cavity,
while the skeleton of the dermal cortex is made up of tri-
radiates, lying parallel to the dermal surface, and longer or '

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 199

shorter oxeote spicules usually projecting more or less at right
angles therefrom. The skeleton of the chamber layer, on the
other hand, consists of irregularly scattered triradiates or
quadriradiates.

In some species, e.g. L.phillipensis and L.australiensis,
the quadriradiates of the gastral cortex are continued inwards
along the larger exhalant canals, which possibly indicates a
folding of the wall of the gastral cavity, though I do not think
it necessarily does so.

In some species, e.g. L. phillipensis, we find many of the
triradiates of the chamber layer retaining their primitive
sagittal character, and with the basal ray pointing towards
the dermal surface. This may be a reminiscence of the time
when they formed part of an articulate tubar skeleton. Sub-
gastral sagittal triradiates may also still be present (figs.
1G 17).

Lelapia.

In this genus of a single species the canal system is un-
known, and the skeleton is peculiar. The skeleton of the
gastral cortex is composed of ordinary radiates; while that
of the dermal cortex is composed of triradiates, quadri-
radiates, and minute oxea. The skeleton of the chamber layer
is composed of large, longitudinally arranged oxea, crossed at
right angles by bundles of tuning-fork shaped triradiates whose
basal rays point towards the dermal surface. For further
particulars I must refer the reader to Mr. Carter’s description
of the sponge (12), as I have never had an opportunity of exa-
mining it myself.

I should, perhaps, mention that Mr. Carter describes the
occurrence of small quadriradiate spicules “on the surface,”
presumably in the dermal cortex. ‘Triradiate spicules are very
apt to develop an apical ray, often so small as to be of no con-
ceivable use, and apparently indicating a mere “sport,” if one
may use the term. ‘Therefore, unless the feature could be
shown to be constant, well established, and characteristic, I
should not be inclined to attribute any great value to it. It

200 ARTHUR DENDY.

certainly appears to be a difficulty in the way of utilising the
presence of subdermal quadriradiate spicules as a family cha-
racteristic in the Amphoriscide, but in the latter case the apical
rays are so constant, and form, in most cases, such an im-
portant part of the skeleton, that we must treat the question
from a different standpoint.

Leucyssa.

In the skeleton of this genus we meet with a parallel case
to that of Sycyssa, only associated this time (presumably)
with a Leuconoid canal system.

The genus is very imperfectly known, and what we do know
about it we owe chiefly to Haeckel (5). Two of the three
species (L. spongilla and L. cretacea) are described from
single specimens, and a third (L. incrustans) is extremely
rare. In all, the skeleton appears to be quite irregular, con-
sisting of a felt-work of oxeote spicules. In L.spongilla the
spicules are spindle-shaped, straight, and smooth. In L.
cretacea they are swollen at one end and pierced like a
sewing-needle; and it is interesting to note that similar spi-
cules are found in Leucandra ochotensis (5). In Leu-
cyssa incrustans (Carter’s Trichogypsia villosa), a
European species which, thanks to the kindness of Mr. Carter,
I have had the opportunity of partially examining (in the dry
state), the skeleton consists of a very dense, irregular felt-work
of spinose oxea.

Grantessa (fig. 18).

In this genus we have an entirely new element introduced
into the composition of the skeleton, in the form of sub-
dermal sagittal triradiate spicules. The existence of
these spicules in certain species has long been recognised, but
their importance from a systematic point of view appears to
me to have been under-rated. They never appear, so far as my
experience goes, in sponges without a dermal cortex, and they
seem to be first introduced as additions to the ordinary
Grautia type of skeleton. Thusin Grantessa sacca (4, 11)
we find all the parts of a typical Grantia skeleton present,

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 201]

including the skeleton of the dermal and gastral cortex and a
normal articulate tubar skeleton; but in addition to these
parts we find also a well-developed layer of sagittal triradiates,
whose oral rays are extended in the dermal cortex parallel to
the surface, while the elongated basal ray projects inwards
for some little distance into the chamber layer. The position
of these spicules is thus exactly the reverse of that of an
ordinary tubar triradiate, the basal rays pointing in exactly
Opposite directions in the two cases.

In Grantessa intusarticulata (4), the anatomy of which
is represented in fig. 18, we find a precisely similar condition,
but the chambers are shorter, and the number of joints of the
articulate tubar skeleton correspondingly fewer. In_ this
species, also, it is highly interesting to notice that in the
youngest portions of the sponge, nearest the osculum, where
the chambers are shortest, the skeleton of the chamber layer
consists solely of the basal rays of the subdermal and sub-
gastral sagittal triradiates, so that we have here a typical
example of the so-called “inarticulate” tubar skeleton. In
other words, we find the two types of tubar skeleton, articu-
late and inarticulate, existing side by side in the same sponge.
In Grantessa poculum, an excellent figure of which, under
the name Amphoriscus poculum, is given by Poléjaeff (8),
the skeleton is almost exclusively “inarticulate,” but I find
from my own examination of the sponge that the inner
triradiates are not always strictly subgastral, but may be
situate some little distance beneath the gastral cortex.

The dermal and gastral skeleton in this genus agrees so
closely with that of Grantia as to require no further com-
ment (compare figs. 9 and 18), unless we specially mention the
fact that in some species of Grantessa the oxeote spicules of
the dermal surface are collected into tufts, which, however,
bear no relation tothe radial chambers. This character, which
is well shown in Grantessa sacca, first gave origin to von
Lendenfeld’s genus Grantessa (11). I do not believe it one
of great systematic value; and I do not hesitate to include
in the same genus species like G. intusarticulata and G.

202 ARTHUR DENDY.

poculum, in which the oxeote spicules are not collected into
tufts.

Heteropia.

This genus bears precisely the same relation to the genus
Grantessa as Ute does to Grantia, the skeleton being of
the ordinary “‘inarticulate’”? Grantessa type, but with the
addition of large longitudinally arranged oxea in the dermal
cortex. For further details I must refer the reader to Mr.
Carter’s description of Heteropia(Aphroceras) ramosa(l15),

Vosmaeropsis (fig. 19).

All three species included in this genus (4) exhibit, like
Grantessa and Heteropia, a well-developed layer of sub-
dermal sagittal triradiates; but as the other parts of the
skeleton vary somewhat in the three cases, it will be advisable
to consider them separately.

Vosmaeropsis macera, the anatomy of which is repre-
sented in fig. 19, shows least deviation from the ordinary
Grantessa type. There are, however, no quadriradiates in
the gastral cortex, but I do not consider this a very important
character. The oxeote spicules of the dermal cortex are also
very variable in their development, and may be almost, if not
quite, absent; but this, again, is not an important character.
Otherwise, the skeleton agrees closely with that of an ordinary
Grautessa in which no distinct articulate tubar skeleton is
developed.

In Vosmaeropsis depressa (4) the skeleton 1s modified, in
accordance with the massive form of the sponge and the
arrangement of the canal system. The bulk of the skeleton is
made up of fairly large, subregular, or slightly sagittal tri-
radiates, scattered without definite order throughout the thick-
ness of the sponge, many having one slightly longer ray
pointing towards the dermal surface. Beneath the dermal
surface, but apparently only on the upper surface of the
cushion-shaped sponge, is a distinct layer of subdermal
sagittal triradiates, with inwardly directed basal rays. The

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 203

dermal skeleton is made up principally of subregular tri-
radiates of various sizes, placed horizontally, but with no
definite arrangement; amongst these very minute slender
oxea are scattered, especially numerous around the osculum.!
Around the main exhalant canals is a layer of small sagittal
triradiates, forming what must probably be regarded as the
gastral skeleton. No quadriradiates are present, and I have
not detected any special subgastral sagittal triradiates. The
skeleton of this sponge bears much the same relation to that
of Grantessa as that of Leucandra does to that of
Grantia.

Vosmaeropsis Wilsoni is remarkable for the enormous
development of the cortex and its contained skeleton on both
gastral and dermal surfaces. The skeleton of the chamber
layer is like that of Vosmaeropsis macera (fig. 19), but the
sagittal triradiates are of unusually large size. A _ well-
developed oscular skeleton, in the shape of a fringe of oxeote
spicules, is also present, but this is not a feature upon which
much stress need be laid in any case.

Heteropegma (fig. 20).

We have seen how, in Grantessa, the very characteristic
skeleton is derived from the ordinary Grantia type by the
mere addition of a layer of subdermal sagittal triradiate spicules.
The skeleton of Heteropegma is also derivable from the
Grantia type by an analogous change, only the subdermal
spicules are quadriradiates and not triradiates. Moreover it is
important to observe that the subdermal quadriradiates of
Heteropegma and other Amphoriscide are not modifications
of subdermal sagittal triradiates, such as are found in the
Heteropide. If they were so, we should expect to find the
basal ray still pointing inwards towards the gastral cavity, and
the additional apical ray lying in a plane more or less parallel
with the dermal surface. As a matter of fact, however, the
three facial rays of the subdermal quadriradiates lie parallel

’ Oxeote spicules are sometimes developed around the osculum when they
cannot be found anywhere else in the sponge.

204 ARTHUR DENDY.

with the dermal surface, and it is the apical ray which is
directed inwards through the chamber layer. Hence the posi-
tion of these spicules is quite different from that of the sub-
dermal sagittal triradiates of the Heteropide, while in the one
case it is the basal ray in the other it is the apical one which plays
such an important part in the support of the chamber layer.’

The anatomy of Heteropegma nodus-gordii is repre-
sented in fig. 20. It will be seen that the dermal cortex is
strongly developed, and has the ordinary Grantia structure,
except that there are no oxeote spicules. The gastral cortex
is extremely thin and its skeleton greatly reduced, consisting
of a number of very small quadriradiates arranged in the usual
manner. The articulate tubar skeleton is also very much re-
duced, consisting of very minute spicules, chiefly quadriradiate.
This reduction in the skeleton of the chamber layer is partly
compensated for by the presence of very large subdermal quadri-
radiates, whose facial rays are extended in the inner part
of the dermal cortex, while the apical ray penetrates through
the chamber layer almost to the gastral surface.

The skeleton of the Victorian Heteropegma latitubu-
lata (4) resembles that of H. nodus-gordii, except in certain
very minute details of spiculation, the tubar and gastral spicules
being still further reduced in size.

Amphoriscus.

In this genus the articulate tubar skeleton, which was still
clearly recognisable in Heteropegma, has disappeared,
although the radial Syconoid character of the canal system
remains. In one species, however (Poléjaeff’s Amphoriscus
elongatus [8]), the subgastral sagittal triradiates still persist.
We always find, as in Heteropegma, a layer of subdermal
quadriradiates with inwardly directed apical rays; and we also
find, in addition to these, in nearly all species, a layer of sub-
gastral quadriradiates with outwardly directed apical rays.”

1 The subdermal quadriradiates are probably derived from the ordinary
triradiates of the dermal cortex by the development of an apical ray. Com-
pare my remarks on the skeleton of Lelapia.

* These spicules are not homologous with the subgastral quadriradiates of

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 205

Hence the chamber layer is usually supported solely by the
apical rays of quadriradiates, which pierce it from opposite
directions. In Amphoriscus cylindrus and A. cyathes-
cus (4, 5) we learn from Haeckel that the dermal cortex is
composed entirely of the facial rays of quadriradiates, there
being no triradiates left. This fact supports the view that the
subdermal quadriradiates of the Amphoriscide are merely modi-
fications of the ordinary dermal triradiates, which have de-
veloped an apical ray without changing their position. In
the case of the subdermal sagittal triradiates of the Heteropide,
on the other hand, the position of the spicules is quite different
from that of ordinary dermal triradiates.

Syculmis.

The skeleton of Syculmis synapta (5), the only known
species of the genus, agrees in the main with that of Ampho-
riscus, but is further complicated by the addition of a root-
tuft of oxea and anchoring quadriradiates, which is, I believe,
without parallel amongst the Calcarea. For further details I
must refer the reader to Haeckel’s monograph.

Leucilla (figs. 21, 22).

In this genus, again, we always find dermal or subdermal
quadriradiates with inwardly directed apical rays. These form
the most constant feature of the skeleton, the other parts
being somewhat variable in different species. A gastral cortex,
constructed as in Grantia, is probably always present, and
sometimes, as in Leucilla uter (fig. 21) and L. austra-
liensis (fig. 22), subgastral sagittal triradiates are still recog-
nisable, indicating probably the derivation of the skeleton
from the Grantia type. All the triradiates of the dermal
cortex may be converted into quadriradiates by the develop-
ment of a long inwardly directed apical ray, as in Leucilla
uter (fig. 21); or some of them may still retain their primi-
tive triradiate character, as in L. australiensis (fig. 22).

Grantiopsis, or of Poléjaeff’s Grantia tuberosa (8), which are only
slight modifications of the ordinary subgastral sagittal triradiates.

206 ARTHUR DENDY.

Dermal oxea, like those of Grantia, may also be present
(fig. 21). The chamber layer is generally further supported
by other quadriradiate spicules, which may have a more or less
definite subgastral position (fig. 21), or be irregularly scattered
through the chamber layer (fig. 22), or by irregularly scattered
triradiates, as in Leucilla prolifera (4).

The skeleton of Leucilla (Leucandra) cucumis, Haeckel
(5), exhibits a further complication of the Leucilla type. The
skeleton of the dermal cortex is made up of an outer layer of
triradiate spicules, as in Grantia, and an inner layer of quadri-
radiates with inwardly directed apical rays. These quadri-
radiates help to form a framework for a remarkably regular
series of subdermal cavities, and are assisted in so doing by a
second, deeper layer of quadriradiates with outwardly directed
apical rays. Within this deeper layer of quadriradiates comes
the chamber layer of the sponge wall, supported by irregu-
larly scattered quadriradiates ; and within this comes the gas-
tral cortex, composed of triradiate spicules. In addition to
the spicules mentioned, large oxeote spicules are found arranged
longitudinally between the dermal triradiates. Illustrations
of the skeleton of this remarkable sponge will be found in
Haeckel’s great work.

Summary.

We find, from our survey of the various genera of Hetero-
cela, that the starting-point for the development of the
skeleton throughout the group (leaving out of account Leu-
cascus, which may have originated independently from the
Homoceela) is the radially symmetrical skeleton of Lycetta.
This primitive radial symmetry is highly characteristic of the
group, and is obviously dependent upon the primitive radial
symmetry of the canal system. The first great change in the
structure of the skeleton is brought about by the development
of a dermal cortex, in which a special skeleton is formed.
The skeleton of the chamber layer of the sponge wall now
begins to vary. This variation is in some cases obviously de-
pendent upon the gradual change of the canal system from the

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 207

Syconoid to the Leuconoid condition. This change brings
about a loss of radial symmetry in the chamber skeleton,
which is transformed from the regular “ articulate” type
of Sycetta, Sycon, Grantia, to the irregularly scattered
condition of Leucandra. Other modifications, however,
also take place, which are obviously not dependent on
the variation of the canal system, and which are conse-
quently of the utmost importance for systematic purposes.
These modifications consist in the development of subdermal
sagittal triradiate or of subdermal quadriradiate spicules,
which characterise the two families Heteropide and Ampho-
riscidz respectively. Both these modifications are first insti-
tuted (in the genera Grantessa and Heteropegma respec-
tively) while the canal system still retains its primitive
radial symmetry and its articulate tubar skeleton (figs. 18, 20),
and in both cases they are retained, while the canal system gra-
dually changes from the Syconoid to the Sylleibid or Leuconoid
type, and the primitive articulate tubar skeleton disappears
(figs. 19, 22). Thus the primitive centrifugal radial sym-
metry of the skeleton is lost as the canal system changes from
the Syconoid to the Leuconoid type, but in the Heteropide
and Amphoriscide it is to a certain extent replaced by a kind
of secondary centripetal radial symmetry, due to the develop- -
ment of subdermal radiates with inwardly directed basal or
apical rays. Indications of the primitive radial symmetry are
sometimes found in the presence of subgastral sagittal trira-
diates after all other traces of the articulate tubar skeleton
have disappeared. The gastral cortex, as might be expected,
is subject to less modification than any other portion of the
skeleton, and does not vary to any great extent throughout
the group. A striking exception to this rule is, however,
found in the genus Utella. Very startling exceptions to
the ordinary rules of skeletal structure are found in the
genera Sycyssa and Leucyssa, apparently due to the loss
of all radiate spicules, and the development in their stead
of oxea.

In Leucascus the skeleton appears never to have

VOL. 35, PART 2,—NEW SER. Q

208 ARTHUR DENDY.

reached the radially symmetrical condition, but to have
remained in the scattered one found amongst the reticulate
Homoccela.

IV.—Tune Histotocy or tHe Hetrerocera CALCAREA.

All my observations upon the histology of the Heterocela
Calcarea have been made upon spirit-preserved specimens,
and I cannot therefore claim for them the same value as
attaches to the recent observations of Bidder (21—24) and
Minchin (25, 26), who were able to study the sponges in the
living condition. Nevertheless even a cursory examination of
my figures 23 to 64 will, I hope, show that there is a good
deal of histological detail to be made out even in spirit-
preserved specimens of calcareous sponges.

With regard to the classification of the tissues, I still main-
tain the opinion which I expressed in my memoir on the
anatomy of Grantia labyrinthica (9)—thatis to say, I still
follow Schulze (27) in considering that the ectoderm of
the larval sponge (at any rate in the Calcarea) furnishes not
only the epithelium of the dermal surface, but also that of the
entire inhalant canal system; that the endoderm lines the
remainder of the canal system (including, of course, the
flagellated chambers) from the prosopyles to the osculum,
and that the mesoderm furnishes all the remainder of the
sponge body.

As to the homology of the three layers in sponges with the
similarly named layers in higher animals, I do not presume to
offer an opinion. It is sufficient for our present purposes
that three layers exist in sponges; and as the names ectoderm,
mesoderm, and endoderm have come into general use for these
layers, and serve admirably to express their relations one to the
other, I naturally adopt them in this place.

The Ectoderm.

This forms a single layer of epithelial cells which lines the
external surface of the sponge, and also the whole inhalant

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 209

canal system, from the dermal pores to the prosopyles. If we
consider the manner in which the inhalant canal system of the
more highly developed Heterocela has been derived from the
Sycetta condition, by the closing in of a space which originally
lay altogether outside the sponge, it is obvious that Schulze’s
view as to the extent of the ectoderm must be correct.

T also believe that Schulze was perfectly correct in his opinion
(7) as to the structure of the ectoderm. In the case of Grantia
labyrinthica I stated that the ectoderm resembles exactly
what Schulze has described in Sycon raphanus, “ consisting
of a single layer of flat, polygonal epithelial cells lining the
dermal surface of the sponge and the inhalant canal system.
These cells are most readily distinguished around the inhalant
canals, where they are less obscured by spicules and other
mesodermal structures than on the dermal surface. The
nucleus is surrounded by the very characteristic granules de-
scribed by Schulze in Sycandra. In my preparations I have
only after some trouble succeeded in making out the boundary
lines between the individual cells, and Schulze himself observes
that it is remarkable that the boundaries of these cells—some-
times so distinct—are not always clearly visible. Nevertheless
I have been able to determine the shape of the cells pretty
accurately, and found them to agree precisely with Schulze’s
drawings ” (9).

Since the above was written I have also described and figured
very carefully the ectodermal epithelium of a Homoceele sponge,
Leucosolenia Wilsoni (1). “It consists of thin, flattened,
plate-like cells, polygonal in outline, and each with a swelling
in the centre where the nucleus is situate. The cell itself
averages about 0°0136 mm. in diameter, and the nucleus,
which is very distinctly outlined and more or less spherical
(or perhaps somewhat flattened in the same manner as the
cell), has a diameter of about 0:0034mm. Within the nucleus
appear a few small, deeply staining granules. Around the
nucleus the protoplasm is highly granular, exactly as described
by Schulze, while towards the periphery of the cell it becomes
gradually hyaline. Adjacent cells are in contact at the edges,

210 ARTHUR DENDY.

and all together form a single-layered, continuous epithelium
over the outside of the Ascon-tube. As a rule, in ordinary
preparations, although the nuclei and granules of the ectoderm-
cells may be clearly enough visible, it is very difficult to dis-
tinguish the outlines.”

In writing the above I regret to say that I overlooked
Metschnikoff’s very precise and clear account (28) of the
ectoderm of Leucosolenia (Ascetta), published long before,
in which he describes and figures an ectodermal epithelium
precisely similar to that previously described by Schulze in
Sycon raphanus, and subsequently by myself in Grantia
labyrinthica and Leucosolenia Wilsoni. In the same
paper (28) Metschnikoff also states that the ectoderm of
Leucandra aspera has the same structure.

Metschnikoff’s latest views on the subject of the ectoderm in
sponges in general are contained in the following very interest-
ing passage from his recent work on “Inflammation ” (29).
After pointing out that the body of the sponge is composed of
three characteristic layers, he observes: “ La couche super-
ficielle, ou l’ectoderme, revét le corps entier de cellules
épithéliales plates, limitées entre elles par des contours qui
deviennent trés nets aprés application d’une solution de
nitrate d’argent. Les cellules mémes sont visiblement con-
tractiles, ce qui s’observe surtout aux bords libres des jeunes
individus, ow on apergoit des prolongements amiboides ap-
partenant aux éléments ectodermiques. La contractilité de
ces cellules joue certainement un role dans le phémoméne
remarquable de l’ouverture des pores nombreux, éparpillés sur
la surface de l’éponge entiére, et apparaissant entre deux ou
plusieurs cellules plates.”

Minchin also (25, 26) concludes that the normal condition
of the ectoderm in Leucosolenia clathrus, the form
specially studied by him, is flattened and plate-like, and he
also ventures upon the generalisation that “the contractile
elements in all cases are the flattened ectodermal epithelium.”
The contractile nature of the ectoderm-cells I do not doubt ;
I believe it to be an important character, and I shall presently

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 211

offer some slight evidence of a similar contractility amongst
the Heterocela.

Taking into consideration all the above evidence, I think
we may regard it as an established fact that the ectoderm,
at any rate in all calcareous sponges, like that of Sycon
raphanus, normally consists of flattened plate-like poly-
gonal cells with centrally placed nuclei. I shall pre-
sently have to bring forward fresh evidence in support of
this view.

A peculiar “ flask-shaped ”-condition of the ectodermal cells
has, however, been described from time to time in various
sponges, and has recently received a good deal of attention
from Bidder and Minchin.

This condition was first described by Metschnikoff (28) in
an Olynthus form of Leucosolenia, and the peculiarity
appears to consist in the nucleus being suspended, as it were,
in an envelope of protoplasm from the under surface of the
plate-like cell. Bidder (21), in criticising my description of
the flattened ectoderm of the Homoceela, observes, ‘‘ Although
this form occurs in the Homoceela, it is, in my experience,
rare. The typical ectoderm (e.g. Ascetta clathrus) I
find composed of onion-shaped gland-ceils containing a nucleus
and granules, and provided with a usually fine duct, the
expanded end of which forms the hexagonal area whose
boundaries are, in the case of most sponges, all that has been
observed.” ! Bidder identifies his “onion-shaped gland-cells”
with the cells above referred to as described by Metschnikoff
in an Olynthus form, and gives a list of species in which he
has observed this condition. In his later papers he brings
forward additional evidence for believing “ that in all groups
of sponges the flask-shaped epithelium does occur” (24). He
also identifies (24) the ‘‘ pendent cell body” of his flask-shaped
cells with the subdermal gland-cells which have been described
by von Lendenfeld in the Keratosa and by myself in Grantia
labyrinthica. This identification is probably correct, but I

1 It appears to me that it is generally much more easy to see the nuclei
than it is to see the outlines of the cells.

212 ARTHUR DENDY.

do not think that Bidder has by any means proved that the
“pendent cell body” belongs to the ectoderm at all. It seems
quite as reasonable to believe that it is the presence of
subdermal gland-cells, in many cases at any rate, which has
led Bidder to believe in the general occurrence of flask-shaped
or onion-shaped glandular ectoderm-cells. Until we have
more light on the subject I prefer to retain my old belief in
subdermal gland-cells, but I am quite open to conviction on
this point, which, with the material at my disposal, I am
incapable of deciding for myself.

The observations of Minchin (25, 26) are also very note-
worthy in this connection. This author explains the flask-
shaped condition as due simply to contraction of the normal,
flattened, plate-like cells, and his observations upon Leu-
cosolenia certainly support this view. Subdermal gland-
cells do not appear to occur, at any rate as a rule, in the
Homoceela, and so far as these sponges are concerned I am
inclined to accept Minchin’s explanation.

I must now describe what few additional observations I have
myself been able to make on the ectoderm of the Heterocela.
These observations support the view that the ectoderm is
composed of a single layer of flattened plate-like cells, which
retain to a greater or less extent the power of contraction. I
believe that the character of this epithelium is, usually at any
rate, the same throughout, from osculum to prosopyles, although
it is much more difficult to detect on the exposed outer
surface of a dermal cortex than on the protected surface of an
inhalant canal. I have observed it most satisfactorily in a
specimen of Vosmaeropsis Wilsoni (4) which I took the
precaution to kill with osmic acid. This sponge is provided
with a very distinct oscular sphincter, which has much the
same structure as that described by Minchin in Leucosolenia
clathrus (25). It is composed of two layers of epithelial cells
with probably a very thin layer of gelatinous mesoderm (ground
substauce) between them. Both layers of cells are regarded
by Minchin in Leucosolenia as ectodermal, and (as a mere
matter of convenience, for we must draw the line between

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 213

ectoderm and endoderm somewhere) I follow him in this
respect, and regard both layers of epithelial cells in the oscular
diaphragm of Vosmaeropsis Wilsoni as ectodermal also.
As there is no structural distinction between the flattened
endodermal cells which line the gastral cavity in Heterocele
sponges and the ectodermal cells, it is impossible to say exactly
where one begins and the other ends.

A portion of the epithelium which covers the upper surface
of the oscular diaphragm in the sponge in question is repre-
sented in fig. 63, drawn from an osmic acid preparation
mounted in glycerine. The cells are mostly polygonal, and
fit tightly together by their edges. Each has the form of a
flat plate, with usually a single nucleus situated in about
the centre. The nucleus is small, and in osmic acid pre-
parations looks clearer and more transparent than the
surrounding protoplasm. The latter contains numerous
granules, larger and more abundant towards the centre of
the cell, which are stained very darkly by the osmic acid.
The majority of the cells are about as broad as they are long,
but some of them are fusiform, like those of the Leucosolenia
diaphragm figured by Minchin in his fig. 12. On the under
surface of the diaphragm a similar layer of cells occurs, but
these are not so deeply stained by the osmic acid, doubtless
owing to want of penetration.

I do not doubt that these epithelial cells are contractile,
and that their contractility serves to open and close the dia-
phragm.

If we trace this epithelium from the oscular diaphragm up
over the lip of the osculum and on to the outer surface of the
sponge, we find that it maintains exactly the same character
throughout, covering the whole outer surface of the sponge
with a layer of flattened cells. The dermal cortex is penetrated
by very numerous inhalant pores (fig. 23), and on reaching the
margin of one of these the ectodermal epithelium turns in,
and is continued as a lining to the inhalant canal system, a
lining which resembles minutely the epithelium of the oscular
diaphragm, and the component plate-like cells of which are

214 ARTHUR DENDY.

extremely conspicuous in sections of osmic acid specimens
(fig. 23).

Sometimes in sections of ordinary spirit-preserved material,
cut by the paraffin method, the flattened cells of the ecto-
dermal epithelium appear separated from one another by
considerable intervals, as is shown in fig. 29, representing a
portion of the epithelial lining of the inhalant canal system of
Grantessa intusarticulata, and in fig. 59, representing
epithelium from a corresponding situation in Sycon Ramsayi.
This appearance is evidently due to contraction of the cells,
which may still remain connected in places by strands of
protoplasm which stretch across from one to the other. Both
ectodermal and endodermal epithelial cells appear to be very
subject to such contraction, and it appears to me very possibly
to indicate a normal contractility in life, such as is described
by Metschnikoff (29).

It has been recently maintained by Bidder (23) and Minchin
(26) that the prosopyles which pierce the walls of the flagel-
lated chambers are formed each by the perforation of a single
nucleated cell; but, while Bidder attributes to these cells an
endodermal origin, Minchin regards them as derivatives
of the ectoderm. For my own part I am disposed to regard
the prosopyles as inter-cellular and not intra-cellular in
nature, and Schulze’s admirable drawings of the anatomy of
Sycon raphanus (7) point to the same conclusion. My own
drawings of the anatomy of Grantia labyrinthica show
exactly the same condition, and in the very numerous sections
of calcareous sponges which I have examined, I have met with
no evidence to cause me to doubt the correctness of the view
that the prosopyles are simply gaps between cells. Bidder,
indeed (21), interprets the remarkable groups of yellow granules
which I described in Leucosolenia cavata (1) as perforated
prosopyle cells. I do not deny the possibility of this view, but
I do not think that the evidence is sufficient to justify it,
and, in any case, the occurrence of such structures (whatever
they may be) is in my experience very rare even in the
Homoceela, and altogether unknown in the Heterocela, of

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 215

which I have examined very numerous species. Therefore,
even if we admit that perforated prosopyle cells occasionally
occur, there is no reason for believing that this is the usual
condition, but quite the contrary. When viewed en face the
prosopyles of calcareous sponges usually appear as sharply
defined approximately circular spaces (figs. 25, 30). Not
infrequently the nucleus of an ectodermal epithelial cell
happens to lie exactly on the margin of this space (fig. 30), but
I do not think there is any reason for regarding this fact as
indicating that the prosopyle itself is of an intra-cellular
nature; while sections such as those figured by Schulze in the
case of Sycon raphanus, and by myself in the case of
Grantia labyrinthica, clearly indicate the contrary.

Before leaving the question of the ectoderm I may con-
veniently describe in this place a very curious condition which
exists in Grantiopsis cylindrica (4). This highly interesting
Australian species is known only from a few fragments,
probably all belonging to the same specimen. These fragments _
are, however, in a remarkably good state of preservation, but
notwithstanding this fact it is a difficult matter to make out
the epithelial cells either on the outer surface of the sponge or
in the somewhat complicated inhalant canal system. On the
other hand, we find in both these situations a very peculiar
layer of minute granules, which, when examined under a Zeiss
F objective, presents in surface view the appearance shown in
fig. 57, and in sections at right angles to the dermal surface
the appearance shown in fig. 56. The granules themselves
appear slightly elongated and very thickly and evenly scattered
over the surface. Embedded in the gelatinous ground sub-
stance of the mesoderm, at some little distance beneath the
dermal surface, are found numerous very distinct subdermal
gland-cells, connected by thread-like processes with the
granular layer, as shown in fig. 56. The granules themselves
do not stain with borax carmine, and present, under the Zeiss
F obj., a very striking resemblance to bacteria. So great
is this resemblance that I submitted a fragment of the
sponge to Mr, Thomas Cherry, Demonstrator in Pathology in

216 ARTHUR DENDY,

the Melbourne University, who kindly made a careful exami-
nation of it for me, with the following results. The granules
were scraped into a glass, dried, stained with fuchsin, mounted
in Canada balsam, and examined under a ;4-inch oil im-
mersion objective (Leitz) with a No. 5 eye-piece. We
then found that they presented the appearance shown in fig.
58, being usually each somewhat dumb-bell shaped, or com-
posed of two ovoid segments placed end to end, though
sometimes the shape was irregular. They averaged about
0:004 mm. in length, and consisted of a fairly darkly stained,
sometimes slightly granular body, with a round very darkly
staining structure, presumably a nucleus, in each end of the
dumb-bell, or, when the shape was irregular, with several such
nuclei. It thus appears that, though probably not bacteria,
the bodies in question are micro-organisms of some kind
which live upon the surface of the ectoderm. Nuclei, pre-
sumably belonging to the epithelial cells, are still visible
amongst the granules, at any rate in the inhalant canals, and
I am inclined to think also on the outer surface of the sponge;
but we have here an excellent illustration of the way in which
the structure of the ectoderm is sometimes obscured by the
accumulation of foreign bodies.

Bidder would possibly interpret the subdermal gland-ceils
of Grantiopsis cylindrica as representing the true ecto-
dermal epithelium. But they do not look like it (fig. 56).
They appear to be but very slight modifications of the ordinary
stellate mesodermal cells, and as such they will be described
on a future page.

The Endoderm.

I have little to add concerning the histology of the endo-
derm to what I have already written on the subject. As already
stated, I consider as endoderm not only the collared cells
which line the flagellated chambers, but also the flattened
pavement epithelium which lines the gastral cavity and ex-
halant canals of the Heterocela. It is, of course, quite
conceivable that this pavement epithelium is formed by an

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 217

ingrowth of the ectoderm from the region of the osculum,
replacing and, as it were, pushing back the collared cells. I
do not, however, consider that there is any reason for supposing
this to be the case. In the case of the Homoceela I have
shown (1) that the lining of the gastral cavity is not always
composed entirely of collared cells, but that the apical rays of
the quadriradiate spicules which frequently project into this
cavity are clothed with a layer of flattened epithelium. I
suggested at the time that this epithelium might possibly be
derived from the mesoderm; but, on the other hand, it might
with equal justice be regarded as indicating a transformation
of collared cells into permanent cells, and we are quite at
liberty to suppose that a similar transformation, only on a
much larger scale, has taken place in the gastral cavity and
exhalant canals of the Heteroccela.

The epithelium which lines the gastral cavity and exhalant
canals resembles minutely the ectodermal epithelium which
clothes the outer surface of the sponge and lines the inhalant
canal system. This will be evident from a comparison of figs.
59 and 60, representing respectively the epithelium from an
inhalant and from an exhalant canal of Sycon Ram sayi, the
differences which exist between the epithelium in these two
situations being obviously very slight and insignificant.
Similarly fig. 64 represents the epithelium from an exhalant
canal of Vosmaeropsis Wilsoni, which, if we allow for the
contraction of the cells due to difference in the method of
preparation, is practically identical with the ectodermal
epithelium from the same sponge represented in fig. 63. In-
deed, unless specially prepared, as with osmic acid, the
ectodermal epithelium, as already pointed out, exhibits a
precisely similar contraction, and in both cases this is such a
constant and well-marked feature that I am inclined to think
that it betokens a normal contractility during life. The
appearance of the endodermal pavement epithelium, as seen in
section, is shown in figs. 27, 28, and 61. The swelling in the
centre of the cell, where the nucleus is situated, is fre-
quently very strongly marked, and causes the epithelium to

218 ARTHUR DENDY.

be very conspicuous in sections taken at right angles to its
surface.

In an undetermined specimen of Leucandra, from Port
Jackson, the epithelial cells lining the gastral cavity just within
the osculum are so much swollen out as to be almost as thick
as they are broad, and they also present a vacuolated blister-
like appearance, as shown in fig. 32. Lower down in the gastral
cavity the epithelial cells have the normal flattened form.

The flagellated chambers of most, if not all Heteroccele
sponges, whether they be of the elongated Syconoid or of the
rounded Leuconoid form, are separated from the exhalant
canals into which they open by very distinct membranous
diaphragms (figs. 25—27, 31). When viewed en face these
diaphragms have usually the appearance shown in fig. 31,
consisting of a thin, transparent membrane, in which are
visible spindle-shaped cells arranged concentrically. The
actual outlines of the cells I have not succeeded in distinguish-
ing, but their elongated character and concentric disposition
are usually clearly indicated by the arrangement of the
granules which surround the nucleus. Collared cells are never
found on either surface of this membranous sphincter, and
whether it consists of one or two layers of cells is a very
difficult question, which I have not been able definitely to
decide. The concentrically arranged, spindle-shaped cells
are obviously muscular, and as such I described them in my
memoir on Grantia labyrinthica. In that paper, however,
I classed them as mesodermal elements. I am now convinced
that they are simply modifications of the cells which line the
exhalant canals, comparable to the muscular cells of the oscular
diaphragm described by Minchin in Leucosolenia, and by
myself in Vosmaeropsis. Fig. 27 shows how the epithelial
cells of the exhalant canal are continued on to the surface of
the chamber diaphragm. Usually they appear very much
elongated in this situation, as shown in fig. 31, but sometimes
the elongation is scarcely visible at all, as shown in figs, 25
and 26. Probably the exact form of the cells depends upon
their state of contraction. The epithelial cells of the exhalant

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 219

canals themselves may exhibit elongation and concentric
arrangement around these canals, but not to the same degree
as in the chamber diaphragms.

The chamber diaphragms, then, I regard as being composed
of endodermal muscular cells, formed by modification of the
ordinary endodermal pavement cells. Whether they consist of
one or two layers of these cells is doubtful; probably of two layers,
with a small amount of gelatinous ground substance between.

It thus appears that, if Minchin is right in regarding the
muscular cells of the oscular diaphragm in Leucosolenia as
ectodermal, and if Schulze is right in regarding the lining of
the exhalant canals in Heterocele sponges as endodermal, then
both the ectoderm and endoderm may give rise to strikingly
similar muscular structures. Possibly this may be an argument
for regarding the muscular cells of the oscular diaphragm in
Leucosolenia as endodermal and not ectodermal; but it
appears to me that this is a question which cannot be decided
in the present state of our knowledge.

Concerning the remaining endodermal cells, the collared
cells which line the flagellated chambers, I have little to say.
I may, however, call attention to fig. 25, representing half of
a flagellated chamber of an undetermined species of Leu-
candra, in which Sollas’s membrane and the collars of the
collared cells are remarkably clearly seen. This figure also
shows the exhalant aperture with its membranous diaphragm,
and two prosopyles. Whether or not Sollas’s membrane is
continuous over the prosopyles I have been unable to deter-
mine, and it appears to me that the point can only be decided
by the examination of living specimens.

I quite agree with Bidder (21) that the endoderm is not only
multiform, but also ‘‘ most proteic,” as, indeed, was long ago
shown by Carter (30) from observations on the living Sycon
(Grantia) compressum, in which he observed the collared
cells becoming ameeboid, and moving about the field of the
microscope. Iam therefore quite prepared to believe that the
existence of Sollas’s membrane may be only a transitory con-
dition, due to special circumstances, of which we are as yet

220 ARTHUR DENDY.

ignorant. That it does exist in many cases is beyond dispute,
and it is extremely interesting to learn that Bidder (21) has
observed in the living Syconraphanus the coincidence of
the flagella of the collared cells with Sollas’s membrane, just
as I described it in Halichondria panicea (81).

A curious illustration of the polymorphic nature of the
collared cells, is afforded by the peculiarly contracted cham-
bers which I described in my memoir on Grantia labyrin-
thica, and indicated in my drawings of the anatomy of that
sponge by the letter « I have now observed the same phe-
nomenon in other Heteroceele sponges—viz. Ute syconoides
and Leucandra phillipensis—and am very glad to be able
to give some more exact details as to the form and arrange-
ment of the collared cells in these cases. Fig. 24 represents
a contracted chamber of Ute syconoides shown in trans-
verse section, and surrounded by four ordinary chambers and
four intercanals. It will be seen that the layer of collared
cells has shrunk away from the tubar skeleton, drawing the
mesodermal tissue after it. The collared cells themselves have
become radially elongated, and, owing to their having a
smaller area to spread themselves over, very much crowded.
It was doubtless this crowded condition which caused me to
describe them as being arranged in more than one layer in the
case of Grantia labyrinthica. Lach cell is more or less
pyramidal in form, and the nucleus is situated in the apex of
the pyramid towards the lumen of the chamber.

In my earlier paper (9) I endeavoured to explain the occur-
rence of these contracted chambers (which exhibit a very
different appearance from chambers in which contraction has
been produced by the action of the preserving fluid) by sup-
posing them to represent old and exhausted chambers in pro-
cess of dying, and destined to be replaced by the development
of new chambers.

Bidder, in his very interesting ‘“ Note on Excretion in
Sponges” (23), has suggested a different explanation. After
describing the accumulation of ‘ spherules” in the bases of
collared cells, he adds, “I believe these basal spherules to be

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 221

stores of nutritious matter. In Leucandra aspera and
Sycon raphanus.... the collar cell, after it has accumu-
lated a certain quantity of spherules in its base, splits off this
base by a transverse fission as a non-nucleated mass of pro-
toplasm, which we may term a ‘ plinth’ (fig. 4); the plinth
then lies between the nucleated distal part of the cell—the
‘column ’—and the mesodermal jelly in 8S. raphanus, or the
thin basement membrane which is all that usually divides
the epithelia in L. aspera. I have observed in the meso-
derm of 8S. raphanus large wandering cells, which I believe
to be generative elements, with pseudopodia attached to these
plinths, aud spherules of the same character as the basal
spherules, both in the wandering cells and in their pseudo-
podia. There can be little doubt that they were feeding on
ths reserve stores of the collar-cells. The division into column
and plinth takes place as a rule at the same time in all or
most of the cells of a chamber. The ‘columns’ or distal
parts appear as small, columnar, nucleated cells, provided with
a small amount of clear protoplasm, rudimentary collars not
united, and flagella (fig. 4a).” Later on, in the same paper,
Bidder observes that he interprets the peculiar structures met
with by me in Grantia labyrinthica as column-and-plinth
chambers violently contracted in alcohol. There is certainly
a good deal to be said in favour of this view, and, examining
the structures in question in the light of Bidder’s remarks, I
have seen in Leucandra phillipensis appearances which
might be taken as indicating the formation of “ plinths.”” The
abundant occurrence of small nucleated cells in the mesoderm
which surrounds the contracted chambers, as shown in fig. 24,
is certainly suggestive either of the congregation of amceboid
cells to feed on the collared cells, or of the collared cells or
segments thereof becoming amceboid and wandering away
into the mesodermal jelly. The occurrence of such cells in the
surrounding jelly is characteristic of the contracted chambers,
and their small size appears to me to be an argument in favour
of regarding them as metamorphosed collared cells rather than
as ordinary ameeboid ones.

222 ARTHUR DENDY.

In Leucandra phillipensis the ordinary ameboid cells
are very much larger than the cells in question in the same
sponge; in fact, compared to the latter and to the collared
cells they are as giants. Curiously enough, I observed one of
these large amceboid cells, shown in fig. 50, apparently feeding,
by means of pseudopodia, on the collared cells. The chamber
to which the collared cells belonged was, however, in the ordi-
nary uucontracted condition, and I could not see any indication
of the division of the collared cells into “ column and plinth.”

On the whole, the peculiar contracted condition, such as is
shown in fig. 24, and which occurs in chambers both of the
Syconoid and Leuconoid types, is perhaps to be regarded as
indicating not so much a process of death as one of re-
juvenescence, and this is, after all, but a slight modification of
my former view. We may suppose that, after perhaps taking
in a large supply of food, the chamber passes into a resting
condition. The collared cells then undergo certain changes
which are not fully understood, but which finally result in
the formation of a new chamber with ordinary active collared
cells. The contraction of the chamber during this process
appears to me to be normal, and not due to the action of re-
agents.

The Mesoderm.

The transparent gelatinous ground substance which forms
the bulk of the mesoderm in all calcareous sponges, appears to
vary only ia the extent to which it is present and the propor-
tion which it bears to other parts of the sponge. It is most
strongly developed in the dermal and gastral cortex (figs. 10
—12). It may also be fairly abundant in the walls of the
flagellated chambers, as in Sycon gelatinosum (figs. 3—5),
but it is usually very sparingly developed in this situation, so
as to be distinctly recognisable only at the points where the
chambers touch one another (fig. 13). Hence in very many cases
the walls of the chambers appear in thin sections to be made
up of two contiguous layers of cells, the collared cells on the
inside and the pavement epithelium of the inhalant canals on

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 2238

the outside. Ido not doubt, however, that there is always a
thin layer of mesoderm between the two, a fact which is
indicated by the development of spicules in the chamber walls.
It is certainly continued into the endogastric septa of
Grantessa erinaceus, which consist of a layer of gelatinous
ground substance, with an occasional stellate cell, covered
on each side by a layer of flattened epithelium (fig. 62).

Concerning the spicule sheaths, which are formed by a
slight concentration of the structureless mesodermal jelly
around the spicules (fig. 53), I have nothing to add.

In my memoir on Grantia labyrinthica I classified the
mesodermal cells which lie embedded in the gelatinous ground
substance as follows:—(1) Ameeboid, (2) Stellate, (3) Glan-
dular, (4) Endothelial, (5) Muscular, (6) Nervous, and (7) Re-
productive. I am still prepared to abide by this classification
so far as the ameeboid, stellate, glandular, and reproductive
cells are concerned. The only muscular cells which have yet
been observed in calcareous sponges I now regard, as already
indicated, as ectodermal or endodermal. The supposed
nervous cells I am now extremely doubtful about, in all cases,
and I am strongly inclined to think that the appearances
described as such are due to the presence of subdermal gland-
cells. This point, however, can only be settled by a very
careful re-investigation of living material. The endothelial
cells I still retain provisionally amongst the mesodermal
elements, although I shall presently give reasons for believing
. that they need not necessarily be regarded as such.

Amcboid and Stellate Cells—I have observed most
beautiful examples of ameeboid cells in Leucandra echinata
(fig. 33) and L. phillipensis (figs. 48—50). They are found
scattered here and there in the mesodermal ground substance
between the flagellated chambers, and, in these two sponges,
are conspicuous by their enormous size. The amceboid cells are
typically distinguished from the stellate cells by their more uni-
formly granular protoplasm, their larger nuclei, and the absence
of the slender, thread-like, radiating processes, which charac-
terise the stellate cells ; the latter are very different in appear-

VOL, 35, PART 2.—NEW SER, R

224 ARTHUR DENDY.

ance from the blunt, rounded pseudopodia which characterise
a typical amoeboid cell. Nevertheless it is often very difficult
to say whether a particular cell should be classed as stellate or
amceboid, especially when that cell is of small size; and I am
inclined to think that no hard-and-fast line of distinction can
be drawn between the two. Thus figs. 389—42 represent a
number of cells from the dermal cortex of Leucandra
phillipensis, which I personally should class as stellate, but it
would be very difficult to prove that they are not ameboid.
Similarly fig. 55 represents four mesodermal cells from the
dermal cortex of Grantiopsis cylindrica, which may be
classified as stellate or amceboid, according to the taste of the
observer. For my own part Iam inclined to believe that even
the most typical stellate cells may be, to a greater or less extent,
amoeboid, and capable of a certain amount of movement. What
appears to be division of the ‘‘ stellate” cells by fission is very
clearly seen in the cortex of Leucandra phillipensis, in
which many of the cells have two nuclei, and also show
signs of division in the form of the cell body (figs. 39,
40, 42).

Glandular Cells.—These, as I have already pointed out
(9), are of two kinds, calcoblasts and subdermal gland-cells. I
can only confirm the account of these structures which I gave
in the case of Grantia labyrinthica. I have recently
observed some very beautiful examples of calcoblasts. They
appear to be most easily recognisable on the inner portions of
projecting dermal oxeote spicules. Figs. 44—47 represent
portions of such spicules from the dermal cortex of Leu-
candra phillipensis. ach spicule has a single nucleated
calcoblast attached to its inner half, and I do not hesitate to
regard these cells as the manufacturers of the material of
which the spicules are composed. At the same time it will
be sufficiently obvious, on comparing figs. 44—47 with figs.
39—42, that the calcoblasts are very similar in appearance to
the ordinary “stellate” cells, which are found embedded in the
gelatinous ground substance in their immediate neighbourhood.

In Grantiopsis cylindrica I have found very large cal-

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 225

coblasts! attached to the rays of the large triradiate spicules
in the dermal cortex. Three of these cells are shown in figs.
52—54, and their characters certainly justify the assump-
tion that they are but slight modifications of ordinary stellate
cells. It appears to me that the calcoblasts, at any rate in the
case of large spicules, must be ameceboid, for, unless they
be so, I cannot understand how the spicules can increase
uniformly in thickness. I have already suggested (9) that
there are probably primary and secondary calcoblasts, the
former being mother-cells within which spicules are formed,
and the latter cells which apply themselves to the surfaces of
already formed spicules and increase the thickness of the latter.
The cells represented in figs. 52—54 are probably secondary
calcoblasts; those represented in figs. 44—47 may be primary.

Minchin (82), in the case of Leucosolenia, found the tri-
radiate spicules to have a nucleus at the extremity of each ray,
and a fourth one at the confluence of the rays. As there are
four of these nuclei they probably indicate the presence of
secondary calcoblasts, for we can hardly suppose that the
spicule is originally formed by more than one cell.

We come now to the subdermal gland-cells, which Bidder
(24) regards as the.‘ pendent cell bodies” of flask-shaped
ectoderm-cells. These cells may occur beneath both the
gastral and the dermal surfaces of the sponge. In Grantia
labyrinthica they are more plentiful beneath the gastral than
beneath the dermal surface, a fact which I associate with the
peculiar shape of the gastral cavity, which causes its surface
to be almost or quite as much exposed as the dermal surface.
A situation beneath the dermal surface, however, appears to be
their normal one, and they are rare elsewhere. To my descrip-
tion of these structures in Grantia labyrinthica I have
little to add. I have found them especially well developed in
Grantiopsis cylindrica (fig. 56), where the connection
with the surface is very long and slender. In Leucandra
phillipensis I have detected them around the inhalant

1 Compare with these the “conjectural calcoblast,” figured by Poléjaeff
(8) in his Leuconia multiformis,

226 ARTHUR DENDY.

canals as well as beneath the dermal surface (fig. 43). In all
these cases they closely resemble the ordinary stellate meso-
derm-cells, and as modifications of such I am for the present
disposed to regard them. That they secrete a slime which
covers the surface of the sponge in life, I do not doubt.

Vesicular Cells.—I propose this name for certain large
rounded cells which occur scattered singly, but in considerable
numbers, immediately beneath the gastral epithelium of
Ute syconoides. The general appearance of these cells is
shown in fig. 37. The body of the cell stains uniformly and
fairly darkly with borax carmine. It is not granular, but
hyaline. The nucleus is small, darkly staining, and situated at
one side of the cell. It appears to have been pushed aside by
the accumulation of fluid in the cell body, as in fat-cells. I
am not aware that cells of this kind have been noticed before
in calcareous sponges, and they appear to me to be most nearly
comparable to the ‘‘cystenchyme” cells of Silicea and
Keratosa—as, for example, Stelospongus (83).

Reproductive Cells.—The only reproductive cells upon
which I have made any observations are the ova, and these
have been so frequently described that I need say little about
them here. Figs. 34 and 36 represent typical calcisponge ova,
consisting each of a large naked body of highly granular
protoplasm, with a large, very sharply defined nucleus and a
distinct nucleolus.

In my memoir on Grantia labyrinthica I adduced evi-
dence for believing that the ova, at a certain stage of their
existence, migrate through the epithelium of the inhalant
canals and hang freely from its surface, so as to be bathed by
the inflowing stream of water, and I regarded this as a special
contrivance for securing fertilisation. In Ute syconoides
I have had the good fortune to observe a very beautiful in-
stance of the manner in which the ovum hangs suspended from
the wall of an inhalant canal, as represented in fig. 35. I was
particularly glad to obtain this confirmation of my previous
observations, as the inhalant canals of Ute syconoides are
much more sharply defined and easily recognisable than those

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 227

of Grantia labyrinthica, and there is no possibility of doubt
as to the exact position of the pendent ovum. Fig. 36 repre-
sents an ovum of the same sponge which has presumably been
fertilised, and has taken up the normal position for development,
just behind the layer of collared cells which lines one of the
flagellated chambers, as shown in figs. 5 and 13.

For an account of the spermatozoa of the Heteroceela I must
refer the reader to Poléjaeff’s work on the subject (34).

Endothelial Cells.—Under this name I have previously de-
scribed (9) the more or less flattened cells which line, in a single
layer, the cavities in which the embryos develop. Fig 38 shows a
typical example of an embryo lying in a cavity lined by such
cells. Schulze (7) attributes to these cells a mesodermal
origin. For the present I adopt this view; but I would like
to point out that they may possibly be ectodermal, for it
is easy to imagine that an ovum, after migrating through the
wall of the inhalant canal, in returning into the mesoderm
may push before it a portion of the ectodermal epithelium, from
which the embryo capsule might be derived. This, however,
is, in the present state of our knowledge, mere speculation.

V.—Tue CLAssiIFICATION OF THE HeETEROC@LA CALCAREA,
witH DraGNosEs OF THE FAMILIES AND GENERA.

We are now in a position to apply the results of our anato-
mical investigation to the classification of the group. I have
found it necessary to forestall to a certain extent what I have
to say here about the classification, in order to satisfactorily
arrange my notes on the anatomy, but this was obviously
unavoidable. In the present section I propose not ouly to set
forth the classification of the group, and to give brief diagnoses
of the families and genera, but also to discuss, so far as
appears necessary, the questions of synonymy and nomencla-
ture.

Order HETEROCCELA, Poléjaeff (8).

Diagnosis.—Calcareous sponges in which the collared cells
are confined to more or less well-defined flagellated chambers.

228 ARTHUR DENDY.

Family 1.—Lervcascipm, Dendy (4).

Diagnosis.—Flagellated chambers very long and narrow,
copiously branched; communicating at their proximal ends
with exhalant canals, which converge towards the oscula;
their blind distal ends covered over by a dermal membrane
pierced by true dermal pores, which lead into the irregular
spaces between the chambers. Skeleton consisting princi- ~
pally of small radiates, irregularly scattered in the walls of
the chambers and exhalant canals, and in the dermal mem-
brane.

Genus 1.—Leucascus, Dendy (4), fig. 1.

Diagnosis.—The same as that of the family.

Remarks.—Leucascus, the sole genus of the family,
appears to occupy a very isolated position amongst the
Heterocela, both as regards skeleton and canal system.
Its skeleton, as already pointed out, has not yet attained to
the typical radiate condition of the Heteroceela, but retains
the irregular scattered character of the reticulate Homoceela.
The genus might, indeed, be easily confounded with the
reticulate section of the genus Leucosolenia (1), but differs
in several very important particulars; viz. (a) The possession
of a distinct pore-bearing dermal membrane, which is found
in no Homocele sponge with which I am acquainted, and is
not even developed amongst the Heterocceela until we reach
the most advanced Sycon condition. The pseudoderm
of many reticulate Homoccela (1) is by no means an homo-
logous structure. (4) The absence of collared cells from the
exhalant canals, which there is no reason for believing to be of
a “ pseudogastral ” nature, as in my type D of the reticulate
Homocela. (c) The flagellated chambers, although, owing
to the massive form of the sponge and the absence of a single
central gastral cavity, they cannot be truly radial, nevertheless
show a marked tendency to become so.

The genus may, perhaps, be best regarded as derived inde-
pendently from a lowly organised type of radiate Homoceela,

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 229

It might, perhaps, be regarded as a degenerate Heteroccele
which has lost its radial symmetry, but the facts do not
appear to me to warrant this hypothesis.

Family 2.—Sycertipm, Dendy (4).

Diagnosis.—Flagellated chambers elongated, arranged
radially around a central gastral cavity, their ends projecting
more or less on the dermal surface, and not covered over by a
continuous dermal cortex. Tubar skeleton articulate.

Remarks.—This family is equivalent to van Lendenfeld’s
(10) two sub-families Sycanthine and Syconine taken to-
gether. As I have already pointed out, his genus Sycantha
is probably but a slight modification of the ordinary Sycon
type. The name Sycettide is adopted partly because the
genus Sycetta is the simplest of the three included in the
family, and partly to avoid confusion with the much more
comprehensive Syconidz of previous writers, such as Polé-

jaeff (8).
Genus 2.—Sycetta (Haeckel [5], emend.).

Diagnosis.—Radial chambers separate from one another,
and without tufts of oxea on their distal ends.

Remarks.—This genus, as here maintained, is much less
comprehensive than was originally intended by its author (5).
Haeckel included in the genus all Syconoid species in which
triradiate spicules alone entered into the composition of the
skeletons, and thus was obliged to group together such struc-
turally different species as S. primitiva and S. stauridia;
while he had to exclude S. conifera simply because some of
the triradiates develop an apical ray, and thus become
quadriradiate. Von Lendenfeld also confined his diagnosis of
the genus to the character of the spiculation (10), but modified
it so as to include S. conifera. Poléjaeff (8) merged the
genus into Sycon. If, however, we include in the genus only
those species which, like Haeckel’s Sycetta primitiva and
S. (Sycaltis) conifera, conform in skeletal characters and
canal system to the diagnosis given above, we shall have left a

230 ARTHUR DENDY.

very natural and well-characterised genus, comprising only
three species, viz. S. primitiva, S. sagittifera, and S.
conifera, all of which are described by Haeckel in his great
monograph.

Genus 3.—Sycon (Risso, emend.), figs. 2—8.

Diagnosis.—The radial chambers are usually more or less
united at places where they come into contact with one another,
and they are always crowned at their distal extremities with
tufts of oxeote spicules.

Remarks.—The most characteristic feature of this genus
is afforded by the tufts of oxeote spicules which crown the
distal ends of the radial chambers. In some species, such as
Sycon boomerang (4) and 8. gelatinosum (4, 5), we find
a transition to the genus Grantia, in the presence of a pore-
bearing dermal membrane stretched between the distal ends
of the radial chambers; but this never forms a cortex which
completely covers over the chambers as in the Grantide, and
it in no way interferes with the characteristic tufts of oxeote
spicules.

The genus as here constituted includes most of Haeckel’s
species of his genus Sycandra (5), but I follow Poléjaeff (8)
in giving priority to the old name Sycon. Familiar European
examples are S. raphanus and S. ciliatum, and I also
include 8S. (Grantia) compressum, on account of the tufts
of oxeote spicules which crown the ends of the chambers.
Poléjaeff’s genus S ycon is, as already pointed out, more com-
prehensive, and includes our Sycetta, while von Lendenfeld
(10) adopts the genus Sycandra in almost the same sense as
Haeckel.

Geuus 4. Sycantha, von Lendenfeld (10).

Diagnosis.—Radial chambers united in groups, with
freely projecting distal cones surmounted by tufts of oxeote
spicules.

Remarks.—lIt appears to me that the only character which
can be relied upon for distinguishing this genus is the grouping

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 281

of the chambers. I have already given my reasons for coming
to this conclusion in discussing the canal system.

Family 3.—Grantip#, Dendy (4).

Diagnosis.—There is a distinct and continuous dermal
cortex, completely covering over the chamber layer, and pierced
by inhalant pores. There are no subdermalsagittal triradiates,
nor conspicuous subdermal quadriradiates. The flagellated
chambers vary from elongated and radially arranged to
spherical and irregularly scattered ones, while the skeleton of
the chamber layer varies from regularly articulate to irregu-
larly scattered.

Remarks.—This family does not nearly correspond with
any which has hitherto been proposed; for, disregarding
differences in canal system, I include therein some genera
with a Syconoid and others with a Leuconoid type, and rely
upon the structure of the skeleton for thus uniting them. I
believe myself to be justified in this course of action by the
great variation of the canal system which has been shown to
exist in closely related forms and even in the same species,
and which I have already discussed in a previous part of this
paper.

The family, as here constituted, is distinguished from the
Sycettidze by the positive character afforded by the develop-
ment of a strong dermal cortex, and from the Heteropide and
Amphoriscide by the negative characters afforded by the
absence of subdermal triradiates or quadriradiates. It is
very difficult to diagnose the family, which is a very compre-
hensive one, so as to exclude the Leucascide, but the latter
present a peculiar combination of skeleton and canal system
which is not to be found amongst the Grantide, and appear
never to have passed, as I believe all the Grantide have,
through a primitive Syconoid stage with radially symmetrical
skeleton.

Genus 5.—Grantia (Fleming, emend.), figs. 9, 10.
Diagnosis.—The elongated flagellated chambers are ar-

PAB ARTHUR DENDY.

ranged radially around the central gastral cavity ; they are not
provided with tufts of oxea at their distal ends, but are covered
over by a dermal cortex composed principally of triradiate
spicules, and without longitudinally disposed oxea. An ar-
ticulate tubar skeleton is present.

Remarks.—The genus Grantia, as here defined, is very
nearly co-extensive with the same genus as employed by
Poléjaeff (8). It appears to me, however, that the genus must
be limited to those species in which no tufts of oxea are
developed on the ends of the radial chambers, in order to define
it with desirable sharpness from Sycon; for, as I have already
shown, we meet with the first indications of a dermal cortex
in the latter genus. Hence I cannot agree with Poléjaeff
in including Sycon (Grantia) compressum. Nor can I
agree with von Lendenfeld’s diagnosis (10), for he expressly
states that a crown of radial “ Thabden” may be present on
each chamber, and his diagnosis is so worded as to exclude
Haeckel’s G. (Sycetta) strobilus and G. (Sycetta) cupula,
which, as was recognised by Poléjaeff (8), undoubtedly belong
to the genus Grantia.

Sub-genus Grantiopsis, Dendy (4), fig. 11.

Diagnosis.—The sponge has the form of a greatly elon-
gated hollow tube, whose wall is composed of two distinct
layers of about equal thickness. The outer (cortical) layer
is provided with a very strongly developed skeleton of radiate
spicules, and is penetrated by narrow, ramifying, inhalant
canals. The inner layer is formed by elongated radial
chambers, arranged very regularly side by side. The skeleton
of the inner layer is very feebly developed. The tubar
skeleton is articulate, and composed of very abnormal sagittal
triradiates whose paired rays are greatly reduced.

Remarks.—This genus is obviously only a very special
modification of the well-known Grantia type, although at
first sight, especially as regards external form, it appears
very distinct. The only species known is the Victorian
Grantiopsis cylindrica (4).

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 233

Genus 6.—Ute (Schmidt, emend.), figs. 12—14, 24.

Diagnosis.—The ends of the elongated radial chambers
are covered over by a well-developed cortex, composed in
great part of large oxeote spicules arranged parallel to the
long axis of the sponge. The tubar skeleton is articulate,
or else composed entirely of the basal rays of subgastral
triradiates.

Remarks.—This genus I maintain in the same sense as
Poléjaeff (8) and von Lendenfeld (10). It now includes five
species, viz. the original European type, Ute glabra,
Schmidt (17), and the Australian U. argentea, Poléjaeff (8),
U. syconoides, Carter (12), U. spiculosa, Dendy (4), and
U. Spenceri, Dendy (4). The spiculation of the dermal
cortex, upon which the genus is founded, is paralleled in
two species of Leucosolenia, L. asconoides, Carter (12),
and L. uteoides, Dendy (16); and also in Leucandra
(Aphroceras) alcicornis, Gray, and Heteropia (Aphro-
ceras) ramosa, Carter (15). Indeed, this character gave
rise to Gray’s family “ Aphrocerasid” (18), which, like other
families founded upon an insufficiency of characters, has had
to be abandoned.

The genus appears to be a natural one, but is obviously very
closely related to Grantia, and, had it not been already well
established, I should have hesitated in attributing generic
importance to a character which is found in so many very
distinct sponges.

Sub-genus Synute, Dendy (18), fig. 15.

Diagnosis.—Sponge compound, consisting of many Ute-
like individuals completely fused together, and invested in a
common cortex, composed largely of huge oxeote spicules
arranged longitudinally.

Remarks.—I at first thought that my Synute pulchella
(13) should stand as the type of a new genus, but I have since
come to the conclusion that the complete fusion of the Ute
individuals, though perhaps unparalleled amongst sponges

234. ARTHUR DENDY.

with a Syconoid type of canal system, is scarcely a character
of generic importance, and I have therefore reduced the
species to the rank of a sub-genus.

Genus 7.—Utella, Dendy (4).

Diagnosis.—Flagellated chambers elongated, arranged
radially around the central gastral cavity. There are no
longitudinally arranged oxea in the dermal cortex, but a layer
of oxeote spicules, longitudinally arranged, lies beneath and
parallel to the gastral surface. The tubar skeleton is articulate.

Remarks.—This genus was proposed for the reception of
Haeckel’s remarkable Sycandra hystrix (5), and Schmidt’s
Ute utriculus (19) may perhaps also be included in it.
The genus is obviously a special modification of the
Grantia type. It is, as I have already pointed out, very
unusual to find oxeote spicules in the gastral cortex of any
calcisponge.

Genus 8.—Anamixilla, Poléjaeff (8).

Diagnosis.—Flagellated chambers elongated and radially
arranged. There is no special tubar skeleton, the skeleton
of the chamber layer consisting of large radiate spicules,
arranged without regard to the direction of the chambers,
and of the outwardly directed basal rays of the subgastral
sagittal triradiates.

Remarks.—The derivation of this remarkable genus from
the Grantia type is still indicated by the presence of the
subgastral sagittal triradiates. It may be compared to a
Grantia in which the ordinary articulate tubar skeleton has
been almost entirely replaced by the invasion of large tri-
radiates from the dermal cortex. The only species as yet
known is the Australian Anamixilla torresi, Poléjaeff
(8), tothe account of which given by Poléjaeff I have nothing
to add. I have, however, ventured to slightly alter the original
diagnosis.

Genus 9.—Sycyssa, Haeckel (5).

Diagnosis.—The flagellated chambers are elongated and

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 235

arranged radially around the central gastral cavity. The
skeleton consists exclusively of oxeote spicules.

Remarks.—The only known species of this genus is
Haeckel’s Sycyssa Huxleyi, from the Adriatic (5). The
genus occupies a very isolated position. The remarkable
skeletal characters upon which it is based have been discussed
on a previous page.

Genus 10.—Leucandra (Haeckel [5], emend.), figs. 16, 17.

Diagnosis.—The flagellated chambers are spherical or sac-
shaped, never arranged radially around the central gastral
cavity, with which (or with the main exhalant canals derived
therefrom) they communicate by a more or less complicated
exhalant canal system. The skeleton of the chamber layer is
composed of irregularly scattered radiate spicules, but it may
still present traces of its derivation from a radially symmetrical
type, in the presence of a few subgastral sagittal triradiates.

Remarks.—This genus, as here maintained, is still a very
comprehensive one, and does not correspond exactly to any
which have hitherto been proposed. It includes many species
which would fall under Poléjaeff’s Leuconia (8), but that
author frankly admits that his Leuconia requires sub-
dividing; and, moreover, Vosmaer (14) has shown that Bower-
bank’s name Leuconia, adopted by Poléjaeff, was previously
occupied bya genusof Mollusca. I therefore agree with Vosmaer
in adopting Haeckel’s name Leucandra, but as that genus was
based entirely upon the presence of certain forms of spicules,
without regard to their arrangement, I cannot accept it in the
sense originally intended. On the same principle, I include in
the genus Haeckel’s species of Leucetta, as I do not believe the
mere absence of quadriradiate or oxeote spicules, or both, to
be of generic significance. Indeed, I was strongly inclined to
adopt the name Leucetta, on grounds of priority, for the
genus as now constituted ; but considering that Leucandra,
as employed by Haeckel and Vosmaer, makes the nearest
approach to the genus as now characterised, and considering
also that the name Leucetta has been adopted by Poléjaeff

236 ARTHUR DENDY.

(8) and Vosmaer (14) in a very different sense, I have thought
it desirable to make use of the name chosen.

I believe that very probably the genus will still require sub-
dividing at some future date, but it will be an extremely
difficult matter to satisfactorily characterise the different sub-
divisions; the irregular nature of the skeleton appears to
me to defy classification. It might be possible to maintain
Gray’s genus Aphroceras for species like Leucandra
alcicornis, in which the dermal cortex is composed chiefly of
longitudinally arranged oxea. Aphroceras would then stand
in the same relation to Leucandraas Ute does to Grantia;
but, unfortunately, we find species intermediate in skeletal
characters between the typical Leucandra and the proposed
Aphroceras, in which we have large oxea in the dermal
cortex, but not arranged with regularity parallel to the long
axis of the sponge, nor yet projecting at right angles from the
surface. I have already expressed my hesitation in maintain-
ing the genus Ute; and as the genus Aphroceras has not,
like Ute, come into general use, I do not care to take upon
myself the responsibility of re-establishing it.

Certain species of Leucandra, as already pointed out in
dealing with the skeleton, still exhibit traces of descent from
aradially symmetrical form, in the presence of subgastral and
other sagittal triradiates; and for the present we may con-
veniently regard the genus as being descended from an ancestral
Grantia type, by modification of the canal system and skeleton
along the lines laid down in an earlier part of this paper.

The true relations of many species of the genus are obscured
by the habit of colony formation by the fusion of many
individuals, which gives rise to irregular, massive sponges, as
in the case of Synute; but I do not think that we can
generically separate these species from those which retain the
more primitive condition, with a single central gastral cavity.

My Leucandra phillipensis (4), the anatomy of which
is drawn in fig. 16, is a typical example of the simpler section
of the genus, while the European Leucandra nivea (5) offers
an illustration of the massive colonial habit. Numerous other

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 237

examples are given in my synopsis of the Australian Calcarea
Heteroceela (4).

Genus 11.—Lelapia (Gray [18], emend.).

Diagnosis.—Canal system unknown. Skeleton of the
chamber layer composed of large, longitudinally arranged
oxea, crossed at right angles by bundles of tuning-fork shaped
triradiates whose basal rays are directed towards the dermal
surface.

Remarks.—This genus was first proposed by Gray (18) for
certain remarkable tuning-fork shaped spicules described by
Bowerbank, but the sponge to which those spicules belonged
was then unknown. Carter (12) subsequently described some
sponges, collected by Mr. J. Bracebridge Wilson in the neigh-
bourhood of Port Phillip Heads, in which similar tuning-fork
shaped spicules were present, and he gave to these sponges
Gray’s name, Lelapia australis. Whether Mr. Carter’s
species is really identical with the sponge to which the spicules
described by Bowerbank and Gray belonged, must remain an
open question. The fact that the spicules in question came
from Australia lends an air of probability to the identification,
but then similar spicules are known in the fossil condition
from deposits as old as the Cretaceans, according to Carter
(12). In any case, the Victorian sponges described by Carter
(12) under the name Lelapia australis must stand as the
types of that species, and the species thus constituted must
stand as the type of the existing genus Lelapia.

Unfortunately we know nothing definite as to the canal
system of Lelapia australis, although it is to be inferred,
from the fact that Mr. Carter places it amongst the ‘‘ Leucones,”
that the canal system belongs to the Leuconoid type. The
structure of the skeleton as described by Mr. Carter is very
peculiar, and it is upon this character alone that the generic
diagnosis must, for the present, be based.

The position of the genus in the system of classification is
necessarily only provisional, and it can only be finally deter-
mined by further research on the canal system. Unfortunately

238 ARTHUR DENDY.

Lelapia australis is extremely rare, and although Mr.
Wilson has for some years past sent me the results of his
dredging expeditions, so far as the sponges are concerned, I
have never yet had the good fortune to meet with a specimen.

Genus 12.—Leucyssa, Haeckel (5).

Diagnosis.—Canal system presumably of the Leuconoid
type. Skeleton composed solely of oxeote spicules irregularly
scattered through the sponge.

Remarks.—This is another of those genera of which we
know scarcely anything but the skeleton with any degree of
accuracy. There can, however, be little doubt that the canal
system belongs to the Leuconoid type. The fact that Haeckel
included the genus amongst his Leucones, and the irregular
character of the skeleton, both point to this conclusion. I
have already discussed the relationship of the genus in speak-
ing of the skeleton. For the present it may perhaps best be
regarded as derived from a Leucandra-like type by loss of
the radiate spicules, and their replacement by oxea. Thereare
three very distinct species included by Haeckel (5) in the genus,
viz. Leucyssa spongilla, L. cretacea, and L. incrustans.
All these species are extremely rare, so that the probabilities
of our obtaining a more exact knowledge of the canal system
of the genus are somewhat remote.

Family 4.—Heteroripa, Dendy (4).

Diagnosis.—There is a distinct and continuous dermal
cortex covering over the chamber layer, and pierced by in-
halant pores. Subdermal sagittal triradiates are present.
The flagellated chambers vary from elongated and radial to
spherical and irregularly scattered. An articulate tubar
skeleton may or may not be present.

Remarks.—The leading characteristic of this family, by
which it is distinguished from all others, is the presence of the
subdermal sagittal triradiate spicules, which, to a greater or
less extent, replace the primitive articulate tubar skeleton.
As regards the canal system, we meet with much the same

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 239

series of variations as in the Grantide; and the shape and
arrangement of the chambers, as in that family, can only be
utilised as an aid in diagnosing the genera.

The family is evidently derived from an ancestral Grantia-
like type, as is clearly indicated by the retention of the primi-
tive radial arrangement of the chambers, and the articulate
tubar skeleton, by the least modified species (fig. 18).

The family, as here maintained, is not nearly identical with
any which have hitherto been proposed. No previous writers
have drawn that sharp distinction between subdermal sagittal
triradiates and subdermal quadriradiates which I have indi-
cated in dealing with the skeleton, and which seems to me to
be of the greatest value for purposes of classification.

I have adopted the name ‘“‘ Heteropide,”’ not because I
regard the genus Heteropia as most typical of the family,
but because a family name derived from the principal genus,
Grantessa, would be liable to confusion with the name
“ Grantide,” already used for the preceding family.

Genus 13.—Grantessa (von Lendenfeld [11], emend.),
fig. 18.

Diagnosis.—The flagellated chambers are elongated, and
arranged radially around the central gastral cavity. The
dermal cortex consists principally of triradiates, and does not
contain longitudinally disposed oxea.

Remarks.—In working over the Australian Heterocela I
met with an extensive series of specimens, belonging to at least
six species,! which evidently formed avery natural assemblage,
characterised by essentially the same peculiarities of skeleton
and canal system. Most of these species had already been
described by various authors under a variety of generic names,
viz. Amphoriscus( Poléjaeff), Grantessa (von Lendenfeld),
Heteropia (Carter), Hypograntia (Carter),and Leuconia
(Carter).

From these names I selected Grantessa for the group of
species in question, for the following reasons :—The name Am-

1 For a list of these species, with synonyms, vide 4.
VOL. 35, PART 2.—NEW SER. s

240 ARTHUR DENDY.

phoriscus is occupied by a distinct genus of sponges, which
ought not to be confounded with the one under consideration ;
the name Leuconia is altogether out of the question, for
reasons already discussed ; and the name Grantessa has pri-
ority over both Heteropia! and Hypograntia.

The genus Grantessa of von Lendenfeld (11), however, is
by no means synonymous with the genus here maintained.
The original type of the genus is Grantessa sacca, and its
author bases the generic distinction upon the presence of tufts
of oxeote spicules projecting more or less at_right angles from
the dermal surface, and arranged without regard to the radial
chambers (10). In his description of G. sacca, however, he
mentions the presence of “ dermal” sagittal triradiates with
an inwardly directed basal ray, but he evidently does not con-
sider this character as of generic, much less of family im-
portance. Attributing, as I do, great importance to the
presence of subdermal sagittal triradiates, and very little im-
portance to the arrangement of the dermal oxea in tufts (which
bear no relation to the radial chambers), I have felt it neces-
sary, while adopting the name Grantessa, to give an entirely
new significance thereto.

Genus 14.—Heteropia (Carter [15], emend.).

Diagnosis.—The distal ends of the elongated radial
chambers are covered over by a well-developed dermal cortex,
consisting principally of large oxea arranged parallel to the
long axis of the sponge.

Remarks.—The name Heteropia was first applied by
Carter to his Aphroceras ramosa (15), and I therefore
consider that species as the type of the genus. The name was
also applied by the same author and at about the same time
(12) to a number of Victorian sponges, but I have already
pointed out that for the latter the name Grantessa must
take priority. Indeed, it is perhaps doubtful whether Hete-
ropia ramosa deserves to be generically separated from
Grantessa, to which genus it bears exactly the same relation

1 Heteropia is also in use for another genus.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 241

as Ute does to Grautia. As, however, there appear to be no
intermediate stages known between Grantessa and Hetero-
pia, the latter genus appears to have just as much right to stand
separately as Ute has, and the same objection does not hold
good as in the case of Leucandra and Aphroceras previously
discussed. Even should the genus have to be abandoned, still
the name “ Heteropide’’ may be conveniently retained for
the family. The only known species of the genus is H.
ramosa.

Genus 15.—Vosmaeropsis, Dendy (4), fig. 19.

Diagnosis.—Flagellated chambers spherical or sac-shaped,
never truly radial. Dermal cortex composed principally of
triradiates, without longitudinally disposed oxea.

Remarks.—This genus stands in much the same relation to
Grantessa as Leucandra does to Grantia. The canal
system of Vosmaeropsis macera (fig. 19) affords an excellent
example of the “Sylleibid” type which characterises von
Lendenfeld’s genera, Polejna and Vosmaeria (10); but I
have already expressed my opinion that the mere presence of
this type of canal system cannot be utilised for purposes of
classification, as it is found in several very distinct forms which
we cannot unite in defiance of the great structural differences
in their skeleton. It might be doubted whether Vosmaer-
opsis should be generically separated from Grantessa, as
this separation depends entirely upon the canal system; but,
although I do not consider the distinction between the
Sylleibid and Leuconoid types of canal system sufficiently well
marked to be utilised for generic purposes, I think we may
very conveniently thus utilise the much greater distinction
between the Syconoid and Leuconoid types, considering the
Sylleibid one, for purposes of classification, as belonging to
the Leuconoid division.

I only know of three species which can be included in the
genus Vosmaeropsis, viz. V. macera (4), V. depressa (4),
and V. Wilsoni (4).

242 ARTHUR DENDY.

Family 5.—Ampuoriscipa, Dendy (4).

Diagnosis.—There is a distinct and continuous dermal
cortex over the chamber layer. Conspicuous subdermal
quadriradiate spicules, with inwardly directed apical rays, are
present. The flagellated chambers vary from elongated and
radially arranged to spherical and irregularly scattered.

Remarks.—In this family, which does not by any means
correspond to von Lendenfeld’s “ Amphoriscine” (10), the
distinguishing characteristic. is the presence of subdermal
quadriradiate spicules with inwardly directed apical rays.
These, as I have already pointed out, are not to be regarded
as homologous with the subdermal sagittal triradiates of the
Heteropide, although they appear to fulfil the same function.
We find in this family exactly the same series of variations in
the canal system as in the two preceding, and, as before, I
have utilised this variation, so far as the Syconoid and
Leuconoid types are concerned, for purposes of generic dis-
tinction, including the Sylleibid type in the Leuconoid.

The family is evidently to be derived from a Grantia-like
ancestral type, by the development of conspicuous and inwardly
directed apical rays by the triradiates of the dermal cortex.

Genus 16.—Heteropegma, Poléjaeff (8), fig. 20.

Diagnosis.—The flagellated chambers are elongated and
arranged radially around the central gastral cavity. There is
a vestigial tubar skeleton of minute radiates. The dermal
cortex is very thick, composed principally of triradiate spicules.

Remarks.—This genus is extremely interesting as showing
how the primitive articulate tubar skeleton, which is on the
verge of disappearance, is gradually replaced by the development
of the subdermal quadriradiates (fig. 20). Poléjaeff (8) makes
no mention of the subdermal quadriradiates in his diagnosis of
the genus, although he was well aware of their presence, and
I have therefore been obliged to draw up a fresh diagnosis.

The genus was founded for the reception of Heteropegma
nodus-gordii from off the Bermudas and Cape York, and

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES, 243

we now also know a species, H. latitubulata (4), very closely
resembling the first, from Victorian waters.

Genus 17,—Amphoriscus (Haeckel [20], emend.).

Diagnosis.—The flagellated chambers are elongated, and
arranged radially around the central gastral cavity. There is
no articulate tubar skeleton, but, in addition to the subdermal
quadriradiates, subgastral sagittal triradiates or subgastral
quadriradiates may also be present.

Remarks.—In his “‘ Prodromus” (20) Haeckel proposed the
generic name Amphoriscus for a very natural group of three
species, all characterised by the Syconoid type of canal
system and the presence of quadriradiate spicules only in the
skeleton. Of the highly characteristic arrangement of the
skeleton, with subdermal and subgastral quadriradiates whose
apical rays poiut in opposite directions, his generic diagnosis
takes no notice. This arrangement, however, is exhibited by
all the species referred by Haeckel to the genus, and affords,
to my mind, a much more suitable foundation for a generic
diagnosis. In his monograph (5) Haeckel altered the generic
name to Sycilla, but retained the old diagnosis (“Sycones
spiculis quadricruribus’’). He also added a fourth species
to the genus.

Poléjaeff (8) retains the genus Amphoriscus, but extends
the diagnosis to include species with subdermal sagittal tri-
radiates, which, in my opinion, must be kept quite separate
in the genus Grantessa. His Amphoriscus poculum and
A. flamma both belong to the genus Grantessa. His
A. elongatus, however, must be retained in the genus
Amphoriscus, and is, indeed, a very noteworthy species ;
for, although the articulate tubar skeleton has been lost, the
subgastral sagittal triradiates still persist, and have not yet
been replaced by subgastral quadriradiates, as in all Haeckel’s
species of the genus. Von Lendenfeld (10) follows Poléjaett
in including in the genus Amphoriscus species with sub-
dermal sagittal triradiates as well as those with subdermal
quadriradiates, while he separates, quite unnecessarily to my

244, ARTHUR DENDY.

mind, those species which happen to have oxeote spicules as a
distinct genus, to which he gives the name Ebnerella.

As examples of the genus Amphoriscus as maintained by
me, I may cite all Haeckel’s species of Sycilla (5) and Polé-
jaeff’s Amphoriscus elongatus (8).

Genus 18.—Syculmis (Haeckel [5], emend.).

Diagnosis.—The flagellated chambers are elongated, and
arranged radially around the central gastral cavity. The
skeleton of the chamber layer is composed of the apical rays
of subdermal and subgastral quadriradiates. There is a root-
tuft of oxea and anchoring quadriradiates.

Remarks.—Haeckel’s diagnosis (5) of the genus was based
entirely upon the spiculation, without regard to the arrange-
ment of the skeleton. As it happened, however, he only knew
one species—the remarkable Syculmis synapta—which
possesses only quadriradiate and oxeote spicules; and there-
fore, although it is necessary to alter the generic diagnosis, it
is not necessary to make any change in the extent of the genus.
Syculmis synapta (5) is evidently only a very special modi-
fication of the Amphoriscus type.

Genus 19.—Leucilla (Haeckel [5], emend.),
figs. 21, 22.

Diagnosis.—Flagellated chambers spherical or sac-shaped,
never truly radial.

Remarks.—This genus occupies a position amongst the
Amphoriscide, analogous to that occupied by Leucandra
amongst the Grantide, and Vosmaeropsis amongst the
Heteropide, being distinguished by the Leuconoid (or Syl-
leibid) character of the canal system. Leucilla uter, as I
have already pointed out, affords an excellent example of the
Sylleibid type (fig. 21), while in Leucilla australiensis
(fig. 22) we meet with a more typically Leuconoid modification.
Even in Leucilla australiensis, however, we often find the
flagellated chambers more or jess elongated, thus showing how

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES, 245

impossible it is to draw a sharp line of distinction between
the two types.

The genus Leucilla was founded by Haeckel (5) for Leu-
conoid sponges in which the spicules are all quadriradiate, of
course without regard to the arrangement of the skeleton.
The first species which he describes is Leucilla amphora,
and I have taken this as the type of the genus, and constructed
the generic diagnosis accordingly. The only other species
described by him, L. capsula, also comes into the genus as
now constituted. It thus appears that our Leucilla is a
much more comprehensive genus than Haeckel’s, and this is
because we include in the genus species with triradiate and
oxeote spicules.

Poléjaeff (8) has attached a very different significance to the
name Leucilla, giving to the genus much the same character
as von Lendenfeld has given to his family Sylleibide ; indeed,
the latter was founded (11) chiefly upon Poléjaeff’s Leucilla.
I revert, however, to Haeckel’s two species, and make their
characters the foundation of the genus.

In my synopsis of the Australian Calcarea Heterocela
(4) I, somewhat unnecessarily I fear, proposed the name
Paraleucilla for Haeckel’s Leucandra cucumis, which is
characterised by the presence of distinct subdermal cavities
supported by a somewhat specialised portion of the skeleton of
the chamber layer. I now regret having taken this step, the
more so as Poléjaeff (8) had previously proposed the name
Pericharax for the same sponge. Poléjaeff, however, also
included in the genus Pericharax his P. Carteri, which is a
very different sponge from Leucandra cucumis, and does not
even come within the limits of our family Amphoriscide.
I do not now think that either Pericharax or Paraleucilla
ought to stand as a distinct genus. Poléjaeff’s Pericharax
Carteri is but a slight modification of the ordinary Leu-
candra type, while Haeckel’s Leucandra cucumis is a
similar modification, only rather more marked, of the Leucilla

type.

246 ARTHUR DENDY.

VI. Tue Origin anp PHyLoGENY oF THE CaLCAREA
HETEROC@LA.

No one will probably dispute the now well-established
hypothesis that the Calcarea Heteroceela are descended from
some one or more ancestral Homocela. The Homoceela are
undoubtedly the more primitive of the two great groups into
which Poléjaeff (8) divided the calcareous sponges, and the only
question which we need discuss is the process by which the
Homoceele gave rise to the Heterocele type.

In my memoir on the organisation and classification of the
Homoceela (1) I proposed to divide the sole genus, Leucoso-
lenia, into three sections, according to the nature of the canal
system, for even in the simplest of sponges the canal system
exhibits a considerable amount of variation. To these sections
of the genus I applied the names Simplicia, Reticulata, and
Radiata. The Simplicia include such simple Olynthus
types as never form colonies, and also those colonial forms in
which the whole colony consists of individuals (Ascon-persons)
which may branch, but which never form complex anastomoses
nor give off radial tubes, so that the individuality of the
different members of the colony is always recognisable. In
the Reticulata the sponge colony forms a more or less
complex network of branching and anastomosing tubes, and it
is no longer possible to distinguish the individual Ascon-persons
of which the colony is composed. The Radiata include such
species as exhibit a radiate structure—the sponge consisting of
a single central Ascon-tube, from which other tubes are
budded off radially.

In all species of Leucosolenia it is supposed that the
whole of the primitive gastral cavity is lined by a layer of
collared cells, and that when, as in my type D of the Reticu-
lata, exhalant canals exist which are not lined by collared
cells, these canals (pseudogasters) really le outside the
sponge, and are probably formed by upgrowth of the colony
around them.

The Olynthus type, consisting of a simple sac lined by

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 247

collared cells, and pierced by an osculum and prosopyles, is
universally admitted to be the ancestral form of all calcareous
sponges. If we look for a more advanced type amongst the
Homoceela, from which the Heterocela may possibly be de-
rived, the radiate section of the genus Leucosolenia at once
suggests itself. The type of this radiate section is Leucoso-
lenia tripodifera (1) ; and Bidder (21) is of opinion that
certain other species, notably Leucosolenia Lieberkuhnii,
should also be included therein. In L.tripodifera we find a
single wide central tube, with a terminal osculum, numerous
prosopyles, and a thin wall lined by collared cells—so far a
typical Ascon individual. From this tube, numerous radially
arranged branches are given off, which themselves branch copi-
ously, and terminate all at about the same level in blind,
rounded extremities, which touch one another, and thus form
an even surface to the whole sponge. Each of the radial
branches repeats, on a smaller scale, the structure of the
parent tube, and each is first formed as a hollow bud or out-
growth from the latter, the youngest buds lying at a little
distance below the osculum. The radial tubes occasionally
anastomose and inter-communicate, like the branches of a
reticulate Leucosolenia, but this appears to take place only
occasionally ; at any rate, it is not a characteristic feature. The
skeleton of this sponge consists of rather slender sagittal tri-
radiates and quadriradiates, and of the “tripod” spicules.
The latter are confined to the distal extremities of the radial
chambers. The sagittal radiates are arranged in a single layer
in the walls of the tubes, the quadriradiates with the apical ray
projecting into the gastral cavities both of the central tube
and its radial branches. As already stated, the spicules are
sagittal, and in the central tube the basal rays point towards
the base of the sponge; while in the radial tubes the basal
rays point towards the blind distal ends, exactly as in the
articulate tubar skeleton of Syconoid Heteroceela.

Here, then, we have a form which makes a very near
approach, both in canal system and skeleton, to the Sycetta
type amongst the Heterocela, and it is from some such

248 ARTHUR DENDY.

radiate Homoccelous form as this that the Heterocela have
probably been derived. I do not by any means wish to single
out Leucosolenia tripodifera itself as the ancestor of the
Heteroceela, for I do not for a moment believe that it is so.
The copious branching of the radial tubes, and the fusion of
adjacent tubes at the points of contact, indicate a more
advanced condition in some respects than is met with in the
simplest Heteroceela (Sycetta). The youngest part of the
sponge, nearest to the osculum, where the radial tubes are still
simple and unbranched, makes a much nearer approach to the
hypothetical ancestral form.

In any case, the simple radiate type of structure, though
perhaps no more highly specialised than certain other types
met with amongst the Homoceela, appears to form the natural
starting-point for all the divers types met with amongst the
Heteroceela, with the probable exception of Leucascus, to
which I shall refer again later on. The simpler forms of
Heteroccela still preserve the radiate character, which becomes
at first even more pronounced in the structure of the skeleton.
It is also a most important and significant fact, that in the
ontogeny of these simple Heterocela the radial tubes
(flagellated chambers) are formed as outgrowths from a central
tube, exactly as in Leucosolenia tripodifera. An excellent
illustration of this mode of formation is given by Korschelt and
Heider (2).

I ought, perhaps, in this place to say something about Dr.
von Lendenfeld’s supposed families “ Homodermide” and
« Leucopside.” Ihave, however, already expressed my views
on this subject in an earlier paper (1), and those views I still
maintain, for I cannot see that there is anything in Dr. von
Lendenfeld’s observations, including those published in “ Die
Spongien der Adria” (10), to justify us in accepting these
families. It would take too long to argue the question here,
and I am quite content to refer the reader to Dr. von Lenden-
feld’s own writings.

I maintain, then, that the ancestral form of the Heteroccela
was a simple, radiate, Homocele sponge, simpler even than

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 249

Leucosolenia tripodifera, and that the radiate Heteroccele
type was developed from this by replacement of the collared
cells of the central gastral cavity by a flattened epithelium and
specialisation of the skeleton, a modification which might
readily give rise to the primitive Heterocele genus Sycetta.
The manner in which I believe the majority of the remaining
genera of Heterocela to have been derived from a Sycetta-
like ancestor, has perhaps been already sufficiently indicated
in dealing with the anatomy and classification of the group.
The genus Leucascus, however, does not appear ever to have
passed through a Sycetta stage, for while the greatly
elongated flagellated chambers may indicate by their arrange-
ment (at any rate in some specimens) a certain degree of radial
symmetry, the skeleton exhibits none whatever; it is quite
as irregular as that of a reticulate Homoceele, or as that
of the most modified Leucandra. Considering the nature
of the canal system, however, I believe that the irregularity
of the skeleton in Leucascus is a primitive condition, and
not, as in Leucandra,asecondary one. Possibly Leucascus
may be either derived directly from a reticulate Homoccele
ancestor, by the formation of a true dermal membrane and
dermal pores, or from a very low type of radiate Homoceele,
in which the skeleton had not yet acquired any radial symmetry.
I am inclined to adopt the latter view.

The opinions as to the origin and inter-relationships of the
Calcarea Heterocela which I have endeavoured to justify
in the foregoing pages, may be conveniently summarised in

the accompanying diagram, which naturally takes the form
of a genealogical tree.

250 ARTHUR DENDY.

LEUCANDRA
LEUCYSSA LEUCILLA
CRANTIOPSIS VOSMAEROPSIS
GRANTIA LELAPIA HETEROPEGMA
SYNUTE GRANTESSA
HETEROPIA \
UTE. SYCULMIS
aa ee AMPHORISCUS
SYCYSSA BY
ANAMIXILLA kn ore
OAS
WAS
> x“
& os
2
YCON
os SYC
fe)
SYCANTHA
SYCETTA
x
3s
> ray
b ©
°,
4 $ LEUCASCUS
Le Ww
2 fo)
13) oO
$ vs
oS
Ke

VIL. Rererence List or Literature.

1. Denpy.—‘“ A Monograph of the Victorian Sponges. Part 1. The
Organisation and Classification of the Calcarea Homoccela, with De-
scriptions of the Victorian Species,” ‘Transactions of the Royal
Society of Victoria,’ vol. ili, part 1.

2. Korscuett and Herppr.—‘ Lehrbuch der vergleichenden Entwicke-
lungsgeschichte der wirbellosen Thiere.’ Porifera.

8. Von LenpEnFELD.—“< The Homoceela hitherto described from Australia
and the New Family Homodermide,” ‘ Proceedings of the Linnean
Society of New South Wales,’ vol. ix, part 4.

4. Dunpy.—‘‘ Synopsis of the Australian Calearea Heterocela; with a
Proposed Classification of the Group and Descriptions of some New
Genera and Species,” ‘ Proceedings of the Royal Society of Victoria,’
vol. v.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 251

5. Harcket.— Die Kalkschwimme.”

6. Sorzas.—Article “Sponges,” in the ‘Encyclopedia Britannica,’
ed. 9.

7. Scuutze.—* Ueber den Bau und die Entwickelung von Sycandra
raphanus,” ‘ Zeitschr. f. wiss. Zoologie,’ Band xxv, suppl.

8. Potisanrr.— Report on the Calcarea of the ‘‘ Challenger ” Expedition.’

9. Denpy.—“ Studies on the Comparative Anatomy of Sponges. III. On
the Anatomy of Grantia labyrinthica, Carter, and the so-called
Family Teichonide,” ‘Quarterly Journal of Microscopical Science,’
vol. xxxii, N.S.

10. Von Lenprenretp.—“ Die Spongien der Adria. I. Die Kalkschwamme,”
‘Zeitschr. f. wiss. Zoologie,’ Band liii.

11. Von Lunpenretp.—“A Monograph of the Australian Sponges,”
‘Proceedings of the Linnean Society of New South Wales,’ vols.
ix and x.

12. CarteR.—“ Descriptions of Sponges from the Neighbourhood of Port
Phillip Heads, South Australia,” ‘Annals and Magazine of Natural
History,’ 1885—1887.

13. Drnpy.— Preliminary Account of Synute pulchella, a New Genus
and Species of Calcareous Sponges,” ‘ Proceedings of the Royal Society
of Victoria,’ vol. iv.

14. VosmaEr.— Porifera,”’ in Bronn’s ‘ Klassen und Ordnungen des Thier-
Reichs.’

15, Cartrer.—“ Description of Aphroceras ramosa,” ‘ Proc. Lit. Phil.
Soc. Liverpool,’ vol. xl, appendix.

16. Denpy.—“ Description of a New Species of Leucosolenia from the
Neighbourhood of Port Phillip Heads,” ‘Proceedings of the Royal
Society of Victoria,’ vol. v.

17. Scumipt.—‘ Die Spongien des Adriatischen Meeres,’ supplement 1.

18. Gray.— Notes on the Arrangement of Sponges, with the Description
of some New Genera,” ‘Proceedings of the Zoological Society of
London,’ 1867.

19. Scumipt. — ‘ Grundziige einer Spongien—Fauna des atlantischen
Gebietes.

20.—HarcKEL.—“ Prodromus eines Systems der Kalkschwimme,”’ ‘ Jenaische
Zeitschrift,’ Band v.

21. Bipper.—Review of “A Monograph of the Victorian Sponges,”
* Quarterly Journal of Microscopical Science,’ vol. xxxii, N.S.

22. BrippER.—“‘ Preliminary Note on the Physiology of Sponges,” ‘ Proceed-
ings of the Cambridge Philosophical Society,’ vol. vi.

252

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

ARTHUR DENDY.

Bipper.—* Note on Excretion in Sponges,” ‘ Proceedings of the Royal
Society of London,’ vol. li.

Brpper.—* On the Flask-shaped Ectoderm and Spongoblasts in
one of the Keratosa,”’ ‘ Proceedings of the Royal Society of London,’
vol. lii.

Mincutn.—* The Oscula and Anatomy of Leucosolenia clathrus,”
‘Quarterly Journal of Microscopical Science,’ vol. xxxiii, N. 8.

Mincuin.—“Some Points in the Histology of Leucosolenia
(Ascetta) clathrus,” ‘ Zoologischer Anzeiger,’ No. 391.

Scuutze.— Untersuchungen ueber den Bau und die Entwickelung
der Spongien. IX. Die Plakiniden,” ‘Zeitschr. f. wiss. Zoologie,’
Band xxxiv.

Metscuntkorr.— Spongiologische Studien,” ‘ Zeitschrift f. wissensch.
Zoologie,’ Band xxxii.

Metscunikorr.—‘ Legons sur la Pathologie comparée de I Inflam-
mation.’

Cartrer.—“ Notes Introductory to the Study and Classification of
the Spongida,” ‘ Annals and Magazine of Natural History,’ 1875.

Denpy.—“ Studies on the Comparative Anatomy of Sponges. IV. On the
Flagellated Chambers and Ova of Halichondria panicea,”’ ‘ Quar-
terly Journal of Microscopical Science,’ vol. xxxii, N.S.

Mrncutn.—‘‘ Note on a Sieve-like Membrane across the Oscula of a
Species of Leucosolenia, with some Observations on the His-
tology of the Sponge,” ‘ Quarterly Journal of Microscopical Science,’
vol. xxxili, N.S.

Dernpy.—* Studies on the Comparative Anatomy of Sponges. II. On
the Anatomy and Histology of Stelospongus flabelliformis,
Carter, with Notes on the Development,” ‘Quarterly Journal of
Microscopical Science,’ vol. xxix, N. 8.

PontsaErr.—‘“‘ Ueber das Sperma und die Spermatogenese bei Sycandra
raphanus, Haeckel,” ‘Sitzb. der k. Akad. der Wissensch. Wien,’
Band lxxxvi.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES, 253

EXPLANATION OF PLATES 10—14,

Illustrating Dr. Dendy’s paper ‘ Studies on the Comparative
Anatomy of Sponges.”

Note.—The majority of the drawings have been made from paraffin sections
of ordinary spirit material, and in most cases a camera lucida has been
employed.

In Figs. 1—22 the collared cells are diagrammatically represented by red
dots, and the outlines of the spicules are drawn in blue.

Reference Letters.

cale. Calcoblast. ¢. c. Collared cell. c.g. c. Central gastral cavity.
ch, di. Diaphragm of exhalant opening of flagellated chamber. d. cor. Dermal
cortex. d.p. Dermal pore. ec/. ep. Ectodermal pavement epithelium. emdé.
Embryo. emb.c. Embryo capsule. ed. ep. Endodermal pavement epithe-
lium. e#.c. Exhalant canal. ez. op. Exhalant opening of flagellated chamber,
Ji. ch. Flagellated chamber. (fl. ch. x. Flagellated chamber in the contracted
condition. g.cor. Gastral cortex. g.g. Apical ray of gastral quadriradiate
spicule, projecting into the central gastral cavity. 7. c. Inhalant canal.
mus. c. Muscle cell. ose. Osculum. ov.Ovum. pros. Prosopyle. s.d.g. Sub-
dermal quadriradiate spicule. s.d.s. Subdermal sagittal triradiate spicule.
s.g. 8. Subgastral sagittal triradiate spicule. s. m. Sollas’s membrane. sp.
Spicule. sp. s. Spicule sheath (formed from the gelatinous ground substance
of the mesoderm). s¢. c. Stellate mesoderm-cell. 7. ov. Tuft of oxeote
spicules at the end of a radial chamber.

PLATE 10.

Fre. 1.—Leucascus simplex. Portion of a vertical section, passing, on
the left, through the osculum. Drawn under Zeiss A, ocular 2.

Fie. 2.—Sycon Carteri. Portion of a transverse (horizontal) section,
showing three of the radial chambers. Drawn under Zeiss C, ocular 2.

Fic. 3.—Sycon gelatinosum. Portion of a longitudinal (vertical) sec-
tion, passing through the distal ends of the radial chambers. Drawn under
Zeiss C, ocular 2.

Fic. 4.—Sycon gelatinosum. Portion of a longitudinal (vertical) sec-

tion, passing through the proximal ends of the radial chambers. Drawn under
Zeiss C, ocular 2.

254. ARTHUR DENDY.

Fie. 5.—Sycon gelatinosum. Portion of a tangential section, cutting
across the radial chambers. Drawn under Zeiss C, ocular 2.

Fie. 6.—Sycon gelatinosum. Portion of a tangential section of the
dermal surface showing the tufts of nail-shaped oxea which crown the distal
ends of the radial chambers, and the pore-bearing membrane stretched between
them. Drawn under Zeiss C, ocular 2.

Fic. 7.—Sycon boomerang. Portion of a transverse (horizontal) sec-
tion, showing one much-branched radial chamber. Drawn under Zeiss A,
ocular 2.

Fic. 8.—Sycon boomerang. Portion of a tangential section of the
dermal surface, showing the tufts of oxea which crown the distal ends of the
radial chambers, and the pore-bearing membrane (containing a few spicules)
stretched between them. Drawn under Zeiss C, ocular 2.

PLATE 11.

Fie. 9.—Grantia extusarticulata. Portion of a transverse (hori-
zontal) section. Drawn under Zeiss A, ocular 2.

Fie. 10.—Grantia Vosmaeri. Portion of a longitudinal (vertical) sec-
tion. Drawn under Zeiss A, ocular 2.

Fic. 11.—Grantiopsis cylindrica. Portion of a transverse section.
Drawn under Zeiss A, ocular 2.

Fie. 12.—Ute syconoides. Portion of a transverse (horizontal) sec-
tion. (The majority of the spicules have been dissolved out by the action of
acid alcohol, but the large oxea of the dermal cortex are shown cut across.)
Drawn under Zeiss C, ocular 2.

Fie. 13.—Ute syconoides. Portion of a tangential section cutting
across the radial chambers. Drawn under Zeiss C, ocular 2.

Fic. 14.—Ute syconoides. Portion of a tangential (longitudinal) sec-
tion, passing above through the dermal surface and below through the dilated
distal ends of the radial chambers. Drawn under Zeiss A, ocular 2.

PLATE 12.

Fic.15.—Synute pulchella. Portion of a transverse (horizontal) sec-
tion, showing the central gastral cavities of three Ute individuals, each sur-
rounded by radial chambers and all together invested in a common cortex.
Drawn under Zeiss A (with the bottom lens removed), ocular 2.

Fie. 16.—Leucandra phillipensis. Portion of a transverse (horizontal)
section. Drawn under Zeiss A, ocular 2.

Fie. 17—Leucandra australiensis. Portion of a transverse (hori-
zontal) section. Drawn under Zeiss A, ocular 2.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 255

PLATE 13.

Fic. 18.—Grantessa intusarticulata. Portion of a transverse (hori-
zontal) section. Drawn under Zeiss A, ocular 2.

Fic. 19.—Vosmaeropsis macera. Portion of a transverse (horizontal)
section. Drawn under Zeiss A, ocular 2.

Fic. 20.—Heteropegma nodus-gordii. Portion of a transverse sec-
tion. Drawn under Zeiss A, ocular 2.

Fie, 21.—Leucilla uter. Portion of a vertical section. Drawn under
Zeiss A, ocular 2.

Fic. 22.—Leucilla australiensis. Portion of a longitudinal (vertical)
section. Drawn under Zeiss A, ocular 2.

PLATE 14.

Fic. 23.—Vosmaeropsis Wilsoni. Portion of the cortical inhalant
canal system, as seen in a transverse section of a specimen killed with osmic
acid. The space occupied by the dermal cortex, between the inhalant canals,
is left blank, and the pore-bearing dermal surface is shown in perspective.
Drawn under Zeiss C, ocular 2.

Fig. 24.—Ute syconoides. Portion of a tangential section cutting
across the radial chambers, showing a chamber in the contracted condition,
surrounded by four ordinary chambers and four inhalant canals, portions of
which only are drawn. Drawn under Zeiss H, ocular 2.

Fic. 25.—Leucandra sp. One of the rounded flagellated chambers cut
in half. One row of collared cells, with collars united by Sollas’s membrane,
is seen round the margin, while the observer looks down upon the exhalant
opening of the chamber and two prosopyles. Drawn under Zeiss F, ocular 2,

Fie, 26.—Leucandra sp. Exhalant aperture of another chamber from
the same specimen as Fig. 25, surrounded by the membranous chamber
diaphragm, in which the nuclei and granules of muscle-cells are seen. Drawn
under Zeiss F, ocular 2.

Fig. 27.—Grantessa intusarticulata. Portion of a section through
the gastral cortex and proximal end of a radial chamber. On the right a
portion of the chamber diaphragm, marking the junction of the radial chamber
with the exhalant canal, is seen in section. On the left a portion of the inner
end of an inhalant canal is seen. Drawn under Zeiss F, ocular 2.

Fic. 28.—Grantessa intusarticulata. Portion of a section through
the gastral cortex, showing two of the endodermal pavement cells in section
and a stellate cell embedded in the gelatinous ground substance of the meso-
derm. Drawn under Zeiss F, ocular 2.

Fic. 29.—Grantessa intusarticulata. Three contracted ectodermal

VOL. 39, PART 2,— NEW SER. T

256 ARTHUR DENDY.

pavement cells from the lining of an inhalant canal. Drawn under Zeiss F,
ocular 2.

Fic. 30.—Grantessa intusarticulata. Portion of the wall of a radial
chamber looked down upon, showing a number of contracted collared cells, and
between them a prosopyle, with the nucleus of an ectodermal pavement cell

ou its margin seen at a slightly higher focus than the collared cells. Drawn
under Zeiss F, ocular 2.

Fie. 31—Vosmaeropsis macera. LHxhalant opening of a flagellated
chamber, surrounded by the membranous chamber diaphragm containing
muscle-cells. Drawn under Zeiss F, ocular 2.

Fic. 32.—Leucandra sp. Small portion of a vertical section through the
region of the osculum, showing four of the blister-like epithelial cells which
line the gastral cavity in this region. Drawn under Zeiss F, ocular 2.

Fie. 33.—Leucandra echinata, var. An ameceboid cell from the meso-
derm. Drawn under Zeiss F, ocular 2.

Fic. 34.—Leucandra echinata (?). An ovum from a cavity in the
mesoderm. Drawn under Zeiss F, ocular 2.

Fic. 35.—Ute syconoides. Section across an inhalant canal (inter-
canal), showing an ovum suspended from its wall. Drawn under Zeiss F,
ocular 2.

Fic. 36.—Ute syconoides. An ovum from behind the wall of a radial
chamber. Drawn under Zeiss F, ocular 2.

Fie. 87.—Ute syconoides. Two vesicular cells from beneath the epi-
thelium of the gastral cortex. Drawn under Zeiss F, ocular 2.

Fic. 38.—Grantessa intusarticulata, Section of an embryo lying in
a cavity lined by epithelial cells (the embryo capsule). Drawn under Zeiss F,
ocular 2.

Fies. 89—42.—Leucandra phillipensis. Stellate mesoderm-cells from
the dermal cortex. Drawn under Zeiss F, ocular 2.

Fic. 43.—Leucandra phillipensis. Subdermal gland-cell beneath the
wall of an inhalant canal, which is cut through on the right. Drawn under
Zeiss F, ocular 2.

Fies. 44—47.—Leucandra phillipensis. Portions of four oxeote

spicules from the dermal cortex, with calcoblasts attached. Drawn under
Zeiss F, ocular 2.

Fies. 48—50.—Leucandra phillipensis. Ameboid cells from the
mesoderm between the flagellated chambers. The one shown in Fig. 50
appears to be feeding by means of pseudopodia upon the collared cells of a
flagellated chamber. Drawn under Zeiss F, ocular 2.

Fie, 51.—Leucandra phillipensis. A group of retracted collared cells,
Drawn under Zeiss F, ocular 2.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 257

Fies. 52—54.—Grantiopsis cylindrica. Calcoblasts from the dermal
cortex. In Fig. 52 the calcoblast is lying upon a ray of a large triradiate
spicule. In Fig. 53 the spicule is partially dissolved by the acid alcohol, and
the spicule-sheath is visible. Drawn under Zeiss F, ocular 2.

Fig. 55.—Grantiopsis cylindrica. Four small mesoderm-cells from
the dermal cortex. Drawn under Zeiss F, ocular 2.

Fic. 56.—Grantiopsis cylindrica, Three subdermal gland-cells, each
connected by a long, slender process with the granular dermal surface. Drawn
under Zeiss F, ocular 2.

Fie. 57.—Grantiopsis cylindrica. Small portion of the dermal surface.
Drawn under Zeiss F, ocular 2.

Fie. 58.—Micro-organisms from the dermal surface of Grantiopsis
cylindrica. Drawn under Leitz =4, oil immersion.

Fic. 59.—Sycon Ramsayi. Contracted epithelium from an inhalant
canal. Drawn under Zeiss F, ocular 2.

Fie. 60.—Sycon Ramsayi. Contracted epithelium from an exhalant
canal. Drawn under Zeiss I’, ocular 2.

Fie. 61.—Sycon Ramsayi. Sections of two epithelial cells from an ex-
halant canal. Drawn under Zeiss F, ocular 2.

Fig. 62.—Grantessa erinaceus. Contracted epithelium of an endo-
gastric septum, formed by outgrowth of the gastral cortex into the gastral
cavity. Drawn under Zeiss F, ocular 2.

Fic. 63.—Vosmaeropsis Wilsoni. Epithelium from the upper surface
of an oscular diaphragm. From an osmic acid specimen mounted in glycerine.
Drawn under Zeiss F, ocular 2.

Fic. 64.—Vosmaeropsis Wilsoni. Contracted epithelium from an exs
halant canal. Drawn under Zeiss F, ocular 2.

(Tey si2
ie ai ree Pry

rie =) ’
od ul

Bs
=

oe wnt ba as go
fi ies

‘Jit : i aq ae oe eke

on ve iad
~ Cate, wa
’ } 7 i A

, Rs a 5 tahoe
i" = he
ag . ¢

(l Aoi

— >
aa)
«? 7 = y ‘ wit en

J = c 700

>

REPRODUCTIVE ELEMENTS IN APUS AND BRANCHIPUS, 259

Some Points in the Origin of the Reproductive
Elements in Apus and Branchipus.

By

J. E. S. Moore, A.R.C.S.,
From the Huxley Research Laboratory, Royal College of Science, London.

With Plates 15 and 16.

Tur mode of generation of the reproductive elements and
their relation to the cells of the parental tissues, is a problem
which has always made heavy demands on the labours of micro-
scopical investigators, whether zoological or botanical; and
although an immense literature has grown about the subject
since Flourens thought embryos were. formed “tout d’un
coup,” at the moment of the fusion of the sexual elements, it
may be safely asserted that not until the last few years has any
very definite knowledge been acquired, and only in a limited
number of cases has it yet reached any high degree of accuracy.

In a certain number, however, we do know what is actually
done during the origin of these cells; and this knowledge is a
priceless gift to the biologist, as there is little doubt that the
apprehension of karyokinesis, both in its relation to the
“ Reductions-Theilung ” and the ordinary division of somatic
cells, has brought him face to face with the actual mechanical
expression of hereditary transmission and the problems con-
nected with it.

The modus operandi of the forces which bring about these
changes, however, or any serious attempt to ascertain whether
they be modifications of ordinary physical phenomena at all,
or whether the whole “ fleeting show” of attractions, repulsions,
and nuclear metamorphoses must be looked upon as something

260 J. E. S. MOORE.

outside the physical domain, as ordinarily understood, is still
a fundamental, though quite legitimate problem for micro-
scopical inquiry.

In attempting to obtain a clear conception of the premises
from which we have to start in such inquiries, the once
absorbing question of the nature of the wide structural differ-
ences often apparent in the male and female cells will be found
to have lost, if not much of its importance, at any rate most
of its original characters. The conjoint labours of Hertwig,
Ishikawa,” vom Rath,* and others, as well as the curious
observations of Weismann concerning the non-specialisation of
certain. spermatozoa, have completely changed the scenes in
this direction; and it is matter for rejoicing that the shifting,
at any rate in this particular, tends towards a simplification in
the gradual banishment of apparent difference in such elements,
and of their associated mechanical complexity, to the rank of a
purely physiological importance.

Opinion to-day is almost unanimous that the ova and
spermatozoa are strictly similar objects, that even the most
modified spermatozoon still carries about with it the dwarfed
representatives of cell structure; and our knowledge of its
development is sufficiently advanced to recognise in the head
the reduced nucleus, the kytoplasm in the tail; and lastly, it
appears probable from Hermann’s,* and more especially from
Fick’s® investigations, that the hitherto enigmatical “ Mittel
Stiick ” is in reality nothing less than the attraction sphere.®

1 “© Vergleich der Ki und Samenbildung bei Nematoden. Hine Grundlage
fiir cellulare Streitfragen,” ‘ Archiv f. mikroskop. Anat.,’ Bd. xxxvi, 1890.

2 «Studies of Reproductive Elements.’

‘ Archiv mikro. Anat.,’ Bd. xl, p. 102.

‘ Archiv fiir mikros. Anat.” Bd. xxxiv, Tafel 3.

* Ueber die Befruchtung des Axolotleies,’’ ‘Anatomischer Anzeiger,’ vii,
pp. 818—821.

6 ‘The results of a re-examination of the facts of Mammalian spermatogenesis
have shown that the centrosomes are incorporated in the spermatozoa in the
position of the Mittelstiick of Amphibia, while a portion of the archoplasm
is applied to the pointed extremity of the head. Field has shown that the

archoplasm ‘‘ Nebenkern” of Echinoderms is incorporated together with the
centrosomes as the Mittelstiick.

oa > wo t

REPRODUCTIVE ELEMENTS IN APUS AND BRANCHIPUS. 261

Although both the eggs and the spermatozoa are cells, and
similar cells, the final karyokineses which produced them
are different from the preceding divisions in the cells of the
genital epithelium, whether male or female. As is now well
known, this change in the divisional phenomena appears in the
extrusion of the polar bodies in the egg, and the “ Reductions
Theilung” in the spermatozoa. The existence of these phe-
nomena constitutes the empirical though important ground
for several well-known theories concerning the physiological
value of the “‘ Reductions Theilung” as a preparatory balancing
of the “hereditary substance” in two cells, whose fusion forms
the starting-point of a succeeding generation.

Nearly all the readings of this riddle actually offered, have
sprung from one of two sources: either they have come back
upon us as the new-clad ghosts of Balfour’s famous interpreta-
tion of the polar bodies; or have accepted as sufficient explana-
tion of the facts, Weismann’s theoretical conceptions of the
necessity of a reduction in the quantity or the quality, or both,
of the hereditary substance (chromatin).

The continuous processes of assimilation and growth in the
resting cells of any tissue, although beyond our actual scrutiny,
offer nothing antagonistic to the generally adopted notion, that
they are the result of the passive influence of an immensely
complicated structure, operating under certain rather limited
conditions. But the final dissolution of these conditions, when
assimilation and growth in the individual can presumably go
no further,! and their re-establishment under more favorable
circumstances, is quite another matter.

In this procedure, the essential material constituents of the
cell are accurately halved and separated, and we may presume
that the complicated mechanical basis of the resting cell’s
activity, or at any rate the power to return to it at some future
time, goes with them. What are the means by which this
material distribution is effected? According to more recent

1 Herbert Spencer suggests that cell division becomes a necessity in virtue
of the continual decrease of the absorptive area proportionally to the growth
of the cell.

262 3.0 Bs 8 MOORE,

investigation, an essential agent seems to be two centres of
an alternately attractive and repulsive nature (centrosomes) ;
but concomitantly with, and independently of, their opera-
tion, other and no less important changes accrue within
the cell, perhaps most notably, the evolution of the fine
chromatic reticulation of the resting nucleus into a limited
number of chromosomes at the nuclear periphery, their
number being constant for any particular species—a fact
making them of great practical importance in all questions
relating to heredity.

Now I wish to call attention to some points in the sperma-
togenesis of Branchipus, which may appear to throw much light
on portions of these successive stages; and although it is at
present hopelessly inadequate to illuminate the greater pro-
blems which arise from it, I have taken extra trouble to be sure
of my ground here, because I conceive that the existing theo-
retical solutions of these problems must one day find powerful
confirmation or the reverse, in a true appreciation of the
character of the processes which underlie the karyokinetic
metamorphosis.

Spermatogenesis.

The male gland in Branchipus is a rather straight biramous
tube extending up the tail, and a short distance further up the
body. If spermatozoa are free in the lower portion all the
stages of spermatogenesis are visible as we pass up. The
difference in phase amongst the cells may be taken to represent
the zones of Hertwig, van Beneden, and Julin.!

The spermatocytes, however, break away from the walls in
groups (fig. 21), their individual components being all in the
same phase. But as this phase, characteristic of each group as
a whole, is not often similar to those on either side of it, we

1 « Nouvelles recherches sur la fécondation et la division mitosique chez
ascaride mégalocéphale,” ‘ Bullet. de l’Académie Royal de Belge,’ 3me sér.,
ce. xix, 1887. “ Befruchtung und Theilung des thierischen Kies,” ‘ Morph.
Jahrb.,’ 1875.

REPRODUCTIVE ELEMENTS IN APUS AND BRANCHIPUS. 263

cannot strictly divide the tube into ascending zones. When
the ordinary somatic division has come to an end spermato-
genesis sets in among the cells lining the hollow of the tube.
These somewhat minute elements rapidly increase in size, and
their nuclei pass into a spirem, with a peculiar and charac-
teristic grouping of their chromatic elements all on one side,
just as in Hermann’s beautiful figures of spermatocytes in
the salamander (fig. 3).!

Sometimes before, and always during the course of these
changes, bodies answering to the centrosomes in all peculiari-
ties except their number, which is abnormally great, make
their appearance in the angular mass of protoplasm at the bases
of the characteristic cells represented in figs. 1—4, a.

Merely for the sake of clearness, and to keep these bodies
out of confusion with the true and enormous centrosomes ap-
pearing later, as well as to separate them from another type of
body, to which I shall have to refer at length, I have provision-
ally collected these bodies under the term pseudosomes.

As the spermatogenesis proceeds, the lop-sided chromatic
arrangement of the spirem rapidly gives place to ten chromo-
somes, all arranged on the nuclear periphery, and these ten
chromosomes in turn become transversely constricted to form
the well-known dumb-bell elements (figs. 8—12), so that we
have ten double or twenty single chromosomes, which rapidly
arrange themselves in the disc-like equatorial plate seen in
optical section (fig. 11).

At this period of the metamorphosis (Flemming’s meta-
kinesis) a number of most remarkable bodies make their
appearance, more or less exclusively related to the cell peri-
phery, but connected one to another and to the inner group of
chromosomes by fine strands, which remain uncoloured by re-
agents ; and, as their relation to these fine threads suggests the
nodal points in a net, I have termed them dictyosomes
(figs. 11—18, d).

The constriction between the dumb-bell-like heads of the
chromosomes becomes more and more pronounced, and they

1 © Arch, f, mikros, Anat.,’ Bd. xxxvii.

264 J. HE. 8S. MOORE,

ultimately separate, passing in opposite directions towards the
relatively colossal centrosomes now occupying the spindle
apices (figs. 12, 17, 19, c, d). I have separated the dictyosomes
from the centrosomes, not because they appear to be in any
way essentially distinct, but because they originate at a later
period in the division, and the two sets of structures might
otherwise be confused.

Respecting the relation between the pseudosomes, centro-
somes, and dictyosomes I shall speak later on.

The two nuclear groups now separate as in fig. 14, and the
first reduction division is completed. The small elements thus
formed never again regain the character of a resting cell,! but
there are appearances of irregular division, resulting in the
formation of two excessively small spheroidal bodies, each
presumably containing the equivalent of five chromosomes
(figs. 15, 16). This procedure must be looked upon as con-
stituting the second “ Reduction Theilung,” and the resulting
elements are the mature spermatozoa.

With the above broad facts of spermatogenesis kept well in
view I proceed to a more minute description of the successive
stages of the karyokinesis related to it, more especially with a
view to determining the nature of the bodies I termed dic-
tyosomes and pseudosomes in the previous description, and
which at first seemed to appear, disappear, and reappear in a
quite bewildering fashion, The original small spermatocytes
are similar in all essentials to the least specialised elements of
the somatic tissues.

When stained with orange, gentian violet, or hematoxylin,
after treatment with Hermann’s or Flemming’s fluid (the best
results were obtained from a combination of gentian violet and
orange), the somewhat triangular cells* present a fine reticulate
appearance, both within and without the nucleus. The meshes
of this reticulum are of fairly equal size in both cases (fig. 1),
and a close examination leaves no doubt that the appearance
(at any rate in these cells) is produced by a vast number of

1 Compare vom Rath and Ishikawa, loc. cit.
? Compare vom Rath’s figs., loc. cit.

REPRODUCTIVE ELEMENTS IN APUS AND BRANCHIPUS. 265

clear globules, kept apart by some non-miscible intervening
fluid ;' in fact, the whole might fitly be described as a foam
structure, or “ Schaumplasm”’ of Biitschli.

I have attempted to give some idea of this appearance in
fig. 1, a resting spermatocyte just previous to its division, but
the result is not nearly so impressive as I could wish.

Nuclear stains affect to a certain extent the intervening fluid
throughout the whole cell, and the stain appears to be related
to excessively fine granules suspended in a clear plasma.
These cyto-microsomes do not appear to be “ varicosities” of
the kytoplasmic strands between the globules, but the stain
appears to affect microsomes suspended in this intervening
fluid. The whole darkened nuclear area suggests a condensa-
tion of this staining material, possibly by its own cohesion.

Outside the nucleus there are usually to be found, on the
side where there is most cell body, and where vom Rath repre-
sents the centrosomes in the resting spermatocytes of Gryllo-
talpa, those dark points, whose appearance corresponds in
everything but number with the centrosomes as ordivarily un-
derstood, and which I collected in the more general descrip-
tion under the term pseudosomes (figs. 1—6, a). Noarchoplasm
is apparent round them, and a close examination suggests that
they are simply the expression of a collection of the above
staining material (microsomes) in the angular spaces between
the spheroids, producing the reticulate appearance (figs. 1—9).

Careful search will, as I have said, often raise the number
of these bodies as high as six or eight. The more we look
the more difficult it becomes to separate the pseudosomes
from the less conspicuous interspaces of fluid between the
globules; both appear to pass insensibly into each other,
The appearance and relation of the more conspicuous are
very striking, as observation of their subsequent behaviour
left no doubt on my mind that they were intimately bound
up, if not with the origin, at any rate with a remarkable
increase witnessed in the centrosomes ultimately occupying

1 When sufficiently high powers are used the appearance is almost identical
with the coarse vacuolation in the ectosarc of Amceba and other protozoa.

266 J. ol 8. MOORE.

the apices of the spindle figure. It will, however, be well
to advance the description of the mitosis a little before dis-
cussing this point. The first nuclear differentiation appears
at one side of the nucleus as a colourless spot (fig. 2), which
grows, driving the chromatic network before it to one side
(fig. 3). The individual chromatin bands become shorter and
thicker in proportion to this displacement, and nearly all the
fine strands of “linin”’ disappear from this area, or, in other
words, the spheroids fuse one with another, the fusion being
produced by the substance of the clear globules breaking
through the walls of intervening fluid one into another. In
fact, this fusion spreads just as in soap froth the larger bubbles
grow at the expense of the smaller, and the continuance of
such a process results in the chromatin being thrown to one
side in the form of a crescent (fig. 3), its threads are naturally
thickened in proportion to their displacement, and the curious
initial figure, which so much struck Hermann in the sperma-
tocytes of the salamander, appears to be a necessary conse-
quence of an intra-nuclear fusion in Branchipus.!

As the intra-globular fluid (with its staining granules) is
between the adjacent spheroids, or in any single instance is
peripherally disposed towards them, it follows that if the fusion
continues until the whole nucleus consists of one or of a small
number of spheroids, the intervening staining chromatin will
be, as it practically is, all on the periphery.

Secondly, the fewer the globules, the larger and fewer the
angular spaces between them (figs. 4,5), and consequently the
more deeply staining intervening matter appears as a limited
number of chromosomes connected by fine striz (linin) ; their
actual number will naturally depend on the size of the
spheroids compared with that of the nucleus.”

A great deal of importance has been attached directly

1 © Arch. fir mikroskop. Anat.,’ Bd. xxxvii, pp. 569—582.

? T do not mean to maintain that there are no other controlling factors in
the formation of a definite number of chromosomes; this cannot be the case
on account of the wide difference in the size of the nuclei of tissues of the

same animal. At the same time we have no knowledge of the reticulum as
related to different cellular dimensions.

REPRODUCTIVE ELEMENTS IN APUS AND BRANCHIPUS. 267

or indirectly to the number of the chromosomes by all the
more recent investigators, and this factor in their origin (in
Branchipus) fully bears out the assumption that they are the
visual expression of the primary constitution of the cell to which
they belong. Nor is this all; for if we believe, as we have
every reason to believe, that the character of the nucleus is the
determining factor of the nature of the cell’s activity, that
curious variation in the number of the chromosomes in cells of
closely allied species would be more intelligible; for although
the frothy structure of the nucleus might be actually or closely
similar, a very slight difference in the cellular dimensions
would, provided the foam structure remained the same,
materially alter the number of the spaces between the globules,
and consequently the number of the chromosomes. It is at
the same time apparent that this number, as well as the
general nuclear characteristics, oscillate within narrow limits
for the same species.!

It is interesting to note in this connection that the characters
of nuclei, in Arthropods and Annelids, have much in common.
They nearly all present the peculiar ball-like chromosomes
during metamorphosis, just as they tend to form a reticulate
nucleus when at rest. In fact, we might say such nuclei
constitute an Annelidean nuclear type.

Again, the characters of the Mammalian nuclei are very
constant, but they nevertheless differ in minor details even
from those of the Amphibia. In fact, the difference between
these two latter is as small as that between them both, and the
former, is great. The comparative study of nuclei is well
worthy of more minute attention; suffice it, however, at the
present moment to point out that such generalisations would
have weight in our conceptions of heredity.

Of the regular occurrence of a peculiar intra-nuclear fusion
in Branchipus the appearances leave no doubt, or that it is
primarily instrumental in bringing about the conversion of the

‘In connection with this see Valentine Hacker, “ Die heterotypische

Kerntheilung im Cyklus der generation Zellen,” ‘ Berichte der Naturforschen-
den Gesellschaft zu Freiburg,’ Bd. vi, 1892, pp. 160—188.

268 J. BE. 8. MOORE.

fine chromatic network of the resting nucleus first into the
lop-sided figure described by Hermann, and probably has a
good deal to do with the origin of the ten chromosomes on the
nuclear periphery (figs. 4—8). But the initial impulse which
starts such a fusion is an entirely different matter. This
might rise from a variety of causes, from a gradual increase of
internal pressure caused by osmotic action, or it might be pro-
duced by some change in that polarity supposed to exist
between the centrosomes lying close to its exterior; and it is
curious to note in this connection that the fusion in Branchipus
does start from that side where the most marked pseudosomes
exist (figs. 3, 4,6,a,a), and, if we may put the same interpretation
on the metamorphosis of other cells, Hermann’s, vom Rath’s,
and possibly Flemming’s figures would be in complete accord-
ance with such a view. I very much doubt, however, if either
of these suppositions will be found to be the explanation of the
origin of the fusion in the first instance. But, if once started,
we have seen that the nuclear metamorphosis during karyo-
kinesis, from the resting stage up to that at which a limited
number of chromosomes exist on the periphery, is, to a certain
extent, the logical consequence of its progress.

I have arranged the succeeding description in the hght of
this conception because, since it has helped us thus far, we
might pre-suppose it useful in the elucidation of other karyo-
kinetic phenomena; and, unless I have done very indifferent
justice to the appearances before me, this supposition should be
fully justified.

The ten ellipso-spherical chromosomes (figs. 5, 7) which
have arisen from an irregular transverse splitting, or, rather,
running into drops of the thickened chromatic network, after
it was brought by the progressive fusion to the nuclear
periphery, become rapidly constricted in the middle to form
the dumb-bell figures characteristic of these and many other
Arthropod nuclei (figs. 7—12).

Each cell now contains ten double or twenty single chromo-
somes, i. e. double the ordinary number (figs. 7, 10, 12); and it
is interesting to compare such nuclear figures and their origin

REPRODUCTIVE ELEMENTS IN APUS AND BRANCHIPUS. 269

with others, like those of Salamander, in which the chromatin
is arranged in a succession of bands or irregular annuli, set
more or less transversely to the long nuclear axis. These
appearances would be produced by a similar fusion of globules,
in such a manner that a few diaphragm-like membranes of the
intervening fluid with its staining microsomes were left across
the long axis of the nucleus. In such a case the chromatin
would inevitably be arranged as it always is in the re-entrant
solid angles. Interesting artificial reproductions of the nuclear
figures may be seen by watching the growth of bubbles, and
I have given in fig. 18 some drawings of the ultimate con-
figurations produced by the growth of bubbles in a fine froth,
The lines of foam left as bands along the position of the
ruptured walls would represent the chromatic loops, and it
will be seen that they show a marked tendency to contract into
more or less rounded bodies.

The existence of nuclei in groups of four or five, each with
their ten dumb-bell chromosomes, gives a very striking appear-
ance to the testes of Branchipus (fig. 21); and while the con-
dition characterises one of the longest phases of the whole
nuclear division, its final metamorphosis occurs with the
utmost rapidity, the cells appearing as if transformed by magic
into a complete spindle figure. Intermediate phases are, how-
ever, to be found, and it appears that the fusion or running
together of the globules continues, breaking through the old
nuclear boundary at several points into the surrounding kyto-
plasmic network (fig. 8), so that the clear mass of nuclear
plasm appears to spread out on all sides (figs. 8, 11, 19).
The result of this is that chromosomes are at last left
hanging in a clear central space by a few irregular strands
of this kytoplasmic network, into which the fusion has not
yet broken (figs. 10—12, 19). These irregular strands are
ultimately reduced to fine threads (figs. 11, 12, 17, 19),
and their peripheral extremities are related to dark bodies
which can be nothing but the pseudosomes of which I have
already spoken in an earlier phase of the metamorphosis (figs.
10, 11, 12, da.). These pseudosomes appear now to have

270 J. E. S. MOORE.

increased in size somewhat, their relation to the spaces of the
intra-globular network being more pronounced, and we are
naturally led to the conclusion that such an increase is brought
about by the massing of the staining material in these angular
spaces, owing to the progressive fusion tending to lessen their
number and increase their size, just as it did with respect to
the chromatin within the old nuclear limits.

Proportionately to the extension of this fusion, the tension
along the achromatic lines, on which the chromosomes are
suspended, becomes greater as the dark points (pseudosomes)
at their peripheral extremities retreat with the vanishing
achromatic network and its contained microsomes towards the
cell’s circumference (figs. 1O—12, 19).

If we now try to realise what is actually taking place, it will
become apparent that the traction towards the periphery
through these points (pseudosomes) along the achromatic
threads, and ultimately upon the chromosomes themselves,
will tend to set itself along some axis across the nuclear figure
as a whole, and the points (pseudosomes) chosen will be those
on opposite sides which have, so to speak, the best foothold in
the periphery. The remaining points (pseudosomes) will tend
to glide as they do (figs. 9, 10, 12) towards the extremes of such
an axis, and a spindle figure will be finally set up (figs. 9, 12,
22, 23). From the figures just referred to, it will be apparent
that the coalescences of the points of attachment of the distal
extremities of the achromatic fibres (pseudosomes) become
marked out as centrosome-like bodies which travel away
towards the cell’s circumference, and finally come to rest on its
extreme margin (figs. 11, 22, 23). In other words, these centro-
somes are virtually derived from a fusion of some of the pseudo-
somes, and these were in turn seen to originally correspond to
the angular-spaces in a network exterior to the nucleus.’

ln cells a trifle more advanced than those represented
in the preceding figures showing (fig. 19) the area of clear

1 Professor Farmer has kindly shown me some preparations of Lilium

which give exactly the same bundles of fibres related to separate granules,
any one of which might be individually considered as a centrosome.

REPRODUCTIVE ELEMENTS IN APUS AND BRANCHIPUS. 271

fluid produced by the massing of the original diffuse staining
material within the nucleus into the small space of the chromo-
somes, by the process I have described—it will be seen that
this space, which in the first stage of the spindle figure repre-
sents the nuclearplasm, and retains the spherical character of
the original nuclear contour, has become very much enlarged
(figs. 11, 12, 23), not only in the direction of the spindle axis,
but laterally all round, so that it appears as a continually in-
creasing irregular area, occupying by far the greater part of
the cell’s substance. Round this irregular space a rind of the
original kytoplasmic reticulum still remains (figs. 11, 12, 19,
23), and it will be noticed that at the junction of this rind and
the clear fluid within (fig. 19) a number of small staining points
exist, related to the angular spaces between the clear
globules and the non-miscible intervening fluid.

These bodies grow continually, and their size marks the pro-
gress of the fusion of the clear central mass of fluid with the
similar constituents of the peripheral rind, just as the thicken-
ing of the chromatin bands was the measure of the fusion pro-
ceeding within the original nuclear limits.

They continually stain more and more deeply with orange
and gentian violet, as the diffuse staining material dispersed
through what remains of the kytoplasmic network is swept
before the progress of the fusion into nodal points until,
simultaneously with its extension through the whole cell, they
are left as some twenty conspicuously dark bodies regularly
arranged on the periphery (figs. 13, 17). Close examination
reveals, however, that the fusion is not really complete, but
that fine achromatic threads connect these bodies one to
another (figs. 13—19) and to the inner group of chromosomes
in the manner described in an earlier part of my paper—a fact
which led me to devise the term dictyosome as expressive
of these peculiar relations.

It will be seen that these dict yosomes appear at a definite
point in the karyokinetic metamorphosis, viz. the later phases
of the spindle figure ; and the cells which present these condi-
tions are comparatively large spherical bodies, which have

VOL. 35, PART 2,—NEW SER. U

O72 J. E. S. MOORE.

become so much altered in their refractive characters that one
is reminded of Flemming’s words when describing a similar
change witnessed in the dividing cells of the salamander :

‘“ Bekommt man unwillkiirlich den Eindruck, als sei die
Zelle wahrend ihrer Theilung durch und durch mit einer be-
sonderen Substanz durchtrankt oder—um mich vorsichtiger
auszudriicken—als besitze sie durch und durch eine besondere
physikalische oder chemische Beschaffenheit.”

This change appears in the spermatocytes of Branchipus to
be the direct result of the collecting of the primarily diffuse
staining material of the resting nucleus into ten chromosomes,
and of that existing in the kytoplasm without into distinct
chromatic bodies (dictyosomes). Both these changes are appa-
rently due to a progressive fusion or running together of the
clear globules which, begun within the nucleus, formed the
chromosomes on its surface, and extending, swept the diffuse
staining material of the cell body together into some twenty
dictyosomes on its circumference. In this brief and necessarily
crude manner I hope to have made the main drift of the investi-
gation up to this point clear. I have dealt with the develop-
ment of the chromosomes in Branchipus, and shown reason to
believe that it is in a measure dependent on a fusion of the
globules which give rise to the reticulate appearance. The
progress of such a fusion would produce the one-sided figure
described by Hermann in the spermatocytes of the salamander,
and tend ultimately to form a limited number of chromosomes
all on the nuclear periphery; and we have seen that during
these changes there exist in the resting spermatocytes those
dark points (pseudosomes) whose appearance corresponds in
everything but number with the centrosomes of previous
authors.

We have seen also that the fusion producing such wide
changes in the nucleus spreads beyond it, leaving the chromo-
somes suspended to the pseudosomes. These pseudosomes
retreat with the remnant of the original network as it vanishes
towards the periphery, and in connection with this motion an
axis tends to be set up round which the spindle figure gathers,

REPRODUCTIVE ELEMENTS IN APUS AND BRANGHIPUS. 273

while at its apices some of the pseudosomes coalesce to build
up the colossal centrosomes,

Lastly, just as the fusion within the nucleus brought about
the massing of the chromatin into a limited number of
chromosomes, so also the extra-nuclear fusion operating in the
same way upon the sparse staining material of the kytoplasm,
without the nucleus ultimately collects this, into chromatic
bodies in the angular spaces between the enlarged globules.
They first appear as an irregular cloud on the outskirts of the
fusion (fig. 23,d), grow enormously in size, and acquire a regular
distribution on the cell periphery. They still, however, remain
connected one to another and to the inner group of centro-
somes by fine threads ; the fact that they thus form, as it were,
the nodal points in a net, suggesting the term dictyosome as
expressive of this peculiar relation. It will, moreover, have
become apparent from the description that there is no genetic
distinction between the pseudosomes, centrosomes, and dictyo-
somes, and my sole reason for using the two new terms is their
successional appearance.

In thus bringing into prominence the existence in Branchipus
of a veritable “ Schaumplasm ” and its inter-activities, I would
observe that I do so with no predisposition to utilise Biitschli’s
conception of such structure as a fundamental interpretation
of some of the phenomena of karyokinesis, either in this or any
other case, but rather the reverse. Nevertheless the observa-
tion that a foam structure is intimately bound up with the
phenomena of karyokinesis on the one hand (even in a single
type) must materially enhance the value of Biitschli’s ingenious
hypothesis that it is sufficient to account for the amceboid
activities of protoplasm on the other.

A very natural objection to the conclusion I have stated may
arise out of the apparent whittling process to which it subjects
the centrosomes, resolving these bodies into nothing more
than the irregular staining material between the globules of a
protoplasmic froth.

I wish, however, while concluding this part of my paper, to
point out that such an objection is only apparent, and not real.

274 J. E. S. MOORE,

In a former essay, while discussing the meaning of the differ-
ence in the component parts of the spheres apparent in the
works of Flemming, Hermann, van Beneden, Boveri, and
others, I remarked, “ Comparison between the spheres and
their constituent parts in various animals might appear pedan-
tic, and, in the present state of our knowledge, unnecessary, if
it were not that some of these parts are probably, as we have
seen, the fleeting expression of metamorphic phenomena; while
others (such as the central body), though dividing, retain their
characteristics unimpaired ;”’ and I have ventured to repeat
this as showing that the great pioneers of this phase of cytology
had already hunted the all-important part of the sphere down
to the narrow limits of the centrosome. And the fact that in
Branchipus six or eight bodies indistinguishable from one
another exist at first, and that these afterwards fuse to aug-
ment the size of the two actually chosen to occupy the spindle
apices, does not prevent anyone from regarding these two of the
six or eight, as endowed with special properties if he pleases,
nor does their relation to the interglobular spaces affect the
point in any way that I can see. Whether two of these bodies
are really to be regarded as different from the rest is a point on
which at present I offer no opinion.

Comparison with the Ovigenesis.

It will be seen that the spermatogenesis of Branchipus
corresponds in the main with that described by vom Rath in
Gryllotalpa, and that the reticulum has disappeared in the
ultimate division altogether (fig. 16). Nowif,as I have shown,
there is reason to believe the reticulum in this particular in-
stance is a mechanical factor in portions of the karyokinesis,
all possibility of such phenomena will come to an end with the
complete fusion of the clear globules, and there is thus a
definite reason why the subdivision goes thus far and no farther,
at any rate for a time.

In the ovigenesis proper—that is, in the metamorphosis
among those cells which directly produce the eggs—there is
nothing special; but among the cells subsidiary to this process,

REPRODUCTIVE ELEMENTS IN APUS AND BRANCHIPUS. 275

many points are worthy ofattention. Scattered through the egg
mother-cells are numerous groups and rows of nuclei, obviously
of a different character from those destined to form the eggs
(figs. 20, 21, 43, &c.). These nuclei are very irregular in out-
line, and of a fine reticulate appearance. They show numerous
figures of direct division (figs. 20, 43). At the same time it is
quite easy to establish a gradational series extending from the
true egg, forming nuclei on the one hand to the irregular
akinetically dividing elements on the other, the latter class
being always intimately and actually concerned in the secretion
of a peculiar slimy substance (fig. 44); and this slime in turn
is ultimately worked up in the lower portion of the tube to
form the ornamental egg-case, so that although in the egg
formation in Branchipus the primitively similar genital cells
(male ova) diverge along two ways, one leading through succes-
sive karyokinesis to the final eggs, they ultimately both co-
operate in the perpetuation of the species by the rest being
bodily transmuted into the ornamental case in which the eggs
are laid. This duality in the ovarian elements is interesting in
the sense that it offers a precise parallelism to the dualism caused
in the spermatic apparatus by the presence of the akinetically
dividing foot-cells, over whose significance so much controversy
has at times been raised. In Branchipus the foot-cells are
more regularly arranged than the above akinetically dividing
elements in the ovary. At the upper end of the gland they
occur at intervals of about ten cells in all directions, and, true
to their female homologues, are more numerous as we descend
towards the genital aperture. Apart from the function of the
foot-cells no one can be in doubt as to their homology with the
above akinetically dividing elements of the ovary ; and the fact
that the latter are intimately bound up with the formation of
the slime that makes the egg-case (slime-cells) seems to me to
remove all doubt from vom Rath’s theory, that in the sperma-
togenesis they are concerned in the secretion of a fluid in which
the spermatozoa are suspended. The key to the whole position
seems to lie in the observation of La Valette St. George, that
the mulberry-shaped masses of the spermatocytes in Blatta are

276 J. E. 8. MOORE:

produced from one cell, whose residual moiety remains, ac-
quiries distinct characters from the rest, and is not converted
into the spermatozoa. From such a starting-point we may see
our way through a gradual evolution to meet the physiological
necessities of the case, to the complex reproductive apparatus
in Branchipus, where two different kinds of cells exist in both
sexes, one to form the eggs or spermatozoa, and one to form
the case or fluid in which these bodies are respectively sus-
pended or enclosed.

Whether akinetic division is really wholly related to the
foot-cells in animals is a controverted question, but from what
I have seen in Branchipus (figs. 20, 21) and elsewhere, I am
inclined to believe that it is not wholly restricted to these ele-
ments, but that there is a general tendency towards the two
methods in the two kinds of cells.

To recapitulate, it will be seen then—

1. That in Branchipus the observations bear out the general
law as to the similarity of the male and female cells, and that
their own specific peculiarities are physiological in origin,
having no morphological significance.

11. The derivatives of the primitive genital cells (male ova)
are of two kinds, one transformed directly into the reproduc-
tive elements, the other into the egg-case or into the fluid in
which the spermatozoa are suspended. Karyokinesis is the
method of procedure in the one—akinesis in the other.

u1. That the divisional phenomena of these cells are inti-
mately related to a protoplasmic structure, which might be
fitly described as ‘Schaumplasma,” and one of the initial
physical impulses towards metamorphosis is a fusion of some
of the intra-nuclear globules ; and a considerable portion of the
complicated karyokinetic figures, with their centrosomes, pseu-
dosomes, and dictyosomes, appear to be the logical as well as
the actual consequence of the continuance of this process.

With the foregoing results of observation as a basis of com-
parison, I made a close examination of the ovigenesis in Apus.
Unfortunately the male of this species is practically unknown,
ten thousand having been collected without a male appearing

REPRODUCTIVE ELEMENTS IN APUS AND BRANCHIPUS. 277

in a single instance. This renders the chances of proper fer-
tilisation very rare, and, as all the specimens are equally pro-
lific, we must look either to hermaphroditism or partheno-
genesis as the means by which the embryonic development
is started. The hermaphrodite character has recently been
ascribed to the genital gland of various species of Apus, and
certainly the appearances which have come under my notice
favour this view.!_ It is, however, immaterial to the line of in-
quiry I have adopted, which method of reproduction actually
obtains. The genital gland is an irregular tube with numerous
diverticula branching out on all sides. The cells lining the
main tube and its numerous ramifications are excessively
minute columnar bodies (figs. 24, 31, and 35—37), and the whole
appearance is far more like that of an Invertebrate intestine
than a reproductive gland. Lach of the epithelioid cells con-
tains a small peculiar nucleus (fig. 36), whose position in the
rod-shaped mass of protoplasm it dominates, varies in concert
with all the nuclei of the same diverticulum. The nuclei oscil-
late backwards and forwards from the extremities of the cells
nearest the lumen of the gland to those nearest the basal
membrane bounding at its periphery. When in the former
position, such protoplasm as remains between them and the
actual glandular cavity is seen to be rapidly degenerating into
masses of slime (fig. 81) ; and, just as in the case of Branchipus,
this slime is ultimately worked up into an ornamental egg-
case. When the nuclei have translocated themselves towards
the bases of the cells, the slime has broken away in streaks
and globules, and many nuclei are seen subdividing themselves
into groups, from whose derivatives the nuclei of the future
eggs are formed (figs. 25, 27—29, 31).

It is thus obvious, that for some reason or other an economy
has been effected in the reproductive apparatus of Apus, and
that there is no such permanent differentiation between slime
and egg-producing cells as is apparent in Branchipus, but that

1 For an account of this see the description of Siebold’s results in the
‘Klassen und Ordnungen des Thier Reichs,’ pp. 960—962. Also H. M.
Barnard’s ‘ Apodide,’

278 J. EivS. MOORE:

the same type of nuclei having gone down to the lumen of the
gland and, so to speak, performed their dirty work them-
selves, travel back again to the more peripheral regions, pro-
ceeding by a series of extraordinary divisions to instal them-
selves directly as the nuclei of the eggs. During these
migrations the nuclei retain their peculiar character little
changed. In all the genital cells, the chromatin is aggregated
into one or two nucleoli (figs. 24, 30, 36), constituting a
nuclear type which represents the extreme term in a series,
whose mean would be represented by the intestinal nuclei of
Carcinus, Idotea, and others described by Frenzel ;! where the
chromatic substance is still, to a certain extent, distributed
through the mass of the nucleus, although some may be
aggregated into massive nucleoli.

In Apus there is no vestige of colouring matter outside the
single chromosome which occupies its centre (fig. 36), the sub-
stance of which is so dense and refractive, that it appears like
a red lens suspended by one or two colourless threads from
the hollow sphere of the nuclear membrane. The intervening
space is entirely filled by a perfectly clear nuclearplasm.

If such nuclei are at the periphery of the gland, and the egg
formation is about to begin, one of these single chromosomes
is seen to elongate just as the nucleolus in Frenzel’s “ Nucleo-
lire Kernbulbirung”’ become constricted in the middle, and
finally separate into two halves, the nuclear membrane being
but slightly elongated in the direction of the fission (figs.
86—41). The two derived chromosomes may in turn divide at
right angles to the first separation axis, and a nucleus with
four chromosomes results (figs. 27, 29, 34, 35), whose mem-
brane is seen to be gradually tucked in at four intermediate
points, and at fig. 27 a final cross-shaped differentiation can be
made out between the indentations.

Along these lines the four quadrants ultimately separate,
giving rise to four nuclei, each with a single chromosome
(fig. 29).

Such groups of four nuclei are always associated with the

1 ¢ Arch. fiir micro. Anat.,’ Bd. xxxix.

REPRODUCTIVE ELEMENTS IN APUS AND BRANCHIPUS. 279

egg formation in Apus, and are seen in all stages bulging out
the membrane of the gland into the body-cavity. But it is by
no means the rule that these tetrad groups are formed in this
manner from one cell. In some cases two nuclei in adjacent
cells divide, and the derivatives nearest the periphery divide
again to form the group (fig. 25), or the group formation may
proceed in a much more irregular fashion out of one or two
secondary divisious of adjacent cells (fig. 27).

The dividing chromosomes within the nuclei at times present
the very curious appearance represented in figs. 34, 35. It will
be here seen, that between the separating, more darkly stained
portions stretches a stainless band, which again suggests that
the stain only affects particles suspended in a clear fluid, and
that this fluid is non-miscible. Moreover it would seem that
these particles tend to run together into chromatic drops,
leaving a clear fluid (paranuclein).

The initial impulse, whatever it may be, which gives rise to
the groups of four chromosomes, and is ultimately concerned
in the formation of the egg, is seen also to affect the surround-
ing nuclei, which divide in the same peculiar manner again and
again, until they form a narrow stalk connecting the original
group of four with the cavity of the gland (fig. 42). The
extreme minuteness to which this subdivision is carried will
be seen (fig. 33), where it will be observed that all trace of
cell membrane is fast disappearing. The minute, free, nuclear
elements then left spread over the surface of the tetrad group
as a thin protoplasmic membrane, in which they rest without
dividing walls of any kind (fig. 42, a).

During all this multiplication of the nuclei, the character of
their division remains precisely the same. In every division
the single chromatic ball passes through the metamorphosis
represented in figs. 836—41, the nuclear membrane contracting
until two precisely similar nuclei are left in the place of one.

It will be obvious that this method of procedure, though on
the face of it approaching akinetic or direct division, is in
reality very different from the process as it appears in Branchi-
pus, or in the other forms in which it has been described. In

280 J. E.'S. "MOORE;

all these the resting reticulate nucleus never passes out of that
condition, but is constricted into two portions, each retaining
its original character. It does, however, as above stated,
bear considerable likeness to the ‘ Nucleolire Kernbulbirung”’
described by Frenzel. This latter mode of division is also
normal to the intestinal cells of Apus. It will, moreover, be
admitted that it bears a superficial resemblance to the frag-
mentation seen in leucocytes. But the division in these
elements is certainly merely a shortened-up karyokinesis,
accompanied by centrosomes and an archoplasmic metamor-
phosis, while no such structures are apparent in either intes-
tinal or genital cells of Apus.

In the sense that no spindle or related parts are apparent in
these cells, their division approaches akinesis. In the sense
that all the chromatin is gathered into a single chromosome it
approaches karyokinesis. Monomeric (or division by single
chromosomes) is the best term I can devise for this method of
nuclear fragmentation, although of its actual affinities I am
still in considerable doubt.

It is well known that in many plant forms, such as the
Myxomycetes, the karyokinesis, although not absent, is passed
through in a reduced condition, and is apparently, exclusively
related to spore formation; and within the last few days Pro-
fessor Farmer has drawn my attention to a very remarkable
mode of spore formation, which he has found in certain liver-
worts of Ceylon, in which it is with the utmost difficulty that
the apparently akinetic formation of the tetrads can be shown to
be in reality a quadripolar karyokinesis ; and, further, it seems
generally agreed that such simplification is a reduction from,
and not an antecedent of, the more complex karyokinetic
division phenomena.

Consider also the nuclear division in the Protozoa them-
selves. It is now known that karyokinesis of a more complex
order, accompanied by an enormous number of chromatic
bands, is normal to some Rhizopods, while in the more spe-
cialised and less primitive Ciliates, this phenomenon is restricted
to the micronuclear elements, being in them so much reduced,

REPRODUCTIVE ELEMENTS IN APUS AND BRANCHIPUS. 281

that it is not without difficulty that it can be recognised
at all.

Such evidence seems to me to favour the conclusion that
the monomeric division in the genital cells of Apus is the
most extreme term known in a progressive modification of the
more primitive karyokinesis through such forms as Frenzel’s
* Nucleolire Kernbulbirung,’ in which the spindle, and the
breaking up of the chromatin into bands or globes, as well as
the resting reticulum, have long since disappeared.

There seems some probability in the assumption, then, that
owing to the introduction of a peculiar method of reproduc-
tion in Apus (parthenogenesis or hermaphroditism), the divi-
sional phenomenon has exhibited a corresponding change, that
the cells of the genital gland are all alike, and can function
both in slime or egg formation as opportunity arises.

The egg nuclei take origin from nuclei containing a single
chromosome, but they ultimately develop a coarse chromatic
reticulum with an external attraction sphere or archoplasm
(fig. 42, 5), and appear much like an enlarged somatic cell.

How such a modification of the original type has arisen it
is not, perhaps, very difficult to see. In sexually produced
species, the nuclei intended for fusion must, so to speak,
balance one another; and if karyokinesis is the original method
of procedure, any tendency in an individual to infringe this
rule in the origin of its reproductive cells would quickly tend
to be eradicated from the race on account of the wide abnor-
malities it produced. Buta parthenogenetic or hermaphrodite
species might please itself as to the manner in which it evolved
its reproductive elements, so long as these contained the
premises necessary to the proper development of the individual.

282 J. E. 8S. MOORE.

DESCRIPTION OF PLATES 15 & 16,

Illustrating Mr. J. E. S. Moore’s paper on “ Some Points in
the Origin of the Reproductive Elements in Apus and
Branchipus.”

PLATE 15.

a = Pseudosomes. c = Centrosomes. d = Dictyosomes. da = Stages
intermediate between dictyosomes and pseudosomes.

Fies. 1—5.—Stages in first division, from the resting cells to the formation
of chromosomes. Flemming’s fluid and gentian violet. a. Pseudosomes.

Fies. 6 and 7.—Formation of the chromosomes.

Fie. 8.—Cell showing the breaking down of the nuclear membrane.

Fics. 9 and 10.—Two spermatocytes, with nearly the same stage of early
spindle figure. da. Bodies intermediate between pseudosomes and dictyo-
somes.

Fie. 11.—Transverse optical section of equatorial plate; fine strands of
protoplasm connecting the chromosomes to the exterior network.

Fic. 12.—Cell a little later, in which the centrosomes have appeared.

Fic. 13.—Older cell, in which the spindle figure seen in section is break-
ing down, and in which numerous bodies have appeared on the surface
dictyosomes.

Fig. 14.—Final division of same.
Fie. 15.—Secondary cell.
Fie. 16.—Spermatozoa.

Fic. 17.—Group of three spermatocytes in division, with numerous (d.)
dictyosomes and (¢., ¢., c.) centrosomes.

Fic. 18.—Figures produced by the growth of bubbles in a fine froth.

Fie. 19.—Spermatocytes, showing the breaking down of the nuclear area.

Fie. 20.—Foot-cell in akinetic division.

Fig. 21.—Portion of spermatic epithelium, with various stages of sperma-
togenesis. F. c. Foot-cells. Sp. Spermatocytes and spermatozoa.

Fic. 22.—Spermatocytes in division, showing the extreme position of the
centrosomes.

Fig. 23.—Ditto, showing the early formation of the dictyosomes round a
spreading clear space.

REPRODUCTIVE ELEMENTS IN APUS AND BRANCHIPUS. 283

Fic.
Fic.
Fic.
Fig.
Fic.
Fic.
Fic.
Fic.
Fic.
Fie.

PLATE 16.

24.—Portion of epithelium of the female gland of Apus.

25.—Single cell with nucleus in partial division.

26.—Ditto.

27.—Irregular formation of tetrad group.

28.—Hlement with four nuclei.

29.—Tetrad formed from a single cell.

30.—Three nuclei of a tetrad, showing micronucleoli (a).

31.—Two tetrad groups in relation to surrounding cells.

32.—Cell with dividing nucleus.

33.—Portion of nuclei of tetrad, showing the minute subdivision of

the nuclei.
Fies. 34 and 35.—Method of division occurring in same nucleoli.

Fies. 36—41.—Stages of division.

Fic.

42.—Tetrad group, the upper cell being the future six “ N srenceen?

a, a, a. Nuclei which have spread over the circumference of the group.
Fic. 43.—Group of cells from the spermatic gland of Branchipus.

Fic. 44.—Ditto, ditto, from the female gland of Branchipus, showing the
relation of the akinetically dividing elements to the formation of slime.

All the preparations, except where otherwise stated, were fixed in Flem-
ming’s fluid and stained with gentian violet.

NOTES ON THE PERIPATUS OF DOMINICA. 285

Notes on the Peripatus of Dominica.
By

E. C. Pollard, B.Sc.Lond.

With Plate 17.

A NuMBER of specimens of Peripatus from Dominica, West
Indies, have recently been handed over to me for description
by Professor Ray Lankester, to whom I owe my best thanks
for intrusting me with this material. The specimens were
collected by Mr. Ramage in Dominica, and sent, some alive,
some preserved in alcohol, to Professor Lankester.

I had, in all, eighty-six individuals, of which fourteen had
been opened in salt solution before preserving, whilst the rest
were preserved whole.

Size.—There are considerable differences in size; the
length, measured without the antennz, varying from a mini-
mum of 17 mm. to a maximum of 50 mm.

The males are, as a rule, much smaller than the females,
and they are also much less numerous. Out of thirty-nine
specimens in which I determined the sex only eight are males,
and of these the largest is only 25 mm. in length.

On the other hand, a good-sized female measures 42 mm.

There are one or two apparent exceptions to this generalisa-
tion, one of the females being only 17 mm. and another only
19 mm. long; but from the fact that I have not found any
males longer than 25 mm., and also that the majority of the

286 K. C. POLLARD.

females are considerably larger than this, I am inclined to
regard the small females as not yet full-grown.

Colour.—My observations as to the colour of this Peripatus
have been made entirely upon specimens preserved in spirit.

The general colour of the body is a reddish brown dorsally,
with a diffuse longitudinal streak of a darker shade extending
down the centre of the back. The median dorsal line is
marked by a well-defined narrow band still darker in colour.
Ventrally the colouring is much paler, being a light grey or
greyish yellow. The colouring of the legs on their dorsal and
ventral surfaces corresponds with the colours of the dorsal and
ventral body surfaces.

The antenne are of a dark red-brown shade, with their
terminal enlargements much lighter, almost flesh-coloured.

This colouring obtains, with slight individual variations,
for all the specimens with one exception. In this unique case
the dorsal surface is piebald, with a pale straw-colour and a
reddish brown. The brown is disposed as a broad collar, and
as two lateral bands just dorsal to the legs; the band on the
right side, however, is only present in the posterior region of
the body. There is a white median line dorsally. The ventral
surface and the legs are pale yellowish white. The antenne
are dark red-brown, with their knobbed terminations pale
yellow or whitish.

This specimen is small, and appears to be a young form in
which the pigment is as yet not completely developed, or it
may possibly be an abnormality.

Ridges and Papille of the Skin.—As in the other
neotropical species of Peripatus, the ridges of the skin are
continued right across the dorsal median line.

The papille of the ridges are arranged in a single file of larger
ones, or two or three smaller ones occur abreast.

As in P. Edwardsii (Blanchard), there are accessory ridges
extending across the dorsal median line, but not reaching far
on either side of it ; and also, as in P. Edwardsii, the diagonal
lines which occur in the Cape species, breaking the surface
into lozenge-shaped areas, are absent.

NOTES ON THE PERIPATUS OF DOMINICA. 287

Many of the papillz, both on the legs and body, are divided
into two main portions, a basal and a more distal part. Of
these the basal portion is cylindrical in form, thus agreeing
with the species from Caracas, and differing from the Demerara
species, in which the basal portion of a papilla is conical
(fig. 1).

The papille vary considerably in shade, some being much
lighter than others, and to this is due the speckled appearance
of the skin.

Jaws.—Fig. 2 shows that the jaws are very similar to
those of P. Edwardsii (4, figs. 25, 26).

The outer blade is provided with a large main tooth, and a
smaller but still well-marked secondary one. On the inner
Dlade there is a large main tooth and seven or eight smaller
ones, of which the first is closely approximated to the main
one, and is considerably larger than the remaining six or
seven, from which it is separated by a wide diastema.

Antenne.—The papillz on the rings of the antennez are
arranged in several rows.

Ambulatory Appendages.—Only one specimen of Peri-
patus from Dominica has been previously described, and the
authorities differ as to the number of legs possessed by it.
Professor Jeffrey Bell states that thereare thirty pairs(1), whilst
Mr. Sedgwick considers that there are only twenty-nine (4).

In my specimens the number of ambulatory appendages
varies from twenty-five to thirty pairs, the great majority
having twenty-nine.

The relationship between the number of appendages and the
sex of the individual is interesting.

Out of thirty-nine specimens in which the sex was ascer-
tained only eight are males, and each of these is possessed of
twenty-five pairs of legs only.

Of the thirty-one females which I have examined, two are
possessed of twenty-six pairs of ambulatory appendages, one
of twenty-eight, twenty of twenty-nine, and six of thirty;
whilst 1 am uncertain as to the number possessed by the two
remaining specimens.

VOL. 35, PART 2,—NEW SER. x

288 E. C. POLLARD.

Of the remaining two, one at least possessed more than
twenty-seven pairs, but the exact number in both cases is
doubtful, since the specimens had been mutilated.

The numbers given above are perhaps better realised when
arranged in a tabular form, thus:

8 specimens with 25 pairs of ambulatory appendages. All males.

2 39 ee Fe i : Both females.
1 » 9 28 ” ry) ” Female.
6 ” 3” 30 ” ” ” All females.
67 oe) ” 29 ” ” ” : Of these 20 were

opened, and all
found to be fe-
males.

9 =~ 5, an uncertain number of legs. Both females.

The male, therefore, seems to be always possessed of twenty-
five pairs of ambulatory appendages ; whilst the female, with
one doubtful exception, has always more than twenty-five
pairs.

There are four foot-pads ventrally on each of the ambulatory
appendages (fig. 3), with the exception of those of the last pair,
which are possessed of two pads only (fig. 4).

At the distal extremity of the foot, close to the claws, there
are three primary papille, two on the anterior margin of the
foot and one on the posterior; but the basal papille are
absent.

The foot-groove, which in P. capensis extends on to the body
surface as far as the median ventral line, is in the Dominican
form continued only a very short distance on to the ventral
surface.

There are no white papille on any of the ambulatory append-
ages.

In his description of specimens of Peripatus from Guiana and
Dominica, Mr. Sclater (3) mentions the occurrence of a
‘“ bladder-shaped appendage”’ attached to the foot-grooves.
Such a vesicle-like structure is very obvious on the legs of
some of my specimens, but it appears to be due simply to
an extroversion of the lining of the groove.

NOTES ON THE PERIPATUS OF DOMINICA. 289

Apertures.—The anus is situated posteriorly between the
legs of the last pair.

The generative aperture in both sexes is found ventrally
between the legs of the penultimate pair.

The segmental organs of most of the appendages open into
the foot-groove close to the junction of the leg with the body.
On the fourth and fifth pairs, however, the aperture is situated
on a papilla between the proximal and third foot-pads (fig. 5).

IntERNAL ANATOMY.—Male Generative Organs (fig. 6).
—The vas deferens of each side is extremely short. Each
passes under the nerve-cord of its own side, and the two then
unite to form a very long coiled ductus ejaculatorius.

A pair of accessory glands are present, but these are much
shorter than those of the other South American forms which
have been described.

The accessory glands open out independently on either side
of the anus, as in P. Edwardsii, not into the ductus ejacula-
torius, as described for P. capensis.

In figs. 10 a—c, three sections through the testis are figured.
Of these the most anterior (fig. 10 a) is through the prostate
(pr. in fig. 6), and shows simply a mass of large nuclei sur-
rounded by a single layer of flattened epithelial cells. The
layer of epithelium round the nuclei of the prostate is a point
to be noticed, since Gaffron (2), in his description of P.
Edwardsii, states that there is in this region no epithelial
covering to the cells, which are simply enclosed in a thin
muscular sac. Gaffron, in fact, makes this a distinction
between the prostate (Schlauchhoden) and the testis proper
(Blasenhoden). This is well seen in his figure (Taf. xxiii, fig.
46), and is clearly described in his own words :—“ Der bedeut-
-samste Unterschied zwischen Schlauch und Blasenhoden ist
jedoch das Vorhandensein eines characteristichen Epithels im
letzterem. Wahrend der diinnhautige Sack des Schlauch-
hodens gleichmassig und dicht angefiillt ist mit grossen Zellen,
die sich nicht zu einem geschlossenen Wandungsepithel anord-
nen, liegt der eben beschriebenen Blasenhodenmuscularis
innen ein sehr regelmissiges polygonales Pflasterepithel.”

290 E. ©. POLLARD.

It is possible, however, that in P. Edwardsii an epithelial
covering to the prostate is also present, but the flattened
cells lying close upon the large nuclei below as seen in
fig. 10 a may have escaped notice.

The next figure (fig. 10 4) is from a section through the
actual testis (fe. in fig. 6). The contents of the testis in this
region consist of numerous large nuclei and whisps or bundles
of spermatozoa. There is a considerable space between these
and the epithelium, probably due to shrinkage. The epithe-
lial wall is well seen, and is similar to that described and
figured in P. Edwardsii by Gaffron. The cells are higher
and the nuclei rounder than in the epithelium of the pros-
tate.

More posteriorly still we get numerous cross-sections of the
coiled duct of the testis (fig. 10 c). The walls of this duct
are composed of cells with large round nuclei, and within its
cavity are numerous spermatozoa. The whole coiled mass is
surrounded by a layer of flat epithelium, and this is continued
over the straight portion of the vas deferens, where we
have accordingly two layers of epithelial cells,—an inner layer
of fairly tall cells with round or oval nuclei, and an outer layer
of flattened cells with rod-like nuclei.

In no region of the testis have I found a muscular covering
external to the epithelium.

Female Generative Organs (figs. 7—9).—On each ovi-
duct is situated a globular receptaculum seminis (£2. S.) con-
taining spermatozoa, and possessed of double ducts as in the
other neotropical forms. Between the receptacula and the
ovary are a pair of sac-like appendages (R. Ov.), the so-called
receptacula ovorum of Kennel, but I have been unable to make
out a vesicle at their distal ends.

Comparison with other Neotropical Species.—The
Peripatus of Dominica, P. Dominicz, resembles the other
neotropical species in—

i. The possession of four spinous foot-pads.

ii. The position of the generative aperture between the legs
of the penultimate pair.

NOTES ON THE PERIPATUS OF DOMINICA. 291

ii. The division of the primary papillz into two portions.

iv. The absence of the dorsal white line.

v. The arrangement of the teeth on the inner blade of the
jaw, there being a considerable gap between the first minor
tooth and the rest.

vi. The presence of the receptacula ovorum and seminis on
the oviducts.

vii. The unpaired portion of the vas deferens is of great
length and much coiled.

viii. The number of legs is not constant.

Compared with P. Edwardsii, to which the Dominican
species is most nearly allied, we find the following special
points of agreement :

i. There are two foot-pads only on the legs of the last pair.

ii. The male has a smaller number of legs than the female.

iii. The basal part of the primary papille is cylindrical in
both.

iv. The jaws and arrangement of the teeth are similar in
the two.

v. There is in each a pair of accessory glands opening on
each side of the anus.

The chief differences between the two are—

i. The number of ambulatory appendages; of these
there are 29 to 34 pairs in P. Edwardsii, and 25 to 30 in
P. Dominice.

ii. The white tubercles which are present on some of the
legs in P. Edwardsii are not found in the species now
described.

Compared with P. Trinidadensis——The Dominican species
differs much more from P. Trinidadensis. Besides the
difference in the number of ambulatory appendages (P.
Trinidadensis having 28—31 pairs), the two forms differ in
the number of teeth on the inner blade of the jaw, the
Trinidad species having a much larger number than the
Dominican. Moreover in the Trinidad Peripatus the basal
portion of the primary papille is conical, whilst in the
Dominican form it is cylindrical.

292 E. C. POLLARD.

Compared with P. torquatus.—P. torquatus is pos-
sessed of a much larger number of legs than the Dominican
form, having 41 or 42. The colouring is also strikingly
different, P. torquatus being characterised by a bright
yellow band behind the head on the dorsal surface.

BIBLIOGRAPHY.

Complete lists of the literature referring to Peripatus have
already been published by Mr. Sclater and Mr. Sedgwick, and
therefore the following list includes only the papers to which
special reference is made in the above description.

1. Bett, F. J—“ Note on a Peripatus from the Island of Dominica, West
Indies,” ‘Ann. Mag. Nat. Hist.,’ 5th ser., xi, 1883.

2. Garrron, E.— Beitrage zur Anatomie und Histologie von Peripatus,”
‘Zool. Beitrage’ (Schneider), Bd. i, 1885.

8. Scrater, W. L.—“ Notes on the Peripatus of British Guiana,” ‘ Proc.
Zool. Soc. London,’ 1887.

4. Sepewick, AA—“*A Monograph on the Species and Distribution of the

Genus Peripatus (Guildiug),” ‘Quart. Journ. Mier. Sci.,’ New Ser.,
xxviii, 1888.

NOTES ON THE PERIPATUS OF DOMINICA. 293

EXPLANATION OF PLATE 17,

Illustrating Miss E. C. Pollard’s paper, “ Notes on the
Peripatus of Dominica.”

Fie. 1.—Primary papilla from skin of Peripatus.
Fic. 2.—Inner and outer blades of jaw.
Fie. 3.—Leg of Peripatus, showing four foot-pads.

Fic. 4.—Last leg of specimen with twenty-nine pairs, possessed of two
foot-pads only.

Fie. 5.—Fourth leg, showing aperture of nephridium between the third
and proximal pads.

Fic. 6.—Male generative organs. Ac. gl. Accessory gland. D. Hace.
Ductus ejaculatorius. WV. C. Nerve-cord. Pr. Prostate. Te. Testis.

Fie. 7.—Female generative organs. Zig. Ligament of attachment. J. C.
Nerve-cord. Ov. Ovary. £&. 8S. Receptaculum seminis. &. Ov. Recepta-
culum ovorum. U¢, Uterus,

Fig. 8.—Enlarged view of ovary and receptacula. Reference letters as in
Fig. 7.
Fic. 9.—Diagram of the same.
Fie. 10.—Transverse sections through the testis.
10a.—Through the prostate.
1046.—Through the testis proper.
10¢c.—Through the coiled duct of the testis. Hp. Epithelium.

Outlines drawn with Zeiss, oc. 4, obj. B ; and details filled in with Zeiss,
oc. 2, obj. D.

Fic. 11.—Transverse section through vas deferens.

STUDIES ON THE PROTOCHORDATA. 295

Studies on the Protochordata.
By

Arthur Willey, B.Sc.Lond.,
Columbia Coilege, New York.

With Plates 18S—20.

II.

The Development of the Neuro-hypophysial System in Ciona
intestinalis and Clavelina lepadiformis, with an
Account of the Origin of the Sense-organs in Ascidia
mentula.

(With Plates 18 and 19.)

Wuitt following the metamorphosis of the larva of Ciona
intestinalis last year (1892) in the Zoological Station at
Naples, I noticed several peculiarities in the behaviour of the
nervous system which apparently could not be reconciled with
the account relating to Clavelina which was given by Hd. van
Beneden and Charles Julin in their work on ‘ Le systéme
nerveux central des Ascidies adultes et ses rapports avec celui
des larves urodéles.’ The first assumption to be made was
that the conditions might be different in Ciona from what
they were in Clavelina, as I had already perceived how much
the general development of Clavelina was modified in the
direction of a compression of the ontogenetic processes. But
on making preparations in toto of larve of various ages of
Clavelina lepadiformis I could see, as far as the hypo-
physis is concerned, nothing at all like the appearance figured
by the above-named authors in their pl. xvii, fig. 11 (‘ Archives
de Biologie,’ t. v, 1884). Seeliger’s figures of the larve of
Clavelina are much more accurate in this respect (see I, 31).

296 ARTHUR WILLEY.

The results, moreover, to which my friend Dr. Johan Hjort
of Christiania had come in his investigations as to the deve-
lopment of the hypophysis and ganglion in the buds of
Botryllus, determined me to study the same question in the
case of the metamorphosing larve of Ciona intestinalis
and Clavelina lepadiformis more closely than I had at
first intended. Dr. Hjort had the kindness to show me his
preparations and to make me thoroughly acquainted with his
results, a preliminary account of which has appeared in the
‘Zoologischer Anzeiger’ (No. 400, 1892) .1

Hjort’s results are of unusual interest, as they place the
contrast between the organogeny in the larva and in the bud
respectively of Botryllus in the clearest possible light.

1. Closure of Neuroporus and Origin of Sense-
organs in Ascidia mentula.

For the earliest stages of the nervous system after the fusion
of the medullary folds I examined the embryos of Ascidia
mentula, as they are much more transparent than those of
Ciona.

Very soon after the commencement of the curvature of the
embryo within the follicle, the curvature being initiated and
necessitated by the outgrowth of the tail, the neuroporus, as
was correctly described by Kowalevsky (I, 21), closes. It
will be seen later that this primary closure of the neuroporus
in the Ascidians is only temporary, and does not occur in
Amphioxus; while what may be called the secondary or final
closure occurs in both the Urochorda and the Cephalochorda.

After the first closure of the neuroporus has taken place,
the nervous system of the Ascidian embryo consists of a
perfectly closed tube lying immediately below the epidermis,
and containing a lumen which is slightly dilated anteriorly,
the neurenteric canal having been obliterated at a somewhat
earlier stage (Pl. 18, fig. 1). In fig. 1 is represented an optical
sagittal section of an embryo of Ascidia mentula at the
stage in which the first trace of the sense-organs appears, in

1 Tn the same number a preliminary note on my own results appeared.

STUDIES ON THE PROTOCHORDATA. 297

the form of a number of pigmented granules which are deposited
in the interior of certain cells of the dorsal wall of the cerebral
portion of the medullary tube. Without entering into a de-
tailed histological account of the sense-organs I will confine
myself to a few points in which I can to a certain extent sup-
plement the classical work of Kowalevsky.

The origin of the sense-organs in point of time seems to
underlie a certain amount of variation. Kowalevsky describes
the otolith as arising first in Phallusia mammillata. In
the embryo from which fig. 1 was drawn the eye was the first
to appear, being represented at this early stage by scattered
rounded pigment granules lying in several—four or five—of
the cells in the dorsal wall of the cerebral vesicle, while in
fig. 8 the otolith and eye appear simultaneously. The otolith,
as well as the eye, first appears in the form of a number of
scattered pigment granules very similar to those which go to
form the eye, but rather larger, and differing from the latter
in their being confined to one cell, while, as I have just men-
tioned, the pigment granules of the eye are distributed among
several cells. Kowalevsky would appear not to have seen the
eye granules at their very first origin. He says (loc. cit.,
p- 117), “Am Grunde der Zellen des hinteren abgesetzten
Theils der Blase [i.e. cerebral vesicle] erscheinen sehr feine
Pigmentkorner,” and he figures them (Taf. xii, fig. 31) outside
the cells which form them. It is possible that this may be
their mode of deposition in the species studied by Kowalevsky —
viz. Phallusia mammillata.

The granules which belong to the eye have at first essentially
the same character and nearly the same size as those which
go to form the otolith, being scattered throughout the body of
the cells in which they lie (fig. 1); but as they increase in
number they become very much smaller, and then lie entirely
at the inner free extremities of the cells (figs. 3—5). The
otolith granules do not tend to increase in number, but retain
their original size until they fuse together (fig. 5).

The otolith cell or otocyst lies immediately in front of and
adjacent to the eye-cells, and, in fact, forms primarily one of

298 ARTHUR WILLEY.

the same series of cells with the latter. This exact primary
relation of the otocyst to the eye-cells was not observed by
Kowalevsky, and it is interesting as showing that two such
different organs as an eye and an ear can arise in the same
way—namely, by a deposition of pigment—from one and the
same sense-tract.

Next begins the migration of the otocyst, which was
discovered by Kowalevsky. This migration is not an active
one on the part of the otocyst, but goes hand in hand with a
change in the histological character of the wall of the cerebral
vesicle, and is therefore, as in so many other cases of change
in the position of organs, a result of the relative differential
growth of parts. This histological change in its turn is
correlated with the expansion of the original slight anterior
dilatation of the nerve-tube into a spacious vesicle. Successive
stages in this expansion of the cerebral vesicle are shown in
figs. land3—5. In fig. 4 the otocyst is seen to have separated
itself from its previous contact with the cells of the optic
region, and now lies at a lower level in the right dorso-lateral
region of the cerebral vesicle. In fig. 5 the interval between
the otocyst and the eye has further increased itself by the re-
duction of the wall of the cerebral vesicle in this region to a
thin and apparently structureless cuticle, the reduction in bulk
being accompanied by an increase in extent.

In this way, therefore, by a local thinning out or cuticulari-
sation of the wall of the cerebral vesicle the otocyst is shifted
from its primary dorsal position to its secondary position on
the floor of the cerebral vesicle. ‘The migration of the otocyst,
therefore, occurs concurrently with the reduction to a thin
membrane of part of the wall of the cerebral vesicle. The
mode in which the otocyst becomes transported from the
dorsal to the ventral wall of the cerebral vesicle was not made
quite clear by Kowalevsky, who says (loc. cit., p. 117), “Die
vordere pigmentirte Zelle, welche wir vermuthlich als einen
Gehorapparat gedeutet haben, schiebt sich von der rechten
Wand der Blase nach unten, so dass sie auf den Boden der
Blase kommt.” From this description it might very natur-

STUDIES.ON THE PROTOCHORDATA. 299

ally be supposed that the migration of the otocyst was an
active one, whereas, as we have seen, it is really passive.

In Clavelina the migration of the otocyst is also effected by
relative growth, but occurs at a very early stage before the
cuticularisation of the wall of the cerebral vesicle (Pl. %, figs.
25 and 29—381).

Meanwhile, in Ascidia mentula, the eye-cells collect
themselves together (figs. 4, 5), and finally, as is well known,
come to occupy the posterior right-hand corner of the cerebral
vesicle (fig. 2).

After the sense-organs have taken up their definite positions,
the stomodeum forms by a median dorsal invagination of the
ectoderm (fig. 2), and shortly afterwards a communication is
established between the base of the stomodzum and the
branchial sac. At the stage at which the mouth breaks
through, Kowalevsky described the formation of an opening
between the cerebral vesicle and the stomodeum. I have
looked for this opening repeatedly in the tadpoles of Ciona
intestinalis, Phallusia mammillata, and Ascidia men-
tula, but have not been able to see it; and, in fact, I will go
so far as to say, in confirmation of Kupffer (8), that the open-
ing as described by Kowalevsky does not exist in the tailed
larvee.

At a later stage, after the commencement of the metamor-
phosis, a corresponding opening is actually formed, the relations
of which are essentially the same as those of the pore described
by Kowalevsky, though with certain important differences ; but,
nevertheless, Kowalevsky’s actual observation was, according
tomy account, erroneous. Thus in the tadpoles of the above-
named simple Ascidians there is no communication what-
ever between the cavity of the cerebral vesicle and the stomo-
dzeum.

On the other hand, as we shall see, in the tadpole of Clave-
lina there is such a communication, which was, however, denied
by van Beneden and Julin. We have, therefore, the curious
circumstance that what Kowalevsky asserted in the case of
Phallusia mammillata was contradicted by van Beneden

300 ARTHUR WILLEY.

and Julin in the case of Clavelina; and although it might
appear at first sight that somebody must be right, it turns out
in fact that all are wrong. I have stated above that a com-
munication between the cavity of the nervous system and
stomodeum is effected in the simple Ascidians at a consider-
ably later stage than that described by Kowalevsky. Remem-
bering the general abbreviation in the development of Clave-
lina, to which attention has already been directed, we should
expect to find that this communication would become estab-
lished at a much earlier stage in Clavelina than, for instance,
in Ciona, and this is exactly what happens.

2. Origin of the Neuro-hypophysial System in Ciona
intestinalis.

Stage I.—In transverse sections through a young tadpole
which has, by convulsive movements of its tail, succeeded in
bursting the egg-follicle, and so entered upon the brief free-
swimming phase of its existence, we find that quite anteriorly
a minute portion of the cavity of the cerebral vesicle is
separated off from the main cavity, and appears in section as
an independent lumen in the thickness of the wall of the cere-
bral vesicle on its left side (fig. 6). A section or two behind
this the small lumen in question is seen to communicate freely
with the cerebral cavity (fig. 7), and farther back still there is
nothing more of it to be seen but simply the plain wall of the
vesicle (fig. 8).

Thus already at this stage a portion of the cerebral vesicle
has begun to be constricted off in the form of a tube, at present
ending blindly in front, and\communicating behind with the
main cavity of the vesicle. It is of a very small extent, and lies
at present entirely in the thickness of the wall of the vesicle ; in
fact, it is only with the utmost pains that it can be detected at
all at this stage. Attention may be drawn here to the cupule
of the otolith shown in fig. 7 in the ventral wall of the cerebral
vesicle. The otolith itself has become separated from the
cupule which normally carries it, in the process of cutting. A
distinct nucleus-like body is to be seen in the interior of the

STUDIES ON THE PROTOCHORDATA. 301

crescent-shaped cell. Kowalevsky says that the nucleus dis-
appears entirely, and the whole cell becomes strongly refractive.
The latter is of course true, but the nucleus apparently does
not vanish, although it is impossible to see it in the fresh state.

I should add that I have examined sections of these early
stages taken in three planes, but have never found an opening
from the neural tube into the stomodzeum in the free-swimming
larva.

Stage II.—Fig. 9 represents a transverse section through
the region of the cerebral vesicle in a larva of the stage imme-
diately preceding fixation. The shape into which the wall of
the branchial sac is thrown by the pressure of the superin-
cumbent expanded cerebral vesicle should be noted. The epi-
thelium of the dorsal wall of the branchial sac is very flat as
compared with that of the lateral walls into which it gradually
passes on either side. The ventral groove of the branchial
sac in fig. 9 is not the endostyle, as follows clearly from a
comparison of the actual position of the endostyle as seen in
surface views (cf. figures accompanying I); but it is possible
that Kowalevsky mistook it for a continuation of the endo-
style—a mistake which is not difficult to make if larve of a
later stage are not examined for comparison, since, after the
withdrawal of the tail, both the enteric and the body cavities
undergo a general distension, which renders the internal struc-
ture and the topographical relation of parts much clearer.

With regard to the cerebral vesicle, the chief point of dif-
ference between this and the preceding stage lies in the fact
that the tube which we saw embedded in the thick wall of the
vesicle, in the form of a minute cul-de-sac, communicating at
its hinder end with the cavity of the vesicle, has now begun to
set itself off more distinctly from the rest of the wall of the
vesicle, forming a considerable prominence slightly to the left
of the middle line. The process of constriction by which the
tube, or, as it may at once be called, the neuro-hypophysial
canal, comes to be entirely separated from the cerebral vesicle
has therefore now commenced. It can be noticed in fig. 9
that the nuclei in the neighbourhood of the tube in question

302 ARTHUR WILLEY.

have begun to arrange themselves in such a way as to give
plainly the appearance of the lumen being surrounded by a
distinct epithelium. Fig. 10 shows the free communication
between the tube which is being constricted off from the
cerebral vesicle, still ending blindly in front, and the cavity of
the vesicle itself. The rudiment of the neuro-hypophysial
tube or canal has at this stage slightly shifted its relative posi- _
tion from that which it occupied in the preceding stage, having
approached more nearly to the dorsal middle line. This shift-
ing can readily be understood by a comparison of fig. 9 with
fig. 6, from which it will be seen that during the transition
from the preceding stage to the one now under consideration
the wall of the cerebral vesicle has become drawn out to a
thin membrane in the left latero-ventral region of the vesicle.
It will be remembered that the migration of the otocyst was
directly traceable to a similar local thinning out of the wall of
the vesicle.

Stage III.—This is the stage at which the communication
between the cavity of the nervous system and the base of the
stomodeum at the point of junction between the stomodzeum
and the wall of the branchial sac is effected (see fig. 2 as to
depth of stomodzal invagination).

The tube, whose constriction from the wall of the cerebral
vesicle we have been following, has now separated itself en-
tirely from the latter (fig. 11), and has meanwhile acquired an
Opening into the stomodeum (figs. 12, 13). The cerebral
vesicle itself has entered upon the process of histolytic disin-
tegration which eventually leads to its entire disappearance.

Thus the neural tube, of which the neuro-hypophysial
canal, so called on account of its later destiny, is merely a
continuation, now opens into the stomodzeum ; but the open-
ing is a perfectly simple one at present, and no appreciable
evagination from the wall of the stomodzeum can be demon-
strated. Later on an evagination does possibly take place,
and the opening which we see at this stage appears to be
carried somewhat further back. The hypophysis has not yet
differentiated itself from the nervous system.

STUDIES ON THE PROTOCHORDATA. 308

What, then, is this pore leading from the stomodzum into
the neural tube? My answer is that it is the neuroporus.
We have already mentioned the fact, which was originally
determined by Kowalevsky, that the anterior opening to the
exterior of the medullary canal closed up before the formation
of the mouth. We now see that some time after the forma-
tion of the mouth—in fact, as soon as it begins to be func-
tionally active and to take in water, which does not occur
until after the fixation of the larva the neuroporus opens out
again, but this time into the stomodzum.

The primary closure of the neuroporus was therefore only
temporary, and comparable to what occurs in the case of many
blastopores and other organs which become temporarily solid—
such, for instance, as the esophagus of the Selachians. What-
ever may be the actual cause of the temporary closure of the
neuroporus in Ascidians, it is perfectly plain that its per-
sistence through the period during which it is closed would be
of no service to the embryo or larva, because the development
up to the time of the formation of the mouth takes place in-
side the egg-follicle, and when the mouth does appear it does
not at first open directly to the exterior, but is covered over
by the so-called testa.

This is a quite different state of things from what we find in
Amphioxus, where the neuroporus does not undergo this
temporary primary closure; but then the embryo of Am-
phioxus leaves the follicle precisely at the stage in which, with
the above-named simple Ascidians, the neuroporus closes.
In Clavelina the neuropore remains open somewhat longer
than in the simple Ascidians mentioned above.

Fig. 13 represents a sagittal section through a young fixed
Ciona of this stage, drawn with a low power to explain the
topographical relations of the various parts. The neural tube,
or neuro-hypophysial tube, as it may now be indifferently
called, is seen to open into the buccal cavity infront. Fig. 12
shows a portion of the neuro-hypophysial tube from the same
section, but drawn under a much higher power (Zeiss J,
water immersion). It commences anteriorly as a well-marked

VOL. 385, PART 2,—NEW SER. Y

304 ARTHUR WILLEY.

tube lined by a columnar epithelium, the lumen of which,
however, becomes irregular as it proceeds backwards, where,
at the point marked g in the figure, its dorsal wall is seen to
consist of a mass of cells in place of a well-defined epithelium.

In fact, this is the first stage in the formation of the cerebral
ganglion of the adult (fig. 12, g).

Beneath the neuro-hypophysial tube lie the remains of the
cerebral vesicle of the larva, now filled with histolytic residue.

Stage IV.—A sagittal section through the neuro-hypo-
physial tract of a young Ciona, about the time-of the budding
off of the two intermediate stigmata from the two primary, as
already described by me, is shown in fig. 14. The section
is slightly oblique. The ganglion has attained a much greater
development than in the preceding stage, and, indeed, seems to
have been exceptionally large in this particular specimen.

The greater part of the lumen of the primitive neural tube
has become obliterated, and the nervous system is now for the
greater part of its extent solid, with a lumen, however, still
persisting in front, which opens anteriorly through the media-
tion of a funnel-like dilatation, which has possibly arisen in
part by evagination from the stomodzeum, into the stomodeal
region of the branchial sac.

Figs. 15—19 are taken from an extremely instructive series
of transverse sections through the central nervous system
of a rather older individual than that from which the sagittal
section of fig. 14 was obtained. Starting from behind, we have
in fig. 15 a section through the now solid nerve-cord, which is
continued into the ‘cordon ganglionnaire viscéral” of van
Beneden and Julin. Advancing gradually from behind for-
wards, we come to fig. 16, where the transverse section has an
irregular contour, and in its lower portion a very distinct
lumen can be seen, while dorsally it consists of a solid pro-
liferation from the dorsal wall of the neuro-hypophysial tube,
the nuclei being arranged peripherally, while the central por-
tion of the solid mass shows the first indication of the later
characteristic “‘ Punktsubstanz.” In fig. 17 the transverse
section has’a very regular and characteristic 8-shaped outline,

STUDIES ON THE PROTOCHORDATA, 305

the lower half of the 8 containing a lumen, while the upper
half is solid. In the next more anterior section (fig. 18) the
upper division of the 8 predominates over the lower, the lumen
of the latter being still of small diameter—in fact, rather smaller
than in the preceding sections; while still farther in front
(fig. 19) the lumen, which is perfectly continuous all along, has
attained a relatively large diameter, while the superjacent solid
portion of the ganglion is correspondingly small, and is distinct
from the dorsal wall of the subjacent canal. In fact, fig. 19
represents a section through the funnel-like terminal dilata-
tion of the neuro-hypophysial canal spoken of above, which
may be called the hypophysial funnel; and the portion of
the ganglion involved in the section is its anterior extremity,
which has come to overlap the posterior portion of the funnel
(cf. the figures of young individuals of Ciona accompanying
“« Studies,”” &c., No. I).

Attention may be drawn to the ciliated prominence in the
wall of the branchial sac behind and below the hypophysial
opening in fig. 14, This is the epibranchial ridge (epibranchial
groove of Julin), which is grooved in many adult forms. It
is directly traceable to the projection caused in the dorsal wall
of the branchial sac by the pressure of the distended cerebral
vesicle in the larva (cf. figs. 9 and 12).

Stage V.—The description which follows applies to the
relations of the neuro-hypophysial system in young immature
adults.

Anteriorly the duct of the hypophysis expands into the large
funnel-shaped dilatation, which in its turn opens into the
branchial sac at the end of a papilliform prominence, which
projects boldly into the branchial chamber, and is continuous
with the epibranchial ridge referred to above. Fig. 20 shows
a section taken a short distance behind the branchial opening
of the hypophysis, and passing through the anterior dilatation,
above which are seen two cerebral nerves. Fig. 21 is drawn
from a section posterior to that of fig. 20, and shows a great
decrease in diameter of the lumen of the hypophysis, while
still farther back the lumen becomes reduced to a minimum

306 ARTHUR WILLEY.

(fig. 22). This temporary obliteration of the lumen of the
hypophysis at this point and at this period of the development
seems to be a constant feature, and extends over one, or at
most two sections of a thickness of about 7 yu. The lumen
then opens out again (fig. 23), and in the posterior region of
the hypophysial tube, which now lies closely applied to, but
at the same time distinct from, the ganglion, glandular tissue
is seen to be developing from its ventral wall! (fig. 24). Here
and there the peripheral nuclei of the ganglion are absent in
the region where the hypophysis is in contact with the latter
(fig. 24). :

We see, therefore, that the hypophysis and the ganglion,
which have been gradually differentiating themselves from the
common neuro-hypophysial tube, have at last separated entirely
from one another by a completion of the constriction of which
we saw the commencement in the preceding stage (fig. 17).

From the appearances presented I am disposed to believe
that the anterior portion of the hypophysis, including the
funnel-shaped dilatation and the duct, as far as the above-
mentioned point of reduction of the lumen, is derived from a
secondary evagination from the stomodeal region of the wall
of the branchial sac ; while thedivision of the hypophysis which
lies behind that point, and from which the gland is developed,
represents the tube derived by constriction from the cerebral
vesicle of the larva in the way described above. The original
opening, therefore, of the neuro-hypophysial tube into the
branchial sac has on this supposition been carried backwards
by a secondary outgrowth from the stomodeum. It is difficult
to bring other than circumstantial evidence in support of this
view, but it may be possible to test its truth on a future occa-
sion, Meanwhile this seems to be a reasonable explanation of
the local and temporary obliteration of the lumen which divides
the proximal from the distal or glandular portion of the
hypophysis.

' Seeliger (‘ Jenaische Zeitschrift,’ xviii, p. 100) described the hypophysis-

gland in Clavelina as arising by the aggregation of free mesoderm-cells. He
does not, however, commit himself unreservedly to this view.

STUDIES ON THE PROTOCHORDATA. 307

In Ciona, therefore, the cerebral ganglion of the adult arises
by proliferation and constriction from the dorsal wall of the
neuro-hypophysial tube.

3. Origin of the Neuro-hypophysial System in
Clavelina lepadiformis.

The description given above as to the origin of the neuro-
hypophysial system in Ciona, together with that which I am
about to give for the same system in Clavelina, will be found
to be considerably at variance with the results obtained by
Seeliger and van Beneden and Julin in the case also of
Clavelina.

My observations were at first entirely confined to Ciona,
and led me to the conclusion, judging from the very explicit
account, accompanied by numerous figures, of van Beneden and
Julin (loc. cit.), that the mode of development of the parts in
question must be different in Clavelina. But when I came to
study the origin of these structures in the latter form to
enable me to make a definite comparison with Ciona, it turned
out that the relations above described for Ciona were not only
essentially the same in Clavelina, but were very much easier to
determine, on account of the larger size of the object.

The stage which van Beneden and Julin took as their point
de départ was much older than that which I shall now
commence with. In fact, their first stage was that at which
the larval nervous system had already attained the climax of
its development.

Stage 1.—The stage from which I find it is desirable to start
in describing the future development and fate of the nervous
system of Clavelina is a very young embryo, with the anterior
neuropore still open to the exterior; no mouth, no atrial
involutions, no pigment in the brain, and before the migration
of the otocyst.

A transverse section through the neural tube, some distance
behind the neuropore, is shown in fig. 25, Pl. 11. The part of
the neural tube extending between the region through which

308 ARTHUR WILLEY,

this section is taken and the neuropore has in transverse
section an approximately round contour, and is quite simple.
Its lumen, which more anteriorly is reduced to a minimum,
gradually widens out until it becomes a transversely elongated
slit, as shown in the section figured. A large cell im the
dorsal wall of the neural tube in fig. 25 can be identified as
the otocyst, although at present it contains no pigment.

The nerve-tube has therefore not yet commenced to swell
out in its anterior region into the remarkably voluminous
cerebral vesicle which appears later. At this stage it is chiefly
desired to call attention to the fact that at a region consider-
ably removed from its anterior extremity the neural tube,
though still simple, possesses a transversely elongated lumen.

Stage I1.—In embryos belonging to this stage the nerve-
tube still opens in front to the exterior by the neuropore. No
mouth is present, but the atrial involutions have put in their
appearance, in the form of the two well-marked longitudinal
grooves which I have previously described (No. I). No
stigmata have broken through. ‘This stage also marks the
first appearance of pigment in the brain, while the otocyst has
attained its ventral position. In fig. 30 an intermediate stage
in the migration of the otocyst is shown, its position there
being lateral, on the right wall of the cerebral cavity.

Figs. 26—29 represent transverse sections through the cere-
bral portion of the nervous system of an embryo belonging
to Stage II. Fig. 26 passes through the neuropore, and was
drawn with a higher power than the succeeding figures of
this series. In fig. 27 the section is taken a little behind the
neuropore, and the regularity of the circumference of the
neural tube is disturbed on the right side (left of the figure)
by a bluntly-pointed protuberance, which becomes still more
prominent as we pass backwards in the series. Meanwhile the
neural tube begins to show a tendency to divide itself trans-
versely into two portions; and, in fact, when we reach the
point from which the section shown in fig. 28 was taken, we
find that here the neural tube is double, and possesses in this
region two distinct lumina.

STUDIES ON THE PROTUCHORDATA. 309

This double character of the neural tube only extends in
this stage through two or three sections. Anterior and pos-
terior to this region it is a simple tube with a single lumen
(figs. 27 and 29). Of the two halves of the neural tube in fig.
28, that on the left side (right of the figure) retains approxi-
mately its present shape, and forms part of the future
hypophysis; while the other (right) division of the neural tube,
which is at present rather smaller than its neighbour, becomes
enormously distended in the later stages, and is converted
into the spacious cerebral vesicle.

We see, therefore, at this stage the first commencement of
the separation of the hypophysis from the rest of the larval
nervous system taking place entirely independently of any
evagination from the wall of the stomodzum, which, indeed,
does not yet exist. It is to be noted also that the formation
of the hypophysis commences here at a much earlier stage than
in the case of Ciona, a fact which is in thorough keeping with
the general character of the development of Clavelina, to which
I have already alluded.

The neuro-hypophysial tube decreases considerably in dia-
meter in the later stages, owing to the absorptiun of the yolk
with which its cells are at first filled (cf. figs. 28, 31, and 40).

Stage III.—At this stage the lumen of that portion of the
neural tube which will give rise to the cerebral vesicle is com.
mencing to enlarge (fig. 81). The neuropore is closed, but
there is still nomouth. The lumen of the neural tube in front
has a more or less round outline, but widens out transversely
behind until, as in the preceding stage, but now in a more
pronounced way, it becomes divided into two. This condi-
tion is shown in fig. 31, from which it will be seen that the
two portions of the neural tube have now reversed the relative
dimensions which they held in the preceding stage,—the one on
the right side, namely, the one that will become, and, in fact,
is becoming, the cerebral vesicle, and which contains the oto-
cyst, being considerably larger than the other division of the
tube, which, as we have already seen, is the rudiment of the
hypophysis. The section drawn in fig. 32 lies slightly poste-

310 ARTHUR WILLEY.

rior to the region from which fig. 31 was taken, and shows
again the posterior communication between the two portions
of the neural tube. In comparing figs. 31 and 382 with figs.
28 and 29 of the preceding stage, it will be seen that the hinder
opening of the hypophysial portion of the neural tube into that
division of it which corresponds to the later cerebral vesicle
lies somewhat farther backwards in the present stage ; whereas
in fig. 29 the two halves of the neural tube are in open com-
munication in the region of the otocyst, in fig. 31 they are dis-
tinct from each other in this region, and their lumina unite
more posteriorly (fig. 32). This fact illustrates the gradual
constriction of the hypophysial tube from the neural tube,
which is taking place from before backwards.

As we trace the series of sections backwards it is found that
the lumen of the nerve-tube gradually becomes again narrower,
showing that its expansion in the transverse direction is con-
fined to a particular region, namely, the region from which
the hypophysis takes its origin.

Stage IV.—At this stage the larva possesses a mouth.
Fig. 33 is a section taken through the cerebral region of the
neural tube of a young larva of this stage at the time of the
first formation of the mouth. It would seem that almost im-
mediately after the mouth has broken through, a communica-
tion is established between the neuro-hypophysial canal and
the cavity of the branchial sac at the base of the stomodzum.
This communication is at first a perfectly plain one, and not
involved with any evagination from the wall of the branchial
sac.

As to the region of the branchial sac into which the neuro-
hypophysial canal opens, it is only reasonable to suppose that
it corresponds to the base of the stomodzal involution. This
also follows from a comparison between the depth of the stomo-
deal invagination before the actual perforation of the mouth
and the level at which the communication between the neuro-
hypophysial canal and the branchial sac becomes effected at a
later stage (cf. Pl. 10, figs. 2 and 18).

It is possible that in Clavelina, as in Ciona intestinalis,

STUDIES ON THE PROTOCHORDATA, 311

a secondary infolding of relatively inconsiderable extent
takes place from the wall of the branchial sac—i. e. from the
base of the stomodeum,—and carries the original branchial
opening of the neuro-hypophysial tube farther inwards in the
same way as I have suggested above for Ciona.

Van Beneden and Julin, on the contrary, whose account of
the origin of the hypophysis differs essentially from that which
I am giving, speak of the entire hypophysis, including the
glandular portion of it, as arising from an endodermic diver-
ticulum of the branchial sac, and Professor Kupffer (10) has
recently seized on this statement to confirm him in his opinion
that the so-called hypophysis of the Ascidians is really nothing
of the kind, but merely a “ Kiemendarmdriise.”’ The whole
development of the Ascidian hypophysis, however, obviously
opposes itself to such a view.

To return then to fig. 33, we see here a further progress in
the distension of the cerebral vesicle, while the section also
shows the posterior opening of the hypophysis into the vesicle.
The division of the primitive neural tube into two does not
extend to its anterior extremity, but the whole of that portion
of the neural tube which in the previous stages lay between the
point at which the neuro-hypophysial constriction commenced
and the neuropore becomes bodily taken up in the service of
the hypophysis, and at its front end comes to open into the
base of the stomodzum as described above.

Stage V.—At this stage the cerebral portion of the medul-
lary tube has assumed its definite vesicular character with the
accompanying local thinning out of its wall, which has been
already referred to in the case of Ciona. The contrast between
the cerebral vesicle and the hypophysial tube in point of size is
now very great (figs. 34, 35). The latter here appears in the
form of a minute lumen in the thickness of the cerebral wall
just as we found it in Ciona.

Figs. 84—37 are taken from a series of transverse sections
which show very clearly the way in which the lumen of the
hypophysis opens posteriorly into the cerebral vesicle. In fig.
37, the most posterior of the sections drawn, there is no trace

312 ARTHUR WILLEY.

whatever of the hypophysis at this stage. The branchial or
stomodzal opening of the hypophysis and its cerebral opening
may be conveniently referred to as its anterior and posterior
openings respectively. The posterior opening of the hypo-
physis now occurs in the region of the cerebral vesicle which
contains the eye—that is, still farther back than in the preced-
ing stage. Figs. 38—42 represent a series of sections through
the cerebral vesicle of a larva which shows the hypophysis in
a rather more advanced stage of development. It now projects
from the wall of the cerebral vesicle, and has a definitely tubular
appearance. Fig. 88 shows the branchial or anterior aperture
of the hypophysis, while fig. 42 shows its posterior communi-
cation with the cavity of the cerebral vesicle. It should be
noted that the cerebral vesicle expands in every direction, not
only laterally, but also in a longitudinal direction, so that its
anterior wall projects far beyond its previous limit, and so
comes to lie side by side with the anterior opening of the
hypophysis, where its wall, as shown in fig. 38, consists
almost entirely of a thin and apparently structureless mem-
brane.

Stage VI.—At this stage the hypophysis no longer opens
into the cerebral vesicle sensn stricto, but its posterior open-
ing has been carried back by progressive constriction to the
region of the prominent ganglionic enlargement of the ventral
wall of the neural tube (figs. 45 and 46) which hes between
the cerebral vesicle and the anterior extremity of the notochord,
and which has been accurately described by van Beneden and
Julin. As described by the latter authors, this ganglion
eventually becomes absorbed and disappears entirely, leaving
the superjacent neural tube in the shape of a solid cordon
ganglionnaire viscéral. After this stage the posterior
opening of the hypophysis becomes closed, and only the anterior
opening into the branchial sac persists.

Fig. 43 represents a section taken slightly posterior to the
anterior opening of the hypophysis, and serves to illustrate the
general topographical relation of the parts. In fig. 44 a very
important point is illustrated—namely, the origin of the

STUDIES ON THE PROTOCHORDATA. ols

definite cerebral ganglion on the left side (right of the figure)
of the cerebral vesicle just above the hypophysis tube. The
continuity of the lumen of the latter can be traced in the
clearest manner from the anterior branchial opening to the
section under consideration. The extreme posterior termina-
tion of the hypophysis was rather difficult to make out at this
stage, and between the sections drawn in figs. 44 and 45 there
seemed to be an interruption in the continuity of the lumen.
This probably is an indication of the eventual closing up of the
hypophysis posteriorly.

As to the actual origin of the cells which compose the
permanent cerebral ganglion, it is undoubtedly correct to say
that they proceed, together with the hypophysis, from the cells
which form the left dorso-lateral portion of the wall of the
cerebral vesicle (cf. figs. 834—37). This becomes especially
obvious by the study of such a series of sections as that from
which figs. 43—46 were taken. In the hinder region of the
cerebral vesicle the boundary line between the hypophysis and
the developing ganglion was by no means so distinct as it is in
fig. 44.

It is clear from the above description that in Clavelina the
formation of the permanent ganglion commences at a relatively
much earlier stage than it does in Ciona—in fact, before the
atrophy of the cerebral vesicle; and we see, further, that it is
from the begiuning a solid structure. Van Beneden and Julin
appear to have mistaken the developing hypophysial tube for the
developing ganglion. Some of their figures coincide fairly
closely with some of mine, but they have interpreted them
totally differently, and, it must be added, to a large extent
erroneously. They agree with Seeliger in saying that the hypo-
physis, including for their part emphatically its glandular por-
tion, is entirely derived from an evagination from the wall of the
branchial sac, to which they give the name of the “ cecum
hypophysaire,” which applies itself against the cerebral
vesicle, but never communicates with it, My observations
show conclusively that this is quite wrong.

314 ARTHUR WILLEY.

Van Beneden and Julin say (loc. cit., p. 353), ‘‘ Le cerveau
de Vadulte procéde du cul-de-sac cérébral.”’” But their ‘cul-de-
sac cérébral,” which they suppose to be entirely transformed
into the adult ganglion, is no other than my neuro-hypophysial
canal; and although, as we have seen, the brain of the adult
does proceed from the same epithelial tract as the latter
structure, yet it is perfectly distinct from it to the extent that
the lumen of the “ cul-de-sac cérébral”’ is and continues to be
throughout the lumen of the neuro-hypophysial canal. Evi-
dently the true origin of the ganglion of the adult escaped the
attention of the Belgian authors. For the rest, 1 may repeat
that their fig. 11, planche xvii, which would appear to be clear
enough to dispel any doubt as to the origin of the hypophysis,
is to me, in that regard, quite unintelligible, and I have been
unable to duplicate it in any of my preparations. Possibly in
the figure in question it is merely the funnel-shaped anterior
dilatation of the hypophysis which has been drawn, its posterior
narrower continuation having eluded observation.

In face of the above statements I was much surprised to
read in an interesting note on the eyes and subneural gland of
Salpa, communicated to a recent number of the ‘ Zoologischer
Anzeiger’ (No. 409, Jan., 1893) by Maynard M. Metcalf, the
following lines :—‘‘ The ganglion of Salpa is homologous with
the visceral portion of the larval Ascidian nervous system.
Van Beneden and Julin have shown that the dorsal wall of
this portion of the Ascidian tadpole’s neural tube proliferates
cells which become the ganglion of the adult, while the thick-
ened ventral wall of the same region gives rise to the subneural
gland.” It is sufficiently clear from what I have said above
that this statement must rest upon a complete misapprehension
on the part of the author as to the results arrived at by van
Beneden and Julin. This is what they say about the subneural
gland (loc. cit., p. 350) :—“ En un point de son trajet [i.e. of .
the cecum hypophysaire] sur un petit nombre de coupes et
seulement dans sa partie antérieure on constate que le plancher
du tube épithélial [i. e. the hypophysis-tube which for them is
derived entirely from an evagination of the wall of the

STUDIES ON THE PROTOCHORDATA. Sip

branchial sac] s’est développé en un petit amas de cellules ;
e’est la ’ébauche de la glande hypophysaire.”’

A communication between the cavity of the central nervous
system and that of the branchial sac in the Tunicata has been
observed in several other cases by previous authors—thus by
Ganin (2) in the case of Didemnum (Diplosoma) gelatino-
sum, Keferstein and Ehlers (see Uljanin, 20) in the case of
Doliolum, Kowalevsky (7) for Pyrosoma, Salensky (14 and
15) and more recently Metcalf (12 and 18) for Salpa, Lahille
(11) and Hjort (5) for Distaplia magnilarva, and Hjort
again for Botryllus.

From the observations of these authors, together with those
which I have recorded above, we may conclude that in all the
Ascidians the lumen of the hypophysis is in all cases at first in
direct communication with the lumen of the central nervous
system. And this forms the great difference, but at the same
time a very suggestive and instructive difference, between the
development of the hypophysis in the Ascidians and in the
higher Vertebrates. In the latter the lumen of the oral por-
tion of the hypophysis does not come into communication with
the cavity of the infundibulum, and this permanent separation
of the two parts of the hypophysis cerebri in the higher Verte-
brates may be compared with the temporary obliteration of the
lumen between the proximal and distal portions of the hypo-
physis which I have described above for Ciona.

Julin’s (6) and Balfour’s (‘Comp. Embryol.,’ vol. ii,
p. 437) suggestion of the homology of the subneural gland
and dorsal tubercle taken together of the Ascidians, with the
pituitary body of the higher Vertebrates, founded on anato-
mical considerations, and especially worked out in great detail
by Julin, may be considered as being borne out fully by the
facts of development as described above. In the Ascidians, as
in the higher Vertebrates, the hypophysis cerebri consists of a
neural portion and an oral or stomodzal portion. The neural
portion of the hypophysis in the higher Vertebrates is the in-
fundibulum or processus infundibuli, and in the Ascidians it
may be that this is represented by the subneural gland. The

316 ARTHUR WILLEY.

oral portion in the Vertebrates is the pituitary body, and in
the Ascidians the proximal portion of the hypophysis, in-
cluding the dorsal tubercle,

ITI.

On the Position of the Mouth in the Larve of the Ascidians
and Amphioxus, and its Relation to the Neuroporus.

(With Plate 20.)

In the first of these “ Studies” I have quoted a sentence of
Kupffer, in which he says that the pronounced dorsal position
of the mouth in the Ascidian tadpole is occasioned by the pre-
sence of what I have called the przoral lobe, which contains
the anterior body-cavity. But in Balanoglossus, where an
homologous anterior body-eavity or proboscis-cavity is present,
the mouth is ventral. So that, according to this point of view,
what causes the mouth to be dorsal in one case causes it to be
ventral in another.

The way out of this dilemma is found as soon as the fact is
recognised that the anterior body-cavity has nothing to do per
se with the position of the mouth, and that at least in the
groups of the Protochordata (Cephalodiscus, Balanoglossus,
Tunicata, Amphioxus) the dorsal or ventral position of the
mouth does not affect the homology of organs which lie in
front of it, for the reason that there is every evidence to show
that the anterior body-cavity in all these forms is not a truly
median structure, but has, either actually or virtually, a bi-
lateral origin.

In Balanoglossus, as shown by Bateson, the anterior body-
cavity arises at first as a perfectly median archenteric pouch.
It becomes, however, in the course of the development, incom-
pletely divided into two by the formation of a mesenchymatous
septum, in which lie the so-called heart, notochord, and pro-
boscis-gland. But perhaps the strongest evidence of the essential
bilaterality of the proboscis-cavity of Balanoglossus is, that
while in most forms there is only one proboscis-pore, namely,

STUDIES ON THE PROTOCHORDATA. 317

on the left side, in B. Kupfferi, as is well known, there are
two such pores—a right and a left.

Returning to the question of the dorsal position of the
mouth in the Ascidian tadpole, I have on a previous occasion
(see this Journal, vol. xxxii, N. 8., 1891, pp. 214—217) put
forward the suggestion that the lateral position of the mouth,
and consequently the unilateral position of the gill-slits, in the
larva of Ampbioxus, was due to the mouth having been shifted
from a primitively dorsal to a lateral position by the secondary
forward extension of the notochord (see Pl. 12). Any attempt
to account for this position of the mouth on principles of
utility to the larva would be futile, because it only occurs
during the period in which the larva is pelagic. On the other
hand, when the young Amphioxus begins to burrow in the
sand at the bottom or near the shore, frequently lying on its
side on the sand, the mouth has already become median, an-
teriorly directed and ventral. The observations which I have
been able to make as to the relations existing between the
mouth, hypophysis, and nervous system in the Ascidians have
raised the above view as to the origin of the asymmetry of the
larva of Amphioxus in my mind from the rank of a tentative
suggestion to that of a demonstrated fact.

Tn the Ascidians, as we have seen, the neuropore opens, or
more correctly reopens, at first directly into the stomodzum.
Later on there is some reason for supposing that an evagina-
tion occurs from the stomodzum which carries the original
neuropore further back.

In Amphioxus the neuropore opens for a long time directly
to the exterior in the dorsal middle line, and then later an in-
vagination of the epidermis occurs, which carries the neuro-
pore some distance inwards. This invagination gives rise to
the so-called “ olfactory pit” of Kolliker, or “ Flimmergrube”
of Hatschek, and into its base, as shown by Hatschek, the
nerve-tube at first opens by the neuropore. Eventually the
neuropore becomes closed, and the olfactory pit is then a ciliated
cul-de-sac abutting against the anterior end of the nerve-tube.

Thus the so-called olfactory pit of Amphioxus bears precisely

318 ARTHUR WILLEY.

the same relation to the neuropore as the dorsal tubercle does
to the neuropore in the Ascidians (cf. fig. 14, Pl. 1). The
only conclusion to be drawn from this is that Kélliker’s olfac-
tory pit in Amphioxus is homologous with the proximal
portion of the hypophysis duct in the Ascidians, while the
glandular portion of the hypophysis is unrepresented in
Amphioxus.

Hatschek (I, 16), if I understand him aright, has curiously
enough suggested that the dorsal tubercle of the Ascidians—
that is, the opening of the hypophysis duct into the stomo-
deum—is homologous with the preoral pit of Amphioxus ;
while the glandular portion of the Ascidian hypophysis, or the
“ Neuraldriise,’ would be homologous with the olfactory pit
(Flimmergrube) in Amphioxus, the two portions of the hypo-
physis being in the latter separated from one another by the
notochord. Judging from a recent publication (8), in which
Hatschek makes the preoral pit of Amphioxus a gill-slit, he
would seem to have somewhat modified his original view,
which was based largely on observations made by Herdman
(4) on Ascidia mammillata, in which, while confirming
Julin’s discovery that in this species the neural gland, besides
having the usual duct running anteriorly to communicate with
the pharynx by the dorsal tubercle, has also a number of short
funnel-shaped apertures into the peribranchial cavity, he adds
that in two specimens examined by him the duct of the hypo-
physis had no opening into the pharynx, the dorsal tubercle
being entirely absent. Herdman, therefore, suggested that
the dorsal tubercle and neural gland represent originally dis-
tinct structures, which in most Ascidians have acquired a
secondary communication with one another. This view, which
receives only the slenderest support from the facts intended to
establish it, is obviously untenable in the light of what has
been said above as to the origin of the respective structures.

In 1870, several years before Ussow discovered the con-
tinuity of the dorsal tubercle of the Ascidians with the sub-
neural gland, Ganin, in studying the development of Di-
demnum (Diplosoma) gelatinosum, found that the cavity

STUDIES ON THE PROTOCHORDATA. 319

of the central nervous system communicated directly with that
of the branchial sac, and said (2, p. 515), “Somit ist die
Flimmergrube [dorsal tubercle] der Ascidien am ehesten mit
dem Geruchsorgane [olfactory pit] des Amphioxus zu ver-
gleichen.”

Schimkewitsch (16) has recently put forward the same
opinion in that he says, “ Der vordere Neuroporus [of Balano-
glossus] entspricht der Flimmergrube der Amphioxus-Larve
(Hatschek) und dem Flimmerausgang der Neuraldriise der
Tunicaten (Julin).”

I consider it, therefore, well established by all this more or
less concurrent testimony that the hypophysis of the Ascidians
is represented in a simplified form by the olfactory pit of Am-
phioxus, both structures communicating during a longer or
shorter period of the development with the cavity of the
central nervous system by means of the neuropore. But
while in the Ascidians the hypophysis opens into the mouth-
cavity, in Amphioxus it opens dorsally to the exterior, and is
separated from the mouth by the notochord.

In Amphioxus the mouth has not merely been forced by the
forward extension of the notochord to forsake its primitive
dorsal position, but it has also, ipso facto, lost its primitive
relation to the hypophysis, by which name we may now
designate the olfactory pit of Amphioxus.

The relation of the mouth to the hypophysis is a remarkably
close and constant one throughout the whole of the Vertebrate
series. There are, however, as might be expected, some ex-
ceptions to the general rule. One of these exceptions is the
well-known case of Petromyzon, where the hypophysis, as shown
by Dohrn (1), Scott (17), and Kupffer (9), arises approximately
in the normal position for the Craniata, and is then secondarily
carried round to the dorsal middle line by the enormous
development of the upper lip which grows out between the
hypophysial involution and the stomodzum.

Another exception is met with in the case of Amphioxus,
where it is not the hypophysis which has been carried away
from the mouth, but the mouth which has been separated by

VOL. 39, PART 2.—NEW SER. Z

320 ARTHUR WILLEY.

the secondary forward extension of the notochord from the
hypophysis. In Petromyzon the whole process can be observed,
while in Amphioxus only part of it, as the notochord grows
forward at a very early stage before the formation of the mouth.

I take it for granted, therefore, that the mouth of Amphi-
oxus was primitively dorsal, and the prime reason of its being
dorsal was, not the presence of a preoral lobe or anterior
body-cavity, but the fact that in the common ancestor of the
Urochorda and Cephalochorda the mouth stood in intimate
relation with the neuroporus, probably through the inter-
mediation of a ciliated funnel or hypophysis. This conclusion
would suit very well with the views of Sedgwick (19) and
van Wijhe (21) as to the primitive respiratory function of the
neural canal, water entering it by the neuroporus which opened
into the mouth, and leaving it by the neurenteric canal.

As to the position of the mouth in the higher Vertebrates,
it is obvious, supposing the above considerations to be correct,
that it has, so to speak, been pushed round to its present ven-
tral or subterminal position by the cranial flexure. ‘This was
first suggested in part by Sedgwick in his well-known paper on
the “ Origin of Metameric Segmentation,” although the mouth
of Amphioxus, whose final ventral position is due to an entirely
different set of causes, was left out of consideration. He says
(18, p. 77), “With a slight change in the shape of the ante-
rior end of the body of the Ascidian larva in Kowalevsky’s
figure, the mouth would be removed from what we call the
dorsal (neural) to what we call the ventral (abneural) surface.
This would involve a flexure of the anterior end of the neural
canal, and, I think, gives a clue to the phylogenetic meaning
of the cranial flexure.”’

As for the higher Vertebrates, my friend Mr. H. B. Pollard
had the kindness to show me in Naples some of his prepara-
tions of Teleostean embryos, in which it could readily be seen
that the hypophysis was morphologically dorsal with reference
to the nervous system, its actual ventral position being due to
its having been carried round the front of the head by the
cranial flexure, just as the optic nerves are morphologically the

STUDIES ON THE PROTOCHORDATA. 3821

first pair of nerves, as pointed out by van Wijhe. This point is
of great importance, and is very strong evidence in favour of
the view that the hypophysis of Amphioxus (i.e. Kolliker’s
olfactory pit) occupies a primitive position, which in the higher
Vertebrates has been shifted to the ventral median line by the
cranial flexure.

Returning to the Protochordata, it follows from what has
been said above that the mouth occupies a more primitive
position and exhibits more primitive relations in the Ascidian
tadpole than it does in Amphioxus. In the larva of Amphi-
oxus, however, the mouth occupies an intermediate position
between that of the Ascidian larva and that of the adult
Amphioxus. Some of the stages in the migration of the mouth
from a dorsal to a ventral position have, in fact, been preserved
to us in the ontogeny of Amphioxus. The mouth of the adult
Amphioxus occupies the same position as that of the craniate
Vertebrates, but gets there by totally different means. It is
extremely interesting to note that there is more than one way
in which the primitive position of such an apparently stable
organ as the Vertebrate mouth can become altered, namely,
either by the cranial flexure or by a forward extension of the
notochord.

Thus we find that in the case of the mouth of Amphioxus
and the higher Vertebrates we have almost identical topogra-
phical relations, established by widely divergent methods. A
similar instance is afforded by the hypophysis, which opens to
the exterior in the dorsal middle line in both Amphioxus and
Petromyzon, but primarily in the former and secondarily in
the latter form.

A question may arise as to the actual way in which the
mouth of Amphioxus could have been originally forced aside
from its primitive position by the advance of the notochord. The
probability is that the actual oral opening was never dis-
placed by the notochord. But the change from a dorsal to a
lateral position of the mouth in the larva of Amphioxus could
be, and undoubtedly has been, effected by a change in the
order of its appearance.

322 ARTHUR WILLEY.

The time or order of formation of certain organs seems to be
very generally subject to a great deal of variation. I have
previously described some such variations in the case of the
secondary gill-slits of Amphioxus, and similar instances are
very easy to find. I therefore suggest that either a slight
delay in the formation of the mouth, or an acceleration in the
anterior development of the notochord—probably the latter,—
introduced in the first place as a variation and subsequently
becoming a fixture, was the method by which the perforation
of the mouth at a point other than the primitively dorsal one
was rendered possible.

The above observations and considerations all tend to show
that the primitive vertebrate mouth, before the cranial flexure
had become an established feature of the vertebrate ontogeny,
had a dorsal or a dorso-terminal position.

In view of this conclusion a genuine difficulty is presented
by the position of the mouth in Balanoglossus, where it is from
the beginning ventral. This difficulty cannot be fully met at
present, but it may be well to point out that the intermediate
stage between a ventral mouth as found in Balanoglossus, and
a dorsal one as it occurs in the Ascidian tadpole, would be
arrived at by a form in which, by a reduction of the preoral
lobe, the mouth came to occupy a terminal position. Supposing
it possible to conceive a common ancestor for all the Proto-
chordata, it would seem to be probable that it had a terminal
mouth. For from such a situation the mouth could be made
to assume either a definitely dorsal or ventral position, accord-~
ing to circumstances, as soon as the paired head-cavities
co-operated to produce the peculiar features and proportions of
a preoral lobe.

Appendicularia is a form in which, together with the
reduction, and indeed apparent absence in the adult of any
trace of a preoral lobe—a state of things brought about by the
purely pelagic life which has been acquired by the organism—
the mouth has come to occupy a terminal position, and thus
shows us that, under certain circumstances, the topographical
relations of the mouth which I have just predicated for the

STUDIES ON THE PROTOCHORDATA. 328

ancestor of the Protochordata could exist, and, moreover, in a
free-swimming animal.

In Sagitta, again, we have a pair of head-cavities which are
very possibly homologous in a certain way with the head-
cavities of Amphioxus, but which do not occur in such a way
as to produce a prexoral lobe, and therefore do not prevent the
mouth from holding its anterior terminal position. The
preoral lobe or proboscis of Balanoglossus, as well as its
homologue which I claim to have identified in the Ascidian
larva, represents a pair of head-cavities analogous to those that
occur in Sagitta—although I do not wish to assert a genetic
relationship between the former and the latter—which, how-
ever, have acquired such a mode of development as to produce
by their fusion a large median lobe in front of the mouth.
The preoral lobe, however, while standing in the way of a
terminal mouth, does not, as I have said above, determine
whether the mouth shall be dorsal or ventral. That is
dependent on other circumstances, such as the mouth coming
into important relations to the central nervous system.

In Balanoglossus and in the Ascidians the two head-cavities
do not appear as such distinctly paired structures in the
ontogeny of the individual as they do in Amphioxus. And in
the latter case, as is well known, they do not fuse together, but
remain distinct, one of them undergoing hypertrophy and
giving rise to the preoral body-cavity and the other to the
preoral pit.

This hypertrophy of the head-cavities in the forerunners of
the Protochordata necessitated a change in the position of the
mouth, and a removal from its primitive situation at the
anterior terminal extremity of the body. Along the line of
descent which led to Balanoglossus the mouth migrated along
the ventral side of the body, and along the line of descent
that led to Amphioxus and the Ascidians the mouth passed
along the dorsal side, but in all cases the identity of the
head-cavities and of the mouth remained unaffected.

I have thus shown a possible means of explaining the
discrepancy between the primitive position of the mouth in the

324 ARTHUR WILLEY.

Ascidian tadpole and the larva of Amphioxus and in Balano-
glossus, which may at least serve as a working hypothesis.
My main object has been to point out, that the fact that the
mouth lies dorsally or ventrally has nothing to do with the
homology of the przoral body-cavity in all the forms in
question. The homology between the preoral celom of
Balanoglossus and the head-cavities of Amphioxus was urged
very strongly by Bateson, but it is most important to
remember, as has been repeatedly pointed out, that the mouth
of Amphioxus, although ventral in the adult, has, as I think
beyond a shadow of a doubt, descended from a primitively
dorsal position in the neighbourhood of the neuropore. In
other words, the mouth in Amphioxus originally possessed the
same topographical relations as it does in the Ascidian tadpole;
and if the preoral body-cavity of Amphioxus is homologous
with the corresponding structure in Balanoglossus, so is the
preoral body-cavity of the Ascidian tadpole. The above
considerations all tend to establish the accuracy of my identifi-
cation of the latter structure.

The diagrams on Pl. 12 will place the whole question here
discussed in the clearest possible light. From these diagrams
it will be at once seen that the mouth of the larva of Amphioxus
Occupies an intermediate position between that which it holds
in the Ascidian larva and in Balanoglossus, but I am very far
from meaning to suggest that phylogenetically it represents an
intermediate stage between these two extremes. On the con-
trary, it certainly does not. There is no evidence whatever to
suppose that the mouth of Balanoglossus has migrated from a
dorsal to a ventral position. As has been said above, it is
probable that both the mouth of the Ascidian and that of
Balanoglossus have attained their present situation from an
ancestral terminal position.

Errata.

In No. I of these “ Studies on the Protochordata” one or
two trifling lapsus calami, which I had not the opportunity
of correcting in the proof, crept into the text. On p. 348, re-

STUDIES ON THE PROTOCHORDATA. 325

ferring to the pyloric gland of Ascidians and the cecum of
Amphioxus, it is stated that they both lie on the left side.
While they are of course essentially median outgrowths of the
alimentary canal, the cecum of Amphioxus usually lies for the
greater part of its extent to the right of the pharynx. Atten-
tion may, however, be drawn to Schneider’s observation (‘ Beit.
zur vergl. Anat. und Entw. der Wirbelthiere,’ Berlin, 1879,
p. 17, foot-note) that he often found it on the left side of the
pharynx.

Finally, on p. 336 (eight lines from bottom of page)
“Trigeminus ” should read “ Facialis.”

ADDENDUM.

Since the above contributions were sent into the press,
several new publications relating to kindred subjects have
appeared.

1. A. Pizon, ina “ Note additionnelle” appended to his long
treatise on the Blastogenesis of the Botryllide (see ‘ Annales
des Sciences Nat.,’ 7me série, t. xiv, p. 374, et seq.), expresses
doubt as to the accuracy of Hjort’s and my results, and states
his own opinion that in the larve of Botryllus, and in the buds
of many other forms, “ lVorgane vibratile est toujours un diver-
ticule de la vésicule endodermique primitive.”

2. Hjort (‘“ Uber den Entwicklungscyclus der zusammen-
gesetzten Ascidien,’ ‘Mitth. Zool. Stat. Neapel,’ x, pp.
584—617, Taf. 37-39) gives a detailed account of his re-
searches on the budding of Botryllus and the metamorphosis
of the nervous system of Distaplia. In the latter case his
observations agree in the most satisfactory manner substan-
tially with mine on Ciona and Clavelina.

3. Davidoff (“ Uber den ‘ canalis neurentericus anterior’
bei den Ascidien,” ‘Anat. Anz.,’ viii, pp. 301—303) objects to
the identification of the subneural gland of the Ascidians with
the hypophysis of the Vertebrates, and agrees with Kupffer
that the latter is homologous with the Tunicate mouth.

4. Van Wijhe (“Ueber Amphioxus,”’ ‘ Anat. Anz.,’ viii, pp.
152—172) agrees with me in regarding the club-shaped gland

326 ARTHUR WILLEY.

as a modified gill-slit, but thinks that its antimere is the larval
mouth which he calls the Tremostoma. This he homologises
with the left spiracle of Selachians. The Tunicate mouth is
for him represented in Amphioxus by the pre-oral pit, which
he calls the Antostoma.

For the rest, van Wijhe records some most important obser-
vations on the peripheral nervous system and on the muscula-
ture of Amphioxus.

5. W.Salensky (“ Morphologische Studien an Tunicaten:
I, Ueber das Nervensystem der Larven u. Embryonen von
Distaplia magnilarva,” ‘ Morph. Jahrb.,’ xx, pp. 48—74)
appears to come to similar results to those already published
by Hjort with regard to the neuro-hypophysial system of
Distaplia. He also comes to a conclusion on which I have
dwelt in the foregoing pages in connection with Ascidia men-
tula. Salensky finds likewise in Distaplia “dass alle Theile
der Sinnesblase: Retina, Linse, Pigmentschicht und Otoli-
thenzelle durch die Differenzirung einer und derselben Epi-
thelschicht der primitiven Gehirnblase enstehen.”

6. W. K. Brooks (“ Salpa in its Relation to the Evolution
of Life,’ ‘Studies from the Biological Laboratory, Johns
Hopkins University,’ vol. v, No. 3, Baltimore, 1893).

At the conclusion of this otherwise interesting memoir, Prof.
Brooks devotes several paragraphs (pp. 199—201) to a criticism
of my “ Studies on the Protochordata” (No. I, ‘ Quart. Journ.
Mier. Sci.,’ vol. xxxiv, Jan., 1893).

In the body of his memoir, Professor Brooks develops, with
great elaboration, the view that “the chordata type was
evolved under purely pelagic influences,” and that Appendi-
cularia is the direct descendant and somewhat modified living
representative of this pelagic archetype.

Then, referring to my work, he says (p. 199), ‘‘ While the
author seems to agree with me in rejecting Dohrn’s view that
the Tunicates are degenerated fishes, he holds that the Ascidians
exhibit, during their development, certain features of resem-
blance to other primitive chordata which are not exhibited by
Appendicularia ; and he believes that these characteristics prove

STUDIES ON THE PROTOCHORDATA. 327

that the Ascidians are more closely related than Appendicularia
to these protochordata.

“ The features upon which he lays most emphasis are these :—
I. The endostyle is at first vertical and pre-oral ; II. The organ
of fixation is a pre-oral lobe, and its cavity is the pre-oral or
anterior body-cavity ; and III. The first four primary stigmata
of Ciona intestinalis are developed from one primitive gill-

“T cannot believe that students of the Tunicata will regard
the first and second of these argumeuts as entitled to the least
consideration.” After this very decided expression of opinion,
Professor Brooks goes on to say, “It has long been known
that the endostyle of Ascidian larve is at first vertical or at
right angles to the long axis, and it is so figured and described
by Seeliger ; but the relative position of organs is so much
influenced by changes in other organs that we cannot attribute
a phylogenetic significance to the position of the endostyle.”
It was certainly so described and figured by Seeliger for
Clavelina, and as a tribute to the excellence of his description
I quoted a considerable portion of it verbatim in my paper
(loc. cit., pp. 331, 332).

The point on which I insisted, however, was that in the larva
of Ciona, a simple Ascidian whose development in comparison
with that of Clavelina is remarkably uncompressed, the endo-
style behaved in the way stated by me, and not as described
and figured by Kowalevsky in the case of a closely allied
simple Ascidian. It is not a light thing to impeach Kow-
alevsky’s accuracy, and I considered it important to call
attention to the actual relations of the endostyle in the larva of
the simple Ascidians, which had not been done before.

With regard to the second half of the above-quoted para-
graph, I will merely point out that the primary position of the
endostyle in the larva is that which it holds prior to the
changes in the arrangement of the other organs, in which it is
subsequently involved.

The method by which the endostyle attains its secondary
and final position has nothing whatever to do with the question

328 ARTHUR WILLEY.

as to whether its primary position has a phylogenetic signi-
ficance. The remarkable constancy of the latter and its
analogy with Amphioxus would seem to indicate that it has.

Professor Brooks says (p. 200), “ Willey’s observations
add nothing to Seeliger’s excellent account of the organ of
fixation”’ (except to show that it behaves very differently in
Ciona from what it does in Clavelina, the differences being of
such a nature as to affect very sensibly the morphological in-
terpretation of the structure); “and he gives no reason for
holding that it is a pre-oral lobe, except that it contains loose
mesenchyme-cells derived from the two lateral mesodermic
bands.” This isa complete misrepresentation. What I chiefly
relied on in forming my opinion as to its morphological value
was its topographical relations. It is the barest statement of
the facts of the case to say that it is a lobe, that it is pre-oral,
and that its cavity is the anterior and pre-oral portion of the
body-cavity. Under these circumstances I confess my in-
ability to understand how the suggestion as to the possibility
of this pre-oral lobe being genetically related to a similarly
placed structure in Amphioxus can, on any pretence, be re-
garded as not being entitled to the least consideration. The
presence of loose mesenchyme-cells in place of a lining epi-
thelium was emphasised by me as a necessary evil common to
the rest of the body-cavity. Professor Brooks, however,
argues as follows :—‘ This (i.e, the presence of loose mesen-
chyme-cells) is equally true of other parts of the body-cavity,
and there is no more evidence that the organ of fixation is a
pre-oral lobe than there is that it is homologous with the jaws
and teeth of sharks.”

“How much,’ I ask—“ how much consideration is this
argument entitled to?”’

“Tf,” continues Professor Brooks, “it is a pre-oral lobe, it
is a ventral one, and it cannot be compared with the dorsal one
of such protochordata as Balanoglossus and Amphioxus.” I
venture to think that the reflections urged in the foregoing
contribution, No. III, will demonstrate that here Professor
Brooks has fallen into an egregious error. The primary

STUDIES ON THE PROTOCHORDATA. 329

topography of the mouth in the larve of the Ascidians and
Amphioxus belongs to one and the same category, while
that of the mouth of Balanoglossus belongs to quite another
category. If it is desirable to speak of bilateral structures as
being dorsal or ventral, the pre-oral lobe of Amphioxus is,
palingenetically, ventral and not dorsal.

In view of the numerous divergent opinions which have
recently been expressed with regard to the correspondence of
parts in the Protochordata and Chordata generally (Kupffer,
Hatschek, van Wijhe, Davidoff), it is obvious how much
depends on a correct estimate of the asymmetrical mouth of
the larva of Amphioxus.

LITERATURE.

Works cited in the first ‘‘ Study ” and again referred to in
the foregoing pages have the numeral I prefixed to them.

1. Doury, A.—Studien. IIT: “Die Entstehung und Bedeutung der Hypo-
physis bei Petromyzon planeri,” ‘Mitth. zool. Stat. Neapel,’ iv,
1882, pp. 172—189.

2. Ganin, M.—‘‘ Neue Thatsachen aus der Entw. der Ascidien,” ‘ Zeit. f.
wiss. Zool.,’ xx, 1869-70, p. 512.

3. HatscHEx, B.—‘ Die Metamerie des Amphioxus und des Ammoceetes,”
‘Verh. Anat. Gesellschaft in Wien,’ 1892, pp. 136—161.

4. Herpmay, W. A.—“ On the Homology of the Neural Gland in the
Tunicata with the Hypophysis Cerebri,’ ‘Proc. Roy. Soc. Edin.,’
xii, 1882-4, p. 145.

5. Hsort, J.—‘Zum Entwickelungscyclus der zusammengesetzten As-
cidien,” ‘ Zool. Anz.,’ xv, 1892, 328—332.

6. Jutin, C.—“ Rech. sur l’organisation des Ascidies simples: sur |’Hypho-
physe,” &c., ‘ Arch. de Biol.,’ ii, 1881, pp. 59—126 and 211— 232,

7. Kowauevsky, A.— Ueber die Entwickelung der Pyrosoma,” ‘ Arch. f.
mikr. Anat.,’ xi, 1875.

8. Kurrrer, C. von.—“ Zur Entwickelung der einfachen Ascidien,” ‘Arch. f.
mikr. Anat.,’ vill, 1872, pp. 358—395.

330 ARTHUR WILLEY.

9.

10.

Ub le

12.

13.

14.

15.

16.

17.

18.

19.

20.

2i.

22.

Kurrrer, C. von.—‘ Die Entwickelung von Petromyzon planeri,”
‘ Arch. f. mikr. Anat.,’? xxxv, 1890.

Kurrrer, C. von.—“ Mittheilungen zur Entw. des Kopfes bei Acipenser
sturio,” ‘S. B. Ges. fiir Morph. und Physiol. zu Minchen,’ Nov.
and Dec., 1891, pp. 107—123.

Laniuie, F.—‘ Rech. sur les Tuniciers des cétes de France,’ Toulouse,
1890, p. 173.

MetcaLr, M. M.—* Anat. and Dev. of Eyes and Subneural Gland in
Salpa,” ‘Johns Hopkins Univ. Circulars,’ vol. xi, No. 97, April, 1892.

Mertcatr, M. M.—“ On the Eyes, Subneural Gland, and Central Nervous
System in Salpa,” ‘Zool. Anz.,’ xvi, 1893, pp. 6—10.

SaLensky, W.—“ Ueber die embryonale Entwickelung der Salpen,” ‘ Zeit.
f. wiss. Zool.,’ xxvii, 1876.

SALENSKY, W.—‘‘ Ueber die Knospung der Salpen,” ‘Morph. Jahrb.,’
ie Wye

ScHIMKEWITscH, W.—“ Ueber die morphologische Bedeutung der Organ-
systeme der Enteropneusten,” ‘ Anat. Anz.,’ v, 1890, pp. 29—32.

Scott, W. B.—“ Notes on the Development of Petromyzon,” ‘Journ.
Morph.,’ i, 1887, see pp. 263—271.

Sepewick, A.—‘On the Origin of Metameric Segmentation,” &c.,
‘Quart. Journ. Micr. Sci.,’ xxiv, 1884, pp. 43,—82.

Sepewick, A.— The Original Function of the Canal of the Central
Nervous System of Vertebrata,” ‘Studies from Morph. Lab., Cam-
bridge,’ ii, 1884, pp. 160—164.

Uusantn, B.—‘‘ Die Arten der Gattung Doliolum,” ‘ Fauna und Flora
des Golfes von Neapel,’ x, 1884.

Van Wisuz, J. W.—“ Ueber den vorderen Neuroporus und die phylo-
genetische Function des Canalis neurentericus der Wirbelthiere,”
‘Zool. Anz.,’ vii, 1884, pp. 683—687.

Witizy, A.—“ On the Development of the Hypophysis in the Ascidians,”
‘Zool. Anz.,’ xv, 1892, pp. 332—334.

STUDIES ON THE PROTOCHORDATA. 331

EXPLANATION OF PLATES 18, 19, and 20,

Illustrating Mr. Arthur Willey’s paper, “ Studies on the
Protochordata.”

Letters for Plates 18 and 19.

ant. p. Anterior opening of neuro-hypophysial canal into branchial sae.
at. Atrial involution. dr. s. Branchial sac. cer. ves. Cerebral vesicle.
cer. ves, res. Histolytic residua of cerebral vesicle. ¢. Hye, or eye-tract.
ect. Ectoderm. end. Endostyle. ent. Entoderm. ent. c. Knteric cavity.
ep. r. Hpibranchial ridge. g. Cerebral ganglion. g/. Subneural gland. hyp. /f.
Funnel of hypophysis. izé. Intestine. m. Mouth. mes. Mesoderm. x. c.
Neural canal. xch. Notochord. 2. hyp. Neuro-hypophysial canal. 2%. p.
Neuropore. of. Otocyst, or otolith. ost. p. Posterior opening of neuro-
hypophysial canal into cerebral vesicle. s¢. Stomodeum. ¢. Kemains of
tail. vac. Vacuolar spaces in cells of opticregion. vise. g. Visceral ganglion
(Rumpfganglion of Kowalevsky).

PLATE 18.

Figures 1—5 relate to Ascidia mentula, and were drawn from living
object with camera lucida.

Fic. 1.—Optical section of young embryo with closed nervous system,
showing first appearance of pigment-granules of eye in dorsal wall of brain.
Zeiss, 3, C.

Fic. 2.—Older embryo, to show depth of stomodeal invagination. Behind
cerebral vesicle is seen the right atrial involution. Zeiss, 8, C.

Fic. 3.—Cerebral vesicle of young embryo (a trifle older than that shown
in Fig. 1), to show primary relation of otocyst to eye-tract. Zeiss, 4, D.

Figs. 4 and 5.—Cerebral vesicles of somewhat older embryos in optical
section, to show stages in the migration of the otocyst. Zeiss, 3, D.

Figures 6—24 relate to Ciona intestinalis.

Fics. 6—8.—Transverse sections through cerebral vesicle of newly hatched
larva. Fig. 6 shows the neuro-hypophysial canal in left wall (right of figure)
of vesicle. Fig. 7 shows its communication with cavity of vesicle. Fig. 8
shows cerebral vesicle posterior to neuro-hypophysial region. Zeiss, 3, E.

Fig. 9.—Transverse section through an older larva of the age of that
figured in I, Plate XXX, fig. 1 (this Journal, Jan., 1893).

Fic. 10 shows posterior communication of neuro-hypophysial canal with
cerebral vesicle—farther back than in Fig. 7. Zeiss, 3, E.

302 ARTHUR WILLEY.

Fie. 11.—Transverse section through cerebral vesicle of newly fixed larva.
Shows disintegration of the vesicle, and complete separation of neuro-
hypophysial canal. Zeiss, 3, E.

Fic. 12.—Sagittal section through neuro-hypophysial system of same stage
as preceding, showing first appearance of adult ganglion as a proliferation in
the dorsal wall of the tube. Zeiss, 3, J, water immersion.

Fic. 13.—Entire section, of which Fig. 12 was a part, to show topographi-
cal relations. Zeiss, 3, C.

Fig. 14.—Sagittal section through more advanced neuro-hypophysial sys-
tem. Posteriorly in the region of the ganglion the section is tangential.
The nervous system proper has become solid. Zeiss, 3, J, water immersion.

Fies. 15—19.—Transverse sections through neuro-hypophysial system,
slightly older than preceding. Fig. 15 is the most posterior section, and
passes through the hinder portion of the ganglion or the anterior extremity
of the “cordon ganglionnaire viscéral,” which is now solid. Figs. 16—18
show the ganglion developing from the dorsal wall of the neuro-hypophysial
canal. In Fig. 18 the remains of the eye are involved in the section. Fig. 19
passes through the anterior extremity of the ganglion, which overlaps the
funnel of the hypophysis. Zeiss, 4, J, water immersion.

Fies. 20—24.—Transverse sections through the hypophysis and ganglion
of young immature adult. Fig. 20 is the most anterior section, passing
through funnel of hypophysis and two cerebral nerves. Fig. 21 is taken
slightly further back, and Fig. 22 passes through the point at which the lumen
of the hypophysis is temporarily obliterated by mutual approximation of the
cells forming its wall. Fig. 23 shows the lumen widening out again posterior
to this point, and finally Fig. 24 shows the origin of the glandular portion of
the hypophysis by cell-proliferation from its ventral wall in its posterior
portion, Zeiss, 2, J, water immersion.

PLATE 19.

All the figures on this plate relate to Clavelina lepadiformis, and all
represent transverse sections.

Fig. 25.—Stage I. Through cerebral region of very young embryo, of an
age corresponding to that shown in Plate 18, fig. 1. Shows transversely
elongated lumen of neural tube in this region. 3, D.

Fies, 26—29.—Stage II. Fig. 26, through neuropore; 3, J. Fig. 27,
just behind neuropore. Fig. 28, through neuro-hypophysial region. Fig. 29,
posterior to this. 2, D.

Fic. 30.—Intermediate between Stages I and II, showing otocyst in right
(left of the figure) wall of cerebral vesicle. 3, D.

Figs. 31 and 32.—Stage Il]. Fig. 31, through region of otocyst, shows
increase in size of cerebral vesicle. Fig. 32 shows the communication between
the hypophysial and cerebral portions of the nervous system. 3, D.

STUDIES ON THE PROTOCHORDATA. 333

Fig. 33.—Stage IV. Shows opening of neuro-hypophysial canal into still
larger cerebral vesicle, between the region of the otocyst and of the eye.
3, D.

Fies. 34—37.—Stage V. Series showing relation of neuro-hypophysial
canal to cerebral vesicle at this stage. Fig. 37 passes through the vesicle
behind the posterior opening of the canal. For the anterior opening into
the branchial sac in this larva see I, Plate XXXI, fig. 28. 2, D.

Figs. 38—42.—Similar series through somewhat older larva, showing the
neuro-hypophysial canal from its anterior opening into the branchial sac to its
posterior opening into the cerebral vesicle. 2, C.

Fies, 43—46.—Stage VI. Fig. 43 shows an entire section slightly posterior
to the anterior opening of the neuro-hypophysial canal, and in front of the
atrial cavities (cf. I, Plate XXXI, fig. 29); 2, D. Fig. 44 is a most im-
portant section, and shows the origin of the adult ganglion in company with
the neuro-hypophysial canal from the left (right of the figure) latero-dorsal
wall of the cerebral vesicle. Figs. 45 and 46 pass through the region of the
visceral ganglion, which later becomes absorbed. 2, J.

PLATE 20.
Letters for Plate 20.

p.¢. Preoral lobe. p.p. Preoral pit or proboscis pore. x. p. Neuropore.
m. Mouth. ed. Endostyle. x. c. Neural canal. xch. Notochord. gt. and
4. Proboscis-gland and heart of Balanoglossus.

Fic. 1.—Diagram of anterior portion of an Ascidian larva (e. g. Ciona)
about the time of fixation. The features possessed by the larva at the stages
immediately prior to and after fixation are thrown into one diagram.

Fig. 2.—Diagram of anterior region of larva of Amphioxus.

Fic. 3.—Similar diagram of Balanoglossus (compiled from Bateson).

a ae yee

: a re
cts WAM aL Sa Bhs! if 7
vee
NOY 7
im 4 ih 4

ah Lae

: ? i 4 é

= 4 ‘

. * ‘ All , “iG
: ‘ fupgalet
7 & Pr 4G war n;

. evan! Jah db 1h ee
; : Cs . » Bick aan Midi tvety ag
oa 7 7 Tid var | ‘ 7
4 : ; ve b re va
. ey : F g a i wy
me - . i i
-_ <

« uy
mn
_ : Ls
_ a
a.
- f -
.
f * .
-
1
.
-
§
;
Ree

DEVELOPMENT OF THE HEAD IN GOBIUS CAPITO. 335

Observations on the Development of the Head
in Gobius capito.

By

H. B. Pollard,
Oxford.

With Plates 21 and 22.

INTRODUCTION.

WHEN attempting to pursue some studies on the Compara-
tive Anatomy of the head in Teleostei, I became aware that it
would be desirable to understand the development of some
form. Gobius capito showed itself to be the most suitable,
and therefore in this species I have investigated and described
stages of development of the brain, mouth, and mesodermal
structures. The work has been carried out during occupation
of the Oxford table at the Naples Zoological Station.

My grateful acknowledgments are due to the University of
Oxford and to my college, Christchurch, for funds, and to the
members of the staff of the Naples Station for their constant
kindness, extending frequently to questions other than those
of material and reagents.

Material and Technique.

Gobius capito lays its eggs in the month of March at
Naples, at which time they can be brought in in unlimited
quantities by the fishermen. ‘The embryos are enclosed in a
tough shell, spindle-shaped, with pointed ends somewhat like
a grain of barley. The eggs are 4—5 mm. long, yellowish in
colour, and are regularly affixed to rocks and bits of pottery,

VOL. 35, PART 3.—NEW SER. AA

336 H. B. POLLARD.

&c., by one end—hundreds together. Apparently they are
deposited by preference on fragments of tufa, with the colour
of which they match. Kept in aquaria the young fish hatch
out at about the twenty-fifth day. Very possibly they grow
faster in their natural habitat. I found that the best way to
obtain the embryos ready for section-cutting was to snip off
the free projecting ends of the egg-shells with scissors, and
quickly shake the embryos out into dilute nitric acid or picro-
sulphuric acid (Kleinenberg). Afterwards when wanted the
embryos may be removed from the yolk in alcohol. [I at first
employed the methods of Henneguy (6), removing the embryos
from the yolk before complete hardening in alcohol, but by
these methods they are very liable to be deformed. Sections
were made by the Thoma microtome, and usually 5 p in thick-
ness. For convenience I used Kastschenko’s “ Beschneider”
(12) in order to obtain good ribbons of sections. The models
were made after Strasser’s method (28).

BRAIN.

One finds by experience that, in order to understand the
disposition of the various organs in the Teleostean embryo, an
exact comprehension of the form of the brain at the separate
stages is necessary. The development of the brain in the
trout has been described by Rabl-Riickhard (19), who investi-
gated the embryos as opaque and transparent objects under
the microscope. Goronowitsch (4) studied the early stages of
the brain in the salmon by means of models. My own ob-
servations have been made by models and projections, and
though in the main I have but to confirm what Goronowitsch
says, yet Gobius shows some not uninteresting variations from
the salmon, and, moreover, some description is necessary in
order later to explain the disposition ofthe mesoderm and mouth.

Fig. 1 represents the brain of Gobius in the first stage
viewed from the side. Scarcely any flexure is visible apart
from the general curve of the embryo. As shown by trans-
verse sections, the extent dorso-ventrally is far greater than
from side to side, the former being perhaps on the average

DEVELOPMENT OF THE HBAD IN GOBIUS CAPITO. 337

double the latter. The figure shows that the brain of Gobius
at this age bears considerable resemblance to the brain of a
human embryo of the third week as depicted by His (7), ex-
cept that the fish does not possess the sharp curve in the
region of the mid-brain. The chorda extends as yet only to
below the cerebellum, and the central cavity of the brain is
but partially formed. This cavity appears first in the region
of the optic stalks, and generally in the ventral half of the
neural cord.

Fore-brain, mid-brain, and hind brain are already indicated
by slight constrictions. In the hind brain can be distinguished
cerebellum and medulla oblongata. The mid-brain already
shows the preponderance characteristic of Teleostei. It is
slightly moulded to the shape of the eyes, which appear to be
relatively larger in Gobius than in Salmo. The fore-brain
differs somewhat from that of other Teleostei, and shows some
peculiarities which deserve special attention. Ventrally a
bulging of the floor indicates the commencement of the in-
fundibulum. Behind this point, as will be shown later, the
premandibular mesoderm is continuous across the middle line.

In front of the infundibulum is a constriction where the
optic chiasma comes to lie subsequently. The foremost ventral
portion of the brain is a rounded prominence which gives rise
laterally to the optic stalks and eyes. At the extreme anterior
end the wall passes inwards to form with the corresponding
upper portion of the brain a well-marked indentation.
The nasal organs lie above and in front of the optic stalks,
both nose and eye being as yet directed outwards and slightly
upwards. That portion of the brain lying between the nasal
organs is the region of the corpora striata. Figs. 5 and 6
represent sagittal and transverse sections of the anterior end
of the brain. In fig. 6, whose position is indicated in the
diagram (a, a, fig. 2), both upper and lower portions of
the brain are cut in section where they separate to form
the above-mentioned indentation. The upper portion, the
corpus striatum, is oval in section, the long axis being dorso-
ventral. The lower is more rounded. In the corpus striatum

338 H. B. POLLARD.

the central cavity is only indicated by the disposition of the
cells, and near the centre are seen several cells characteristic-
ally in mitotic division. In fig. 5, which is approximately in
a median vertical plane, the corpus striatum and lower ex-
tremity of the brain are seen to be separated by a fairly well-
defined line extending from the indentation to the median
cavity between the optic stalks. The ectoderm in this section
is artificially deformed. It should continue over the yolk
exactly abreast of the indentation. The characteristic mitoses
are again seen near the central cavity. Where the latter is
fully formed—that is, in the immediate region of the optic
stalks—it is seen to be partially occupied by structures (Gp.)
which Henneguy (6) terms “ globules parablastiques,’ but
which Goronowitsch regards as cell débris due to the manner
of formation of the central cavity. The same structures are
seen in section in fig. 7, 0.

The account here given confirms the observation of Gorono-
witsch made on the salmon, that the neural axis ends approxi-
mately between the optic stalks. The indentation in Gobius
can but be regarded as arudimentary neuropore.

The brain at this first stage 1s approximately in the same
condition as the salmon brain of the sixteenth day.

On the following day the brain of Gobius shows the same
characters, but with a slight general advance. The eyes are
becoming more rounded, and the optic stalk is being separated
from the eyes by a distal constriction. The lens is formed and
cut off from the ectoderm. Thus the development of the eyes
in Gobius proceeds more rapidly than in the salmon. No
choroidal fissure is as yet formed. The indentation which
marked the neuropore is less perceptible, though the exact
point where it occurred is still to bedetermined. The cavities
of the brain are well formed, and the vertical extent is less in
proportion to the breadth. The corpora striata, which in the
previous stage, as shown in sections (figs. 6, 7), were only part
of the dorso-lateral wall of the neural tube, are now well-
marked thickenings, pushing the optic stalks downwards, and
giving rise to a slight cranial flexure.

DEVELOPMENT OF THE HEAD IN GOBIUS OAPITO. 339

On the third day important changes have taken place.
There is a considerable cranial flexure, inasmuch as the optic
staiks are distinctly ventral. The eye has rotated correspond-
ingly, and a choroid fissure has been formed. The nose is
turned downwards and forwards. The pineal organ is a small
solid knob. The cavity of the fore-brain is formed, and, as
Goronowitsch states for the salmon, is a dorsal outgrowth from
the cavity. The mid-brain has become very broad, and
possesses a large cavity, which, however, is being invaded by
the tori longitudinales, which appear as thickenings of the
dorso-lateral wall. The cavity of the medulla oblongata
shows the well-known beaded constrictions. ‘The infundi-
bulum is large, and its cavity is narrow butdeep. As yet there
is no indication of lobi inferiores or of the formation of the
saccus vasculosus. Behind the infundibulum the isthmus and
base of the cerebellum are raised from the yolk, leaving a space
occupied by loose mesodermic tissue. The wall of the brain
from the pineal organ to the infundibulum—that is, the region
of the “‘vordere Endnaht ”’ of His—is epithelial in the median
line.

At this stage nerve-fibres are first differentiated at the peri-
phery of the brain.

Fig. 4 represents a model of the brain at the fifth day. The
model has been cut in halves at the middle line in order to
show the relation of the cavities. Viewed externally it shows
practically the same characters as are depicted by Goronowitsch
for the salmon embryo of the thirtieth day.

Posteriorly, to the leftin the figure, the medulla oblongata
is cut in section. The conditions of the first day have been com-
pletely reversed, in that the brain in this region is broader
than deep. The fourth ventricle is roofed in by thin membrane.
The roofing membrane passes into the cerebellum, which at
this stage passes near the middle line insensibly into the mid-
brain. The cavity of the mid-brain is now mainly filled by the
massive torus longitudinalis. In the figure the commissures
present at this stage are represented in red. Where the mid-
brain passes into the fore-brain the posterior commissure occurs,

340 H. B. POLLARD.

In front of it is situated the pineal organ, which has a slight
cavity. Below the pineal organ lies the cavity of the fore-brain ;
below the latter, the massive corpus striatum. The cavity
of the fore-brain is continuous with the recessus opticus. A
slight distance in front of the ventral limit of the recessus
opticus is seen the anterior commissure ; behind it the optic
chiasma, for the optic nerves are formed between the third and
sixth day, dating from the first day described. The infun-
dibulum is seen to be very large. Below it the hypophysis is
already formed and in situ. The cavity of the infundibulum
shows two smaller cavities proceeding slit-like outwards. The
anterior becomes later the cavity of the saccus vasculosus.
The posterior is the cavity of the lobus inferior. The axis
of the brain-cavity may clearly be seen to be bent like a shep-
herd’s crook, and to terminate in the region below and behind
the anterior commissure. Examining a transverse section in
the region of the mid-brain, a very clearly marked crucial lumen
is seen to divide the brain into four quadrants. The upper
laterals evidently correspond to the “ Fligelplatten” of His,
and the lower laterals to the “ Grundplatten.” The nucleus of
the oculo-motor nerve lies in the lower lateral quadrants, and
the upper forms the torus longitudinalis.

The characteristic crucial lumen may be traced con-
tinuously forward round the curve of the cranial flexure
in such a manner as to indicate that the brain axis ends at the
above-mentioned point. In a model of the cavities these
relations are especially clear. Thus in its morphological dis-
position the corpus striatum corresponds in the fore-brain with
the torus longitudinalis, while the wall of the infundibulum
corresponds with the region which gives rise to the oculomo-
torius in the mid-brain.

The wall of the infundibulum, however, does not give rise
to motor nerves.

The position of the neuropore and termination of the neural
axis in Vertebrates has been the subject of great discussion.
The various views have been well summed up by Kupffer in a

DEVELOPMENT OF THE HEAD IN GOBIUS CAPITO. 341

recent paper (14) on the development of the sturgeon. Von
Baer, Dursy, and His regarded the infundibulum as the
terminal point. Reichert, Kélliker, and Mihalkovics sought the
point in the region of the optic stalks. Goette concluded that
the pineal organ represented the neuropore. Van Wijhe found
that in birds the last point of connection of the ectoderm and
brain representing the neuropore lay at the middle of the sac
of the fore-brain.

Orr (17) found that inthe frog the “ anterior medullary
fold” representing the end of the floor of the neural canal lay
above the optic chisma in later stages. Kupffer himself
describes in the sturgeon a very definite structure, the lobus
olfactorius impar, at first hollow, and communicating with the
exterior, and in all respects resembling the neuropore of
Amphioxus; and from the agreement of his conclusions with
those of Van Wijhe he regards the question as settled for all
Vertebrates.

In Teleostei, however, as we have seen, there is no definite
structure corresponding to the lobus olfactorius impar. The
wall of the brain passes continuously over the spot, and only
comparatively late (fig. 4, z) can it be determined which
region is identical with that structure. In Teleostei the
neuropore and endof neural axis seem to have been
situated at the level of the optic stalks, and below the
subsequently formed anterior commissure. This con-
clusion is confirmed by the exact observations of Orr on the
frog.

In his latest writings (8 and 9) His speaks of the neuropore
as a “ Nabelartige Unterbrechung der vordere oder frontale
End-nath.” Further he says, “The total extent of the
neuroporic cleft stretches in all craniate Vertebrata from the
position of the basilar ridge through the region of the sub-
sequently formed recessus infundibuli, chiasma, and recessus
opticus, between the nasal organs and along the lamina
terminalis to its dorsal extremity.” Considering the several
conflicting and yet exact observations, this view of His seems
the only reasonable one to take.

342 H. B. POLLARD.

Mourn anv Hypornysis.

The mouth and hypophysis in Teleostei formed the subject
of Dohrn’s first two studies on the ancestral history of the Ver-
tebrate body. He stated that the mouth appeared first as a pair
of lateral openings of the alimentary canal to the exterior, and
that the hypophysis arose as a bilateral outgrowth of the upper
wall of the alimentary canal in front of the endodermal pouches
which form the mouth.

This account was criticised by Hoffmann (10), who described
the hypophysis as developing from a thickening of the lower
layer of the ectoderm in front of the mouth opening—that, in
fact, in Teleostei, just as in other Craniata, the pituitary body
is an organ derived from the epiblast of the stomodeum.
Subsequently Dohrn, in his fourth “study,” modified his first
views to the effect that there exists an ingrowth of ectoderm.

I find that Professor Dohrn’s figures are correct, except as
far as the interpretation of the endoderm is concerned. Exa-
mination of earlier stages gives the true explanation. With
Hoffmann’s figures and statements my own conclusions on
Gobius do not agree, though from individual pictures by
Henneguy (6) and Goronowitsch (4) the development in Salmo
appears to be the same as in Gobius.

Dohrn’s account was confirmed by Miss Platt (18) for
Batrachus tau. On the other hand, all doubleness of origin
of the mouth is denied by McIntosh and Prince (16) for pelagic
Teleostei. Nor, according to the same authors, is a stomodzeum
or involution of epiblast present.

Figs. 6—10 represent sections through the mouth and hypo-
physis regions of my first stage of Gobius capito embryos.
Fig. 2 gives a diagrammatic view of the general relations.

Examining first fig. 9, whose position is indicated on the
diagram by the line d—d, it is seen that the section is some-
what oblique, cutting the eye on the left near its posterior edge,
and on the right nearer the centre of the lens. In the middle the
brain is cut in section between the mid- and fore-brain, where
it is most compressed from side to side. Below the eye, on

DEVELOPMENT OF THE HEAD IN GOBIUS OAPITO. 343

the left, occurs a considerable ingrowth of ectoderm,
which represents the maximum ingrowth of the mouth at this
stage. At this spot it cannot be perceived exactly how far the
ectodermic ingrowth extends. In fig. 10, which represents a
section 25, further posteriorly (5 sections), the limits can be
sharply defined, and the ectodermic ingrowth is seen to be
connected only with the ectoderm of the skin by a neck, while
towards the middle line it touches another solid mass of cell,
representing a solid forward growth of endoderm.

In fig. 9 ectoderm and endoderm are fused. Where the cells
border on the yolk they form a more or less definite layer. In
all the sections anterior to that represented in fig. 10 this layer
can be perceived, and it separates brain and mesoderm from the |
yolk. Itis very distinct on the right hand in fig. 9. It can be
seen to pass insensibly into the external ectoderm, which forms
the lens of the eye. In the same figure the outer layer or
* Deckschicht ” of the ectoderm has been separated from the
lower layer. Under the centre of the eye no marked solid
ingrowth of the ectoderm occurs.

Fig. 8 is a drawing of a section further forward through the
posterior limit of the optic stalk. Its position is indicated on
the diagram by the line c. Here, again, is seen solid ectoderm
(Hyp.), occupying a portion of the angle between the eye and
brain. Further forward (fig. 7, on the left side) the ectoderm
is indistinguishable from the mesoderm surrounding the eye.
This mesoderm extends forward below the eye-stalks. At the
anterior end of the embryo the thickening of ectoderm (Hyp.)
is continuous with the ectoderm of the nose.

From these observations one may learn that the embryo at
its anterior end is separated from the periblast and yolk by
continuous ectoderm, one cell thick in the median region, and
that two paired thickened ingrowths occur, the anterior pro-
ceeding inwards from the region of the nose to below the optic
stalk, and the posterior proceeding directly towards the middle
line below and behind the eye. The anterior ingrowth, which
is to a certain extent continuous with the ectoderm of the nose,
gives rise, as we shall see later, chiefly to the hypophysis ; while

344 H. B. POLLARD.

the posterior thickened ingrowth, which is also continuous
with the ectoderm of the lens, gives rise to the portion of the
mouth lying in the region of the articulation of the jaws.

Considering now the endoderm at this stage, it is found
that the alimentary canal is only tubular in the auditory
region of the head. The lumen extends anteriorly to a short
distance behind the termination of the chorda dorsalis, then
passes outwards to open at the only gill-slit formed as yet.
This gill-slit is the hyoid, or the aperture behind the
subsequently formed operculum. By German authors it is
termed the ‘‘ Kiemendeckelspalte.”’” In front of the chorda
the outlines of the endoderm are almost impossible to
discover, but it appears to extend forward as a solid layer,
becoming in the median line very thin and unilaminar.
Laterally, however, in front a more distinct prolongation
meets the posterior ingrowth of ectoderm, and fuses with it,
as shown already in figs. 9 and 10. Curiously enough, when
this prolongation meets the ectoderm it also proceeds upward,
and fuses with the mesoderm surrounding the eye.

In fig. 2 these conditions are represented diagrammatically.
On the right of the figure only the brain and eyes and a
portion of the premandibular mesoderm are drawn. On the
left, ectoderm and endoderm are intact. The embryo is
supposed to be viewed from the ventral side, that is through
the yolk. The edge of the ectodern, where it passes over the
yolk, is cut in section. The blue tint shows its extent.
Towards the centre of the embryo, where the ectodermic layer
is one cell thick, it passes without demarcation into the
corresponding endodermal layer. The dotted blue line indi-
cates the extent of the above-described thickened ingrowth of
the ectoderm. This figure also shows how these ingrowths are
moulded to the shape of the brain and eyes. Consequently
the doubleness of origin must probably be regarded merely as
an embryonic feature without any phylogenetic significance.

The endoderm is tinted yellow. Posteriorly beneath the
chorda the lumen is indicated, and the dotted line proceeding
to the hyoid cleft (Hy.) shows the communication of this

DEVELOPMENT OF THE HEAD IN GOBIUS CAPITO. 345

lumen with the exterior. ‘The dotted yellow line (which
towards its anterior limit must be regarded as lying below the
blue line) represents the upward prolongation of endoderm,
which fuses with the mesoderm surrounding the eye (fig. 2,*).

Figs. 3 and 11 show the condition two days later. Fig. 11
represents a sagittal section passing through the eye, the lens
at its anterior border—the mid-brain and ear. The eyes and
brain are now separated from the yolk by various organs.
The mouth and hypophysis ingrowths are still solid, but at
this stage they are no longer two separate involutions. The
posterior portion is surrounded by condensing mesoderm, which
forms upper and lower jaws, and is separated from the yolk not
only by the forward-growing lower jaw, but also by the
pericardium and its cavity, which are in process of extension
forwards. From the eye the united ingrowths are separated by
the mesoderm of the upper jaw, and by the suborbital ganglion
of the lateral line. The ectoderm is specially thickened at the
angle where it passes over the yolk, beneath the choroid
fissure in fig. 11.

In the same section are seen the mandibular artery (Bi. v.)
behind the lower jaw, the rudimentary hyomandibular cleft
(1. M.), and the hyoid cleft (Hy.).

Following the same series of sagittal sections towards the
middle line, the ectoderm is found to be perfectly continuous
to the middle line. Beneath the infundibulum two layers of
cells may be distinguished, the upper consisting of cubical
cells closely packed together, and the lower of much flattened
cells, which form a flat epithelium next to the yolk. This
lower layer was the only one present in the first described
stage. The cells of the layer bordering on the infundibulum
can but be derived from the anterior thickened ingrowth of the
previous stage. They are the cells destined to form the
hypophysis.

Transverse sections put the conclusions derived from this
series of sagittal sections beyond doubt.

The ectoderm of the angle of the mouth has fused indis-
tinguishably with the endoderm, which also now gives rise to a

346 H. B. POLLARD.

rudimentary hyomandibular cleft (H. M., fig. 11). As in the
former stage, the lumen of the alimentary canal does not
extend further forward than to the level of the hyoid clefts.
The T-shaped lumen is constant for about four days of
development.

Fig. 3 is a diagrammatic representation of the conditions at
this stage. Ectoderm, endoderm, and mesoderm are tinted, as
in fig. 2, and the brain and other organs, which are represented
as being solid, divided in the middle line so as to show to some
extent the vertical relations. The view is, as in fig. 2, from
the ventral side.

The ingrowth of ectoderm is seen to start from below the
recessus opticus, choroid fissure, and posterior region of the
eye, and it is now throughout of more uniform thickness than
in the earlier stage. As stated above, the ectoderm is more
condensed below the infundibulum, and is thicker where it is
embraced by the maxillary and mandibular processes. In the
figure the maxillary process is indicated by the dotted line.
The yellow dotted line represents the rudimentary hyoman-
dibular gill-pouch. Ectoderm and endoderm pass into one
another without demarcation.

On the fourth day the hypophysis becomes more rounded,
though still more elongated than in its later stages, and at the
same time it is becoming cut off from the mouth. The cells
of the ectodermic ingrowth which forms the mouth are
arranging themselves in layers along the upper and lower jaws,
so as to give rise to a central cavity. An epithelium continuous
with the epithelium of the hollow alimentary canal becomes
differentiated from behind forwards. In subsequent stages
the hypophyis becomes separated from the mouth, and a mem-
brane is found to surround it. Thus the condition depicted
in the majority of Professor Dohrn’s figures is arrived at.

MESoDERM.

The origin, extent, and fate of mesodermic structures are by
no means easy to determine in Teleostei, on account of the
fusions of the three layers and the indefiniteness of the cells.

DEVELOPMENT OF THE HEAD IN GOBIUS CAPITO. 347

No head cavities seem ever to be present in Gobius, though in
Syngnathus a premandibular cavity occurs. Fig. 9 shows the
premandibular mesoderm fairly defined on the right side
(Pmd.mes.). In this section it passes continuously across the
middle line below the brain, to join the corresponding mass on
the opposite side. This is the bridge of mesoderm, so well
known from Selachian embryology, which lies behind the in-
fundibulum ; at its median side and in front the premandibular
mesoderm fuses with the mesoderm surrounding the eye. One
point of fusion is shown on the left side in Fig. 9. A similar
phenomenon is described by Kupffer (18) for Petromyzon. This
author considers the cells lying round about the eye to be
chains of nerve-cells.

Behind the eye, layers in the mesoderm cannot be dis-
tinguished (fig. 10) at this stage.

The mesoderm surrounding the eye and corresponding meso-
derm in other regions has been made the subject of special
investigation by Goronowitsch (5). As far as my researches
go they confirm his statements. ‘This mesoderm is derived, as
it seems, from the “ Ganglienleiste,” or “neural ridge” of
Marshall, whose réle, according to Goronowitsch, has been
entirely mistaken by many embryologists. At the same time
the fusions which occur between this mesoderm and the
premandibular mesoderm make it difficult, at least in Teleostei,
to decide whether ‘‘sclerotomic”? elements derived from the
premandibular mesoderm may or may not take part in the
formation of the skeletal structures which arise from this tissue
much later.

The position of the premandibular mesoderm is shown in the
diagram fig. 2, which also shows the extent of the ccelom,
which, as is well known, forms the pericardial cavity. The
celom actually borders on the mouth involution—that is, it
extends almost to below the eye. The somatopleure (So. pil.) is
thin, but the splanchnopleure forming the inner pericardial wall
is thickened. On the third following day the premandibular
mesoderm is taking its position round the eye so as to form the
eye muscles, which later are supplied by the oculomotorius.

348 H. B. POLLARD.

The bridge of mesoderm still persists behind the infundibulum
(fig. 8) in such a manner as to lead one to suppose that
formerly the eye muscles extended across the middle line,
possibly when no cartilaginous skeleton was as yet formed.

The pericardial cavities have extended to the middle line
and fused, and the heart is formed, though as yet solid (fig. 3).

The development of the pericardial walls in Teleostei seems
to me to have an important bearing on some theories of the
Vertebrate head. According to Van Wijhe (24), in Selachii
the cavities of the visceral arches, whose walls give rise to the
muscles of the gills and jaws, are continuous with the pericardial
cavity, but not (at least in the posterior arches) with the
cavities of the somites. Therefore, concludes Van Wijhe, the
muscles of the gills and jaws, which are voluntary, are homo-
logous with the muscles of the intestine, which are involuntary,
and not with the body muscles, which are voluntary. Van
Wijhe further makes this one of the cardinal points for the
interpretation of the nerves. In Teleostei the somites are, as
is well known, solid, and the cclom with its walls, the
somatopleure and splanchnopleure, is sharply defined. Now
the coelom in Teleostei is perfectly continuous far forward
(fig. 2), but its walls, the somatopleure and splanch-
nopleure, the “Seitenplatten” of Van Wijhe, do not
give rise to the gill and jaw muscles. Yet it does not
seem possible to deny the general homology of the gill and jaw
muscles in Selachii and Teleostei.

THE Jaws.

Fig. 11 shows the first condensation of mesoderm around
the mouth, and it is seen that one Anlage gives rise to
skeletal elements of both upper and lower jaws. The
condensation of mesoderm extends forward above the ectodermic
ingrowth of the mouth some little distance below the eye.
Below the mouth the lower jaw at this stage extends only a
very slight distance forward, though further towards the
middle line. In fig. 3 the structures are represented schema-
tically. The dotted line extending forward below the eye

DEVELOPMENT OF THE HEAD IN GOBIUS CAPITO. 349

shows the region of condensation of the mesoderm. In later
stages the tissue at the point of deepest ingrowth of the ecto-
derm of the mouth becomes less compact, so that then two
Anlagen may be distinguished. The upper grows forward, and,
with the differentiation of cartilage, gives rise to trabecule,
intertrabecule, and the tissue underlying the maxilla, as
described by Stéhr (22). The lower gives rise to the lower
jaw. I hope to make the subsequent modifications which give
rise to the adult conditions the subject of a special study. The
fact that both upper and lower jaw structures arise as one
paired Anlage at the angle of the mouth appears to have interest
in view of the theories as to the presence and number of gill-
bars having relation to the mouth.

SUMMARY.

1. The neural axis in Gobius terminates at a point near the
optic stalks, precisely as stated by Goronowitsch for the
salmon. An indentation and the disposition of the cells at
this spot indicate a rudimentary neuropore. In later embryos
a characteristic crucial lumen, which may be traced con-
tinuously from behind forwards, gives another indication of
the termination of the axis, and also shows that the corpora
striata belong to the upper part of the wall of the brain.

2. Mouth and hypophysis arise as solid ingrowths of ecto-
derm. That the earliest and maximum ingrowths are inde-
pendent and paired is probably an embryonic feature due to
the mode of development of the brain and eyes. Nothing has
been observed to indicate that the mouth was ever anything
but a mouth.

3. Skeletal structures of both upper and lower jaws arise
from one condensation of mesoderm round the ectodermic
ingrowth of the mouth.

4. The jaw muscles do not arise from the “ Seitenplatten ”’ —
a point of considerable theoretical importance.

NAPLEs,
March, 1893.

350 H. B. POLLARD.

15.

16.

17.

BIBLIOGRAPHY.

. BorEer.— Mesoderm in Teleosts,” ‘ Bull. Mus. Comp. Zool. Harvard,’

Xxiil, 2.

. Dourn.— Studien zur Urgeschichte des Wirbelthierk6rpers,” I, II, 1V,

‘ Mitth. a. d. Zool. Sta. zu Neapel,’ 1881, 1884.

. Frorrer.—“ Zur Entwickelungsgeschichte des Kopfes;” Anat., 2te

Hefte, Abth. Ergebn., 1891.

. GoronowitscH.—“ Das Gehirn u. die Giakistettn von Acipenser

ruthenus,” ‘Morph. Jahr.,’ 13, 1887-8.

. GoronowitscH.—“ Die axiale u. die laterale (A. Goette) Kopfmetamerie

der Vogelembryonen,” ‘ Anat. Anz.,’ vii, 1892.

. Hennrcuy.—“ Recherches sur le développement des Poissons osseux,”’

‘Journ. de l’Anat. et de la Phys.,’ 1888 (with literature).

. His.—* Die Formentwickelung des menschlichen Vorderhirns,” ‘ Abh. der

Math. Phys. Classe der Konig]. Sach. Ges. der Wiss.,’ xv, Leipzig, 1889.

. His.— Zur allgemeinen Morphologie des Gehirns,”’ ‘Arch. f. Anat. u.

Phys.,’ Anat. Abth., 1892.

. His.—“ Die Entwickelung der menschlichen u. thierischen Physiogno-

mien,” ‘Arch. f. Anat. u. Phys.,’ Anat. Abth., 1892.

. HorrmMann.—‘‘ Zur Ontogenie der Knochenfische,” ‘ Arch. f. mikr. Anat.,’

Xxiil.

. Hort.—‘‘ Development of the Teleostean Brain,” ‘ Zool. Jahr.,’ Morph.

Abth., iv, 1890.

. Kastscoenxo.—‘‘ Ueber das Beschneiden mikroskopischer Objecte,”

‘Zeitschr. f. Mikroskopie,’ v, 1888.

. Kuprrer.—* Die Entwickelung von Petromyzon planeri,” ‘ Arch. f.

mikr. Anat.,’ 35, 1890.

. Kuprrer.—‘ Studien z. Vergleichenden Entwickelungsgeschichte des

Kopfes der Kranioten, Acipenser sturio,’ Minchen, 1892 (with
literature).
Kuprrer.—‘ Mitth. z. Entwickelungsgeschichte des Kopfes bei Aci-
penser sturio,” ‘ Sitz. Ber. Ges. Morph. u. Phys.,’ Miinchen, 7, 1892.
M‘Intosu and Prince.—* On the Development and Life Histories of
the Teleostean Food and other Fishes,” ‘ Trans. Roy. Soc. Edinburgh,’
xxxv, 1890.

Orr.—‘‘ Note on the Development of Amphibians,’ ‘Quart. Journ.
Micr, Sci.,’ xxix, 1889.

DEVELOPMENT OF THE HEAD IN GOBIUS CAPITO. 351

18. Pratt, Jura B.—* Further Contribution to the Morphology of the
Vertebrate Head,” ‘ Anat. Anz.,’ vi, 1891.

19. Rasi-Rickuarp.— Zur Deutung u. Entwickelung des Gehirns der
Knochenfische,” ‘Arch f. Anat. u. Phys.,’? 1882, Anat. Abth.

20. Rasi-RiickHaRD.— Das Grosshirn der Knochenfische u. seine Anhangs-
gebilde,” ‘ Arch. f. Anat. u. Phys.,’ 1883.

21. Rast-RicKHarp.—*‘ Das Gehirn der Knochenfische,”’ ‘ Deutsche medi-
cinische Wochenschrift,’ 1884.

22, Stéur.— Zur Entwickelungsgeschichte des Kopfskeletes der Teleostier,”
‘ Zeitschr. der Med. Fac. Wiirzburg,’ ii.

23. Strasser.— Ueber die Methoden der plastischen Reconstruction,” VI,
‘Zeitschr. f. Mikroskopie,’ 4, 1887.

24, Van Wisue.—‘ Ueber die Mesodermsegmente u. die Entwickelung der
Nerven des Selachierkopfes,’ Amsterdam, 1882.

25. Witson, H. V.—“ The Embryology of the Sea-bass (Serranus atra-
rius),” ‘ Bull. of the U.S. Fish Commission,’ 1891.

EXPLANATION OF PLATES 21 AND 22.

Illustrating Mr. H. B. Pollard’s paper, ‘‘ Observations on the
Development of the Head in Gobius capito.”

Lettering of Figures.

Aud. Kar. Bl. v. Blood-vessel. C, str. Corpus striatum. (Cd. Cerebellum.
Ch. Chorda dorsalis. Ch. jiss. Choroid fissure. Ch. opt. Optic chiasma.
Comm. ant. Anterior commissure. Comm. post. Posterior commissure. ect.
Ectoderm. Hud. Endoderm. F%. Fore-brain. Gp. Cell débris (Glob. parabl.—
Henn.). H. M. Hyomandibular cleft. Hy. Hyoid cleft or bar. Hyp. Hypo-
physis. Inf. Infundibulum. JZ. if. Lobus inferior. MM. Mouth. WM. od.
Medulla oblongata. Md. Mid-brain. Md. Lower jaw. Mes. Mesoblast.
N. Nose. Np. Neuropore. Opt. Eye. P. Pineal organ. Pc. Pericardial
cavity. Pmd. mes. Premandibular mesoderm. &. opt. Recessus opticus.
S. orb. Suborbital ganglion (i. e. buceale). Sace. vasc. Saccus vasculosus.
So. pl. Somatopleure. Sp. pl. Splanchnopleure. 7. /. Torus longitudinalis.
4 Vent. 4th ventricle.

Fie. 1.—Model of brain at the first stage, viewed from the side. The nose
is partially represented.

VOL. 35, PART 3.—NEW SER. BB

352 H. B. POLLARD.

Fie. 2.—‘ Schematic” representation of brain—ectoderm (blue), mesoderm
(red), and endoderm (yellow)—viewed from below. First day.

Fic. 3.—“ Schematic” representation of brain. Third day. The struc-
tures are represented also as cut in halves in the middle line vertically.

Fic. 4.—Model of brain at fifth subsequent day. Left half viewed from
inside. Commissures represented in red.

Fic. 5.—Sagittal section of anterior end of embryo of first day. (x 450
approx.)

Fic. 6.—Transverse section through same region. Slightly oblique.
(Xx 450 approx.) ;

Fic. 7.—Transverse section through optic stalks. Slightly oblique. (x 450
approx.)

Fie. 8.—Transverse section behind optic stalks. Slightly oblique. (x 450
approx.) :

Fie. 9.—Transverse section through the eyes. Slightly oblique. (x 450
approx.)

Fic. 10.—Transverse section behind the eye. Slightly oblique. (x 450
approx.)

(Figs. 6—10 are from the same embryo.)

Fic. 11.—Sagittal section through eye and ear. Third day.

ON THE HEAD KIDNEY OF MYXINE. 353

On the Head Kidney of Myxine.
By

J. W. Kirkaldy,
Somerville Hall, Oxford.

With Plate 23.

In a former number of this Journal Professor Weldon has
given an account of the structure of the pronephros of
Bdellostoma (9), but, so far as I know, that of Myxine is
undescribed, excepting for some observations made by Professor
Wilhelm Miller (8). Professor Weldon has very kindly
furnished me with some specimens of Myxine glutinosa,
in order that I might examine the head kidney. I should
like here to express my thanks to him for this kindness,
and also for advice and assistance given me throughout the
investigation, which was conducted in his research laboratory
in University College, London.

The head kidneys of Myxine are situated towards the anterior
part of the body, a little behind the external gill-slit, and a
little in front of the fore-end of the mesonephron. Lach is an
elongate organ, showing on the surface a much convoluted
tuft of tubules (fig. 1, p. ¢.), the pronephric tubules of Miller,
and projects into the pericardial cavity just beneath the heart,
being also connected with the anterior extremity of the post-
cardinal vein. I have been unable to trace any connection
between the pronephros and the segmental duct.

Of the material upon which I have worked, some individuals
contained large ova, whilst others did not, and the pronephros
of an animal containing ova differs in many respects from that

354 J. W. KIRKALDY.

of one without. I will describe first the anatomy of an organ
from a Myxine in the latter condition.

Examined by means of transverse sections, the head kidney
is found to be in very intimate relation with the post-cardinal
vein (fig. 2). The greater part of it is actually lodged in the
vein, whilst the more superficial tubules are embedded in the
vascular wall, or lie free in the pericardial cavity. At the
posterior end a giomerulus is present (fig. 2, gl.), extending
along the inner side of the head kidney for about one fourth
of its length. Posteriorly it is enclosed in a sheath of its own,
but towards the front end this becomes indistinct, and the
glomerulus tissue is interwoven with that of the pronephros.
The organ seems, however, to be in a state of reduction, and
the vascular supply cannot be clearly made out from any
sections. The tubules, which are caught in section in every
direction, open into the pericardium on the one hand by means
of funnels (fig. 2, f.), and on the other are connected
ultimately with a large duct (fig. 2, c. d.), the central duct (9),
which, just as has been shown for Bdellostoma, sometimes
divides into two or three great anastomosing branches, and then
again becomes single.

In Myxine, however, as Miiller states, the central duct
gives off, on the side away from the tubules, outgrowths
containing glomeruli (fig. 4, ¢.). In sections towards the
hinder end of the organ I find that the glomerulus may, for a
short distance, quite fill up the central duct, so that no lumen
is visible, and then this reappears a few sections further on
(fig. 2, gl. d.). Blood-vessels pass across the vein to break up
inside the capsule into the characteristic capillary loop. The
efferent vessels return to the capillary network between the
central duct and the wall of the vein (fig. 2, ¢.n.).

The central duct does not reach to the anterior extremity of
the organ ; it stops short, and only a bunch of tubules projects
forward. The walls of the duct consist of tall columnar cells
of granular protoplasm, each containing a large oval nucleus of
coarser granulation (figs. 8 and 6). The cells have rather a
ragged appearance at their fore-edge; this is emphasised by an

ON THE HEAD KIDNEY OF MYXINE. 3505

aggregation of the protoplasmic granules, which seem to be
cast off from the cells, along with the mucus apparently
excreted by them. Externally a well-defined basement mem-
brane (9) gives support to these cells.

The walls of the tubules differ little from those of the main
duct. The columnar cells are not as tall, but they consist of
the same granular protoplasm, with the same large nuclei and
the same jagged look at the free edges. They also rest upon a
basement membrane, but they show distinct traces of ciliation
throughout the whole length of the tube.

The lumen, which always contains more or less mucus,
decreases somewhat in diameter towards the nephrostome, and
here the columnar cells are continued in to the flat cells of the
pericardial epithelium. In between the tubules are numerous
blood-capillaries. As in Bdellostoma, the peripheral tubules
are aggregated into lobuli, which are invested with pericardial
epithelium ; or, again, isolated tubes may be provided with a
separate investment (fig. 7). In these cases also the capillaries
are always present, lying beneath the epithelium and around
the tubes.

Sections through the pronephros of the other Myxine
present a very different appearance. The anterior tubules for
their whole length, and also the distal region of the others, are
unchanged, but as these more posterior tubes reach their inner
extremity their character is much altered. The lumen is
diminished in diameter, or even obliterated, and the tall
columnar cells are no longer recognisable: some have become
much attenuated—in fact, fibrous, with a dwindled nucleus ;
whilst others have a greatly swollen nucleus, surrounded with
a small layer of protoplasm. A section through this region of
a tubule is shown in fig. 8. Within the vein no definite
central duct or glomerulus can be made out. The position
occupied by these structures in specimens, as already described,
is here taken up by a mass of tissue, consisting of a reticulum
of protoplasm, whose fibres are nucleated, and in whose meshes
are small cells with small nuclei, and larger cells with very
large, round, and deeply staining nuclei (fig. 7). Blood-

356 J. W. KIRKALDY.

vessels enter the tissue and break up into fine capillaries,
which are caught in section in every direction. The blood-
capillaries surrounding the tubes are still distinct, and so is
the pericardial epithelium encircling the outermost parts; but
that connected with the inner regions, as also the vascular
wall, where it is much involved in the branching tubes, is
considerably altered. Where at all recognisable the pericardial
epithelium has lost its flat delicate appearance, and has
become thickened and heavy-looking : the wall of the vein
is also considerably thickened and much broken.

In sections of head kidneys of any stage I find a small
quantity of lymphatic tissue (4, 8) lying external to the wall
of the post-cardinal vein; but though in the later stages it
comes into close relation posteriorly with this ‘‘ central mass,”
I do not think the two are connected in origin. The “ central
mass ”’ would appear clearly to be derived from the breaking
down of the central duct of earlier stages.

Grosglik has already pointed out (4) that the recorded
observations on the pronephros of Myxine (8) and Bdel-
lostoma (9) were probably made on developing, and not on
mature specimens. A study of my sections confirms this
view, and also the opinion that the Myxinoid head kidney is
undergoing reduction.

In the younger stages the organ consists of numerous
tubules opening at one extremity into the pericardium by
means of funnels, and at the other into a main duct, in
connection with which are glomeruli.

As the animal reaches maturity a change takes place,
beginning at the hinder end of the organ and working for-
wards, affecting first the central part and the ends of the
tubules adjacent. The change consists in the formation of
a mass of tissue much resembling lymphatic tissue, which
would appear to result from the breaking down of the walls
of the ducts, the constituent cells furnishing the fibrous
groundwork of the new tissue, and also cells with large
nuclei, which I am inclined to regard as formative cells,
since in the most reduced organs I have examined they have

ON THE HEAD KIDNEY OF MYXINE. 357

disappeared, their place being taken by many smaller cells.
The whole tissue is richly supplied with blood-vessels pro-
vided by the vascular supply to the glomeruli and the
capillary network into which this breaks up, and answers
with great exactness to the description of the mass of
lymphatic tissue in the anterior region of the Ganoid ex-
cretory system given by Balfour (1).

It would seem, therefore, that the head kidney of Myxinoids
may be regarded as a stage in the phylogenetic reduction of
this organ—a reduction which continues in the Pisces until
the tubular structure entirely disappears.

As regards the relations of this organ with the supra-renal
bodies, the absence of nerve-structures would seem at first
sight to exclude the possibility of any connection between
the two, which, indeed, is the view advanced by Emery (8).
A consideration of the subject, however, shows that this is
not an insuperable objection.

In his description of the supra-renals of Elasmobranchs,
Balfour! states that these organs are derived partly from the
mesoblast and partly from the sympathetic ganglia, and that
in fact the two constituents remain distinct in this group
throughout the life of the animal; whilst Mitsukuri has
demonstrated the compound origin of these organs in the
Mammalia (6).

It may, therefore, be concluded that the pronephros in
Myxine represents the mesoblastic part of the supra-renal
bodies, which have been shown by Professor Weldon (10)
to be derived from the anterior part of the mesonephron in
the higher Vertebrata.

1 *Comparative Embryology,’ vol. ii, p. 664.

3858 J. W. KIRKALDY.

BIBLIOGRAPHY.

. Batrour.—“‘ The Head Kidney in Adult Teleosteans and Ganoids,”

‘Quart. Journ. Micr. Sci.,’ 1882.

. Batrour, and Parker, W. N.—‘‘ The Structure and Development of

Lepidosteus osseus,” ‘ Philosophical Transactions,’ 1882.

. Emery.—‘ Studi intorno allo sviluppo ed alla morfologia del rene dei

Teleostei,’ ‘Mem. Acad. Lincei,’ vol. xili, p. 43 (translated into
French in ‘ Arch. Ital. de Biolog.,’ t. ii, p. 185).

. GroseLix.—‘ Zur Morphologie der Kopfniere der Fische,” ‘ Zoologischer

Anzeiger,’ 1885.

. JANOVZIK.— Bemerkungen iiber die Entwickelung der Nebenniere,”

‘ Archiv fiir mikroscopische Anatomie,’ 18838.

. Mitsuxur1.— Supra-renal Bodies in Mammalia,” ‘ Quart. Journ. Mier.

Sci.,’ 1882.

7. Mitr, J.—‘ Vergleichende Anatomie d. Myxinoiden.’

8. Miter, W.—“ Urogenitalsystem des Amphioxus und der Cyclostomen,”

10.

‘ Jenaische Zeitschrift,’ 1878.

. Wetpon.—“On the Head Kidney of Bdellostoma,” ‘ Quart. Journ.

Mier. Sci.,’ 1884.
Wetpon.—* Supra-renal Bodies of Vertebrata,” ibid., 1885.

ON THE HEAD KIDNEY OF MYXINE. 359

EXPLANATION OF PLATE 23,

Illustrating Miss J. W. Kirkaldy’s paper, “On the Head
Kidney of Myxine.”

6. c. Blood-corpuscles. 4. m. Basement membrane. 4. v. Blood-vessel.
c. Capsule. c.d. Central duct. c.. Capillary network. ¢. w. Capillary
wall. /#. Funnel. g/. Glomerulus outside the vascular wall. g/. d. Glome-
rulus of the central duct. 7. Lumen of tubule. m. Mucus. m.s. Anterior
end of mesonephron. x. Nucleus. p.c. Pericardial epithelium. p. ¢. Pro-
nephric tubule. s. Cilia. .d. Wall of central duct. «. v. Wall of vein.

Fic. 1.—Head kidney of Myxine glutinosa.
Fic. 2.—Diagrammatic longitudinal section.

Fic, 3.—Transverse section through the head kidney to show the relations
of the central duct and the pronephric tubules.

Fig. 4.—Transverse section, showing glomerulus.
Fic. 5.—Glomerulus enlarged.

Fie. 6.—Longitudinal section through a funnel, showing the columnar
animal, to show the central mass.

Fic. 7.—Transverse section through the head kidney of a more mature
epithelium passing into the pericardial epithelium.

Fic. 8.—Tubule from the same organ, with modified epithelial cells.

ac apy ad Meni SF icy GL hei ‘Guhl wale a i> Vee ihe ot =e e
Mt 2 ° j au yo. an ane Tr aie wLE | Rh \igt

‘ne exe A

icglal nar TL each ecient > ry elit un

i ayy _ ae ary "W walle 7 aoe:

Ne i AF” Nabe mee
obtain! «4 a got jardingll ag A Sone.
{iit Pa 26 oh wif iagt PERO eye 7 tse

an

Uy aa qlailgidte avenge y WNab) Ae BT iay
t

iL a > seh alas Udo Fa a ee

la

en
‘

i d at i) ae mr Lid Tene

7 ' aj ; 7 oh ee its! (eee
Or Vee i iit P nl Al abe Wi 7 hd spay dj filet) abi ‘of
; 2 Pala \ 1) GAP sS) pn =) Ane ote i
E : “™ - aie 4
| 5 s:nted OG Reena
Te)! a j i i Hart ree, +62)
‘ "i ob us “t
ay Ne wlan 2 = ; Mahi vee Mey
F . Ri rte, ai i ee Seile. | Pie 2 fe
j i i DAE TGA BER Oat: ae ea. a ae
s.
.

REPORT ON A COLLECTION OF AMPHIOXUS. 361

Report on a Collection of Amphioxus made
by Professor A, C. Haddon in Torres Straits,
1888-9,

By

Arthur Willey, B.Sc.Lond.

Tus collection, comprising some ninety specimens of Am-
phioxus, was obtained by Professor Haddon between the months
of September, 1888, and January, 1889, inclusive, and was
placed by him in the hands of Professor Lankester, who kindly
entrusted it to me for examination.

The specimens were in a good state of preservation, and a
careful study of them has revealed several facts of interest.

They were taken from different localities (Flinder’s Entrance,
Mabinag, &c.), at depths varying from six to thirty fathoms,
the sea bottom being here, according to Professor Haddon’s
records, largely composed of fine broken shells.

They all belong to one species of Branchiostoma,! namely,
B. cultellum = Epigonichthys cultellus, Peters.

This species can be distinguished at a glance by the great
development of the dorsal fin, which presents a striking appear-
ance on account of the unusual height to which it projects
above the body. The height of the fin gradually increases
from behind forwards, reaching its climax in front in the
region of the seventh and eighth myotomes. The distinctive

1 As it is not only out of the question, but undesirable, to relinquish the
name “Amphioxus,” although ‘“Branchiostoma” has the priority, the
latter, following the precedent set by Giinther, is retained as the systematic

name of the genus, while the former may be regarded rather as a colloquial
name.

362 ARTHUR WILLEY.

characteristic of the dorsal fin is accurately portrayed in the
figure of the species given by Peters ; but he erred seriously in
asserting thé absence of a caudal fin, and in describing the
anus as occupying a median ventral position.

Giinther! perceived this error on the part of Peters suffi-
ciently to justify him in releasing the species from the cum-
brous genus to which the latter author had assigned it.

Asa matter of fact, the caudal fin and anal opening have
the same relations here as in the European species, only the
former structure is much thinner and more delicate in B.
cultellum, and can be practically annihilated by unfavorable
reagents.

32 44 $3
| !
i :

BKK LM =

KKK CK Ql ie =

| ch.
E j

=a

, I
™m. Zm.

Reena

Explanation—Branchiostoma cultellum from left side. ch. Anterior
and posterior extremities of notochord. e. Eye. m. Mouth. at. Atrio-
pore. az. Anus. /. m. Left metapleur.

Continuity of right metapleur with mesial ridge of ventral fin is indi-
cated, as is also the lateral position of anus and the well-marked
caudal fin.

In many of Professor Haddon’s specimens, particularly in
the sublimate and osmic acid preparations, the caudal fin,
which is perfectly hyaline and of extreme tenuity, is very
finely preserved, and in all cases the anus lies on the left side.

The number of myotomes in B, cultellum is about fifty-

1 Gunther (8, p. 33), criticising the account given by Peters, says, “The
position of the vent is, at least in some of our specimens, rather lateral than
median. Whether these differences are owing to the better state of preserva-
tion of our specimens, or related to the difference of locality, I am not pre-
pared to decide.” Professor Haddon’s collection decides this point in favour
of the former explanation of the differences in question.

REPORT ON A COLLECTION OF AMPHIOXUS. 363

two, but it appears to vary within narrow limits, namely, from
fifty-one to fifty-five. Giinther formulates them with reference
to the position of atriopore and anus as follows :
oz + 10 + 10 = B2,
or 31 + 11 + 10 = B2.

As the position of these apertures stands in no causal
relation to any particular myotomes, it is difficult and perhaps
impossible to give an unvarying statement with reference to
this point.

In one or two instances I have counted as follows :—
32 + 12 + 8 = 52, the anus being placed at the base of the
forty-fourth myotome (cf. accompanying figure). In another
case, in which I counted fifty-five myotomes, the formula was
838 +114 11=55. Still, in several other instances I have
counted as many as fifty-four myotomes.

The average length of the specimens in Professor Haddon’s
collection may be placed at 2°5 cm., although several indi-
viduals measured upwards of 3 cm., two of them attaining a
length of nearly 3°5 cm. Peters stated the length of his
specimens, which came from Moreton Bay, to be 2°3 cm.
Beyond a few measurements Peters gives no numerical data
whatever, and no account of the internal organs.

There are from twenty-four to twenty-seven ventral fii-
chambers between atriopore and anus. External examination
from the ventral aspect gives rise to the impression that the
ventral fin in this species contains paired fin-rays, such as are
known to occur in B. lanceolatum. This optical effect is
due to the fact that the caudal fin is continued forwards as a
mesial ridge below the ventral fin-chambers, and so produces a
double appearance in the latter in a surface view of the ventral
aspect. As in B. lanceolatum, the fin-chambers are always
single median spaces; but while in the former each of them
coutains a pair of gelatinous fin-rays suspended from its dorsal
wall,in B.cultellum the ventral fin-chambers are desti-
tute of fin-rays.

The absence of ventral fin-rays has recently been also
observed by E. A, Andrews in a new species of Amphioxus

364 ARTHUR WILLEY.

from the Bahamas; but in this case the author adds the curious
fact that the ventral fin space is also absent. In this respect,
therefore, the new species differs essentially from B. cultel-
lum, in which separate fin-chambers are emphatically present
in the ventral fin.

In B.lanceolatum the right and left metapleural folds
gradually decrease in size behind the atriopore, and, while
converging together, finally die out on either side of the mid-
ventral line below the anterior portion of the ventral fin. In
B. cultellnm this is not the case. Here the left metapleural
fold gradually dies away behind the atriopore, whilst the right
fold does not die out, but is continued behind the atriopore into
the mesial ridge, which lies below the ventral fin-chambers,
and represents, as mentioned above, the forward continuation
of the caudal fin.

This behaviour of the right metapleur has also been pre-
viously signalised by Andrews in the Bahama species.

In the latter, again, according to Andrews, the buccal cirri
are “smooth and united by the hood membrane for the greater
part of their length.” In B. cultellum, on the other hand,
the cirri are distinctly free throughout their entire length,
except, of course, at their bases, which are embedded in the
margin of the oral hood ; at the same time they appear to be
without those papilliform prominences which are a well-
known feature in the European species.

I have been unable to detect an olfactory pit in B. cul-
tellum, but will not undertake to say that it is always absent.
It appears, however, according to Andrews, to be absent in the
Bahama species.!

1 That this is a distinct species is shown once for all (in the absence at
present of the detailed illustrated description which I am informed by
Professor Andrews will shortly appear in vol. v of the ‘Studies from the
Biol. Lab, Johns Hopkins University’) by the formula of the myotomes as
given by Andrews, viz. 44 + 9 + 13 = 66. Length 13—16 mm.

[Since the above was in type Andrews’ memoir has appeared in the
‘Bulletins of the Biol. Lab. of the Johns Hopkins University,’ vol. v,
No. 4, under the title ‘An Undescribed Acraniate: Asymmetron luca-
yanum,.”—EpIToR].

REPORT ON A COLLEOTION OF AMPHIOXUS. 365

One of the most remarkable features in the internal organisa-
tion of B. cultellum is the fact that the gonads occur as a
unilateral series of pouches confined to the right
side of the body.

Singular to say, it agrees in this respect also with the
Bahama species, and it should be clearly pointed out that the
first discovery of this curious form of asymmetry in Amphioxus
is due to Professor Andrews. The fact of its occurrence also
in B. cultellum, which is a distinct species with a very
different geographical distribution, is of some interest.

In view of what we know as to the asymmetry of the larva
of Amphioxus, I cannot, however, agree with Andrews in re-
garding the above species as being generically distinct from
the European species, although it might perhaps be legitimate
from a systematic point of view to create a new sub-genus for it.

I have paid particular attention to this asymmetry of the
gonads in B. cultellum, and find that it occurs invariably. A
great number of specimens in the collection were mature, and
they were all examined with great care. In some cases so
large were the gonads, extending from one side to the other
across the middle line, that a casual glance would lead to the
impression that they were paired in the usual way, but a dis-
section or a careful examination with a lens always revealed
their unilateral disposition.

All observations were controlled by transverse sections, and
in those which passed through the greatly distended gonads
the pharynx was found to be pressed tightly against the left
wall of the atrium, almost the entire available space being
usurped by the gonads.

In view of the fact of the unilateral disposition of the gonads,
both in B. cultellum and in the species from the Bahamas,
it is important to note that often in specimens of Amphioxus
from the Mediterranean the gonadic pouches of the right side
ean be observed to preponderate greatly over those of the left
side, as if the hypertrophy of the former led to the reduction
or incipient atrophy of the latter. I am not aware that this
observation has ever before been recorded.

366 ARTHUR WILLEY.

In B. cultellum, as well as in the Bahama species, the
predominance of the right gonads over the left has been accen-
tuated to such a degree that the latter have been entirely lost.
Whether or not rudiments of the left gonads appear at any
time in the development remains to be decided.

Such instances of unilateral asymmetry as that described
above are always of interest, since they are obviously ceno-
genetic deviations from the normal, which can in a measure be
satisfactorily accounted for.

Since, according to Boveri’s observations, the development
of the reproductive organs of Amphioxus takes place after the
metamorphosis of the larva, it is evident that the occasional
partial asymmetry of the gonads (in respect of size) which I
have noted above in the Mediterranean species, and the com-
plete asymmetry of the gonads in B. cultellum, &c., must
belong to a different order of phenomena from the remarkable
asymmetry of the pharynx in the larva of Amphioxus. More-
over, Boveri has shown that the perigonadial ccelom is a deri-
vative of the myocele, and the myotomes are not involved in
the larval asymmetry.

It would seem, in fact, that the absence of the antimeres of
the right gonadic pouches in the Bahama species and in the
Australian species is due to considerations of economy of
growth and accommodation to a limited space. Provision
being made in the economy of the organism for a certain
bulk of gonads, the onus of this can either be shared equally
by the two sides, or a greater proportion can be assigned to
one side, or finally the entire mass can be confined to one
side. It is difficult to give a priori a reason why it should
always be the same side which is affected by this unequal
growth.

The above would seem to be the correct mechanical explana-
tion of the asymmetry in question, the hypertrophy of the
gonads of one side necessarily leading to the final atrophy of
those of the other side. Whether it is correlated physiologi-
cally with any greater locomotor activity on the part of the
species in which it occurs must remain an open question,

REPORT ON A COLLECTION OF AMPHIOXUS. 367

According to Andrews the Bahama species “ swims free in the
evening, both at Bimini and in Nassau Harbour.”

There is thus every reason to suppose that the above-
described unilateral asymmetry of the gonads in certain species
of Amphioxus belongs, broadly speaking, to the same category
as, for example, the well-known asymmetry of the female
genital glands and the lungs of snakes, in which the respective
organs of the right side usually predominate over their anti-
meres on the left side, the latter being often rudimentary ; and,
again, the female reproductive organs of birds, in which the
left ovary and oviduct predominate over the right, the latter
being either absent or rudimentary.

There are seventeen to twenty unpaired gonadic pouches in
B. cultellum. When there are as many as twenty the
first one lies at the base of the 9-10th myotome. In one
specimen, taken from a bottle labelled “ Mabinag, Oct. 24th,
1888,” there were quantities of free ova in the atrial chamber
derived from the discharge of the anterior eight or nine gonadic
pouches, while the nine posterior pouches still remained intact.
Another specimen, in which the hypertrophy of the unilateral
gonads was carried to an extraordinary pitch, was taken on |
Dec. 24th, 1888.

This, therefore, may be taken to indicate the time of spawn-
ing of B. cultellum, which is somewhat later than is the case
with the Mediterranean species. In fact, it would appear as
though the spawning of B. cultellum commences at about
the time of the year at which that of B. lanceolatum
ceases.

I am not aware of any observations on the habits of B. cul-
tellum, but the special elaboration of the dorsal fin would
seem to point to the fact of its being, like the Bahama species,
an active swimmer.

As for the significance to be attached to the continuity of
the right metapleur with the mesial ridge of the ventral fin, it
seems to show that the metapleural folds or ridges are, after all,
structures of the same nature as the median longitudinal
ridges which constitute the fins of Amphioxus. If the meta-

VOL. 30, PART 3.—NEW SER. co

368 ARTHUR WILLEY.

pleura had been entirely different in their nature from the
median fins it is not very likely that one of them would have
undergone concrescence with the ventral fin. As shown by
Lankester and Willey, the metapleural folds arise as solid
longitudinal thickenings of the integument, which are at first
largely ectodermic in origin (the ectoderm-cells assuming a
columnar character), while the cutis subsequently takes part in
their formation. Eventually a lumen (schizoccele) appears in
the ridges.

The right metapleuron is in advance of the left in order of
appearance, and in front of the pharyngeal region of the larva
it curves sharply inwards towards the middle line, in which it
gradually dies away on the ventral surface of the snout.

There is, in fact, no essential difference between the mode
of origin of the metapleura and of the median fins as integu-
mentary ridges, and it is possible that in the above-mentioned
cases, in which the right metapleuron is continuous with the
ventral fin (z. e. the mesial ridge in connection with it), they
actually arose in continuity in the first instance.

If, then, it is necessary to admit the intrinsic similarity in
the nature of the metapleural folds and the median fins of Am-
phioxus, we are led back to the theory of Thacher with
reference to the origin of the paired limbs of Vertebrates.

Balfour, as is well known, was the first to discover the con-
tinuous lateral fin-ridges of the Selachian embryo in 1876;
and it is curious to note, in the light of what has just been
said as to the ectodermic thickenings which prelude the
formation of the metapleura in Amphioxus, that they also (7. e.
the Selachian fin-ridges) consist in longitudinal thickenings of
the epiblast. From his observations on the embryonic de-
velopment of the Selachians, Balfour came to the conclusion
“that the limbs are the remnants of continuous lateral fins ;”
but he did not suppose that the continuous lateral fins were
represented in Amphioxus.

At about the same time, and quite independently, Thacher
was led by observations on the adult forms and on the skeleton
of Selachians and Ganoids, &c., to a belief in the homodynamy

REPORT ON A COLLECTION OF AMPHIOXUS. 369

of median and paired fins. As to their phylogeny, he said,
“ As the dorsal and anal fins were specialisations of the median
folds of Amphioxus, so the paired fins were specialisations of
the two lateral folds which are supplementary to the median
in completing the circuit of the body. These lateral folds,
then, are the homologues of the Wolffian ridges in embryos of
higher forms.”

Shortly afterwards Mivart also came independently to the
conclusion that the paired and azygos appendages of Verte-
brates were fundamentally of the same nature. Subsequent
paleontological researches have only confirmed this view.

The point, however, which is at issue on the present occa-
sion is whether or not the primitive continuous lateral fins are
represented by the metapleural folds of Amphioxus. From
what has been said above there seems to be good reason to
expect that this question will sooner or later be answered by
the consensus of morphologists in the affirmative.

Reference should be made here to an interesting feature in
the geographical distribution of Branchiostoma, which does
not appear to have received sufficient attention.

The areas of distribution of the species of Branchiostoma
are, as a rule, separated from one another by such wide inter-
vals of space that it is extremely surprising to find an instance
of the overlapping of two specific areas. Such an in-
stance apparently occurs in the Torres Straits.

According to Dr. Giinther specimens of Branchiostoma
Belcheri were obtained by Dr. Coppinger from the sea around
Prince of Wales Island, Torres Straits, while during the same
voyage (H.M.S. “ Alert”) B. cultellum was obtained from
the neighbouring Thursday Island. B. Belcheri, Gray, which
is characterised chiefly by the presence of sixty-four or sixty-five
myotomes, and is more elongated than B. lanceolatum, was
first obtained by Sir E. Belcher, during the cruise of H.M.S.
“Samarang,” from the coast of Borneo.

Professor Haddon does not appear to have obtained any
specimens of Amphioxus from Prince of Wales Island ; but the
fact that his large collection, taken from several different

370 ARTHUR WILLEY.

stations in Torres Straits, does not contain any other species
than B. cultellum is noteworthy, and it would be a matter
of interest to determine the exact relations to one another,
and limitations in the distribution of the two species of
Branchiostoma which have been recorded from the Torres
Straits.

A great deal of work has still to be done in connection with
the geographical distribution of Amphioxus, and this not only
with regard to its distribution on the face of the earth, as to
which we can hardly hope for a speedy settlement of the ques-
tion, but even as to its distribution in more restricted pro-
vinces, as, for example, the coast of Europe. Although there
is only one species, there are undoubtedly several varieties of
European Amphioxus which differ from one another in point
of size. Thus the Messina Amphioxi average larger than
those from the Gulf of Naples, while both of these varieties
would appear to come far short of that found on the coast of
Brittany, which is said to attain a length of no less than
8 cm.

An extensive series of measurements of European Amphi-
oxus from different localities would be certain to yield
important results

It is worthy of note that differences in size among indi-
viduals or species of Amphioxus are not causally related to the
number of myotomes. A single instance will suffice to illus-
trate this point. The Bahama species, according to Andrews,
has sixty-six myotomes, with a length of only 13 to 16 mm.
B. lanceolatum has sixty myotomes (sometimes fifty-nine
and sometimes sixty-one), with a length of 4 to 6 and even
8 cm.

The relation of size to physical or organic environment is a
subject for investigation.

In conclusion it may be said that Professor Haddon’s collec-
tion, which I have had the privilege of examining, has enabled
the specific characters of Branchiostoma cultellum to be
definitely ascertained, and has brought to light several interest-
ing features in its organisation.

REPORT ON A COLLECTION OF AMPHIOXUS. 371

REFERENCES.

. Anprews, EH, A.A—“ The Bahama Amphioxus” [preliminary description ],
‘Johns Hopkins University Circulars,’ vol. xii, June, 1893, p. 104.

. Batrour, F. M.—‘A Monograph on the Development of Elasmobranch
Fishes,’ London, 1878 (portion relating to paired limbs previously
published in ‘ Journal of Anat. and Phys.,’ vol. x, 1876).

. Boveri, Tu.—‘ Uber die Bildungsstatte der Geschlechtsdriisen und die
Entstehung der Genitalkammern beim Amphioxus,” ‘ Anat. Anz.,’ vii,
1892, pp. 170—181.

. Gtntuer, A.—“ Branchiostoma” (synopsis of genus; specimens of B.
cultellum, collected by Dr. Coppinger in the neighbourhood of
Thursday Island), in the ‘ Report on Zoologicat Collections of H.M.S.
* Alert,”’ pp. 31—33, London, 1884.

. Lanxester, E. R., and Witiny, A.—‘‘The Development of the Atrial

Chamber of Amphioxus,” ‘Quart. Journ. Micr. Sci.,’ vol. xxxi, N. S.,
1890, pp. 445—466.

. Mivart, St. G.—‘“ Notes on the Fins of Elasmobranchs, with Considera-
tions on the Nature and Homologies of Vertebrate Limbs,” ‘ Proc.
Zool. Soc.,’ 1878, pp. 116—120. (Reference taken from ‘ Zoological
Record,’ vol. xv.)

. Peters, W.—“ Epigonichthys cultellus, eine neue Gattung und Art
der Leptocardii” (specimens collected by Dr. Studer in Moreton Bay
during voyage of S.M.S. “ Gazelle”), ‘ Monatsberichte der k. Preuss.
Akad. der Wiss.,’ Berlin, 1876, pp. 822—327, with plate.

. Tuacner, J. K.—‘‘ Median and Paired Fins, a Contribution to the

History of Vertebrate Limbs,” ‘ Trans. Connecticut Academy,’ vol.
iii, No. 7, pp. 281—310, February, 1877.

ae

- . - “ , - : - . - . 5 . e - ~
aise F 7 Pgs Wt a
F as

+The
« ee a j it

bls

p ast 5 ial im fal
iy é : : : > a Ss AA , ¥ Pits. 4
s : he =
y iA Rese Pew tre
: ~ . , 2 oe
7 y = tavP ‘ ety) Sy i” 1?
7 ree r : -
* tik a tert ae

THE ORIENTATION OF THE FROG’S EGG. 373

The Orientation of the Frog’s Egg.
By

T. H. Morgan, Ph.D.,
Associate Professor of Biology, Bryn Mawr College ;

AND

Ume Tsuda,
Teacher in the Peeress’ School, Tokio, Japan.

With Plates 24 and 25.

if

Tue classical experiments of Pfliiger on the segmenting
frog’s egg, and the important conclusions drawn by Roux
from a study of the same egg, have made it very desirable to
have an accurate knowledge of the relation existing between
the early segmenting egg and the position of the embryo with
respect to the egg.

The interpretation of certain embryos in which the blasto-
pore has failed to close, recorded by Roux and Hertwig, will
likewise depend on the normal position of the embryo on the
egg. Pfluger, Roux, and Hertwig have come to the conclusion
that the embryo forms over that portion of the unsegmented
egg which is normally directed downwards, i. e. over the white
hemisphere. Schultze supports the old view, that the embryo
lies on the upper or black hemisphere.

Pfliiger based his conclusion on the evidence obtained by
actually following the dorsal lip of the blastopore in its migra-
tion over the white hemisphere. Roux based his conclusion on
evidence obtained by destroying definite portions, both of the
segmented and unsegmented eggs.

374 T, H. MORGAN AND UME TSUDA.

Hertwig’s conclusions were based on the evidence furnished
by certain abnormalities, while Schultze’s conclusion rests on
a study of normal development.

It seemed to me at first, from a study of eggs developing
normally, that it was impossible that the embryo should lie
entirely over the white hemisphere. Schultze pointed out that
Roux’s earlier results are contradictory in themselves, and I
had reached the same conclusion from a careful reading and
re-reading of Roux’s earlier papers. I was prepared, there-
fore, to find some truth in each view, and expected to find the
embryo forming partly over the black, partly over the white
hemisphere. I was then not a little surprised to find that our
studies led to the conclusion that the embryo is formed over
part of the white hemisphere of the egg. In the main point,
therefore, I am in agreement with Pfliiger and Roux, although
not entirely so, for I hope to be able to show the extent
of the white hemisphere of the unsegmented egg, covered by
the blastopore, to be somewhat different from that affirmed by
Pfliger and Roux.

Our work in relation to the orientation of the embryo has
covered the ground somewhat more extensively than that of
any previous author, since we have made use of the methods
employed by all of them.

Our results will be considered under three headings:

lst. Normal development and location of blastopore.

2nd. Results obtained by injury to definite portions of
the early embryo.

8rd. Results obtained from embryos whose development
had been modified by artificial means.

A word of personal explanation ought to be added. The
senior author is responsible for Sections 1, 111, 1v, and v of the
present paper. The work recorded in these was done in the
spring of 1893.

Section 11 is the record of the results obtained by Umé
Tsuda while a student in the Biological Laboratory of Bryn
Mawr College. This work was done during the winter of
1891-2; the account written in the spring of 1892. Only

THE ORLENTATION OF THE FROG’S EGG. 375

very slight alterations have been made in this portion prepa-
ratory to publication.

IT.

In studying a series of eggs of the early stages of Rana
temporaria from the segmentation period to the beginning
of the formation of the blastopore, a few points in regard to
the peculiar development of the pigment and the orientation
of the dorsal lip of the blastopore have been noted, and are
here given briefly.

The eggs, which had been previously hardened and preserved
in 80 per cent. alcohol, were studied chiefly by surface views
with a dissecting microscope. The study of the segmentation
of the early stages only verified previous accounts. The first
cleavage furrow divided the egg into two equal parts; the
second is at right angles to it; the third or horizontal furrow
is much nearer the upper or pigmented pole, thus forming four
small pigmented cells in the upper and four large cells in the
lower hemisphere.! The four cells of the upper half then
each divide, thus forming eight cells; but the division from
this point becomes quite irregular, both in the upper and lower
halves of the egg. 1 found a number of eggs in what clearly
seemed a twenty-four cell stage—eight cells in the lower and
sixteen in the upper; but I could not verify the fact in the
living egg.

A curiously abnormal egg of eight cells was found where the
horizontal or third furrow was entirely wanting, and the fur-
rows of the next division, which would normally have divided
the egg into sixteen cells, had cut through from the upper
pole, reaching down about two thirds of the distance to the
lower pole. On sectioning the egg eight nuclei were found
corresponding to the number of segmentation furrows. No

1 T have found a number of probably abnormal eggs from one lot in which
the first furrow divided the egg into unequal parts, one large and the other
much smaller. Also a number of eggs of the four-cell stage, where neither
the first nor second furrow had met at the lower pole.

376 T, H. MORGAN AND UME TSUDA.

nuclei were found in the yolk portion of the lower half of the
egg, which would normally have been separated from the upper
cells by the third furrow.

In addition to the division of the cells of the segmenting
egg from the surface a delamination of the cells begins about
the thirty-second cell stage. The horizontal and vertical sec-
tions at this stage show elongation of the nuclei at right
angles to the plane of division. In the sixty-fourth cell stage
the delamination can be easily seen to have taken place by
the dissection of an egg under a hand-lens or a dissecting
microscope.

A careful study of the segmentation of the cells around the
lower pole in the advanced segmentation stages has shown
that the greatest irregularity exists.

In many cases, however—and I have reason to believe in
nearly all cases,—the cells lying nearest the lower pole, and
especially the four cells which are around the point where the
first and second furrows intersect, remain larger than the sur-
rounding cells. In the later stages of many eggs I have dis-
tinctly made out four cells, which I think are without doubt the
ones grouped around the lower pole. Figs. vii, v111, 1x, where
cells marked 4 and a are much smaller than the surrounding
ones, might seem to oppose such a conclusion ; but in the later
stages, at the time of the formation of the blastopore (figs. x, x1,
x11), there is a certain regularity in the grouping of the larger
size cells around the point which, from other indications,
I should judge to be the lower pole; and hence I believe that
such a cell as cell a, fig. vi11, though smaller at this particular
stage than the cells surrounding it, does not develop so rapidly
later on. I have not been able to section the eggs at these
stages to find out what relation the real size of the cells bears
to the apparent size from surface views. At present I see
nothing against the hypothesis that the portion of the egg
around the lower pole in the late as well as in the early seg-
mentation stages is the most retarded portion of the developing
egg. The group of cells that remain largest always bear a
certain relation in position to the pigment that leads me to

THE ORIENTATION OF THE FROG’S EGG. 877

believe that this is undoubtedly so, and as yet I have seen no
indication that would tend to a contrary conclusion.

It has been noted that with the splitting off of cells from the
upper corner of the yellow cells of the lower hemisphere new
ectoderm-cells are formed, and it had been generally supposed
that with this growth a continuous formation of pigment took
place, the black. pigment gradually growing down over the
whole egg. I have found that the growth of pigment does not
in any way correspond to the growth of new ectoderm-cells,
but, on the contrary, there seems to be a great variation in the
amount of pigment found in various lots of eggs of the same
stage procured at different times. Some eggs of the two- or
four-cell stage have the pigment covering more than two thirds
of the egg, while others at this point of development have only
a black cap of pigment extending down to the third furrow.
However, all the eggs of the same stage from one cluster, and
hence laid by the same frog, are alike in the quantity of pig-
ment. The amount of pigment in the egg seems a variation
dependent on conditions previous to the beginning of segmenta-
tion, and due to individual difference in the adult frog. There
is, of course, some formation of new pigment in the later seg-
mentation stages, and a most marked and rapid change in the
amount of pigment formed at the time of the first appearance
of the blastopore.

In all the eggs, from the earliest stages up to the blastopore,
there is one marked peculiarity of the pigment. There is
always a greater deposit of pigment on one side of the egg than
on the other; and if we judge the exact position of the lower
pole from the crossing point of the first two segmentation
furrows, we find that the pigment is not only denser, but it
comes down much nearer to the lower pole on one side than on
the other. As this was found to be the case in greater or less
degree, and in some eggs very markedly, in all the stages up to
the beginning of the blastopore, and, moreover, in eggs procured
at several different times and places, I judged it could not be
an accidental variation.

I have tried, therefore, to find out—

378 ', H. MORGAN AND-UME TSUDA.

I. Whether the pigment bore any fixed relation to the first
furrow, and by this to follow out approximately the first furrow
in the later stages.

II. The relation of the dorsal lip of the blastopore to the
pigment and to the first furrow.

I have examined a large number of eggs in the early stages,
in order to ascertain the exact appearance of the pigment, and
to compare it with the later stages. When the egg is looked
at on the under and non-pigmented side, with the lower pole
turned uppermost, the line of pigment, extending, as it does,
nearer to the lower pole on one side than on the other, has a
crescentic outline. The pigment zone or band is not usually
visible on the opposite side (see fig. v1). The same crescent-
shaped appearance of the pigment is easily followed in the later
stages up to the formation of the blastopore, at which time there
is a rapid growth of pigment downward towards the lower pole.
In order to ascertain what relation the first furrow bears to
this crescent-shaped band of pigment, I examined 119 pre-
served eggs, as well as a number of living ones, in the stage
when the second furrow had begun to come in from the upper
pole and was about to meet the first furrow on the lower side.
I made my observation on the eggs just before the two sides of
the second furrow had met and intersected the first furrow at
the lower pole, but when they were near enough to meeting, so
that the point of the lower pole could be approximately judged.
In this way I was able to distinguish the first furrow from the
second, which would not have been possible after the second
cleavage was completed, and at the same time I could guess
approximately the position of the lower pole, the meeting-
point of the first and second furrows.

The first furrow does not seem to cut the pigment zone bi-
laterally ; nor does it, on the other hand, always divide the egg
into a lighter and a darker half. Out of the 119 eggs
examined, in 76 cases the first furrow cut through a little
to one side of the central point of the crescent, only
approximately dividing the pigment.

Figs. a, B, and c show diagrammatic views of the lower

THE ORIENTATION OF THE FROG’S EGG. 379

pole of different eggs. The line m x represents the first furrow,
and a, 6, c, d, the four cells formed by the first and second
segmentation furrows. In Fig. a the dotted line x y passes
through the centre of the crescent-shaped pigmented area, and
divides the egg symmetrically. The first furrow lies to the
right of it. |

[Ones AN ies ah

In 80 cases out of 119 the second furrow seemed to divide
the pigment more equally (see Fig. 8B), and the second furrow
is a little to the left of the imaginary line of symmetry 2 y.

TGA:

In the remaining 13 cases the pigment seemed to extend
as much on one side as on the other, and the line 2 y

380 T. H. MORGAN AND UME TSUDA.

in this case would be as near to the first furrow as to the
second (Fig. c). It will be seen on examining the above dia-
grams that the apparent variation of pigment in the three
cases depends on very slight differences. A little shifting of
the pigment or an addition to one side or the other would
change Fig. a to Fig. c ors. Judging by numbers, a would
seem to be the more typical one.

One thing remains obviously unchanged in all the eggs.
Of the four cells into which the egg is divided, one cell, a, has
always the greatest amount of pigment; and cell 4, which is
opposite to it on the other side of the egg, is the lightest.
This is true in every case, and these two opposite sides can be
distinguished in eggs far advanced up to the end of seg-
mentation. The cells marked ¢ and d are intermediate, being
neither so dark as @ nor so light as 8.

The pigment on the darker side of the egg not only extends
much farther down, but I have observed on dissecting the egg
in the upper hemisphere that in some cases the pigment
extends inward almost to the centre of the egg (blastula), and
to about twice the depth of the opposite side.

The interesting point in connection with the two opposite,
the darker and the lighter, sides of the egg, on which I have
dwelt at such length, is that the less densely pigmented half
of the egg very early in the segmentation shows signs of a more
rapid development and growth than the darker and pigmented
side. This is true of the cells of the upper hemisphere as well
as the lower. Moreover it is on the side that shows this
advance in development that the dorsal lip of the blasto-
pore makes its appearance.

I first noticed the unequal growth of the cells in an egg of
about ninety-six cells, sixty-four in the upper and thirty-two
in the lower half. There was a decided retardation of one side
of the egg. In later stages the lighter half seems often two
stages ahead of the other side. Figs. 1v, v, show the unequal
development in the early stages. Figs. 1v, v (camera draw-
ings), are surface views of an exceptionally fine egg, in which
the unequal development is seen to what might appear an ex-

THE ORIENTATION OF THE FROG’S EGG. 381

aggerated degree. Very few eggs at this early stage in the
development show such a marked difference of the two halves.
In some eggs of this stage the segmentation seems equally
advanced on both sides, but these are rather the exception than
the rule. I have never found a single case in any stage
where the pigmented side was in any way in advance of the
lighter side. On the contrary, the reverse is true in almost
every egg towards the close of segmentation, and in most
cases a superficial glance will reveal the fact plainly.

Sections of the egg parallel to the third furrow show the
cells smaller over one hemisphere than over the opposite, and
prove at least that there is an unequal development of two sides
of the egg, and that the difference which exists between them
is no superficial one.

A careful examination of the early blastopore stages of the
egg with reference to the pigment and to the unequal develop-
ment shows conclusively that the blastopore makes its
first appearance on the less pigmented and further
developed side of the egg, and, moreover, at a short
distance only from the group of large cells around
the lower pole.

I have examined over a hundred eggs at this stage, and my
best observations have been made on eggs in which the de-
velopment of the yolk-cells, as compared with the rest of the
egg, was retarded, so that the outlines and size of the cells, as
well as the unequal development of the two sides, could be
plainly seen by surface views with a dissecting microscope. In
some eggs it was very easy to follow out the outline of the
yolk-cells around the lower pole after the formation of the
blastopore, though the development was often too far ad-
vanced to make this out satisfactorily. But wherever I could
follow out the cells it was plain that the region around the
blastopore was in advance of the opposite side. In most
cases it is difficult also to orient the egg as regards the pig-
ment after the appearance of the blastopore, though I had a
number of specimens where this could be done. ‘Towards the
close of the segmentation period pigment rapidly forms over

382 T, H. MORGAN AND UME TSUDA.

the area where the blastopore is about to appear, so that a line
of dark pigment is distinctly seen in sharp contrast to the
lighter cells lying next to it. The change is so rapid that it
is often impossible to orient the sides of the egg. In some
cases, however, when the blastopore has only just appeared,
and before the pigment increases to any extent, it is easy to
see that the blastopore is forming on the previously lighter
side of the egg, as well as on the side which is segmenting
most rapidly. In spite of the dark pigment formed just above
the blastopore there is often a distinct light area on one side
ofthe are. This area probably corresponds with the cell marked
b, the least pigmented cell, which lies opposite to the centre
of the pigment crescent and opposite to the dark cell, a, in the
diagrammatic figures A, B, C.

Figs. x—x1iI are views of favorable specimens, and show
distinctly a cluster of large cells, presumably those around the
lower pole. Fig. x is before the appearance of the blasto-
pore. The crescent-shaped area of pigment is distinctly seen,
the pigment coming much nearer the group of large cells on
one side than on the other. It is on this pigmented side that
the cells are largest. In the centre are a large cell and three
smalier ones, which probably are the four cells nearest the
lower pole. The unequal segmentation is also shown in
fig. x1, where the blastopore has already formed on the side
where the cell division is more advanced. In fig. x11 there are
four cells distinctly larger than the surrounding ones, between
which probably run the first and second furrows. It will be
noted how much nearer the pigment approaches these cells on
the side marked m than on the opposite side, where the blasto-
pore appears. To the right and left of the blastopore the
pigment is less dense than on the opposite side, though it is
rapidly forming just above it. If it is granted that the four
cells are around the lower pole, and that this is the point
where the first and second furrows intersect, the exact rela-
tion of the blastopore to the lower pole can be easily ascer-
tained. Fig. x1 is a side view of the same egg, in which the
position of the supposed lower pole is shown. It is very near

THE ORIENTATION OF THE FROGS EGG. 383

the line of pigment on one side, as we should expect, while the
blastopore on the opposite side is less than one third of the
distance from the lower to the upper pole.!

It has been almost conclusively proved by previous ex-
periments and observation that the plane of the first furrow
in the case of the frog divides the egg into halves correspond-
ing to the right and left sides of the embryo ; and this study of
the blastopore does not contradict, but would tend to confirm
the fact. Although the are of the blastopore is often not
opposite the centre of the crescent of pigment (as it is in
fig. x11), this is easily accounted for by the distribution of the
pigment as shown in Figs. a, B, c (text). If we suppose the
second furrow rather than the first to cut through the centre
of the crescent (Fig. 8B), we should have the pigment much as in
fig. x1, allowing for the formation of some new pigment.

Ill.

The eggs of two species of frogs were used for the ex-
periments recorded below. Eggs of Rana temporaria were
found on the morning of March 25th. These had not as yet
segmented. The eggs of another species (not determined)
were brought to the laboratory on April 4th. These had just
begun tosegment. Since much of the experimental work was
done on these eggs, it was first necessary to find out whether
the facts recorded in the last section were also applicable
here.

A study of representative stages showed the same distribu-

1 It has been thought by some investigators that the blastopore formed
much higher up in the egg, but it needs only a superficial study to show that
this at least isimpossible. The cells within the blastoporic ring are non-pig-
mented and yolk-cells. A study of the surface view of the early stages shows
that the pigment from an early stage often extends down on one side two
thirds of the side, and on the other one half of that side: The blastopore
forms lower down than the pigmented area, and this would make it at least
halfway down the egg from the upper pole, and much below the plane of the
third furrow. We see in fig. x11 that if the lower pole, as I have attempted
to show, is marked by the large cells the blastopore appears below even the
equator of the egg.

VOL. 385, PART 3.—NEW SER, DD

384 T. H. MORGAN AND UME TSUDA.

tion of pigment as found in the eggs of Rana temporaria, but
the eccentricity in its distribution in respect to the axis of
cleavage was greater than in eggs of Ranatemporaria. The
egg looked at from above (with one pole of the axis turned
directly upwards) showed on one side a distinct white crescent,
as seen in fig. 1.

The most interesting fact is that in the thirty-two-celled
stage a very decided irregularity of the segmentation spheres
of the upper portion of the egg is to be found.

This is readily seen in the three figures of the same egg
drawn in figs. 1—111._ The first of these (fig. 1) shows the egg
from above; eight cells lie along a line (four on each side) that
corresponds presumably to the first cleavage plane. These upper
eight cells are all approximately the same in size in this egg.
The eight cells forming the zone around the egg below the
upper eight, and which are sister cells with the latter, show a
difference in size on opposite sides of the egg, as shown in
figs. 11,111. The difference may be seen from above, but still
better by a study of the opposite (lateral) sides of the egg.

The lighter side of the egg is shown in fig. 111, in which the
border line of black pigment extends only for a short distance
over the side of the egg. The dark side of the egg is drawn in
fig. 11, and here the pigment extends much further into the
lower hemisphere. On the light side of the egg (fig. 111) the
cells of the second and third zones are smaller than the cor-
responding cells on the opposite side of the egg (fig. 11). Un-
fortunately the four-, eight-, and sixteen-celled stages of these
eggs were not preserved, so that I am unable to say how far
back this difference in the two sides may be present.

This led to a re-examination of the eggsof R. temporaria.
Here I found that at the eight-celled stage in most eggs one
of the four upper cells is somewhat smaller than its upper
vis-a-vis. It was also found that this smaller cell is the cell
nearest to the highest point reached by the white crescent ;
therefore it must have come from that cell (now) of the lower
pole that contains the least pigment. At the sixteen-celled
stage those cells on the side of the egg nearest the upper limit

THE ORIENTATION OF THE FROG’S EGG. 385

of the white are also smaller than those opposite to them.
Sections of the hardened eggs, made with a scalpel through
the plane of these smaller cells and their opposites, showed that
these cells are ndt only smaller superficially, but in the third
dimension as well.

Undoubtedly, then, from the eight-celled stage onwards the
distribution of larger and smaller cells on the dark and light
sides of the egg is present, and I have been able to push back
a step farther the differences noted for the later stages in the
preceding section.

Whether or not astill more careful examination of very
favorable material would find the same difference present in
the four-celled stage I am unable to say, but it seems not
improbable that such a difference exists.

A study of the method of gastrulation of the egg of the
unknown species shows that the first traces of the blastopore
appear on the light side of the egg within the white cells.
Presumably the pigment has here also extended farther over
the sides of the egg than at first. The outlines of the cells in
the region of the blastopore are at first polygonal. Dark
pigment appears in the walls of the cells, producing the dark
line seen in surface view. Certain of the cells pull in from
the surface, leaving only their outer small pigmented ends
exposed. These cells subsequently pull in all together to
form the beginning of the archenteron by invagination.
The cells dorsal to the blastopore become narrow and elongated
from above downwards. The light cells, below the point of
invagination, retain their polygonal outline.

The changes that take place in the overgrowth of the dorsal
lip of the blastopore will be recorded below. First let us
examine the embryo when first outlined on the egg. Some-
times the outlines of the medullary folds may appear before
the yolk-plug has entirely disappeared—at other times not
until after this change has taken place. Careful measure-
ments of the embryo at this time show that the embryo ante-
rior to blastopore covers in length about one third of the
periphery of theegg. The relative length of the embryo

386 T. H. MORGAN AND UME TSUDA.

to the egg is shown in Pl. 25, fig. 5. In some cases the
embryo measures a little more than one third of the periphery,
in others a little less. Very quickly after the appearance of
the medullary folds the embryo increases in length, and the
proportions of the egg change, so that in order to determine
accurately the length of the medullary folds as compared with
the egg they must be measured when their outlines are just
indicated by darker pigment.

The suckers appear in front of the medullary folds, and
arise at about the same time: a dark line of pigment marks
their position. This cresceutic line of pigment—the beginning
of the suckers—is not quite halfway around the egg from the
blastopore, i. e. it is nearer to the dorsal lip of the blastopore
than to the ventral.

Experimental Investigation.—Loss of time and ma-
terial was caused at first by attempts to do experiments that
proved to be impossible; also many results were valueless,
because the eggs experimented upon were not watched con-
tinuously. I cannot too strongly emphasise this point, that
unless each egg is carefully followed, from the moment of
injury to the time of preservation, the results become uncertain
and of little value. I have seen an injured point completely
heal over, and the extra-ovate of Roux plough a long furrow
over the surface of the developing egg; consequently any
conclusion drawn from the end result without a knowledge
of the intermediate stages would lead to error, and I cannot
but think that some of Roux’s earlier experiments that seem
to be so contradictory may have been caused by some such
changes.

Futile attempts were made to remove as much as half the
yolk and protoplasm from the fertilised egg. Such eggs col-
lapsed completely. Equally unsuccessful were attempts to
add the yolk removed, by means of a hypodermic syringe, to
another egg.

At the two-celled stage, just as the four-celled stage was
beginning, attempts were made to suck out with a syringe all
of the protoplasm from one hemisphere, in order to determine

THK ORIENTATION OF THE FROG’S EGG. 387

whether the remaining hemisphere would develop a perfect
. half-sized embryo or half an embryo. Many eggs went to
pieces, both during and subsequent to the operation. Others
partially rounded up and continued to develop, but the
greater number of these died later. The few embryos that
formed the medullary folds were very imperfect, but, as each
egg stuck had not been carefully followed during the stages of
segmentation and gastrulation, I hold these results to be
valueless. They show, however, I believe, the possibility of
carrying out the experiment successfully.

In several eggs at the eight-celled stage one of the black cells
was killed by pricking, so that its contents ran out. Such
eggs developed, and in the blastular stage defects were found
in the black hemisphere. In other eggs one of the white cells
was stuck, and, later, defects were found either in the white
hemisphere or just within the edge of the dark area. In these
cases no record was kept of the position of the particular cell
removed; hence the results are of little or no value, and I think
the same statement will apply to the similar experiments of
Roux. Now that it seems to be possible to recognise, even in
the eight-celled stage, the relationship between particular
blastomeres and definite portions of the later embryo, more
successful results ought to be obtained.

The consistency of the yolk in the eggs of the two species is
different. That of R. temporaria is more fluid, and the egg
collapsed more easily than in the other case. Owing to this
difference the eggs of the unknown species were far more
favorable for experiment, and the following results were made
on these eggs.

In order to determine the extent of overgrowth of the lower
pole by the blastopore a large number of experiments were
made by slightly sticking the white cells below the blastopore.
By using a very fine and sharp needle an exceedingly small
injury could be made, so that only a few small cells protruded
from the surface of the egg. These, however, gave a definite
landmark for orientation, The determination of the extent
of overgrowth by injury to the lower cells has a great advan-

388 T., H. MORGAN AND UME TSUDA.

tage, it seems to me, as compared with the more common
method of injuring the upper or black cells.

Owing to the great thickness of the lower wall of yolk-bearing
cells there is no chance of breaking into the segmentation
cavity or archenteron. Asthe white cells seem to be the more
passive cells during development, injury to them has less
serious consequences for the developing embryo.

The eggs were stuck at the time when the blastopore first
appeared, and a sketch made in each case to indicate the dis-
tance of the point of injury from the blastopore. A series of
these eggs were prepared in which the injuries were at varying
distances from the dorsal lip of the blastopore that had just
appeared. A series of figures were drawn from time to time
to show the relations between blastopore and point of injury.
Moreover duplicates of each lot were followed. Inasmuch as
all the experiments gave similar results, I think any doubt as
to abnormality caused by the operation is removed.

If the egg (embryo) be turned with its white area uppermost
at the time when the blastopore first forms, so that the blasto-
pore just appears above the horizon, it will be found that the
white area does not cover quite a hemisphere of the egg. A
border of dark pigment appears around the periphery of the
white, as shown in outline by Pl. 25, fig. 9. The primary pole
of this white area (hemisphere) lies not quite in the centre of
the white, but nearer to the side where the blastopore has
appeared, as shown in figs. 1, 3, and 4. The “centre” of the
white area does not, therefore, correspond with the “lower
pole.”’

The experiments here recorded were made on Rana, sp. ?

The x shows the point where the egg was stuck.

Experiment I (figs. 1O—12).—Egg in which the _blasto-
pore had just appeared. Pricked at 4 p.m. on opposite side of
white area, i. e. nearly a hemisphere away from blastopore.
At 8 p.m. the blastopore has become more arched, and the dis-
tance between the point injured and the dorsal lip of the
blastopore is much less than at first. The dotted line running
out from the ends of the blastopore marks the rather sharp

THE ORIENTATION OF THE FROGS EGG. 389

line of separation of the black and white, and also indicates
the subsequent line of invagination of the remainder of the
blastopore. It will now be seen (fig. 11) that the point of in-
jury lies just outside of the pigmented line. At 12 midnight
the blastopore had grown much smaller (fig. 12), and the point
of injury was outside of the blastoporic rim. The point of in-
jury is now at less than half its former distance from the
dorsal lip of the blastopore.

Experiment II (figs. 18, 14).—Egg at blastula stage, had
been kept overnight on ice to retard rate of development. At
9 a.m. the blastopore had appeared. Egg stuck in white at
a point not quite so far from the blastopore as in the last case.
At 4.30 p.m. the outlines of the whole blastopore to be seen,
but the point of injury, as before, is still outside of the blasto-
poric rim, and is nearer to the dorsal lip of the blastopore than
at first.

Experiment III (figs. 15, 16).—In this egg the blasto-
pore appeared at first as a vertical pigmented line (fig. 15),
which soon extended laterally into the usual crescent. The
point of injury was nearly the diameter of the egg from the
blastopore. At 4.80 p.m. (fig. 16) the crescentic blastopore
was much nearer to the point of injury. Later stages not
followed.

Experiment IV (figs. 17—19).—Stuck at 8 a.m. at
far edge of white, fig. 17. Blastoporic crescent already
formed. At 4.30 p.m., fig. 18, dorsal lip of blastopore
much nearer to defect. At 8 p.m. circular outline of blasto-
pore present, and defect lies just within the edge of the
blastopore.

Experiment V (figs. 20, 21).—Stuck at 4 p.m. quite near
to the blastopore. Blastopore had already formed a crescent.
At 8 p.m. the dorsal lip of blastopore had nearly reached the
defect.

Experiment VI (figs. 22, 23).—In this experiment the
lower pole was not stuck until the circular outline of the blas-
topore was formed (fig. 22). The dark line of the crescent
marks the dorsal lip of blastopore. The dotted line marks

390 T., H. MORGAN AND UME TSUDA.

the boundary line, between black and white, ready to invagi-
nate. The injury was made nearly in the centre of the blasto-
poric plug, somewhat nearer to the dorsal lip. In a later
stage, when the yolk-plug is smaller, the defect still lies near
the centre of the yolk-plug (fig. 23). It seems relatively a
little nearer to the dorsal lip than at first. The blastopore,
therefore, must close in after its circular outline is formed at
a nearly equal rate from all points. In this case the injury
was so small that the overgrowth of the blastopore could not
have been in the least retarded.

The experiments recorded above are taken from a series of
twenty-one recorded cases, and will serve as types for the rest.
All the results point unmistakably to the conclusion that there
is an overgrowth of the lower white cells by the lips of the
blastopore. Moreover the experiments show the extent of
overgrowth of the blastopore and the relative amount of
overgrowth of the dorsal and ventral lips respectively.

Examining the results more in detail, we find, if we
assume the point of injury to bea fixed point, that the dorsal
lip of the blastopore moves over the white to the extent illus-
trated in fig. 24. This figure is made from data of fig. 10, &c.,
keeping the point of injury in the same position. The dia-
meter of the circle (representing the outline of the egg) equals
27 mm. The distance between the blastopore and the injury
equals 24 mm. From 4 p.m. to 8 p.m. the dorsal blastoporic
lip has moved through 8 mm., and is therefore now 17 mm.
from the defect. At 12 p.m. the distance travelled through
since 8 p.m. is 7 mm. The dorsal lip is now 10 mm. from the
defect. So far the blastopore has passed through 15 mm. of
the 24. As the defect lies outside of the point of closure of
the blastopore by 2 mm., the blastopore now measures 8 mm.
Assuming that from this time onwards the blastopore grows
together at an equal rate towards its centre, the dorsal lip will
pass over about 4 mm. more of the white. In this time the
dorsal lip has moved through 20 mm. of the white area. The
ventral lip has passed through 4 mm.

The region in front of the blastopore covered by the over-

THE ORIENTATION OF THE FROG’S EGG. 391

growth (20 mm.) is less than the diameter of the circle (27
mm.). Comparing this with the length of the medullary folds
when they first appear, the area overgrown is found to be some-
what lessin length. If we deduct from the length of the embryo
the thickness of the medullary folds at their anterior border,
we find that the length of the two regions corresponds almost
exactly. In other words, the connection around the anterior
end of the medullary folds lies just in front of the point where
the blastopore first formed, and the area overgrown by the
dorsal lip equals the length of the medullary folds between the
anterior connection and the blastopore.

A few corrections should be made; the measurements just
given apply only to the flat surface, while the embryo lies over
a spherical surface. As the measurements of the overgrowth
and the measurements of the embryo are both projections into
the same plane, no gross error will come into the calculation.
The rate of overgrowth is not quite the same in all the ob-
servations, but approximately so. Even the extent of over-
growth is variable, and we have seen that the length of the
embryo formed is also variable.

The first overgrowth of the dorsal lip of the blastopore is
more rapid than the later growth ; that is, the approach to the
point of injury is faster at first. After the blastopore has
completed its circular outline the process of overgrowth (or
withdrawal) of the yolk-plug is much slower.

I have assumed the point of injury to be the fixed point,
and the approach of the blastopore to be due to the movement
of the latter. We might have assumed that the overgrowth
was due to a forward movement of the whole of the white area
passing beneath the blastoporic lip. The end result would be
the same in either case, the process different. It is not an
easy question to decide, but to any one following the process
in the living egg it will be clear that the change is due to the
movement of the blastopore lips, and not to the white area.
The condition of the cells in the white area points to a relative
stability and inertness, while the reverse is true for the dorsal
lip of the blastopore. The method of invagination of the

392 T. H. MORGAN AND UME TSUDA.

anterior, lateral, and posterior edge of the blastopore points to
the same conclusion.

I believe, however, that the details of the actual process of
concrescence of the blastopore has not as yet been accu-
rately worked out. The migration of cells that takes place
during the process has not been determined. Whether or not
the dorsal lip rolls in as it grows over, or whether its ex-
posed edge always carries the same cells, has not been shown.
Both experimental and structural evidence must be brought to
bear on the problem before its solution will be possible.

A series of ten experiments were made by sticking the
embryo (when the blastopore first appeared) at the apex of the
black pole. Other experiments involved sticking at the apex
of the black and white in the same egg. The latter experi-
ment ought to settle the question as to what portion was the
active agent in the overgrowth. Unfortunately the experi-
ments did not give satisfactory results, nor were the results
uniform. Injury to the delicate roof of the segmentation
cavity may have helped to produce poor results. Failure to
find in the later stage the point injured, shifting of the extra-
ovate if large, and the difficulty of determining the exact
apex, may all have had a hand in the matter. Only two such
embryos are drawn, although other as definite records were
also obtained.

Experiment VII (fig. 5).—Egg when blastopore had just
appeared was stuck at apex of black pole. When the medul-
lary folds appeared the injury was found on the ventral surface
of the body, as shown by the x in the figure. The defect was
at about equal distances from the anterior end of the medullary
folds and from the blastopore.

Experiment VIII (fig. 6).—Injured as in last. Defect
appeared at point 180° from blastopore, therefore some distance
in front of the anterior end of the medullary folds.

Both of these results show that the embryo does not form
over the black pole, but why in these cases the defects are
at such different distances from the blastopore I do not
know.

THE ORIENTATION OF THE FROG’S EGG. 393

REVIEW oF LITERATURE.

There are certain statements made in the papers of Pfliger,
Roux, Schultze, and Robinson and Assheton that bear directly
on the results given above. Pfliiger records that in one set of
eggs the blastopore first appeared at 6 a.m. At 1] a.m. the
blastopore was broader, with the corners turned down. The
blastopore had left the equator of the egg and approached to
the lower pole. At 12.30 p.m. the blastopore was semicircular,
and had pushed further towards the lower pole. At 1 p.m. it
was circular, and now it lay at the opposite point of the white
hemisphere from which it had started.

An examination of the relation of the pigment shows that
the egg as a whole has had no part in this overgrowth of the
lower pole, i.e. no rotation of the whole egg has taken place.

At 2.15 p.m. the yolk-plug was smaller, and the blastopore
has continued to move in the same direction. At 4.15 the
blastopore is narrower still, and its diameter equals about one
eighth the diameter of the egg; it has moved even further,
and is in the region of the equator of the egg, but at the
opposite point of the equator from which it started in the
morning.

These observations point conclusively to the view that ‘the
opening of Rusconi passes from a point on the equator lying
in the meridian of the egg over to the opposite point of the
equator through the lower white hemisphere, and the egg-axis
during the period has not changed its position.”

The arc travelled is not quite 180°, but is certainly more
than a right angle,—variable, however, in different eggs. The
overgrowth is due to a process of invagination.

From 4.15 p.m. till 7.45 p.m. the egg as a whole rotates in
the opposite direction along the same meridian. Due to this
true rotation more than one half of the (new) upper hemisphere
is covered by those cells that overgrew the blastopore, and
which therefore have a lighter colour than the cells of the
primary upper hemisphere. From this clearer portion in front

394 T. H. MORGAN AND UME TSUDA.

of the anus of Rusconi develops the anlage of the central
nervous system.

Pfliiger adds, “In order to avoid a misunderstanding I
must say that I do not by any means believe that the whole
anlage of the central nervous system is a derivative of the
white hemisphere. Since the lighter substance of the white
hemisphere is directly continuous with the lighter substance
of the black, it is possible that the anterior portion of the
medullary plate corresponding to the brain, and even to the
upper portion of the neck, may form in the black hemi-
sphere.”

There are two statements only in the foregoing account
from which I should dissent. In the first place it seems rea-
sonably certain that the blastopore does not originate in the
equator of the egg, but at some distance below it. In the
second place Pfliiger believes that the blastopore, as it en-
croaches on the yolk-plug, moves as a whole further along the
meridian of migration. This means that after the ventral in-
vagination of the blastoporic rim has formed, the ventral lip
still moves upwards towards its nearest equatorial point.
This migration of the whole blastopore is stated in the text,
and is definitely shown in the series of diagrams drawn to
illustrate the process of overgrowth.

I have attempted to show that the overgrowth of the dorsal
lip itself is sufficient to account for the length of the medullary
folds; also that the posterior lip of the blastopore, after its
formation by invagination, closes by a forward growth. There
is, therefore, no evidence for such a migration as Pfliiger sup-
poses, and if the circular blastopore after its formation does
move further upwards it must be due to a slight rotation of
the egg as a whole in this direction. But the statement that
it does continue to move must be re-examined in living
eggs.

It is difficult to give any adequate summary of Roux’s
results. In his later papers he is not always consistent with
his earlier views. Schultze’s damaging criticism of some of
Roux’s earlier conclusions Roux has not answered, although

THE ORIENTATION OF THE FROGS EGG. 395

he has ably replied to other parts of the criticism at great
length. It is needless, however, to criticise Roux without
repeating his experiments. This, no doubt, will come in time,
and it is somewhat surprising that so little has been done by
other workers along the lines laid down by Roux. We may
here confine our criticism to those points that are connected
with the present ground covered. We may pass over the ex-
periments of Roux in which one of the first eight cells was
killed in order to determine its position in relation to the
embryo. These experiments, as Schultze says, contain direct
contradictions.

In Roux’s paper published in the ‘ Breslauer Aerztliche
Zeitschrift,’ No. 6, March 22nd, 1884, it is stated, “ Kine
weitere hierher gehorige Beobachtung machte ich an den
Eiern vom Wasserfrosch (Rana esculenta, s.viridis). Bei
dieser Species stellt sich die Hiaxe nicht senkrecht sondern
der Art -schief ein, dass bei der Ansicht von oben neben dem
hier braunen oberen Pol an einer Seite noch ein mondsichel-
formiger Saum des hier gelbweissen unteren Poles zum
vorschein kommt. Die erste Furchungsebene steht wie bei
R. fusea senkrecht, ist aber so orientirt, dass sie dieses Bild
symmetrisch theilt, wies blos moglich ist, wenn sie zugleich
durch den hochsten Punkt der gelben Randsichel und durch
die schief stehende Eiaxe hindurch geht. Durch die schiefe
Einstellung der Hiaxe zur Richtung der Schwerkraft wird hier
also auch schon die Richtung der ersten Furchungsebene und
mit ihr die Richtung der kiinftigen Medianebene des Embryo
noch vor der Theilung bestimmt.

“ An diesjahrigen Eierstockeiern von Rana escul. sah ich
trotz der noch mangeln den Entwickelungsfahigkeit schon
diese Einstellung beim Schwimmen im Wasserglas eintreten.
Sofern die gleiche Hinstellung reifer Kier im Wasser sich
nach der Befruchtung nicht andern, wirde hier also schon
im unbefruchteten Hie die Lage der Medianebene und das Oral
und Aboral neben dem Dorsal und Ventral bestimmt sein;
womit alle Hauptrichtungen des Embryo bereits vor der
Befruchtung gegeben waren.”

396 T, H. MORGAN AND UME TSUDA.

Schultze maintains the same view for the brown frog, but
Roux, in a later publication, denies this for this species.

From Roux’s conclusions, published in the ‘ Archiv fiir
Mikros. Anat.,’ vol. xxix, 1887, the following quotation is
taken :

1. The unfertilised frog’s egg has determined one main
axis of the median plane of the embryo. This results from
the bipolar arrangement of the yolk material, and corresponds
to the direction of the egg-axis passing from the black to the
white pole, i. e. to a ventro-dorsal direction of the real, a
cephalo-caudal direction of the virtual, embryo.

2. From the innumerably different meridional planes which
can pass through the egg-axis, that one corresponds to the
median plane of the embryo that lies in the direction of the
copulation of the two pronuclei.

3. The plane of copulation of the pronuclei is not in any pre-
determined meridian, but may be determined by localised
[artificial] fertilisation.

4. The side of the egg where the sperm enters forms the
ventro-caudal side of the embryo; the opposite side corresponds
to the dorso-cephalic.

In the body of the same paper Roux says that his experi-
ments show conclusively that there does not exist a latent bi-
lateral construction of the frog’s egg. He further adds that
in this year he “‘ was fortunate enough for the first time to
follow out with success the process of fertilisation in Rana
esculenta, and to observe that in this species a peculiar
typical change of position of the egg-axis takes place, i. e. the
black hemisphere sinks down 20°—30° towards the
side of entrance of the spermatozoon.”

Hence the first line of cleavage that passes through the
upper pole of the egg and the line of the entering spermato-
zoon would also pass through the apex of the white crescent.
Logically no fault can be found with this ingenious explana-
tion; but how explain Roux’s earlier observation, that un-
fertilised eggs of Rana escul. also show the white cres-
cent? Moreover the distribution of the pigment in the unfer-

THE ORIENTATION OF THE FROG’S EGG. 397

tilised normal egg seems to be such as not to allow a secondary
orientation described by Roux. While, therefore, we cannot
deny or refute Roux’s statement at present, it seems to me
that this point must be carefully examined by other workers
before its acceptance will be possible.

In 1888 Roux records the results of new experiments ‘ with
improved methods” to determine the relationship of the embryo
to egg-axes. If the blastula were injured at the apex of the
black pole the defect was found on the ventral side of the
embryo. Roux says that he had previously found that if the
blastula was stuck at the equator on the blastopore side the
defect appeared in the middle of the medullary folds, and he
had concluded that the head half of the embryo was formed in
the upper half of the egg, i.e. the embryo was placed vertically.
The researches of this year (1888) show that this defect was
not a primary phenomenon, but that it represented a later
change where a “ reparation” had taken place.

Roux injured the first anlage of the dorsal lip of the blasto-
pore, and found the defect to lie in the cross-connection at the
anterior end of the medullary folds. Injury to the blastula or
young gastrula at one side of the equator produced a defect in
the middle of the medullary folds. Injury to the young gas-
trula at a point opposite the gastrula crescent produced defects
in the caudal region. Injuries in the middle of the white area
gave no defect in later embryo.

These experiments of Roux’s are of great importance, for
if true they show the method by which we must regard the blas-
topore to be closed. I shall return to this in the final section
when speaking of the general problem of concrescence.

Roux concludes that the embryo lies over the lower hemi-
sphere, and that the dorsal lip of the blastopore moves through
170°. His figures (‘ Anat. Anzeiger,’ 1888) show the head
end of the embryo near that point of the equator at which the
embryo first appeared. The anterior connection of the medul-
lary folds lies just above the equator upon the black hemi-
sphere. From this region the embryo stretches over the lower
pole for 170°.

398 T. H. MORGAN AND UME TSUDA.

In these figures I believe Roux represents the early embryo
as extending over too great an extent of surface of the sphere.
Moreover, it seems, as I have said, most probable that the
blastopore does not start at the equator of the egg, but some
distance below that circle.

Schultze’s conclusion that the embryo lies over the black
hemisphere may be dismissed, as it is completely contradicted
by well-determined facts.

Finally, Robinson and Assheton make certain statements as
to the method of closure of the blastopore that call for notice.
Apparently at the outset they have orientated the embryo
wrongly, for they state, “The segmentation cavity has a roof
which ultimately becomes the anterior wall of the gastrula;
for the anus, which marks the posterior end of the embryo,
appears at the opposite pole of the ovum,—that is, in the floor of
the segmentation cavity.” Again, they say, ‘ For during the
formation of the blastopore the epiblast does not grow over the
yolk-cells enclosing them by a process of embolic invagination.”
This statement is intended to apply rather to the extension of
the epiblast over the sides of the embryo, and as such is per-
fectly correct. But, in addition to this process of delamina-
tion, there is a decided and extensive overgrowth, as we have
seen, of the dorsal lip of the blastopore enclosing the yolk em-
bolically. Further, the statement of Robinson and Assheton
that no portion of the archenteron in the anura is formed by
invagination is certainly incorrect, as I hope to show in a later
paper.

They continue, “ According to some former accounts, to
which we have made reference above, the anus of Rusconi has
been said to diminish in size by the gradual coming together
of each portion of the blastoporic rim simultaneously. This
we believe to be incorrect. The anus of Rusconi gradually
diminishes in size by the concrescence of the ventral part of
the lateral lips.” In their diagrams, to show the method by
which the ventral lip of the blastopore comes together, they
show the right and left sides applied to one another, and sub-
sequently fused. Later they say, “ We infer ... . that the

THE ORIENTATION OF THE FROG’S EGG. 399

anus of the frog, although apparently a new perforation, is
really a reopening of the temporarily closed portion of the ori-
ginal blastopore.” Now I doubt exceedingly whether con-
crescence of the ventral lip of the blastopore takes place in
any such way as the authors’ diagram indicates. The cells
from the sides may come later to the middle line, but not by a
process of apposition of the latero-posterior walls of the
blastopore. Rather the cells stream or migrate to the median
line below the surface, while the surface grows continuously
from behind forward. Hence we cannot speak of a reopening
of the blastopore, as nothing was left behind to reopen; but
we must speak of a perforation at the point where the blasto-
poric lips first began to concresce in a sense different from that
used by the authors.

IV.

Roux and Hertwig have given accounts of embryos in which
the blastopore had failed to close. Roux figures an embryo
with a hemisphere of white exposed, and the embryo lying as
a thickened zone around the border line between the black and
white hemispheres. Hertwig has not figured such extensive
exposures of yolk, but describes stages with varying amount
of blastopore unenclosed.

I attempted to produce such embryos artificially, and after
a great number of attempts that gave no favorable result found
at last a method that made it possible to produce such embryos
at will.

Embryos in which the blastopore had just appeared were
put into the following solutions, with the results recorded :

VOL. 35, PART 3.—NEW SER. EE

400 T., H. MORGAN AND UM& TSUDA.

Hydrochloric acid, 3, 7 per cent. . . Died.
Sodium hydroxide, 3, percent. . . Some died, in others the blasto-
pore closed.
Corrosive sublimate, 45 per cent. . Died.
Curari (weak solution) . . Nearly normal.
Quinine, ‘02 gr. to 500 c.c. HO . - Normal.
Morphine oH . Nearly normal.
Strychnine (only partially Guaeed) . Normal.
Alcohol, 10 per cent. - ; . Developed very little.
, ee . . - More than last, but died (un-
closed blastopore).
55 ors Developed partially (closed).
Sodium chloride . erms. e 500 ¢ c.c. H, ) Gave the desired result. Blasto-
= } strength of sea water pore open.
. jf Pie Phase - Normal.
39 ” 3 ” 9 - Died.

Of the solutions given only one gave the desired result,
although the alcoholic solutions seemed to have a similar ten-
dency. Theseries in which ‘6 per cent. salt was used produced
the embryos to be described below. This happened both for
the frog’s and the toad’s eggs, and was repeated with similar
results. Success depends on using exactly the right amount of
salt. Too much kills; too little does not affect the embryo.
To ensure success a series of trials should be made approxi-
mating to the ‘6 per cent. solution.

In a second series of experiments the recent suggestion of
Herbst was followed. Embryos were placed in solutions of
salts of barium, calcium, sodium, and potassium. ‘Three sets
of each solution were used, one stronger, one weaker, and one
the same strength as the ‘6 per cent. solution.

Although in some of these solutions embryos with large
blastopores were produced, no particular relation was found
between the formation of abnormalities and the series of com-
pounds used. The best results were again in the sodium
chloride.

Fig. x1v shows au embryo as seen from in front. A narrow
pigmented line marks the position of the suckers. Between
this and the white a thick fold of ectoderm marks the anterior
end of the medullary folds. The fold continues on each side

THE ORIENTATION OF THE FROG’S EGG. 401

along the border line between the black and white hemispheres.
These lateral medullary folds can be traced for only a short
distance.

Between the anterior connection of the medullary folds and
the white is found a‘small plate of ectodermal cells. The same
embryo seen from below is shown in fig. xv. We see that the
extent of white exposed corresponds to the whole of the lower
white area,—in fact, to somewhat more of the lower hemisphere
than would be finally enclosed by the normal blastopore; for
in the normal egg the far edge of the white, where it shades off
into the black, does not normally ‘become involved in the
closure of the blastopore. This was shown definitely in the
experiments made by sticking the lower pole, and is corrobo-
rated by the fact, that in these abnormal embryos the far side
of the large blastopore contains much more pigment than
does the ventral surface of the yolk-plug of the normal embryo.
Hence any statement as to the extent of the white closed over
by normal embryos, based on these abnormal embryos, will give
an erroneous conclusion. This, I believe, Roux has drawn.

The embryos produced in the salt solution may be examined
at each stage of their development, and the exact method by
which the blastopore forms be followed out. This gives a de-
cided advantage over the haphazard finding of embryos already
formed. In watching such embryos one sees that the blasto-
pore extends from its point of origin differently in these em-
bryos than in normal embryos. Instead of the corners of
the blastopore extending downwards to produce a deep cres-
cent or horseshoe-shaped outline, they extend laterally around
the border line between black and white. Hence results, I
believe, 2 more extensive enclosure of the lower hemisphere
than under normal conditions.

In fig. xvi is drawn another embryo, differing from the last
only in the? greater extent of the medullary plate lying in the
black hemisphere. This is due without doubt to the greater
extent to which the dorsal,lip of the blastopore has crept over
the white.j| Figs. xvir and xvi11 are drawings of embryos
where the overgrowth of the dorsal lip has_been carried farther

402 T. H. MORGAN AND UM£& TSUDA.

than in the last case, so that not only the anterior connection,
but also the anterior end of the medullary folds, lie on the
black portion of the egg. This embryo shows conclusively
that the extent of closure of the blastopore is far more than
the normal, for if the embryo really had covered so much of
the sphere as the whole of the white area, and as much of the
black as the anterior end now occupies, it would have covered
nearly two thirds of the sphere, and not one third, as in the
normal embryo.

Fig. xtx is from an embryo in which the dorsal lip of the
blastopore has grown over the lower pole to the extent indi-
cated by the medullary folds. When looked at from below—
i.e. with the white area turned up—vwe see still a large ex-
posure of white, but the posterior extension of the blastopore
is not completed, again verifying the statement made above,
that the exposure of the white is in these eggs abnormally
extensive.

In fig. xx is drawn the posterior end of an embryo much
more advanced than the last. Quite a large exposure of yolk
is present, but not nearly so much as in the other cases.
Anterior to the blastoporic plug the medullary folds have met
to form a closed tube. Posterior to the blastopore, as seen in
the figure, a deep groove is present, and this groove is formed
by the posterior medullary folds. The ventral lip of the blas-
topore has, therefore, grown over the white to the extent indi-
cated by the medullary folds. Whether this forward growth of
the ventral lip is unusually extensive I do not know, nor have
I any records to show whether in the earlier stages so much
of the white was enclosed as in the preceding cases; but,
judging ‘from the length of the embryo and from other facts, I
think we may safely conclude that the area enclosed was less.

Other embryos, with still less exposure of white, need not
be figured at present; Hertwig’s description seems to cover
such cases.

Serial sections were made through these embryos. In the
embryo shown in figs. xrv, xv1, &c., sections add little know-
ledge to that formed from surface views. The most noticeable

THE ORIENTATION OF THE FROG’S EGG. 403

structure is the large archenteron that begins at the edge of
the black in the mid-dorsal line, and extends as far forwards
as the level of the suckers. The medullary folds in these
embryos have not as yet rolled in to form the (half) nerve-
cords. Longitudinal section of the embryo drawn in fig. x1x
shows that at the ventral lip of the blastopore only a very
slight depression is present.

Conclusion.—His supposed concrescence of the Vertebrate
embryo to take place by apposition of the sides of the germ
ring, and due to this process the embryo grew posteriorly.

Balfour believed this to be untrue, and that the posterior
end of the embryo grew in length by a process of intussuscep-
tion in front of the last segment of the body.

Roux’s experiments by sticking the border between black
and white point directly to a process of concrescence of some
sort. If Roux’s experiments are accurate we must suppose
that the cells that will later form the central nervous system
are already laid down along the black-white border. These
cells must come to the middle line as the blastopore gets
smaller. The closing of the blastopore from before backwards
would then be due not to a backward extension of all of the
material of the dorsal lip over the yolk, but would
take place by new tissue coming up to the middle line from
the sides and placing itself with or behind the cells already
present in the dorsal lip.

I should not regard this, even if it took place, as apposition
in the strict sense of the word, nor is it intussusception in
Balfour’s sense. It would not be intussusception because
new tissue is coming in continually from the sides to mingle
or mix with the cells already present and multiplying in the
dorsal lip of the blastopore. Nor would it be apposition in
His’s sense, because the lateral borders of the blastopore are
not laid down side by side, since the blastopore does not close
by actual apposition of its lateral rim. I shall not here
attempt to formulate a theory of overgrowth, but merely to
point out the apparent bearing of the evidence furnished by
this experiment of Roux. It will, I think, be possible to de-

4.04 T. H. MORGAN AND UME TSUDA.

termine experimentally exactly the process that takes place in
the dorsal lip of the blastopore, and then we shall be prepared
to formulate more definitely a theory of overgrowth.

Bryn Mawpe, Pa., U.S.A.,
May 22nd, 1898.

DESCRIPTION OF PLATES 24 & 25,

Illustrating Messrs. T. H. Morgan and Umé Tsuda’s paper
on ‘The Orientation of the Frog’s Egg.”

FIGs. IV, V, VI, VII, VIII, IX, X, XI, XII, X11, Rana temporaria.
Figs. I, I, II, XIV, XV, XVI, XVI, XVIII, XIX, xx, Rana, sp. ?

PLATE 24.

Fie. 1.—View of 32-celled stage from above.

Fig. 11.—View of same stage from dark side.

Fig. 111,— View of same stage from light side.

Fie. tv.—View of blastula (about 150 cells), dark side.

Fic. v.— View of same, light side.

Fig. vi.—Four-celled stage, lower pole; M—J, first furrow.

Fic. vi1.—About 100-celled stage, lower pole.

Fie, vi11.—About 100-celled stage, lower pole.

Fic. 1x.—Earlier stage than last.

Fic. x.—View of lower pole of egg at end of segmentation.

Fig. x1.—Harly blastopore stage. Pigment has somewhat shifted its
earlier distribution, lower pole.

Fic. x11.—Harly blastopore stage, lower pole.

Fic. x111.—Early blastopore stage, side view.

Fig. xiv.—Embryo with unclosed blastopore, seen from in front.

Fic. xv.—Same from below.

Fic. xvi.—Embryo with unclosed blastopore, seen from in front.

Fig. xvi1.—Embryo with unclosed blastopore, seen from in front.

Fie.
Fic.
Fig.

Fic.
Fie.
Fie.
Fic.
Fic.
Fic.
Fic.

view.

Fie.

view.

Fic.
Fic.

text.

Fic.
Fic.
Fie.
Fie.
Fic.
Fie.
Fie.
Fie.
Fic.
Fic.
Fic.
Fie.
Fig.
Fic.

THE ORLENTATION OF THE FROG’S EGG. 405

XViiI.— Same as last, side view.
x1x.—Embryo with large blastopore, seen from dorsal side.
xx.—Embryo with large blastopore, seen from behind.

PLATE 25.

1.—Diagram normal egg, lower pole.

2.—Diagram same egg, side view.

3.—Diagram normal egg, lower pole.

4,—Diagram same egg, side view.

5.—Diagram to show defect x produced by sticking apex blastula.
6.—Diagram to show defect x produced by sticking apex blastula.
7.—Diagram normal embryo to show length of medullary folds, side

8.—Diagram same embryo to show length of medullary folds, dorsal

9.—Diagram to show lower white area when blastopore appears.
10.—Diagram egg stuck opposite edge of white from blastopore, see

11.—Diagram same, later stage, see text.

12.—Diagram same, later stage, see text.

13.—Diagram egg stuck far from blastopore, see text.
14.—Diagram same, later, see text.

15.—Diagram egg stuck far from blastopore, see text.
16.—Diagram same, later, see text.

17.—Diagram egg stuck nearer to blastopore, see text.
18.—Diagram same, later, see text.

19.—Diagram same, later, see text.

20.—Diagram egg stuck near blastopore, see text.
21.—Diagram same, later, see text.

22.—Diagram egg stuck centre of early blastopore, see text.
23,.—Diagram same, later.

24.—Diagram with injury taken as a fixed point x to show relative

advance of dorsal lip of blastopore.

‘

7

aa ’ ¥ ;
i sy ofl, ye ul Ole riya ate i wit hae
ak aa : : ms.

a
Ot tie - ital: ere 1 iecainntsi nf
: J it Ag Au isi Bh Se ws adalat b ¢ ial si nM

Soe u
: he *- is
Pk urs
Sd ik
J - ‘
= Al j . f Be) L, } ean pai he {ft = af bin
( Uae whi Pe j ht) i< Nil 0 uy
AOR Pi, loi ynith eee
. ig < ‘ i
1 é * j r O
. et
]
i
= “
Ne
= at
Pee ‘
. 4 = ae
: B.
* - ~ : H m
i il 1
> or
&4 7 iat -
=)
i 56 var
, 7 *
> ‘
- mm!
“ae 2
. . y
|
Ay ‘
' =
2 ’

FOSSIL MAMMALIA FROM THE STONESFIELD SLATE. 407

On the Fossil Mammalia from the Stonesfield
Slate.

By

E. 8. Goodrich, F.L.S.,
Assistant to the Linacre Professor of Comparative Anatomy, Oxford.

With Plate 26.

Ever since the discovery, made some eighty years ago, of
Mammalian remains in the Stonesfield Slate near Oxford,
these fossils have excited the interest of naturalists, and have
been the subject of much discussion amongst geologists and
palzontologists both in England and abroad. Nevertheless
something still remains to be described in the few specimens
which exist ; and, while one of them has not yet been figured
at all, others have been only inaccurately represented. I
therefore propose to write a short history of each fossil, as far
as it is known, giving figures when necessary ; and to sum up
the most important results which have been reached with
regard to them by previous authors, together with some
remarks as to the bearing of the facts ascertained by a careful
study of the teeth belonging to these remains on the general
question of the origin and homology of the cusps of Mam-
malian teeth.

Besides the fragment of the multituberculate form Stereo-
gnathus, which I shall mention later, there are only twelve
undoubted fossil Mammalian remains from the Stonesfield
Slate at present known; ten of these are lower jaws, two are
limb bones. Six of the fossil jaws are in the Oxford Museum,
one in the York Museum, one in the private collection of

408 E. S. GOODRICH.

Mr. Parker of Oxford ;! the two remaining jaws and the limb
bones are in the British Museum.

Through the kindness of Professor Green and of Professor
Lankester, who placed the Oxford fossils in my hands for the
purpose of displaying them in a museum case in a manner
more worthy of their interest and value, I have had the oppor-
tunity of examining and handling our six specimens. I am
much indebted to Dr. Henry Woodward for allowing me to
examine the two British Museum fossils, and_to Mr. Parker
for lending me his. To the authorities of the Museum at
York I must express my thanks for lending me the excellent
specimen in their keeping; but more especially to Professor
Lankester, who spared himself no trouble in obtaining for me
this privilege, and who has further given me much help during
my researches.

It may here be mentioned that I have been able, by care-
fully working away the matrix with sharp needles under Zeiss’s
dissecting microscope, to expose new cusps, and in some cases
new teeth, in five of the jaws described. My figures, which I
have endeavoured to make as faithful as possible, differ, there-
fore, considerably from those previously published.

The formation from which these fossils were obtained be-
longs to the Lower Jurassic period; whence the great interest
attached to them, for at the time of their discovery they were
by far the earliest known remains of warm-blooded Vertebrates,
—being, in fact, the first Mesozoic Mammalia obtained. Since
then, as is well known, remains of a few fossil Mammalia have
been found both in England and elsewhere in strata belonging
to the Triassic age; as, for instance, Microlestes in England,
Dromatherium in America, and Tritylodon in South Africa.

The two fossil limb bones mentioned above have been
figured and described by Professor H. G. Seeley (29). As,
unfortunately, they afford no clue to the relationship of the
animals whose jaws are described below, there being no proof

1 Mr. Parker also has in his possession a toothless fragment of a jaw
which may perhaps be Mammalian,

FOSSIL MAMMALIA FROM THE STONESFIELD SLATE. 4.09

that these bones belonged to any of them, I need only men-
tion that Seeley considers that they are “ limb bones indicating
a generalised insectivorous type, modified from a Monotreme
stock in the direction of the Marsupial plan.”

Genus AMPHITHERIUM.

Four of the fossil jaws appear to belong to thisgenus. Three
of these, two of which are type specimens, are in the Oxford
Museum; the fourth is in the British Museum.

Amphitherium Prevostii, Blainville.
Type specimen, P). 26, fig. 1.

A left ramus of the lower jaw, seen from the inside; in the
Oxford Museum.

In 1824! Dr. W. Buckland, the well-known geologist, first
announced the discovery of the remains of mesozoic Mam-
malia in his. paper ‘ on Megalosaurus’ (5). He there mentions
‘‘two portions of the jaw of the didelphys or opossum,” which
he refers “to this family on the authority of M. Cuvier, who
has examined it.” From Mr. W. J. Broderip we learn that
these rare fossils were obtained as follows :—‘‘ An ancient
stonemason living at Heddington . . . made his appear-
ance in my rooms at Oxford with two specimens of the lower
jaws of mammiferous animals, embedded in Stonesfield slate.
. . . One of the jaws was purchased by my friend Professor
Buckland, who exclaimed against my retaining both” (4).
The fossil purchased by Broderip himself is the type specimen
of Phascolotherium (see below) ; Buckland’s fossil is the type
specimen with which we are now concerned, He placed it in
the Ashmolean Museum, whence it has come with the other
specimens of the Buckland collection to its present home, the
Oxford University Museum. The exact date at which it was
purchased I have been unable to ascertain for certain ; it was
probably about 1814: the date given by Zittel is 1812 (82),
but I know not on what authority.

Cuvier, who visited Oxford in 1818, says of these fossils

1 Not 1823, as is generally stated to be the case.

410 E. S. GOODRICH.

that “lors d’une inspection rapide que j’en pris a Oxford, en
1818, [ils] me semblérent de quelque Didelphe ;” and adds in a
note that the jaw of Amphitherium “ est celle d’un petit car-
nassier dont les macheliéres resemblent beaucoup 4 celles des
Sarigues, mais il y a dix dents en série, nombre que ne montre
aucun carnassier connu” (8). This note was written after
the examination of some careful drawings of Buckland’s fossil
and of the type specimen of Amphilestes (see below) sent to
him by Prévost, who was then travelling in England.

These announcements of the discovery of Mammalian re-
mains in stone belonging to the Mesozoic age created a great
sensation amongst the paleontologists of the time, and it was
not for more than twenty years afterwards that the opinion of
the great French naturalist was generally accepted. Some
contended that the fossil did not really belong to the slate,
others that the strata in which they were found were not of
the Mesozoic period; while others, again, urged that the jaws
were those of a reptile, or even of a fish. All doubt having
been set at rest with regard to these points, it will not be
necessary to enter here in detail into the arguments used on
either side.

Prévost, in 1825, on his return from England, where he had
carefully examined the specimen of Amphilestes now at York,
and “le fameux Didelphe”’ in Buckland’s collection, published
the first detailed description and figure of this the type speci-
men of Amphitherium (25). He describes the teeth as having
tricuspid crowns, and two distinct roots in alveoli, concluding
that the fossil was Mammalian in confirmation of Cuvier. As
to its relationship, Prévost considered that it was probably
‘un mammifére carnassier insectivore qui pouvait offrir quelque
analogie avec les Didelphes, mais qui appartiendrait 4 un genre
inconnu.”

Agassiz, in 1835, mentions the Stonesfield fossils in a short
note (1). He considered that the remains were not sufficient
to allow of a certain determination of their affinities, but
drew attention to the resemblance of the teeth (especially
those of Phascolotherium described below) to those of certain

FOSSIL MAMMALIA FROM THE STONESFIELD SLATE. 411

seals possessing tricuspid molars. Owen mentions (20) that
Agassiz proposed the name Amphigonus for Amphitherium, in
the German translation of Buckland’s Bridgewater Treatise
(which I have not seen).

Dr. Buckland, in 1836, gave a rough figure of this jaw, to-
gether with enlarged drawings of two of the teeth (6).

Two years later M. de Blainville published his “ Doutes sur
le prétendu Didelphe de Stonesfield ” (2), in which he tried to
prove that the fossil in question, of which he reproduced Buck-
land’s figure, belonged to a reptile. In this paper he laid
considerable stress on the fact that “ une portion de machoire
inférieure, rapportée de Stonesfield par M. Brochant de Villiers
et ses éléves MM. Elie de Beaumont et Dufrénoy, et qu’on
avait supposée appartenir au méme Didelphe,” had been proved
to be reptilian, and accepted as such even by Cuvier. Blain-
ville mistook the mylohyoid groove,' well marked in our speci-
men, for a suture indicating that the jaw was of a compound
structure: a similar mistake was made in the case of the other
jaws. He proposed the generic name Amphitherium, which
has since been generally adopted.

Buckland then took with him to Paris the type specimen of
Amphitherium and the second fossil of the same species now
in the Oxford Museum, which will be described below. He
showed these to M. Valenciennes, who made a careful study
of them, publishing a detailed account mainly confirming the
results of Cuvier and Prévost, in which the name Thylaco-
therium Prévostii is proposed (81). Unfortunately Blain-
ville was not convinced, as he did not see the fossils, for he tells
us that “le jour ou M. le docteur Roberton voulut bien
m’inviter 4 passer la soirée chez lui avec M. Buckland, je
partais pour la campagne” (3). He therefore only brought
forward “ Nouveaux doutes”’ (8), in which we learn that “ M.
Buckland lui-méme a exposé le probléme et les piéces sur les-
quelles il repose a l’investigation des naturalistes allemands
reunis en congrés 4 Fribourg, en Brisgaw, au mois de sep-
tembre dernier ”’ (1838). In his contention as to the saurian

‘ For a full discussion of the mylohyoid groove see Osborn (14),

412 E. S. GOODRICH.

character of these remains, Blainville was opposed in Paris by
Duméril (9) and Geoffroy Saint-Hilaire (27), whilst in Eng-
land he was supported by Professor Grant (11) and by Ogilby.
The latter took up a more impartial position, and considered that
they were not justified by the evidence in pronouncing whether
the fossils were mammalian or reptilian, arguing in favour of
saurian affinities that the molars and premolars could not be
distinguished, that the canine and incisors occupy five twelfths
of the dental line, that the incisors (in the type Phascolothe-
rium) are nearly in the same straight line as the grinders, and
that the condyle is below the level of the crown of the teeth (18).

On returning to England, Buckland entrusted to Owen
these “ bones of contention,” as the latter calls them, for the
purpose of making an exhaustive study to ascertain their true
nature. In 1838 Owen read two papers on this subject before
the Geological Society (18), and in 1842 his full treatise was
published with carefully executed figures (19). Owen drew
attention to such Mammalian characters as the convex articular
condyle, the broad, high, and curved coronoid process, situated
immediately in front of the condyle and resembling that of
the opossum, and the separate angle inflected as in Marsupials
and some Insectivora. He then described the teeth in detail,
comparing the six molars of Amphitherium to those of Didel-
phys, and the four premolars to those of Didelphys and Talpa.
For the first time Owen definitely pointed out that the molars
of these animals closely resemble each other, belonging to the
type of tooth now known as the tritubercular-sectorial. “ An
interesting result of this examination is the observation that
the five cusps of the tuberculate molares [of Amphitherium]
are not arranged, as had been supposed, in the same line, but
in two pairs placed transversely to the axis of the jaw, with the
fifth cusp anterior, exactly as in Didelphys, and totally different
from the structure of the molares in any of the Phoce, to which
these very small Mammalia have been compared” (18). It is
surprising, after this, to see Professor Osborn, fifty years later,
claiming the “discovery”’ of tritubercular molars in Amphi-
therium (15).

FOSSIL MAMMALIA FROM THE STONESFIELD SLATE. 413

In subsequent works Owen again describes and figures this
jaw (20, 21, 23); a “diagram” of it is given by Professor
Phillips in his ‘ Geology of Oxford’ (24).

In 1888 Professor H. F. Osborn published his interesting
memoir on the ‘ Mesozoic Mammalia of the Old and New
Worlds, in which he treats of the Amphitherium jaws (14).
Unfortunately the description of the molars is quite erroneous,
owing to the author’s considering that the cusps were all in
the same line. “The fact is,” says Osborn, “ these crowns of
the molars consist of an elevated anterior and median cusps,
followed by a low posterior heel, and with an internal
cingulum rising into the low cusp on the inner face of the
median cusp.” At this time Osborn had only seen drawings
and published figures of the Oxford jaws, but shortly after,
during a second visit to England, he examined them himself,
and corrected his mistake in a subsequent publication (15),
finding the external cusp described half a century before by
Owen,

I have nothing to add to these descriptions, excepting the
fact that Owen was wrong in considering the median projec-
tions in molars 4 and 5 as being the external cusp (which he
found in molars 2 and 6); they are simply the surface where
the internal cusp has been broken off, and I have exposed in
these two teeth the external cusp itself, which was previously
hidden (see figure).

The second specimen in the Oxford Museum (PI. 26,
fig. 2) is a left ramus, seen from the inside.

Valenciennes was, I believe, the first to mention this fossil
(31), which was one of the two brought over to Paris by Buck-
land in 1838. The French author, however, erroneously con-
sidered it was “ de la méme espéce que celle décrite et figurée
par M. Broderip, son Didelphys Bucklandi” (Phascolo-
therium).

Owen figured and described it correctly (19—21, 23) as
a left ramus containing one molar behind, separated (by a gap
wide enough to allow for four molars) from two entire and two

414, E. 8. GOODRICH.

broken premolars.1 In front of these Owen detected the
sockets for six teeth.

Osborn has lately (15) discovered a young molar emerging
from the jaw just behind and within the posterior molar.

The British Museum specimen (Pl. 26, fig. 4) is a
portion of the right ramus, inner view.

This fossil has been partially figured by Osborn in his first
treatise (14), who took it to belong to a left ramus; but in
his later paper (15) he recognised its true character by the
presence of the “ double internal cusps [of the molars], by the
cingulum upon the premolar, and by the faint mylohyoid
groove near the lower border,” adding in a note that “it
would be well to run the risk of injuring one of these molars
to expose the external cone.”

The jaw, which is figured here complete and in detail for the
first time, although much broken is really the most instructive
of the four specimens extant of this genus, as regards the
structure of the molars. In front are the traces of four teeth,
followed by an entire premolar. The latter possesses a late-
rally compressed crown bearing one large cusp, a very small
anterior cingulum cusp, and a posterior heel. On the whole
this tooth is very similar to the premolars of the other Amphi-
therium jaws. The molars, of which there are five remaining,
when I first examined them displayed only the two nearly
equal] and pointed cusps of the inner margin, and a simple low
posterior heel (as shown in the figure by the third and fourth
molars); by carefully clearing off the matrix I have exposed
the large external cusp (protocone) in the first, second, and
fifth molars (see woodcuts, figs. 1 and 2). In the latter tooth
I have also exposed the full extent of the posterior heel,
which is seen to rise on the external margin into a pronounced
cusp; the internal cusp is hardly distinguishable from the
general margin of the heel at its postero-internal angle. The
heel is somewhat narrower than the body of the tooth at the

1 The posterior premolar bears a small internal cusp (= the deuterocone
of Scott P).

FOSSIL MAMMALIA FROM THE STONESFIELD SLATE. 415

level of the more anterior external cone, and a “basin,” if
it can be said to exist at all, is small and shallow as far as I
have been able to ascertain.! It should be noticed that this

ities, a Ere 2:

62.0.

Fie. 1.—Inner face of a molar of the right ramus of the lower jaw of Am-
phitherium Prevostii.

Fic, 2.—Hypothetical representation of the outer face of the corresponding
molar of the left ramus. a. t. c. Anterior internal cusp (paracone).
ex.c. External cusp (protocone). #. Heel. wm. ¢. c., Median internal
cusp (metacone).

The teeth are represented as if entirely freed from the matrix.

is the first complete tooth of the tritubercular-
sectorial type which has been seen in the Stonesfield fossils.

Amphitherium Oweni, Osborn.

The type specimen in the Oxford Museum; a right
ramus with the external surface exposed (PI. 26, fig. 3).

The type and only specimen of this species, obtained with
the other jaws from Dr. Buckland’s collection, was found pro-
bably about 1845. It was first described and figured by Owen
in his ‘ History of British Fossil Mammalia’ in 1846 (20),
and subsequently in his Monograph (23) and Odontography
(21) under the name of A. Prevostii. This specimen, perhaps
the most perfect of all the Stonesfield jaws, shows very clearly
the angle which, according to Owen, exhibits a degree of inflec-
tion which is “ less than in any of the known Marsupialia, and
does not exceed that of the Mole or Hedgehog” (20). He
describes the six posterior molars as quinquecuspidate, but

1 Judging from other jaws, we may presume that the heel is, if anything,
better developed in the anterior molars.

VOL. 35, PART 3,—NEW SER. FF

416 BE. 8. GOODRICH.

these teeth are (at ail events at the present time) so mutilated
that it is scarcely possible even to recognise their tritubercular
character. The outer cusps have been entirely broken away, and
only traces of the two inner cusps and the posterior heel remain.

An outline figure of this fossil has been given by Phillips
(24). Professor Osborn in his first paper (14) separated this
specimen from the genus Amphitherium with the name of
Amphitylus Oweni, under the impression that the molars
were tricuspidate.! There can now be no doubt that this view
is erroneous, and that the teeth are really, as Owen described
them, of the Amphitherium pattern. Zittel (82) follows the
latter in including it in the species A. Prevostii; but I have
retained it as a separate species, chiefly on account of some
considerable differences exhibited between the shape of the jaw
of this fossil and that of the specimens described above. The
coronoid process is straight above and more pointed at its
posterior extremity; the condyle is more slender, the notch
between it and the coronoid process being more pronounced ;
the angle is rather larger, and produced farther back. The
premolars, as exemplified by the third and only entire one in
this jaw, differ somewhat from those of A. Prevostii; the
cusps are more rounded, the main cusp is situated not near
the centre of the tooth, but well forward, and the swelling on
the fangs, also rather large in the molars, is strongly developed.
The molars are unfortunately too broken to compare in detail
with those of the previous species.

The question of the lower dental formula of this genus
has been purposely left until after the description of all the
specimens. Owen, who evidently saw the second Oxford Am-
phitherium (Pl. 26, fig. 2) in a more perfect condition than
that in which it now is, gave the formula 1. 3, c.1, pm. 6, m.6
(19), describing eight sockets in front of the anterior broken
premolar. There can be little doubt, however, that he over-
estimated the number of teeth anterior to the four premolars,
In front of these can still be seen the broken roots of what was

1 Osborn has since changed his mind, and now, I believe, includes this spe-
cimen in the genus Amphitherium.

FOSSIL MAMMALIA FROM THE STONESFIELD SLATE. 417

probably a small premolar, and the next two sockets are most
likely those of a double-fanged canine, such as we find in many
other mesozoic Mammalia. Four incisors account for the
anterior sockets. On the other hand, the type specimen (PI.
26, fig. 1) has six molars, and four molars beyond which the
jaw is broken; so that on combining the two we get the for-
mula i. 4, c. 1, pm. 5, m. 6.1 The dentition of A. Oweni
conforms perfectly to this; we find four incisors, a double-
fanged canine, followed by eleven teeth, of which five were
probably premolars and six molars. This is also the conclu-
sion reached by Osborn (15) for the dental formula of Amphi-
therium.

Genus PHASCOLOTHERIUM.

There are three jaws belonging to this genus—the type
specimen in the British Museum, a specimen in the Oxford
Museum, and one in Mr, Parker’s collection. They are all
placed in one species.

Phascolotherium Bucklandi, Broderip.

The type specimen, aright ramus with the inner surface
exposed, in the British Museum.

This is the fossil mentioned above as having been obtained
by Mr. Broderip, together with the type specimen of Amphi-
therium, about 1814. It must, therefore, be the other of the
“two portions of the jaw of the Didelphys” which Buckland
tells us were seen by Cuvier (5). This jaw was lost for a time,
but, on being found again, was described and figured by
Broderip in the ‘ Zoological Journal’ of 1828 (4). He there
described the teeth as consisting of seven grinders, one canine,
three incisors, and the alveolus of a fourth, naming the jaw
Didelphys Bucklandi.

It has been figured by Buckland (6), by Blainville (2), by
Owen (19—21, 23), and lastly by Osborn (14).

1 The second Oxford specimen is the only jaw which shows definite signs

of having had five premolars; A. Oweni affords no certain evidence in this
respect.

418 E. S. GOODRICH.

In his writings on this jaw Owen describes the rounded and
sweeping outline of the lower margin, the wide recurved coro-
noid process resembling that of the “ zoophagous Marsupials,”
the notch between the coronoid process and the condyle being
especially like that in Thylacinus. Other resemblances he also
finds to Thylacinus, namely, in the inflected angle and the
molars, whilst the condyle is said to be more like that of
Dasyurus and Didelphys. A well-marked mylohyoid groove
is present, and the dental foramen is far forward, as in other
related Mesozoic mammals. Owen claims that the “ molars”
(grinders or cheek-teeth) of Phascolotherium resemble those of
Thylacinus in number (seven in both cases) and in shape, both
possessing three main cusps in a line and two accessory fore-
and-aft cingulum cusps, very small in the case of the living
Marsupial. He could distinguish no difference between pre-
molars and molars, and looked upon the grinding teeth as
being in a simple condition in which the two varieties were
not differentiated.

All observers have noticed the peculiar pitting of the surface
of the grinders of Phascolotherium.! Osborn, in his first
paper, also held the view that the seven grinders could not be
distinguished into premolars and molars: “A close study of
m. 1 shows that it possesses in miniature the characteristic fea-
tures of the other molars, three cusps and a basal cingulum”
(14). Inthe ‘ Additional Observations’ (15), after examining
all the specimens of this species, he writes, “‘ The jirst tooth
behind the canine has a main cusp like that of the posterior
molars, and an internal cingulum horizontal and rising in two
points, instead of showing the sweep downwards and back-
wards which is so characteristic of premolar cingula. The
accessory cusps are either covered with matrix or broken off.
. . . . The chief interest lies in the main cusp [of the second
tooth], which is loftier and more pointed than the protocone

1 It is interesting to notice that the teeth of Phoca barbata, which
resemble them so closely in shape (as pointed out by Agassiz, 1), also exhibit
a pitting of the surface, which might lead one to believe that they are both
adapted to some similar kind of food.

FOSSIL MAMMALIA FROM THE STONESFIELD SLATE. 419

of the third tooth, which in turn has all the characteristics of
a molar.” Osborn concludes that the formula may be pro-
visionally written i. 4, c.1, pm. 2, m. 5. In assigning this
formula to Phascolotherium I entirely agree with Professor
Osborn, and can only say that the difference between the first
two grinders and the succeeding molars is perhaps even more
marked than he has described.

The Oxford Museum specimen; a left ramus, seen
from the inside (PI. 26, fig. 8).

This specimen has been figured in outline by Phillips (24)
and mentioned by Osborn (15). Mr. Lyddeker, in the British
Museum catalogue (12), erroneously referred it to the genus
Amphilestes.
~ Behind are four molars in very good preservation, showing
the marked cingulum rising in the two internal points charac-
teristic of the genus; in front and behind the cingulum forms
small, sharp cusps. I have lately exposed from the matrix an
entire incisor, and as far as possible the sockets of the other
teeth which are missing. This jaw is, therefore, now of some
use in making out the dental formula. Posteriorly we have
the four molars already mentioned, and immediately in front
of these are two sockets, presumably belonging to the first
molar. Then come two pairs of sockets, of which the anterior is
much the smallest, in which were the two premolars. In front
of these, again, is the large alveolus of the canine, preceded
by two sockets for two incisors. Next comes the incisor
recently brought to light, which in shape is stouter at the base,
more closely resembling the incisors of Thylacinus than seem
to do those of the type specimen. Beyond this tooth the jaw
is slightly damaged ; the first incisor has probably been broken
away. The dental formula of this jaw would then agree with
that given by Osborn for the type specimen, and adopted here.

Mr. Parker’s specimen; a portion of the right ramus
with the inner surface exposed (Pl. 26, fig. 9).
This, the third and last specimen of Phascolotherium, is the

420 E. S. GOODRICH.

hinder portion only of the ramus, with three molars in situ.
It has been figured in outline by Phillips (24), and mentioned
by Osborn (15). As the latter notes, it is remarkable for the
great development of the coronoid process. The articular con-
dyle is slender and very well preserved ; and the broken angle,
which I have recently cleared from the matrix, can be clearly
discerned. From the fractured surface we may conclude that
the angle was very much inflected, and exceedingly thin and
flat near its point of attachment; unlike the same process in
living Marsupials, it was situated entirely behind the dental
foramen. In front of the teeth are the impressions in the
matrix of the two anterior molars.

Genus AMPHILESTES.

Of this genus there are three specimens, included in one
species, two of which are in the Oxford Museum, and one in
the museum of the Philosophical Institution at York.

Amphilestes Broderipii, Owen.

The type specimen in the museum at York; a left ramus
with the inner surface exposed (Pl. 26, fig. 5).

Valenciennes was the first to mention this fossil in 1838.
He says, “ Une autre machoire, que je crois étre de cette
derniére espéce [Phascolotherium Bucklandi], fait partie
du cabinet de M. Sykes” (81). The Rev. H. Sykes pre-
sented it to the museum in which it now rests. Valenciennes,
“W@aprés le dessin qui a été envoyé par M. Phillips 4 M.
Cuvier,” mistook it for a right ramus seen from the outside.
Owen, in the same year, described it briefly under the name of
Amphitherium Broderipii,! giving an elaborate but in
some respects erroneous and misleading figure (19), which has
since been frequently copied in his own and other writers’
works (20—28). In this, and in all his subsequent
figures, Owen represents the angle of the jaw as separate and
hardly inflected (as in Amphitherium), and the coronoid pro-

1 Owen remarks that it appears to be closely allied to Phascolotherium.

FOSSIL MAMMALIA FROM THE STONESFIELD SLATE. 421

cess as angular and small. Phillips (24) and Osborn (14)
followed him in this; but in his second paper Osborn recog-
nised the fact that the angle is “ precisely as in Phascolothe-
rium” (15),—that is to say, it is inflected and confluent with
the condyle.

Owen described the molars with their three high-pointed
cusps in a line (much more pointed than those of Phascolo-
therium), their well-developed internal cingulum rising in two
fore-and-aft cusps, and one internal median cingulum cusp.
The premolars have no cingulum, and possess one main median
cusp with a small cusp in front and behind. There are five
molars in situ behind, separated by a gap for one tooth from
three premolars; in front of these all the teeth are missing
except one incisor, which I have exposed from the matrix.

The jaw is much damaged posteriorly, but the impression of
the condyle and coronoid process is still visible in the matrix.
The process is seen to resemble in shape that of Phascolothe-
rium, although it is relatively smaller.

The first specimen in the Oxford Museum; a left ramus
with its outer surface exposed, reversed (Pl. 26, fig. 6).

Professor Phillips first figured this jaw in outline in his
‘Geology of Oxford ’ (24), probably soon after its discovery. By
mistake he put one more molar in the figure than the number
exposed in the fossil. Osborn also mentions this specimen (15).

The coronoid process, condyle, and angle are unfortunately
broken. I have exposed an entire fifth molar behind the four
molars previously visible, the broken base of an anterior in-
cisor, and have rendered what remains of the other incisors
more distinct. The molars have no cingulum on the outer
surface, but the sharp lateral cingulum cusps are well seen,
especially the hinder one.

The second specimen in the Oxford Museum; a portion
of the right ramus seen from the outside (PI. 26, fig. 7).

1 Apparently this eminent and keen-sighted geologist ‘ could see through
a stone wall” better than most people, for the fifth molar which he thus
figured has now been brought to light, twenty-two years after (see figure
here given).

422 E. S. GOODRICH.

This fossil has never been figured before; it was probably
obtained about 1875. The coronoid process and the condyle
have been broken; there are eight consecutive teeth present—
five molars and three premolars. These teeth resemble in
every particular those of the foregoing specimen.

The dental formula of this genus has hitherto been very
difficult to settle, but the working out of the new teeth and of
the sockets of the missing teeth has rendered the task easier.

Owen, who only studied the York specimen, assigned to it
the formula i. 3, c. 1, pm. 6, m. 6 (28). He considered that
the tooth missing between the molars and premolars had been
a molar. In front he made out sockets for the three additional
premolars, the canine, and the incisors. However, these so-
called sockets were mere undulations of the edge of the matrix,
not corresponding to the true sockets which I have now exposed.
Mr. Lyddeker (12) gives the probable formula as i. 4, c. 1,
pm. 4, m. 7, which was adopted by Osborn in his first paper
(14). After personally examining the three specimens Osborn
gives the formula “ with considerable certainty as follows :—
i. ? 8,c.1, pm. 4, m.6” (15). He tells us that “ the Oxford
specimens show that there were but siz molars instead of
SEVEIINTS x2 2 In fact, one .... specimen. .. . shows but
five molars. If this specimen be adult, as seems improbable,
it may represent a new genus transitional between Amphilestes
with six molars and Triconodon with four” (15). I must con-
fess that I am quite unable to see how one specimen with five
molars and another with four (at that time) can even without
“certainty” lead to the adoption of the formula m.6. The
ingenious speculation as to the intermediate genus seems to be
quite unnecessary when we now know that all the existing
specimens possess five molars, and show no signs of having had
more or fewer. The premolars are more difficult to deal with.
If we measure the distance between the last molar! and the
first premolar in the first and second Oxford specimens we find
that the eight consecutive teeth occupy exactly the same space

1 In these measurements I have found it convenient to measure from the
middle of one tooth to the middle of another.

FOSSIL MAMMALIA FROM THE STONESFIELD SLATE. 423

(10 mm.) in both cases. Measuring now the space occupied by
the eight posterior teeth in the York specimen (including the
gap as one tooth), we find that it is the same as in the others.
There can, therefore, be little doubt that the missing tooth was
a premolar, the fourth. Returning to the first Oxford speci-
men—in front of the three premolars is a space with two
sockets, which were evidently occupied by the first premolar,
‘still present in the York jaw. In front of the first premolar
we have a region which must be carefully compared in both
fossils. As this region in the York specimen is broad and flat,
and cannot well be seen from the side, I have given a figure of
it more from above (Pl. 26, fig.5 a). The new and only
incisor present appears to be the fourth, and three sockets are
visible in front of it. Between the incisor and the first pre-
molar are two large sockets, somewhat difficult to account for.
If the first of these represents a missing canine, and the second
a small premolar, we would have one more post-canine tooth
in this jaw than in the Oxford specimen (see above). On the
other hand, if they both belong to a large double-fanged canine,!
we would then apparently have only three incisors in the Oxford
jaw, assuming that here also the canine occupied two sockets.”
On examining the latter specimen closely it is seen that the
anterior extremity of the jaw is broken ; moreover the distance
from the extremity to the second premolar in the York Am-
philestes is 2 mm. longer than in the Oxford specimen. There
is, then, no great difficulty in supposing that in the latter the
first incisor has been broken away.

Provisionally the formula of Amphilestes Broderipii
may therefore be written i. 4, c. 1, pm. 4, m.5.

As for the systematic position of these Stonesfield fossils,
the remains are too scanty to allow us to form any very
definite opinion. Owen considered Phascolotherium to be

' Double-fanged canines are not uncommon amongst Mesozoic Mammalia.

* It is to be noticed that in both the jaws these sockets touch one another,
whilst there is a small space on either side of them. Should the York fossil
prove to have possessed ten teeth behind a single-fanged canine, it would
have to be separated from the Oxford specimens.

424. E. S, GOODRICH.

marsupial and nearly related to Thylacinus; Amphitherium,
on the other hand, he places nearer the Insectivora, and in
one work (20) puts it actually in that group. Lydekker classes
all these Stonesfield Mammalia in one family, the Amphithe-
riide (including Amblotherium, Achyrodon, and Peramus) ;
but there can be no doubt that Amphitherium should be widely
separated from the triconodont group (Phascolotherium and
Amphilestes). Osborn includes the latter with Triconodon
and its allies in one group, the Triconodonta; Amphitherium
he places with the living Polyprotodont Marsupials in the
group Trituberculata, which also includes Amblotherium and
its allies. Zittel adopts this arrangement, which certainly
seems to be the best and safest yet proposed.

Stereognathus ooliticus, Charlesworth.

The only specimen of this interesting Multituberculate
form has been purchased for the Museum of Practical Geology
(London), where it now rests. This fossil, a fragment of a
jaw containing three molars, came from the collection of the
Rev. J. Dennis, of Bury, and was originally described by
Charlesworth in 1854 (6a) as a piece of the lower jaw with
molars having “six similar cusps arranged in two rows”
(transverse). Three years later Owen described and figured
Stereognathus in detail (21a). The teeth, with their three
longitudinal rows of two cusps, he considers somewhat re-
sembled those of certain Ungulata, concluding that they
belonged to a “ diminutive form of the great Ungulate order
of Mammalia.’”’ Marsh (12a) suggested that the fossil belongs
to the upper jaw, as no Multituberculate teeth of the lower
series are known to possess more than two longitudinal rows
of cusps. The specimen is, however, too fragmentary to
enable us to come to any definite opinion on this point; and
its exact systematic position, therefore, must remain some-
what uncertain. It is generally classified in the family
Plagiaulacide (14, 32).

[! I take the opportunity of Mr. Goodrich’s publication to record once
again the existence of another Stonesfield jaw, which I obtained from a

FOSSIL MAMMALIA FROM THE STONESFIELD SLATE. 425

On tHE PrimitivE MAMMALIAN Mo.ar.

Before closing this paper I should like to make a few
remarks with regard to the “ Tritubercular theory,” which has
been so zealously put forward by several eminent American
palzontologists, and which has been so generally accepted in
Europe. There are two important questions involved in the
discussion: firstly as to the character of the primitive mam-
malian molar; secondly as to the origin and homology of the
particular cusps. As before, only the teeth of the lower jaw
will here be dealt with. Professor Osborn, in his illustrations
of the theory of the origin of the tritubercular molar, has made
large use of the Mesozoic mammals found in England ; one can
therefore stand on firm ground while criticising his conclusions
and his interpretations of the facts.

school-fellow thirty-two years ago. I took the specimen to Professor Huxley
at Jermyn Street, who cleared it from matrix and came to the conclusion
that it was a second example of Stereognathus. The jaw was a fragment,
but there were four molars present instead of three, the number in the type
specimen. My first appearance in the field of scientific literature was as the
author of a letter to the ‘ Geologist,’ vol. iv, 1861, recording the discovery of
this jaw. Much to my annoyance at the time my signature appeared as that
of a Mr. KH. Ray, residing at the town of Lanbeater. Whether such a town
exists, or whether the name was a pure invention on the part of the printer’s
devil, I have never ascertained.

Further disappointment fell on me in connection with this jaw. Before
Huxley had had time to figure or describe it, I took it home for a few days
in order myself to make a lithograph of it. The anguish and despair of a
schoolboy on finding that he had broken to powder the treasure the
detection of which had been the pride of his life may be imagined. The
crowns of the molars, so carefully cleared from the matrix by Huxley, were
rubbed from the specimen—wrapped though it was in cotton wool—in my
pocket. I carried back the now mutilated and comparatively worthless
specimen to Jermyn Street, and speechless placed it in Huxley’s hands, who
was only a little less grieved than I was. It was put aside in some cabinet
in the “den” then tenanted by the Naturalist to the Survey, and has never
been seen since, though searched for some twelve years later. There is a
possibility that in the course of time it may turn up either at Jermyn Street
or at the College of Science, South Kensington, whither my revered friend
and master—for so he became from the day when he took my Stonesfield jaw
in hand—migrated in 1870.—E. Ray LanxestTEr. |

426 E. S. GOODRICH.

It is weil known that Osborn (16, 17), Cope (7), and other
advocates of the theory assume that the primitive mammalian
molar was represented by a simple ‘‘reptilian cone,” the
** protoconid,” forming the Haplodont type of tooth. The
protoconid subsequently acquired a cusp in front and behind
(‘‘ paraconid” and “ metaconid” respectively), giving the Tri-
conodont type. This tooth with three cusps in a line would
then have become converted into the Tritubercular type of
molar by the movement outwards of the median large proto-
conid, and inwards of the anterior paraconid and posterior
metaconid. A small posterior “heel” then became developed,
which subsequently formed an external cone, the “ hypoconid,”’
and an internal cone, the “entoconid,” yielding the Tri-
tubercular sectorial type of molar seen in most orders of
Mammalia.

Let us now examine the facts. If the primitive mammalian
molars were simple cones, we should expect to find a gradual
approximation to this condition on comparing the more
specialised living mammals with the more primitive, the later
fossils with the earlier. But, as a matter of fact, we find on the
contrary that these simple molars are only seen in such highly
modified forms as the Cetacea, and that no early mammalian
fossils whatever have been yet discovered possessing them,
whilst multituberculate forms increase in number the lower we
search. Dromatherium and Microconodon, animals having
teeth with one large and several accessory cusps, presumed to be
intermediate between the Haplodont and the Triconodont types,
are possibly not mammals at all, but reptiles; at all events,
they differ so widely in structure from any known mammal,
living or extinct, that they can afford little certain information
on this question. The Reptilia themselves cannot be said to
support the theory, for the most mammalian of all known
reptiles, Galesaurus planiceps, has molars with three
distinct cusps. On the other hand, the most reptilian of all
living mammals, Ornithorhynchus, has multituberculate teeth.
Dealing next with the Triconodont type of molar, we find here
again that it does not occur amongst primitive forms, but in

FOSSIL MAMMALIA FROM THE STONESFIELD SLATE. 427

the highly developed Carnivora, and amongst the more spe-
cialised Phocide. These animals possess molars of a perfect
Triconodont pattern of obviously secondary origin.

Another point on which Professor Osborn lays great stress
is the origin of the Tritubercular from the Triconodont type
by the assumed movement outwards of the protoconid (the
median cusp), the paraconid and metaconid occupying the
inner angles of the “ primitive triangle” so formed. Phasco-
lotherium, Spalacotherium, and Amphitherium represent,
according to that author, three stages in the process. How-
ever, I can see no traces of the beginning of such a movement
in the first species. Spalacotherium has simple tritubercular
molars in which the heel is not developed,! somewhat similar
(as observed by Mr. Lydekker, 12) to those of the golden
mole, Chrysochloris inaurata, and leading perhaps to the
Stylodon type. As for Amphitherium, the molar shows no
trace whatever of being more primitive with regard to this
supposed movement of the protoconid ; the heel, although not
very large, is quite normally developed,’ like that of many
other Mesozoic, Tertiary, or receut forms.

The examination of existing forms only confirms these
results: cusps may disappear, and cusps may arise, but the
relative position of the protoconid and the two inner cusps is
always essentially the same. Embryology also shows that in
the development of molars the cusps arise in those positions
which they will occupy in the adult tooth (Topinard, 30;
Rose, 26).

Mr. W. B. Scott, in a quite recent paper (28), has clearly
shown that in the premolars the protoconid remains in place
while another cusp, the ‘ deuterocone,” is formed on its inner

1 There is no evidence that the teeth of Spalacotherium are of a primitive
intermediate type, even’ should they prove to have been evolved as suggested
by Osborn. The angle of the jaw of this genus has always been described
as inflected and confluent with the condyle; one specimen in the British
Museum, No. 47,799, shows that it was really separate and little inflected, if
at all.

2 The small cusp figured by Osborn in the molar of Amphitherium in front
of the paraconid (17) does not exist in any of the specimens.

428 HE. S. GOODRICH.

posterior surface. So averse, however, is he to damaging
Osborn’s theory, that he is actually driven to the conclusion
that similar cusps in precisely the same situations in the pre-
molars and the molars are not homologous, but of entirely
different origin !!

Another argument—perhaps the strongest of all—against
the assumption that the Triconodont tooth represents a stage
in the evolution of the Tritubercular sectorial type is afforded
by the consideration of the occurrence of the latter in the
various orders of Mammalia, and the probable phylogeny of
the Ditrematous mammals. That the Tritubercular sectorial
type was that of the lower molars of the ancestors of all the
placental mammals can scarcely be doubted in the face of the
mass of evidence collected by Cope, Osborn, Schlosser, and
others. All the various forms of molars met with amongst the
different orders, as adaptations to special food or methods of
feeding, can be referred back to this primitive type. Now the
Tritubercular sectorial molar occurs also amongst the Marsu-
pials; we must, therefore, conclude that the common ancestor
of both Placentals and Marsupials possessed teeth of this type ;
for we cannot assume without evidence that this complicated
yet definite pattern arose independently in the two cases: the
resemblance between the arrangement of the cusps in these
teeth in the two groups is not vague and general, but definite
and detailed.

We have, on the other hand, conclusive evidence that the
Triconodont type of tooth has independently arisen in at least
two widely separated groups, namely, the Phocide and Carni-
vora, and the carnivorous Marsupials (Phascolotherium, Thy-
lacinus, &c.) ; in the former certainly, and in the latter most
probably, by the reduction of the cusps of the primitively
Tritubercular sectorial molar.

1 Speaking of the Insectivora, Scott naively remarks that ‘in this group,
strange to say, the oldest member yet discovered exhibits the most com-
plicated premolar structure ” (28).

2 Dr. Forsyth Major (10) does not favour this view. However, even in

the teeth he figures, traces of the Tritubercular sectorial plan may be easily
detected.

FOSSIL MAMMALIA FROM THE STONESFIELD SLATE. 429

The evidence is, therefore, strongly in favour of the view
that the common ancestor of marsupial and placental mammals
had teeth with many cusps of the Tritubercular sectorial pattern.
What, then, was the pattern of the molars of the ancestors of
both Monotremes and Ditremes? As yet we can give no
definite answer to this question ; but one thing seems extremely
probable, namely, that they were of an indefinite multituber-
culate pattern, which gave rise, on the one hand, to the elabo-
rate multituberculate teeth,! and on the other to the Trituber-
cular sectorial. Thus the development of two longitudinal
rows of three cusps would give rise to the type of lower molar
common amongst the Multituberculata; the fusion of the two
anterior of these cusps or the loss of one would yield the Tri-
tubercular sectorial tooth common amongst the Marsupials
and Placentals; while the loss of the inner cusps would result
in the formation of a Triconodont molar. The conclusion
reached is, therefore, that the primitive mammalian molar
bore a crown with several cusps.

List oF REFERENCES.

1. Acassiz.—Short note in the ‘Neues Jahrbuch fiir Mineralogie und
Geologie,’ vol. iii, 1835.

2. BLAINVILLE, Henrnt Ducrotay pz.—* Doutes sur le prétendu Didelphe
fossile de Stonesfield, &c.,” ‘Comptes rendus Acad. Sci.,’ 1838.

8. Buanvitte, Henri Ducrotay pz.—* Nouveaux doutes sur le prétendu
Didelphe de Stonesfield,” ‘ Comptes rendus. Acad. Sci.,’ 1838.

4. Broperip, W. J.—‘ Observations on the Jaw of a Fossil Mammiferous
Animal found in the Stonesfield Slate,” ‘Zool. Journal,’ vol. iii,
1827-8. Also in the ‘ Annales des Sci. Nat.,’ vol. xiv, 1828.

5. Buckuanp, W.—‘‘ Notice on Megalosaurus,” ‘ Trans. Geol. Soc.,’ 2nd
series, vol. i, 1824.

6. BucxL~anp, W.—‘ Bridgewater Treatise,’ 1836.
6a. CuHaRLEswortH, E.—‘ British Association Report,’ 1854.

1 It must not be forgotten that these teeth have only been found in forms
with a reduced dentition.

430 E. S. GOODRICH.

7. Corr, E. D.—*‘ The Mechanical Causes of the Development of the Hard
Parts of Mammalia,” ‘Journ. of Morph.,’ vol. iii, 1889.

8. Cuvier.—‘ Ossements Fossiles,’ vol. v, 1824.

9. Dumurit.—‘ Comptes rendus,’ 1838.

10. Forsytu, Masor C. J.—‘“‘On some Miocene Squirrels, with Remarks
on the Dentition and Classification of the Sciurine,” ‘ Proc. Zool.
Soc.,’ 1893.

11. Grant.—‘ General View of the Characters and Distribution of Extinct
Animals,” ‘Thomson’s British Annual,’ 1839.

12. LypEKKER, R.—‘British Museum Catalogue of Fossil Mammalia,’
part v, 1887.

12a. Marsu, O. C.—‘‘ Notes on Mesozoic Mammalia,” ‘ Amer. Naturalist,’
vol. xxv, 1891.

13. Ocitpy, W.—“‘ Observations and Relations of the Presumed Marsupial
Remains from the Stonesfield Slate,” ‘Proc. Geol. Soc.,’ vol. iii,
1838-42.

14, Osporn, H. F.—‘‘ The Structure and Classification of the Mesozoic
Mammalia,” ‘Journ. Acad. Nat. Sci. Philadelphia,’ vol. ix, No. 2,
1888. Abstract of the above, ‘Proc. Acad. Nat. Sci. Philadelphia,’ 1887.

15. Osporn, H. F.—“‘ Additional Observations upon the Structure and
Classification of the Mesozoic Mammalia,” ‘Proc. Acad. Nat. Sci.
Philadelphia,’ 1888.

16. Ossorn, H. F.— Evolution of Mammalian Molars to and from the
Tritubercular Type,” ‘American Naturalist,’ 1888.

17. Osnory, H. F.—‘‘ History and Homologies of the Human Molar Cusps,”’
‘ Anat. Anzeiger,’ vol. vii, 1892.

18. Owrn, R.—“On the Jaws of the Thylacotherium Prevostii
(Valenciennes) from Stonesfield,” p. 5, and ‘ Description of the
Remains of Marsupial Mammalia from the Stonesfield Slate,” p. 17,
‘ Proc. Geol. Soc.,’ vol. ili, 1838.

19. Owrn, R.—‘‘ Observations on the Fossils representing Thylacotherium
Prevostii (Valenciennes), &c.,” and “Of the Phascolotherium,”
‘Trans. Geol. Soc.,’ vol. vii, 2nd series, 1842.

20. Owrn, R.—‘A History of British Fossil Mammals and Birds,’ 1846.

21. Owen, R.—‘ Odontography,’ 1840-5.

21a. OwEn, R.—‘‘ On the affinities of the Stereognathus ooliticus,
&e.,” ‘ Quart. Journ. Geol. Soc.,’ vol. xiii, 1857.

22. OwrEN, R.—‘ Paleontology,’ 1861.

23. Owrn, R.—‘ Monograph of the Fossil Mammalia of the Mesozoic
Formations,’ Paleontogr. Soc., 1871.

24,

25.

26.

27.

28.

29.

30.

31.

32.

FOSSIL MAMMALIA FROM THE STONESFIELD SLATE, 431

Paiuirs, J.—‘ Geology of Oxford and of the Valley of the Thames,’
1871.

Privost, C.—‘ Observations sur les Schistes calcaires oolitiques de
Stonesfield, &.,”? ‘Annales Sci. Nat.,’ vol. iv, 1825.

Rész, E.—* Uber die Enstehung und Formabinderungen des men-
schlichen Molaren,” ‘ Anat. Anzeiger,’ 1892.

Saint-HinairE, Grorrroy.—“ De quelques contemporains des Croco-
diliens fossiles,” ‘ Comptes rendus,’ 1838.

Scott, W. B.—“ The Evolution of the Premolar Teeth in the Mammals,”
‘Proc. Acad. Nat. Sci. Philadelphia,’ 1892.

SEELEY, H. G.—‘“‘ Note on a Femur and a Humerus of a Small Mammal
from the Stonesfield Slate,” ‘Quart. Journ. Geol. Soc.,’ vol. xxxv,
1879.

TorinarD, P.—“ Evolution des molaires et prémolaires chez les Pri-
mates,”’ ‘ Anthropologie,’ 1892.

VALENCIENNES, A.—‘“‘ Observations sur les machoires fossiles des couches
oolithiques de Stonesfield,” ‘ Comptes rendus,’ 1838.

ZitTEL, K. A.—‘ Handbuch der Palaeontologie, Palaeozoologie,’ Bd. iv,
1892.

VOL, 55, PART 3.—NEW SER. Ga

432 E. S. GOODRICH.

EXPLANATION OF PLATE 26,

Illustrating Mr. E. S. Goodrich’s paper “On the Fossil
Mammalia from the Stonesfield Slate.”

Reference Letters.

a. Angle. a.c. Articular condyle. c. Canine. c.p. Coronoid process.
*. incisor. m. Molar. m.g. Mylohyoid groove. pm. Premolar.

Fic. 1.—Left ramus of Amphitherium Prevostii, inner surface
exposed. Type specimen in the Oxford Museum. x 5.

Fie. 2.—Left ramus of Amphitherium Prevostii, inner surface
exposed. Second specimen in the Oxford Museum. x 5.

Fie. 8.—Right ramus of Amphitherium Oweni, outer surface exposed.
Type and only specimen in the Oxford Museum. xX 5.

Fic. 4.—Right ramus of Amphitherium Prevostii, inner surface
exposed. Inthe British Museum. x 5.

Fic. 5.—Left ramus of Amphilestes Broderipii, inner surface
exposed. Type specimen in the York Museum. x 5.

Fic. 5a.—Anterior portion of the same fossil seen more from above.
x 1b.

Fic. 6.—Left ramus of Amphilestes Broderipii, outer surface exposed.
First specimen in the Oxford Museum. To facilitate comparison the figure
has been reversed. X 5.

Fie. 7.—Right ramus of Amphilestes Broderipii, outer surface
exposed. Second specimen in the Oxford Museum. x 5.

Fic. 8.—Left ramus of Phascolotherium Bucklandi, inner surface
exposed. In the Oxford Museum. xX 3.

Fic. 9.—Right ramus of Phascolotherium Bucklandi, inner surface
exposed. In Mr. Parker’s collection. x 3.

The lines above the figures indicate the actual length of the fossils.

A POLYNOID WITH BRANCHIZ. 433

A Polynoid with Branchie (Eupolyodontes
Cornishii).

By

Florence Buchanan, B.Sc.

With Plate 27.

A SINGLE specimen of an interesting Polychete was pre-
sented to the British Museum a short time ago by Mr. V. H.
Cornish, of the cable-ship “ Mirror,’ who had obtained it off
the mouth of the river Congo. It wasshown to me by Professor
Bell, who was good enough to suggest that I should describe
it, and Dr. Giinther has kindly sanctioned my doing so.

The worm is evidently a Polynoid belonging to the sub-
family Acoétide. It is remarkably large even for that sub-
family, the specimen, although incomplete, measuring over a
foot in length, its breadth exceeding one and a half inches,
and its depth being nearly half an inch. It is, unfortunately,
somewhat mutilated, the alimentary canal being torn out, so
that the pharynx, which is a characteristic feature of the group,
cannot be diagnosed. The head and greater number of seg-
ments present are, however, complete externally, and while
showing clearly the genetic position of the worm, present also
characters of interest only slightly developed in other members
of the group, and which therefore have not hitherto received
sufficient attention.

Before describing the worm itself I think it will be advisable
to review briefly the characteristics of the sub-family, and to
enumerate the few known species belonging to it, especially as

AB34. FLORENCE BUCHANAN.

most of them are described in scattered journals, and some
have been overlooked by later writers on the group.

The Acoétidx may be defined as elongate Polynoids, with the
elytra alternating regularly with dorsal cirri throughout the
body, except for the second and third pairs which are on
consecutive segments, the 4th and 5th respectively. The dorsal
surface of the body is generally transversely grooved, the grooves
being very fine and close together, often quite obliterating the
segment boundaries. The prostomium bears two large pupil-
lated eyes, generally on well-developed peduncles ; there may
be in addition smaller eyes or pigment spots behind them.
There is a single median prostomial tentacle and a pair of
lateral ones: the former is sometimes rudimentary or even
absent ; when present it springs from the posterior part of the
prostomium. The paired prostomial tentacles are also occa-
sionally absent ; when present they generally arise from the
ventral surface of the prostomium. ‘There is a pair of palps,
usually very large and well developed. The parapodia of the
buccal segment have moved forward so as to lie in front of the
mouth; they consist each of a basal part bearing two peri-
stomial tentacles corresponding to the dorsal and vental cirri
of the other parapodia, and sometimes also bearing chete.
The parapodia of the following segments are either uniramous
or biramous with the notopodial lobe very small ; each one con-
tains a much-coiled dorsal chetal sac, the ‘‘ spinning gland ”
of Eisig (6), producing numerous exceedingly long, fine, silky
capillary chete, which probably help to form the tube in
which the creature lives ; occasionally, however, the sac may be
shorter, and the chet produced in it more like ordinary chet,
projecting from the sac instead of being kept insideit. There
is a median ventral longitudinal ridge protecting the nerve-
cord; it is bounded on each side by a deep furrow, and widens
in front just behind the mouth. The pharynx is exsertile,
papillose on the anterior margin; the jaws large and horny,
armed with two central and many lateral teeth.

The known species of the sub-family are only fifteen in
number, and of most of these only single and incomplete speci-

A POLYNOID WITH BRANCHIA. 435

mens have been seen; only two, the Mediterranean Polyo-
dontes maxillosus and the Northern Panthalis Oerstedi,
have been found by more than one observer, but even they
are not very abundant. I will enumerate the species in the
order of their foundation :
1817.—Polyodontes maxillosus (Ranzani), Audouin and
Edwards. Mediterranean.

Described and figured by Claparéde (4), who
gives its synonymy and refers to previous descrip-
tions and figures.

1832.—Acoétes Pleei, Audouin and Edwards. Mar-
tinique.

Single specimen, described and figured by the
founders (2), and further described by Quatrefages
(15). Grube (8) refers it to the genus Polyo-
dontes.

1841.—Polyodontes Blainvillei (Costa), Claparéde.
Mediterranean.

Single specimen, imperfectly described by Costa
(5), who calls it a “Sigalion.” Referred to the
genus Polyodontes by Claparéde (4).

1855.—Polyodontes gulo, Grube. Red Sea.

Single specimen, described and figured by the
founder (8).

1855.—Eupompe Grubei, Kinberg. Near Guayaquil.
Single specimen.
Panthalis Oerstedi, Kinberg. British and Scan-
dinavian coasts.
Panthalis gracilis, Kinberg. Near Rio Janeiro.
Single specimen.
All three described (11 and 12) and figured (12)
by their founder.
1855.—Acoétes lupina, Stimpson. South Carolina.
Imperfect description by Stimpson (16).
1876.—Eupanthalis Kinbergi, McIntosh. Adventure
Bank.
Described and chete figured by the founder (18).

436 FLORENCE BUCHANAN.

1877,—Panthalis bicolor, Grube. Congo.
Two specimens differing greatly from one another,
described but not figured by Grube (9).
1878.—Panthalis melanotus, Grube. Philippine Islands.
Panthalis nigromaculata, Grube. Philippine
Islands.
Described but not very well figured by Grube
(10).
1885.—Eupompe australiensis, McIntosh. Off Cape
York, Australia. ;
Described and figured by the founder (14).
1887.—Euarche tubifex, Ehlers. Off Carysfort Reef,
West Indies.
Described and figured by founder (7).
1887.—Eupompe indica, Beddard. Mergui Archipelago.
Described and head figured by the founder (8).
The new species which may now be added to the list bears
most resemblance to the Polyodontes gulo described by
Grube (8), and it has, indeed, certain characters in common
with it in which they both differ from all the other known
species. Like P. gulo and no other member of the group
there are no long well-developed palps,! and the eye peduncles
are lateral instead of being anterior, and fused with the sides
of the prostomium, thus giving the prostomium a very broad
appearance (figs. 1 and 2). The paired prostomial tentacles,
when present, in all other Acoétide with pedunculate eyes,
arise from the ventral surface of the prostomium, or rather
from the base of the anteriorly placed eye-stalks, and just in
front of the palps (cf. fig. 10). Here and in P. gulo there
are two small tentacles springing from the anterior (and slightly
ventral) surface of the prostomium, which probably represent
them (figs. 1 and 2, ¢.). Behind these (fig. 2) and springing
from the base of the laterally placed eye-stalks are two other
very minute tentacles, which probably represent the palps of

1 McIntosh does not mention the palps at all in his Eupanthalis, but I
- conclude that he would have done so had they been greatly reduced in size
or absent,

A POLYNOID WITH BRANCHIA. 437

the other species although extremely reduced. Their relation
to the eye-stalks suggests for a moment their homology with
the paired prostomial tentacles (‘‘ antenne ” of authors) rather
than with the palps, but if this were the case we should have
not only to regard the palps as altogether absent, but we
should also have to explain the presence of an extra pair of
prostomial tentacles in front with no homology in other forms.
A comparison of the arrangement of the different prostomial
appendages! in the sessile-eyed forms shows that there also,
as in P. gulo and the new worm, the paired prostomial ten-
tacles arise close to the anterior edge of the prostomium, while
the only other paired prostomial appendages, the palps, arise
close behind them and are developed to their usual extent.” I
think, therefore, that we may conclude that the relation of any
of the prostomial appendages to the eye-stalks is a secondary
one, while their relation to the prostomium is constant.? The
parapodia of P. gulo are not figured, but from the description
they seem to resemble in arrangement those of the new
species.

The only characteristic points of difference between P. gulo
and the new worm is that while in P. gulo there is no trace of
a median prostomial tentacle here there is one, although only
a very rudimentary one ; and that the few dorsal papillze on the
parapodia of P. gulo, some of which are described as elon-
gated to cirri, are here enormously developed and very numer-
ous and arborescent, resembling in appearance the branchiz
of other Polychetes. Both these points, however, seem to me
to be only of specific importance, since they are characters
which vary also in other members of the group. While the
other characters of the prostomium, so much alike in these
two species, but differing so markedly from all the other forms,
seem to mark them off from all the others as a separate genus,
for which I propose the name Eupolyodontes, calling the

1 «© Prostomial appendages ’”’ = 1 median and 2 paired “ prostomial ten-
tacles” + 2 “ palps.”

2 Compare Ehler’s figure of head of Huarche tubifex (7).

* Not wishing to spoil the specimen I was unable to examine microscopi-
cally the structure of the different pairs of prostomial appendages,

438 FLORENCE BUCHANAN.

new species, after the name of its discoverer, E. Cornishii. I
would give the following as a definition of the genus and
of its contained species :

1. Genus EvpotyopontTEs.

Acoétide with peduncles of eyes arising laterally from the
base of the prostomium, and fused with it on either side; short
antenne or paired prostomial tentacles arising from the ante-
rior margin of the prostomium or slightly ventral to it; median
prostomial tentacle rudimentary or absent, arising from the
posterior part of the prostomium when present; palps small,
no longer than the antenne, situated very close to or on the
bases of the eye-stalks. Dorsal surface of body very finely
rugate transversely and segment boundaries thus obliterated.
Parapodia with papille on the dorsal surface, which may be
filamentous or even arborescent. Parapodia of buccal segment
not chetiferous.!

Sp. 1.—E. gulo, Gr. [Polyodontes gulo, Grube (8), ‘Arch.
f. Naturg.,’ xxi].

Eupolyodontes with antenne arising from the anterior edge
of the prostomium, but with no median prostomial tentacle.
Parapodia with minute papillze, sometimes elongated, two to
five in number on the elytra-bearing segments, six or seven on
the others. Only one acicle to each parapodium ; chete of three
kinds :—a comb of short stiff chetz slightly curved at apex, a
fine bundle of bipinnate ventral chet, and a dorsal bundle of
long delicate capillary chete, forming the thick silky thread
of the ‘‘ spinning gland.” Hab. Red Sea. Living in tubes.

Sp. 2.—E. Cornishii, n. sp.

EKupolyodontes with a minute median prostomial tentacle
situated on the posterior part of the prostomium, and just in
front of a slightly raised part of the back which forms a kind

’ This may turn out hereafter, when new species are discovered, to be only
of specific value.

A POLYNOID WITH BRANCHIA. 439

of “caruncle.” Prostomium slightly bilobed, paired pro-
stomial tentacles or antennz arising one from each lobe just
below the anterior edge ; palps smaller than the antenne, each
with a minute swollen basal piece. Parapodia, both those
bearing elytre and those bearing dorsal cirri, with a very
large number of filamentous and arborescent branchia-like
looking structures along the anterior and posterior border of
each, beginning on the posterior border of the 6th chetiferous
parapodium, where there is only a single bifurcate filament ;
three segments further back they are present on both anterior
and posterior border of the parapodium, and are already
numerous ; they increase in number and size and amount of
branching for the next few segments, and are best developed
on the parapodia of the 15th to the 50th segments ; they then
decrease in number and size, and become more papilliform.
In structure each filament is hollow, its cavity being probably
an extension of celom; the cells of the epidermis are laden
with yellow granules, which look like excretory products, and
there is a very thick cuticle. Only one acicle to each para-
podium. Chetz as in E. gulo (see figs. 8B and 8c), with in
addition certain ones with double-brush shaped tips! (fig. 8 a),
scattered amongst the comb of stiff chetz dorsally. The
“spinning gland” is well developed in every segment after the
first few, being long and coiled and occupying the cavity of the
parapodium, and opening on its dorsal surface; the long fine
capillary chetz produced by it are of a silky golden colour,
generally retracted but readily drawn out an inch or two (fig. 4,
cap). Elytrasmooth, anteriorly flat, those of the 2nd parapodium
(the first pair) overlapping one another, but the rest well to
the side, leaving the whole of the dorsal surface except for the
parapodia exposed, small (relatively to the size of the animal)
and scarcely imbricate ; posteriorly they are swollen and pear-
shaped, each being attached to the parapodium by a stalk
(fig. 5). Dorsal cirri short (no longer than the branchize
where these are well developed), outside the elytra. Ventral

1 Resembling those of Hupompe Grubei, Kinberg (12), more than those
of any other of the species of which the chete have been figured,

4.40 FLORENCE BUCHANAN.

cirri rather shorter, those of the 2nd pair of parapodia being
larger than the rest.

Pharynx and jaws not present in the specimen.

Colour (in spirit): of the eye-stalks dark blue-black ; of the
prostomium itself dark, but not quite so dark ; of the dorsal
and ventral surface of the body dark brownish, the parapodia
somewhat lighter, and the ventral ridge below the nerve-cord
also of a lighter colour ; ‘‘branchiew” darker in colour than the
rest of the parapodium.

Length of single specimen, consisting of ninety-two seg-
ments, but incomplete posteriorly, 32°5 cm.; breadth, including
parapodia, 4°2 cm.; of the dorsal surface of the body alone
2 cm.

Hab.—Single specimen, obtained off the mouth of the river
Congo, about thirty-five miles from land, at a depth of from
forty-three to fifty-seven fathoms, from a bottom of mud and
weed. The colour of the water where it was taken was of a
uniform reddish orange.

It is probably tube-forming, although no tube was found with
it. The various points are illustrated in the figures (1—8).

With regard to the other fourteen species of the sub-family
(or rather thirteen, as the Acoétes lupina of Stimpson is
probably the same as A. Pleei), reference to the list given
on p. 435 will show that six genera have been formed for them.
Grube (8) has long ago disposed of one of these by placing
Acoétes Pleei in the genus Polyodontes. Beddard (8)
has recently proposed to throw the genera Eupompe and
Panthalis into one. I agree with him, but would go further,
and place provisionally both these genera in one genus with
Polyodontes, bearing in mind that closer acquaintance with
the different species and the discovery of new ones will pro-
bably lead hereafter to a new subdivision into genera, but
probably not—it seems to me, at least—coinciding with what
we now know as the genera Polyodontes, Eupompe, and Pan-
thalis. ‘The number of genera needed in any group of animals
depends entirely on which different forms and how many of
them happen to be known at the time. When only three

A POLYNOID WITH BRANUHIA. 441

species! of this sub-family were known, all about equally dis-
tant from one another, it was quite enough to have only one
genus for them all, as Grube proposed. But the greater the
number of species made known the less likely are they to
remain equally distant from one another, and they then fall
naturally into groups, only to be reunited when all the inter-
mediate forms are known. In my opinion the sub-family of
the Acoétide falls now, in the present state of our knowledge,
or rather of our ignorance, into three groups, which we may
call genera. One of these, the genus Eupolyodontes, I
have already defined. The other stalk-eyed forms (Polyo-
dontes maxillosus and Blainvillei; Acoétes Pleei;
Eupompe Grubei, australiensis, and indica; Pan-
thalis Oerstedi, gracilis, melanotus, and [in part] bi-
color) I would propose to put together in the genus Polyo-
dontes, defining this genus then in its widest sense as follows:

2. Genus PoLyopontzs.

Acoétide with peduncles of eyes arising from the front of
the prostomium, and meeting, or nearly meeting, one another
in the middle line in front; median prostomial tentacle well
developed, paired ones present in all except P. (E.) indica
aud P. (P.) melanotus, and arising from the ventral surface
of the prostomium at the base of the eye-stalks; palps large
and well developed, arising close behind the paired prostomial
tentacles ; papille sometimes present on the parapodia, but
not developed to any great extent (represented in P. (A.)
Pleei, P. (H.) australiensis, P. (E.) Grubei [on elytra-
bearing feet only], and P. (P.) bicolor ??*). Parapodia of
buccal segment sometimes chetiferous (at least in P. maxil-
losus, P. (A.) Pleei, P. {E.) Grubet, and P. (P.) Oerstedi;
but not in P. (E.) indica, P. (E.) australiensis, or P. (P.) bi-
color; in the other species the fact is not mentioned either way).

I have the less hesitation in placing these species of different

1 At present, in this sub-family, it is easy to speak of these “ different
forms ”’ as “ species.”

2 If I understand Grube’s description aright they would be here on the
ventral surface of the parapodia, and not on the dorsal,

442 FLORENCE BUCHANAN.

genera together into one genus, as I have been able to compare
one species of Eupompe (EH. australiensis) which was in
the British Museum with a type specimen of Polyodontes
maxillosus which Professor Bell kindly procured for me from
Naples. Although the last-mentioned worm has been several
times described, none of the figures of its head show very well
the relations of the tentacles to the prostomium, and I have
therefore figured the head from above and below (figs. 9 and 10).
The remaining species of the Acoétide (Eupanthalis
Kinbergi, Euarche tubifex, and (?) Panthalis nigro-
maculata, and (?) part of P. bicolor) I would place pro-
visionally, but only provisionally, in a third genus, which
would bear the name Eupanthalis, defining it as follows:

38. Genus EvUPANTHALIS.

Acoétide with sessile eyes, four in number; three prostomial
tentacles, except (?) in E. tubifex;! otherwise like Polyo-
dontes.

Although it seems simplest to make one genus for all the
sessile-eyed forms, I have a good deal of hesitation in doing so
on account of Grube’s description (9) of what he calls two
forms of Panthalis bicolor, coming, by the way, from the
same locality as the specimen sent by Mr. Cornish. Grube’s
two specimens agree in colour, and the parapodia are alike;
moreover he found them in the same bottle, which he seems
to think important. But while in the one the eyes are
pedunculate and apparently anterior, the palps very large,
the paired tentacles beneath the median one and the elytra
large, in the other the eyes are sessile, the palps shorter, the
paired tentacles on the front margin of the prostomium and
the elytra much smaller. Grube has already remarked that it
would be very strange and quite unheard of in this family of
Polychetes to find such very different forms of a single species,
and he is not quite convinced of it himself. If it were so it

1 McIntosh’s remark that there is “no” median “tentacle in the spe-

cimen ” seems rather to imply that there may once have been one which has
been lost by accident.

A POLYNOID WITH BRANCHIA. 443

would be exceedingly interesting, as it would suggest that
other sessile-eyed forms might be but second forms of other
species with pedunculate eyes; but I think evidence is wanting
of the fact that the two specimens described by Grube as P.
bicolor do really belong to the same species. Unfortunately
neither of them is figured at all. Panthalis nigromaculata,
which I have also placed with a (?) in this genus, would appear
from Grube’s figure of its head to have quite sessile eyes. In his
description of it, however, he speaks of them as on protuberances.

Besides throwing light on the intrinsic relationships of the
sub-family, the new worm also, it seems to me, increases the
probability of the existence of a relationship between the
whole family of the Polynoidz and the family Amphinomide.
The Acoétide, in common with certain other sub-families of the
Polynoide, resemble the Amphinomide in the forward move-
ment of the first pair of parapodia. The new Acoétid re-
sembles them further in another peculiarity of the head which
I have already mentioned, namely, the ridging of the dorsal
surface of the head behind the median tentacle. The resem-
blance may be only superficial, but one is certainly reminded
by it at once of the “caruncle” of the Amphinomide, which
is sometimes little more than a raised part of the dorsal sur-
face of the head. Another and more striking point of resem-
blance, at first sight at least, is the presence of the arborescent
or filamentous, branchia-like looking structures on the para-
podia, and this brings me to what I consider the most in-
teresting point about the new worm. I have already men-
tioned the position of these filaments and referred briefly to
their structure in diagnosing the species. Their relation to
the parapodium is shown in figs. 4 and 5, a single tuft of
them in fig. 6, and a transverse section of one of them in
fig. 7. The state of preservation they were in makes their
minute structure difficult to interpret, and I cannot be at all
certain whether the central cavity is really an extension of
celom or a large blood-vessel,—that is to say, whether there
is a true space between the epidermis and wall of this central
cavity or not. I am inclined to think that there is no cavity

444, FLORENCE BUCHANAN.

in the filament besides the central cavity, and that there is
connective tissue between this and the epidermis which has
not been preserved, except for a few nuclei (fig. 7 c.¢.). The
space would then be extension of ccelom, and I believe it to be
lined by a definite epithelium, although the nuclei indicating
this are few and far between. (One is shown at n. in fig. 7.)
The clot inside the cavity is more like a celomic clot than a
blood clot. If this central cavity be ceelom, I cannot be certain
of there being blood-vessels going to the filaments at all (unless
certain small structures, seeming to lie in the wall of the
central cavity and marked “ d/. (?)” in fig. 7, represent them),
aud the filaments cannot be termed “branchie” in the
ordinary sense of the word. The extreme thickness of the
cuticle would also seem to indicate that their function is other
than respiratory, and the peculiar character of the epidermis
helps to show what this function is. Although, owing to the
method of preservation, it is scarcely possible to distinguish
cell outlines, nuclei of the epidermis cells are here and there
visible, and grouped around them and apparently densely
loading all the epidermis cells are numerous yellow concre-
tions, some of them refringent, others with a somewhat darker
appearance, and often massed three or four together. These
resemble so closely the concretions of nephridial cells and of
the cells of other renal organs described by Eisig in the Capi-
tellide, and behave in the same way towards chemical reagents
in as far as I have been able to test them, that I think there
can be little doubt of their excretory significance. LHisig has
shown how in the genus Capitella, where the nephridia appear
not to open to the exterior at all, the excretory products are
stored in the epidermis cells, only to be got rid of when the
animal changes its skin, and, as is well known, numerous
Arthropods normally store their excretory products. The fila-
ments, then, on the parapodia of Eupolyodontes Cornishii
would seem to be special organs for storing the excretory pro-
ducts, and perhaps also for forming them.!

1 As far as I am aware nothing is known about nephridia or excretory
organs of any sort in the sub-family Acoétide.

A POLYNOID WITH BRANCHIA. 44.5

In spite, however, of their being so unlike respiratory organs
in structure, their outward resemblance to the “ branchie” of
other Polychztes, and especially to those of the Amphinomide,
struck me so forcibly that I was led also to examine micro-
scopically the structure of these for the sake of comparison ;
and the results are, I believe, sufficiently interesting to warrant
a mention of them here, although I must defer a more detailed
description and more numerous figures to a future publica-
tion.

I have examined by means of sections the branchie of a
Euphrosyne, of two or three Amphinomes, of Chloeia flava,
of Eunice gigantea, Diopatra neapolitana, Arenicola
marina, and a few others. The thickness of the cuticle,
although most marked in the Amphinomids, is remarkable in
all. In none of them nor on any part of them is the epidermis
ciliated. Very minute concretions, nothing like so large as in
Eupolyodontes Cornishii, are present in the branchiz
of the Euphrosyne, one of the Amphinomes, in Areni-
cola marina, and, although here they are present in other
parts of the epidermis as well, in Eunice gigantea. Cla-
paréde (4, p. 110) has already remarked on the thickness
of the cuticle, and the absence of blood-vessels and of axial
cavity in the branchie of Euphrosyne Audouini, and
speaks of them throughout as ‘“ prétendues_branchies.”’
Schmarda shows, however, that in E. polybranchia there
is a vascular network penetrating into the final ramifications,
and in the Euphrosyne of which I cut sections, and which I
believe to be E. borealis, there were certainly two vessels
traversing the main stem of each branchia, breaking up into a
capillary network in the filaments. We have, then, within
the same genus forms with vascular and with non-vascular
* branchie.” In most of the Amphinomes of whose branchize
I cut sections the filaments appeared also to be solid. There
was, however, in one of them at least (fig. 12) a central part
very little blocked up by connective tissue. Between this and
the epidermis is retiform connective tissue (c. ¢.), and in this on
either side of each filament is a blood-vessel (4/.), giving off

446 FLORENCE BUCHANAN.

numerous branches all lying in the connective tissue. Only
quite at the extremity of the filaments the connective tissue
and vessels seem to have disappeared altogether. In Chloeia
(fig. 13) the two vessels have increased enormously in size, and,
except for being connected with one another at intervals at the
tip of each filament, give off no branches. No central cavity
is distinguishable, all the space which is not blood-vessel under-
neath the epidermis being occupied by retiform connective
tissue (c. ¢.). In the other Polychetes examined the branchiz
were more normal in structure, containing an afferent and
efferent vessel lying close under the epidermis, but in a well-
developed extension of ccelom.!

From the above facts I conclude that the so-called “ bran-
chie” of polycheetes do not necessarily serve only as respiratory
organs, and indeed may even have no respiratory function at
all (some species of Euphrosyne) ; and in the sense that we call
them “ branchie,’’ on account of their representing the respi-
ratory organs of allied forms, I claim to be able to apply the
same term to the branching processes on the parapodia of
Eupolyodontes Cornishii, When they are not respiratory,
or at least not mainly respiratory in function, they may have to
do with excretion, serving to store the excretory products, and
probably, in the case of Amphinome at least, and those forms
with blood-vessels immediately underlying the epidermis and
with concretions in the epidermis cells, to form them from
the blood.

In conclusion, I should like to draw attention to the minute
structure of the filaments on the parapodia of the only other
Acoétid possessing them which I have been able to examine,
namely, Polyodontes(Eupompe)australiensis. Although
I think there can be no doubt, from their position in relation
to the parapodium, as to their representing the more numerous
filaments on the parapodium of Eupolyodontes Cornishii,

1 Only in the Diopatra it was difficult to be certain of the blood-vessels, as
the blood did not clot at all, and the two vessels in each of the filaments, each
of them subdivided by connective-tissue partitions, would not be taken for

blood-vessels, were it not for Claparéde’s statement that there is an afferent
and efferent vessel in each filament.

A POLYNOID WITH BRANCHIA. 44,7

their structure is very different, as will be seen by comparing
the figures of two sections through one of them (figs. 11 a and 8)
with fig. 7, There is apparently no central cavity, nor is there
anything looking at all like blood-vessel; the epidermis cells
are flattened and contain no concretions: the only point of
resemblance is the thickness of the cuticle. The substance of
the filament near the tip seems to consist of concentrically
arranged connective-tissue fibres, in which lie a few large
clear cells with large distinct nuclei. Five of these (fig. 11)
are arranged radially round a common centre, and their
appearanceis extremely suggestive of the so-called “gill-glands”’
recently described in a Crustacean by Mr. Allen (1) (where,
by the way, we have also an instance as shown by Kowalewsky
of a branchia exercising some excretory function besides its
normal function). It is true there is nothing to be seen here
representing the duct described by him ; but one could perhaps
scarcely expect to find it, even if present, in material not pre-
served with a view to histological work, and also the plane of
the sections might not be favorable for showing it. Nearer
the base of the filament (fig. 114) the whole space beneath the
epidermis seems to be occupied by retiform connective tissue,
except for a curious mass of what seem to be concentrically
arranged connective-tissue fibres near the centre. The struc-
ture of these filaments bears most resemblance to that of
the branchize of Euphrosyne—in as far as I have been able to
examine them—amongst Polychetes. Here also, in the “ E.
borealis (?) ” at least, we have similar large cells embedded
in connective tissue near the apex of the filament, although
not radially arranged round a common centre as in the
Eupompe filament, and the large cells in a special swell-
ing at the apex of the branchial filaments of so many Euphro-
synes (including E. Audouini) are well known, although
I do not know that sections of them have ever been described.
The rest of the substance of the branchial filament is occupied
by connective tissue, in which, however, in E. borealis (?)
there are wall-less blood-vessels, though apparently there are
not even these in HE. Audouini.
VOL, 39, PART 3.—NEW SER. H H

448 FLORENCE BUCHANAN.

It would be interesting if some one within reach of the other
specimens of the sub-family Acoétide with “branchial”
filaments would examine and report on their structure.

List or MEMOIRS REFERRED TO.

(1) Atten.—“‘On the Minute Structure of the Gills of Palemonetes
varians,” ‘Quart. Journ. Mier. Sci.,’ vol. xxxiv, pp. 75—84.

(2) AupouiIn and Epwarps.—“ Classification des Annélides,” ‘ Annales
des Sciences naturelles,’ le sér., xxvii, 1832, pp. 435—438, pl. x, figs.
7—14.

(3) Brpparp.—“ Report on Annelids from the Mergui Archipelago,”
‘Journ. Linn. Soc.,’ xxi, 1887 (1889), pp. 256—258, pl. xxi,
figs. 1 and 3.

(4) Craparnpr.—‘ Les Annélides Chétopodes du Golfe de Naples,’ 1868,
pp. 892—396, pl. iii, fig. 2.

(5) Costa.—*‘ Description de quelques Annelides nouvelles du golfe de
Naples,” ‘ Ann. Sci. Nat.,’ 2e sér., xvi, 1841, p. 269, pl. xi, fig. 1.

(6) Htstc.—‘ Monographie der Capitelliden.’

(7) Huters.—‘ Results of Dredging of the U.S. Coast-Survey Steamer
‘Blake,’ ‘ Report on Annelids,’ 1887, pp. 54—56, pls. xii and xiil.

(8) Grupe.— Beschreibung neuer oder wenigbekannter Anneliden,”
‘Arch. f. Naturg.,’ 2lter Jhrg., 1855, pp. 83—90, pl. iii, fig. 2.

(9) Grube.—‘‘ Anneliden Ausbeute S8.M.S. ‘ Gazelle,’ Monatsber.,” ‘ Ber-
liner, Akad.,’ 1877, pp. 517—519.

(10) Gruzse.—“* Annulata Semperiana,” ‘ Mém. Acad. St. Pétersbourg,’
7e sér., xxv, 1878, pp. 48—52, pl. iv, figs. 1 and 2.

(11) Krvzere.—‘ Ofvers. af. k. Vet. Akad. Fish,’ 1855, pp. 386, 387.

(12) Kinprerc.— Fregate Hugenies Resa,” ‘Zool. Annulater,’ pp. 24—26,
pl. vii, figs. 34 and 35 (published part); pl. x, figs. 59—61 (unpub-
lished part).

(18) McInrosu.— Annelida of the ‘ Porcupine’ Expedition,” ‘ Trans. Zool.
Soc.,’ ix, 1876, pp. 404, 405, pl. Ixxii, figs. 12—15.

(14) McIntosu.—“ ‘Challenger’ Reports,” ‘ Zoology,’ xii, pp. 135—139,
pl. xxi, figs. 4 and 5.

(15) QuatrEeFaGcES.—‘ Histoire naturelles des Annélés,’ i, p. 216.

(16) Srrupson.—‘ Synopsis of Marine Invertebrates of Grand Manon,”
‘Smithsonian Contributions to Knowledge,’ Washington, 1853, p. 36.

A POLYNOID WITH BRANCHIA. 44,9

EXPLANATION OF PLATE 27,

>

Illustrating Miss Florence Buchanan’s paper, “ A Polynoid
with Branchie (Eupolyodontes Cornishii).”

Fie. 1.—Dorsal view of head! and first ten segments of Hupolyodontes
Cornishii. The elytra are represented as turned aside from their normal
position to show the parapodia. Very slightly enlarged only. ¢. One of the
paired prostomial tentacles.

Fie. 2.—Ventral view of head and first five segments.

Fie. 3.—Dorsal view of three segments some five or six inches from the
anterior end. The elytra left in their normal position on the left side only.
Turned aside to show the “branchie ” on the right.

Fig. 4.—View from bebind of a non-elytriferous parapodium from about
the 50th segment. d.c. Dorsal cirrus. ac. Acicle, seen projecting behind
the cut end of sp. g/., the spin-gland. d.s. Dorsal cheta. sp. Spines or
short stiff chete of comb. v.s. Ventral chete bundle. cap. Silky thread
formed of capillary chete of spin-gland, drawn out from the aperture of the
gland.

Fic. 5.—Similar view of an elytriferous parapodium taken from a more
posterior segment, showing the swollen elytron. The branchial filaments have
become much fewer in number.

Fie. 6.—A single “branchial” tuft from a segment where they are well
developed.

Fic. 7.—Transverse section of one of the “branchial” filaments. c. Cuticle.
ep. Ejpidermis laden with concretions. ¢.¢. Nucleus of connective tissue (?).
c.cav. Central cavity. 2. Nucleus in its wall. 47. (?) A blood-vessel (?).

Fie. 7A.—A portion of the epidermis of the same section, enlarged to show
the nucleus (z.) of an epidermis cell.

Fic. 8a.—Tip of a dorsal cheta.

Kic. 88.—Tip of one of the spines.

Fic. 8c.—Tip of one of the cheetee of ventral bundle. There is not always
such a marked difference at the apex as there is in the one here figured.

Fic. 9.—Dorsal view of head and first segment of Polyodontes maxil-
losus, much enlarged.

1 “Head” here is used to include the prostomium and first or buccal
segment which is fused with it.

450 FLORENCE BUCHANAN.

Fic. 10.—Ventral view of head and first four segments of the same. Palps
turned aside to show the underlying paired prostomial tentacles.

Figs. 11a and 113.—Two transverse sections of a “ branchial” filament of
Eupompe australiensis.

Fic. 12.—Transverse section of a branchial filament of an Amphinome
(Eurythoé). 47. Blood-vessels. Other letters as in Fig. 7.

Fic. 13.—Transverse section of one of the filaments of a branchia of
Chloeia flava.

SOME BIPINNARIA) FROM THE HNGLISH CHANNEL. 451

On some Bipinnarie from the English
Channel.

By

Walter Garstang, M.A.,
Fellow of Lincoln College, Oxford, Naturalist to the Marine Biological
Association.

With Plate 28.

Tue genus Bipinnaria was defined by the elder Sars (3)
in the following terms :—“ Corpus gelatinosum longum cylin-
drico-depressum, pinnis duabus, una postice terminali cordi-
formi, altera triangulari in medio corpore. Os appendiculis
seu brachiis lanceolatis circumdatum.” To this genus he
referred the single species asterigera, “ appendicibus seu
brachiis 12 circa os.”

Few zoologists meeting with this passage for the first time
would, I imagine, recognise in it the description of a starfish
larva; for the Bipinnaria of the text-books is invariably of
that simpler, commoner, and smaller type which formed the
basis of Johannes Miiller’s classical researches. The original
Bipinnaria of Sars was a remarkably elongated creature,
fully one inch in length, with a crown of polyp-like tentacles
at the oral end of the body, a bilobed fin at the other extremity,
and a median ventral fin placed transversely ; attached to the
tentaculate end of the body was a small five-rayed starfish.
Sars himself was so puzzled by the animal that he referred it,
not without some natural hesitation, to a special group of
Acalephe.

452 WALTER GARSTANG.

In October, 1846, a swarm of these Bipinnariz visited
Bergen harbour, and formed the material for some researches
by Drs. Koren and Danielssen (1). These naturalists pointed
out the real nature of the connection between the starfish and
the Bipinnaria. Two preserved specimens of these Bipin-
nariz subsequently came into the hands of Johannes Miller,
and an account of them is included in the second of Miiller’s
memoirs (2).

So far as I am aware, no specimens of this interesting larva
have been captured since 1846. During the month of August
this year, however, I have taken near Plymouth a number of
Bipinnaria larve which resemble Sars’s larva in many points.
At the same time they show some differences which seem to
be important. It may eventually be proved that these dis-
similarities are due to a different stage of development ; or, on
the other hand, the Plymouth larve may be shown to belong
to a type of Asterid distinct from that to which Sars’s larva
belongs. Up to the present time, however, I have not suc-
ceeded in finding specimens in which there is any trace of
approaching metamorphosis, so that the identification of the
larva has not been possible.

1. Structure.

A glance at the figures on Plate 28 reveals at once the
fact that the Plymouth larve are constructed upon a plan
which is intermediate between that of Sars’s Bipinnaria
asterigera and the more common type of starfish larva.
They agree with the former in exhibiting a great development
of the pre-oral lobe; while they resemble the commoner type,
and at the same time differ from Sars’s larva, in the less con-
centrated arrangement of the paired ciliated arms.

The primitive circumoral ciliated ring is divided as usual
into two distinct portions, one of which borders the ventral
side of the pre-oral lobe (figs. 1 and 2, a. c. o. 6.) ; while the
other (p. c. o. 6.) not only borders the dorsal side of the pre-

1 Metschnikoff, I find, made use of this larva in his researches on intra-
cellular digestion, Vide ‘ Quart. Journ. Mic. Sci.,’ xxiv, 1884, p, 99.

SOME BIPINNARI® FROM THE BNGLISH CHANNEL. 458

oral lobe, but is continued backwards down each side of the
body to the posterior end, where it bends first ventrally, then
anteriorly, to pass between mouth and anus across the mid-
ventral region of the body. The mouth is encircled by its own
adoral band (fig. 2, a. b.). The anterior or pre-oral ciliated
ridge (cephalotroch of Lankester) is produced into three pro-
cesses—a median arm anteriorly (p. v. a.) and a pair of lateral
arms posteriorly (7. a. v. a. and counterpart). The posterior
or post-oral ciliated ridge (branchiotroch of Lankester) is pro-
duced into eleven processes—a median arm anteriorly (p. d. a.),
and five pairs of lateral processes. Of the lateral processes
three pairs are dorsal, one pair is posterior (7. py. 7. a. and
l. p. J. a.), and one pair is ventral in position (r. p. v. a. and
counterpart).

Each of the anterior unpaired arms arises from the apex of
a prolongation of the prz-oral lobe, which may be termed the
pre-oral appendage (cf. fig. 1). This appendage possesses a
very characteristic form. It is broad from side to side and
compressed dorso-ventrally, the degree of compression in-
creasing with the age of the larva. Down each side of the
appendage runs a groove which is continuous with the lateral
depressions of the body of the larva, and separates the anterior
prolongations of the pre-oral and post-oral ciliated bands from
each other. The pre-oral appendage bifurcates after a certain
distance into a dorsal and a ventral part, the two anterior un-
paired ciliated arms, of which the dorsal is the larger. The
point of bifurcation (fig. 1, #.), where the pre-oral and post-
oral ciliated bands also diverge from each other, represents the
morphological apex of the pre-oral lobe. This is clearly shown ©
by a comparison of the larva with Tornaria; it is at this spot
—between the anterior extremities of the two ciliated bands—
that the apical plate with its tuft of cilia and pair of pigmented
sensory pits is situated in the larva of Balanoglossus. The
dorsal and ventral pre-oral arms (p. d. a., p. v. a.) are broad
and flattened, and have lancet-like or tongue-shaped termina-
tions. The dorsal arm bends backwards over the dorsal side
of the body, while the ventral arm has an equally pronounced

A454 WALTER GARSTANG.

curve ventrally. The edges of the dorsal and ventral arms
are bordered by the prolongations of the post-oral and pre-
oral ciliated bands respectively. The length of the pre-oral
appendage and of the unpaired arms which arise from it in-
creases with the age of the larva, and increases relatively to the
general growth of the body. At the stage represented in fig.
1 the body of the larva has almost attained its full develop- -
ment; but the pre-oral appendage, on the other hand, is not
fully grown. The increase in length of this appendage is
most marked in the case of the stalk and of the dorsal arm.
I am not in a position to say whether the pre-oral lobe ever
reaches the extraordinary dimensions it possesses in the case
of Sars’s Bipinnaria, but such a development is quite within
the bounds of possibility.

The larva drawn in fig. 1 was not quite my largest specimen.
In the preserved condition its dimensions were as follows :

Length from apex of prostomium (2) to tips of posterior arms 1°S mm.

Length of anterior dorsal arm from apex of prostomium (2)

to tip . : ; ; A , 0:Siaes

Maximum length, therefore . : c ; ue oi leans

The thickness (dorso-ventral) of the pre-oral appendage was
0-2 mm.

The internal structure of the larva may be gathered from
an examination of fig. 1. The alimentary canal is perfectly
normal, and needs no special description. On each side of it
lies one of the lateral enterocceles (7. e.), which fuse together
immediately in front of the cesophagus to form the unpaired
pre-oral enteroceele (p. e.). The latter is attached to the
ventral wall of the pre-oral appendage for the greater part of
its course; but at the base of the unpaired ventral arm it
breaks away from the wall and sends a short process into the
dorsal arm. The left enterocele communicates with the
exterior by means of a “ water-tube” and pore, which occupy
their usual position on the left side of the stomach. The re-
stricted area occupied by the lateral enteroccles posteriorly
points to the conclusion that these larvee have by no means yet
reached the final stage in their development.

SOME BIPINNARIZ ROM THE ENGLISH CHANNEL. 455
The dorsal wall of the prz-oral appendage is lined by a dis-
tinct longitudinal sheet of elongated mesenchymatous muscle-
cells, which are especially well developed in the region of the
bend formed by the origin of the anterior dorsal arm from the
stalk of the pre-oral appendage.

2. Habits.

The larve were captured a few miles south of the Mewstone
at Plymouth. They were taken in tow-nets worked, as a rule,
just above the sea bottom; once or twice they were found in
the surface layers of the sea. Many of the forms taken with
the larve were Atlantic animals which rarely visit our shores,
but have been unusually abundant this year, probably owing to
the prolonged calm and higher temperature of the Channel
waters. This fact leads me to imagine that the adult of this
form is a starfish living in deep water off the entrance to the
English Channel, between the Bay of Biscay and the south
coast of Ireland.! I invariably found the specimens of this
Bipinnaria lazily swimming in the upper layers of water
in the tall clear glass jars into which I have tow-netted
material transferred on its arrival at the laboratory. The
larvee are almost perfectly transparent and colourless, with the
exception of the tips of the posterior lateral arms, which are
yellow, and the alimentary tract, which is tinged with yellow
and pale brown; but the slight coloration and opacity which
the larve exhibit are quite sufficient, when associated with
their relatively large size and their remarkable mode of swim-
ming, to render them conspicuous objects amid the various
forms of life associated with them.

The mode of swimming is quite unique so far as my experi-
ence goes. During locomotion, which is usually in an upward
direction, the pre-oral lobe is anterior, and the body itself is
held quite rigid. Movement is effected by seemingly indolent

1G. C. Bourne records having taken ‘‘ several very large Bipinnaria
larvee and several later stages in Asterid development” during his cruise
off the south-west coast of Ireland, July, 1890. ‘Journ. Mar. Biol. Assoc.,’
1890, p. 320.

456 WALTER GARSTANG.

but regularly repeated strokes of the dorsal arm of the prz-oral
lobe in an antero-posterior direction over the back of the larva.
Fig. 1 represents the position of the dorsal arm shortly before
the completion of a stroke. It is obvious that movements of
the broad flat dorsal arm in the direction mentioned must
propel the larva forward. The strokes of the arm occur with
wonderful constancy and regularity, except during short
intervals of rest; 1 counted the number of strokes during three
consecutive minutes, and found them to be 80, 81,81. The
dorsal arm of the prz-oral lobe is the only part of the body
engaged in this swimming movement. It might be imagined
that the ventral arm of the pre-oral appendage would also be
used for swimming, but I observed nothing to indicate that
such was in reality the case. The ventral arm and all the
paired processes of the body were entirely inert during loco-
motion. The ventral arm may perhaps be called into play at
a later stage of development, or its function may possibly be
simply that of counteracting any tendency to rotation; in
other words, it may take the part of an anterior rudder.

38. Relations.

It has been already mentioned that the Plymouth larva ex-
hibits points of resemblance to the common type of starfish
larva on the one hand, and to Sars’s Bipinnaria asterigera
on the other ; it may, indeed, be regarded as a type intermediate
between these two forms.

The body of the larva is quite normal, and the arrangement
of the ciliated bands differs from that found in common
Bipinnariz only on account of the special development of
the pre-oral lobe. There is, however, a difference between the
two types of larvee in the form of the dorsal arms. The common
type possesses two pairs of dorsal processes, known in the
nomenclature of Agassiz as the dorsal oral and the dorsal anal
pairs of arms; in the Plymouth larva three pairs of dorsal pro-
cesses are present. It will be seen, however, in fig. 1 that the
two anterior pairs of dorsal processes of the Plymouth larva
arise from a single base on each side, and that this common

SOME BIPINNARIZ FROM THE ENGLISH CHANNEL. 4097

base has the same position as the base of the dorsal oral pair
of arms in the more familiar type of Bipinnaria. The two
anterior pairs of dorsal processes in the former represent,
therefore, the single pair of anterior arms of the latter. I have
accordingly made a distinction in this case between the terms
ciliated processes and ciliated arms; in the larva under con-
sideration the two anterior pairs of dorsal processes clearly
represent one pair of bifid arms (figs. 1 and 2, 7. a. d. a. and
l.a.d.a.). The significance of this distinction will be seen
shortly.

Before proceeding to compare the Plymouth larva with
Bipinnaria asterigera I must point out a discrepancy
which has been revealed by a study of the literature upon
the latter form. Miller’s account of the Norwegian larva
differs from those of Sars, Koren, and Danielssen with regard
to the number of lateral processes. It has already been stated
that Sars characterised his Bipinnaria by the possession of
twelve arms round the mouth. It is clear from the later
descriptions that these appendages correspond with the whole
series of paired arms in other Bipinnariz, although in Sars’s
larva they are crowded together in a group at the posterior
end of the body. Koren and Danielssen agree with Sars in
saying that there are six pairs of appendages, and they give a
figure showing the arrangement of tlie arms. The most ante-
rior pair corresponds exactly with the pair of anterior ventral
arms of other Bipinnarie, as it arises as a pair of processes
from the pre-oral ciliated ridge. The remaining five pairs
arise from the dorsal and posterior sides of the larva, and are
clearly processes from the post-oral ciliated ridge; they are
therefore comparable to the five pairs of processes from the
same ridge in the Plymouth larva. Miller, however, not only
mentions that there are six pairs of processes from the post-
oral ridge, but he also gives a figure of them. He does not
contrast this number with that given by the Norwegian natu-
ralists, so that it is quite possible that he was mistaken. It is
certainly not likely that any difference in this respect really
existed ; for the two specimens examined by Miiller were

458 WALTER GARSTANG.

taken at the same time and from the same place as those
described by Koren and Danielssen. If I am right in this
supposition there is an exact correspondence between the
Plymouth and the Norwegian larve in the number of ciliated
processes. The similarity between the two forms may be traced
even to details. Koren and Danielssen state that there are
three pairs of processes from the back of their larva, and this
is also the case in the Plymouth larva; further, one of Miiller’s
figures (l.c., Taf. ui, fig. 1, Nos. 6 and 7) shows that the two
anterior dorsal processes arise from a common base on each
side, and they therefore represent a single bifid dorsal arm
similar to that described in the Plymouth larva. The paired
ciliated arms of the Plymouth form differ from those of the
Norwegian simply in their size and arrangement, being smaller
and less densely crowded together in the former than in the
latter.

But in the case of the unpaired pre-oral arms there is a
marked difference between the two forms. In the Norwegian
larva the dorsal arm is bifid at its end, and expanded into a
broad bilobed fin; in the Plymouth larva it is broadly lanceo-
late, and tapers to a point at its extremity. This is the only
structural difference between the two larve, but its significance
cannot be definitely determined until it is possible to trace the
later development of the Plymouth larva. The much greater
elongation of the prz-oral lobe and the greater length of all the
paired ciliated appendages in the Norwegian form are corre-
lated with its mode of locomotion ; this has not been described
with much exactness, but it has been stated that during life
all the ciliated arms, both pre-oral and post-oral, are in a state
of incessant agitation. In the Plymouth larva, as already
mentioned, locomotion is carried on by means of the prz-oral
lobe alone.

It is in this fact that the chief interest of the Plymouth
larva seems to me to reside. A great development and special-
isation of the prze-oral lobe for locomotive purposes is found
again in Balanoglossus, although all traces of ciliated bands
are lost in the course of its development from Tornaria.

SOME BIPINNARIZ FROM THE ENGLISH CHANNEL. 459

This loss has obviously been brought about by the assumption
of burrowing habits after the ancestor of Balanoglossus
ceased to lead a pelagic life. The existence of Asterid larve
in which the pre-oral lobe is specially developed for the pur-
pose of swimming in the open sea seems to me, when the close
similarity of Tornaria with asimple Asterid larva is taken
into account, to point to the idea that the pelagic ancestor of
Balanoglossus was also provided with a muscular and flex-
ible pre-oral lobe bounded by the two ciliated bands, and used
for swimming. In Balanoglossus, after the adoption of life
upon the sea bottom, the pre-oral lobe, developed under
pelagic influences, was used for moving about in the mud,
and rapidly lost its primitive ciliated bands under the new
conditions.

LITERATURE.

1. Koren Er Dantetssen.—“Sur la Bipinnaria asterigera,” ‘Ann.
des Sciences Naturelles,’ 1847 (3), vii, pp. 347—352, pl. vii.

2. Mtiizr, Jonannes.—‘ Ueber die Larven u. die Metamorphosen der
Echinodermen,” ‘ Abh. d. k. Akad. Wiss.,’ Berlin, 1850, pp. 81—94,
pl. ii; 1851, pp. 61—65, pl. vii.

3. Sars, M.—‘ Beskrivelse og Iagttagelser over nogle merkelige eller nye
i Havet ved den Bergenske Kyst levende Dyr.,’ Bergen, 1835, pp.
37, 38, pl. xv.

4.60 WALTER GARSTANG.

EXPLANATION OF PLATE 28,

Illustrating Mr. Garstang’s paper on “Some Bipinnarie
from the English Channel.”

Fig. 1.—Bipinnaria from Plymouth, viewed from right side, enlarged.
The specimen was preserved, but retained its natural form. All outlines
determined by the camera lucida. Ciliated bands.—a. 4. Adoral band.
a.c.o.6. Anterior (pree-oral) section of primitive circumoral band (= cepha-
lotroch of Lankester). p. c. 0. 4. Posterior (post-oral) section of same
(= branchiotroch of Lankester). Ciliated arms on the anterior (pree-oral)
band.—p. v. a. Pree-oral ventral arm. 7. a. v. a. Right anterior ventral arm.
Ciliated arms on the posterior (post-oral) band.—p. d. a. Pre-oral dorsal
arm. r.a.d.a. and /.a.d.a. Right and left anterior dorsal arms, each
bifid. r.p.d.a.and /.p.d.a. Right and left posterior dorsal arms. r.p./.a.
and/.p./.a. Right and left posterior lateral arms. 7.p. v. a. Right posterior
ventral arm. a. Anus. m. Mouth. p.e. Pre-oral enterocele. 7.e. Right
lateral enterocele. a. Morphological apex of pre-oral lobe.

Fic. 2.—Diagram of same from ventral side, to show the course of the
ciliated bands. Letters as above.

OCTINEON LINDAHLI. 461

Octineon Lindahli (W. B. Carpenter): an Un-
described Anthozoon of Novel Structure.

e By

G. Herbert Fowler, B.A., Ph.D.,
Demonstrator in Zoology, University College, London.

With Plates 29 and 30.

Tus interesting form, dredged during the cruise of H.M.S.
“Porcupine” in 1870, was originally studied by Dr. W. B.
Carpenter, from whose drawings two plates were lithographed
for publication in the ‘ Philosophical Transactions’ of the Royal
Society ; but unfortunately no account of the investigation
was found among his papers. The specimens were then en-
trusted for description to my honoured teacher, Professor
H. N. Moseley ; and on the last occasion that I saw him he
talked to me with his customary infectious enthusiasm about
this Anthozoon, on which he was atthe moment engaged. He,
however, was unable to finish the work, and the late Dr. Philip
Carpenter then asked me to undertake it. Professor Moseley
left only a description of Dr. Carpenter’s figures and an intro-
duction to his own paper; the latter, although only a rough
draft, is printed here without alteration of any kind:

“In their ‘Report on Deep Sea Researches,’ carried on
during the months of July, August, and September, 1870, in
H.M. Surveying Ship ‘ Porcupine,’ Dr. Carpenter and Mr.
Gwyn Jeffreys (‘ Proc. Roy. Soc.,’ 1871, p. 159), in their nar-
rative of their first cruise, state as follows:—‘The most
remarkable novelty here [off the south coast of Spain, not far
from Cape St. Vincent, in 364 fathoms] obtained was a large

462 G. HERBERT FOWLER.

collection of thin sandy discs, from 0°3 to O04 of an inch in
diameter, with a slight central prominence; for these proved,
on subsequent examination, to contain an entirely new type of
Actinozoon, extraordinarily flattened in form and entirely
destitute of tentacles. Dr. Carpenter, by whom this curious
organism will be described, has assigned to it the name of
Ammodiscus Lindahli. Further on, off Cadiz, the same
curious form was obtained in depths of 227 ‘and 386 fathoms ;
and on the next page of the report it is again stated that the
Ammodiscus will be worked out by Dr. Carpenter.”

“ Apparently nothing further has been published concerning
this undescribed Actiniarian. Although Dr. Carpenter devoted
much time to the investigation of its structure, and made a
large set of microscopical preparations of its various parts,
and although he frequently mentioned it and dwelt on its im-
portance to friends, and to myself amongst the number, he
never found time to write an account of his results for pub-
lication. He, however, did prepare two quarto plates in his
scrupulously careful manner of excellent figures of the animal,
both entire and dissected in various ways, so as to display its
most important internal structure and arrangement. Unfor-
tunately he left no manuscript of any kind—not even a descrip-
tion of the plates; and the only clue to his views as to the
component structures is to be derived from the series of micro-
scopical preparations,! which is fully labelled throughout. The
two plates, the whole series of preparations, and a considerable
supply of specimens of the animal in spirits and in the dry con-
dition were placed in my hands by Dr. Carpenter’s family, with
a request that I should work out a memoir from these mate-
rials embodying his results for publication, adding from my
own observations anything that might be required.”’

“Difficulties of manipulation: sand.”

‘It is singular that Dr. Carpenter applied the term Ammo-
discus to this new form, as this name had been conferred on an
arenaceous Foraminifer in 1861. It is necessary that a new
generic title should be adopted, and I propose to call the form .

" These have not come into my possession.—G. H. F.

OCTINEON LINDAHLI. 463

Octineon, from the characteristic octameral disposition of its
large retractor muscles.”

The animal is, therefore, Octineon (Moseley, MS.) Lin-
dahli (W. B. Carp.).

As, in addition to the foregoing introduction, Professor
Moseley left only a description in general terms of Dr. Car-
penter’s figures (utilised for those of the figures which have
been redrawn for the present paper), it will be obvious that the
whole investigation had to be begun afresh for a third time.
The difficulties of manipulation to which Professor Moseley
refers proved to be considerable, the animal being too small
aud too brittle to allow of much dissection, while the sand par-
ticles, which not only cover the external surface thickly, but
are carried deep into the body of the animal by the invagination
of part of the body-wall, ruin alike razor and section.

No specimens of Octineon have been recorded besides those
of the “ Porcupine” expedition; but Professor Lankester
informs me that he has dredged an organism of the same
appearance on the Terebratula ground beyond Capri. I for-
warded one of the “ Porcupine” specimens to my friend Dr.
Paul Mayer at Naples, with the request for others in an
expanded condition ; but this ground is seldom visited, and the
search for Octineon has not as yet been successful.

The normal form of the animal in contraction (fig. 1)
appears to be that of a flat disc from which a truncated cone
projects centrally upwards. All the specimens were in this
state of extreme contraction. In diameter the disc varies
between three and seven sixteenths of an inch, and the total
thickness, including the cone, varies between one and three
sixteenths of an inch.

Whether from damage to the animal, or from some other
cause, the outline of the disc is often most irregular, and the
position of the cone becomes thereby excentric (figs. 3, 4, 5).
A few specimens are rather plano-convex and less flattened; so
far as I have seen, these are the specimens in which generative
organs are being developed.

VOL, 35, PART 3.—NEW SER. I

464 G. HERBERT FOWLER.

The whole exterior of the animal is densely covered by particles
of micaceous sand, Foraminifera, and sponge spicules, which
are firmly embedded in the mesogloea of the body-wall, after
the manner characteristic of the Zoanthez. This exterior is
divisible into two surfaces, of which the lower plane surface
appears to correspond to the limbus or pedal disc of ordinary
Actiniaria, but is obviously never attached in adult life to
stones or other large objects ; the upper and more convex sur-
face, and the cone, correspond to the lower part of the column.

At the truncated apex of the cone (fig. 1) is the opening
through which oral disc, tentacles, &c., have been withdrawn.
It exhibits in many specimens eight grooves which project
slightly towards its centre, and which are doubtless due to the
contraction of the eight retractor muscles. This opening leads
into a tube which passes vertically downwards, and which has
a sandy coating like that of the exterior (cf. figs. 10, 23) ; this
appears to be the upper part of the column pulled inwards and
downwards in retraction, and passes into the oral disc (fig. 23)
from which spring the tentacles; below this, again, is the
stomodeum. The relative position of the various regions in
retraction is, therefore, that of an ordinary Actinia, and is
effected in much the same manner, viz. by retractor and
sphincter muscles, although these are markedly different from
the corresponding structures in Actinia.

It is probably attributable to the tough character of the
body-wall that the preserving fluid (apparently spirit) has been
unable to penetrate, and that the cells of the interior have
consequently been for the most part reduced to a pulp in
which their length and distribution are alone recognisable.

In scraping off the sandy particles from the exterior surface
of the organism, a necessary preliminary to section-cutting,
but frequently detrimental to the anatomy, it is noticeable
that the sand tends to come away in little sheets, in which the
particles are cemented together by a gelatinoid substance,
staining uniformly both with carmine and with hematoxylin,
and formed apparently of an outer layer of mesoglea. No
ectoderm-cells were traceable anywhere on the exterior, a

OCTINEON LINDAHLI. 465

similar condition holding also in the Zoanthez ; presumably,
therefore, the secretion of (?) mesogloea for the adhesion of the
sand is effected by wandering cells from the endoderm, but I
have not been able to detect their presence.

When stripped of the incrusting sand the animal has the
appearance represented in fig. 2; the surface of the mesoglea
is pitted by impressions of the sandy particles, and lines indica-
tive of the mesenteries radiate towards the central cone. ‘The
external body-wall thus exposed consists of a thick mesoglea,
lined internally by a cubical endoderm, and is provided
with a single layer of endodermal circular muscle-fibres. This
description applies also to that part of the column which is
turned inwards and downwards in retraction, but must be ex-
tended by the fact that there is a profusion of “ mesodermal ”
circular fibres in this region, intermingled with a few longi-
tudinal fibres which are doubtless used in expansion; the
mesogloea is extremely thick. Just before this inturned part
passes into the oral disc the sand particles are no longer
adherent, the ectoderm is met with for the first time, and, like
the mesogloea below it, is thrown into six folds or ridges
(fig. 13).

The oral disc is, in complete retraction, pulled outwards and
downwards into pockets which contain the tentacles, as is so
frequently the case in Actiniaria and Madreporaria, this with-
drawal being effected by the mesenteries. I have not been
able to assure myself of the exact number of the tentacles
either by sections or dissections. In addition to the difficulties
of investigation already mentioned, the exact anatomical
relations are often very hard to make out, because, owing to
the great vertical length of inturned body-wall plus oral disc
plus stomodzum, the two latter are generally bent round in
a J-shape in most cases (fig. 7), rendering the interpretation of
sections troublesome. So far as I can make out (and I infer
that Carpenter and Moseley were of the same opinion) there
are twelve tentacles, arranged as is shown in fig. 9, with the
addition of a twelfth tentacle on the left of the upper directive
mesenteries. This would amount to an entocelic tentacle

466 G. HERBERT FOWLER.

over each of the six chief pairs of mesenteries, and six addi-
tional (? entoccelic) tentacles between them. In the histo-
logical condition of my specimens no difference is apparent
between the stomodzum and the oral disc. The contortion of
the specimen was so considerable that I am unable to speak
to the presence of siphonoglyphs.

It is in the arrangement of the mesenteries and their muscles
that the chief interest of Octineon consists. Their exact
number appears to vary considerably in different specimens:
generally there are about thirty-five to forty-five ; but as they
are almost all devoid of filaments and of generative organs,
their exact number probably has no particular significance.
They are generally, but not always, in pairs of approximately
equal breadth.

There are, however, always twelve larger mesenteries, which
we may safely term “ primary,’ which appear to reach the
stomodeum, and to extend downwards to be inserted on the
pedal disc. In spite of its Zoanthid habit of forming a sandy
incrustation of foreign bodies embedded in the mesoglea,
Octineon is therefore a Hexactinian, as possessing twelve
primary mesenteries; but the next point that I wish to bring
out is, that of these twelve Hexactinian mesenteries only eight
carry the extraordinarily powerful retractor muscles, and that
these muscles are arranged in the manner characteristic of a
third group, the Edwardsiz. We have therefore, in Octineon,
an Actiniarian with the characteristic habit of a Zoanthid, with
the twelve mesenteries of a Hexactinian, and the eight muscles
of an Edwardsid.

The section, fig. 11, is taken near the apex of the cone, and
exhibits the twelve primary mesenteries. From the central tube,
which is the inturned body-wall, they are cut off by the large
stoma represented in fig. 23. In these mesenteries at this
height, as in all the inferior mesenteries, the fact that a single
layer of weak muscle-fibres occurs on the extremely thin meso-
elcea lamina is recognisable. (An obvious exception is notice-
able in the mesenteries 3, 3, and will be referred to below.)

There are in all four types of mesenteries present (fig. 12) :-—

OCTINEON LINDAHLI. 467

(i) The majority have very thin lamine of mesoglea; very few
of them ever reach the stomodzum, and then only as minute
ridges on its surface which do not stretch across the ceelenteron.
They carry neither filaments nor generative organs, and their
musculature is so slightly developed that it is impossible to
make out whether they form “pairs”’ according to the ordinary
standard or not. (ii) Of the twelve primary mesenteries, two
(5, 5) are only distinguishable from the former set by their
greater length, by the longer ridge which they make on the
stomodzum, and by their position with regard to the other
primary mesenteries. ‘They are devoid of filaments, generative
organs, and well-developed musculature. (iii) Two others
(3, 3), of similar length to those last mentioned, and, like
them, represented on the stomodzum only by a central ridge,
have at the free edge of the peripheral part a considerable
swelling produced by plication of the mesoglea for the attach-
ment of muscle-fibres. They therefore carry a special pair of
muscles, but are devoid of filaments (and generative organs ?).
Judged by the direction of the fibres, which run nearly verti-
cally in retraction, and are at right angles to the retractors,
these muscles appear to be “ depressors,” tending to flatten the
animal when retracted. They have much the same appearance
as in fig. 11 for their whole course throughout a transverse
series of sections. (iv) The remaining eight mesenteries are
of a remarkable type, the key to which appears to lie in the
fact that the retractor muscle-fibres are shifting off from the
mesenteries, and becoming separate strong muscles with belly
and “ tendon.”

The sections schematised in figs. 14 to 18 are made from
camera drawings of the same mesentery and muscle at different
heights, and are from a “transverse” series, that is, in a plane
parallel to the pedal disc. They should be compared with the
restoration in fig. 23, it being borne in mind that the latter for
clearness’ sake represents the animal in a slightly less con-
tracted condition. The mesentery consists, not of. a single
lamina of mesogloea with pleatings for the increased adhesion-
surface of muscle-fibres, the usual Hexactinian condition,

4.68 G. HERBERT FOWLER.

but of a peripheral plate (a) at the side of which is fastened a
lateral plate (4); the latter unites below with a central plate (¢)
which projects radially outwards from the oral disc and stomo-
dum, and is of course merely that part of the mesentery
which is on the central side of the stoma or perforation repre-
sented in fig. 23. The peripheral and central parts are alone
shown in fig. 11, a section taken above the level at which the
lateral plates grow out. In the thickened and slightly
muscular free edge of the peripheral plates it is easy to recognise
the homologue of the muscular thickenings of the mesenteries
numbered.

Continuous centrally with (c), laterally with (a) and (6), and
inferiorly and peripherally by a sort of tendon (d) with the
pedal disc, is the huge retractor muscle, the general relations
of which certainly suggest that it is shifting off from the
mesentery. This interpretation is even more strongly suggested
by figs. 19 to 22, which are taken from a “ vertical” series,—
that is, parallel to the plane of section in fig. 23, and read from
without inwards. They are, therefore, approximately trans-
verse to the muscle itself, and exhibit its relations to the pedal
disc. The whole structure is of considerable interest, and, if
the Anthozoan origin of coelomate animals be accepted, may
even be taken to indicate the method of evolution of primitive
muscles (aggregates of muscle-fibres) at the sides and from the
walls of the archenteric coelomic pouches.

The mesenteries which exhibit this curious shifting of the
muscles are eight in number, and, as already mentioned, they
are grouped in the same manner as in the Edwardsie (fig. 12) ;
a pair of “dorsal” directives is followed on each side by two (not
a ‘“pair’’ of) mesenteries, with their muscles on the “ventral”
sides; and these face towards a pair of “ ventral” directives.
They carry what appears to be a trilobed filament ; the ova lie
in the usual position, embedded in the mesoglea lamina
(fig. 22).

PHYLOGENETIC CONSIDERATIONS.

The anatomy of Octineon having been described, its syste-

matic position demands consideration. That it is a. Hexac-

OCTINEON LINDAHLI. 469

tinian, with twelve primary mesenteries, of which two opposite
pairs are directives, is fairly certain, in spite of the simple
character of the remaining mesenteries. But the structure of
its retractor muscles is so unlike that known in any other
Anthozoan, that it must stand as the type of a new family, the
Octineonidz, characterised by a specialisation of the retractor
muscle-fibres into a muscle separate, or nearly so, from the
mesentery; probably also by other of its characters.

The next question for discussion is—Does the structure
of Octineon yield any indication of its phylogenetic history,
or of that of other Anthozoa? Two important papers! have
appeared of late on the interrelations of the various groups of
Anthozoa, both possessed by a central idea; namely, that the
Anthozoa having a larva with eight mesenteries, the muscles
of which are in many cases arranged as in Edwardsia, the
Edwardsie are the modern representatives of the starting-
point from which all our present groups originally diverged.
The premiss of the argument, however, has been contradicted
by van Beneden? so far as relates to the Cerianthee ; it is
true of about half of the recorded cases of Hexactinian deve-
lopment, and is an inference only in the Zoanthee.

While-admiring the brilliant ingenuity with which McMur-
rich and Boveri have worked out their suggestive theory, I
must say that its weakness seems to lie in the absence of
evidence as to what constitutes homology among mesenteries
and siphonoglyphs, on which it is mainly based.

In the case of the Hexactiniz, with which we are alone con-
cerned at the moment, there may be an homology among the
first six pairs of mesenteries when two of these are directives ;
at any rate, there exists no present reason to deny it to them.
But the stage with twelve mesenteries is preceded by a stage
with eight, a resting stage in the development of some dura-

1 Bovert, “Ueber Entwicklung und Verwandschaftsbeziehungen der
“Aktinien,” ‘Zeit. wiss. Zool.,’ xlix, 461. McMurnricn, “ Contributions on
the Morphology of the Actinozoa: iii, The Phylogeny of the Actinozoa,”
‘Journ. Morph.,’ v, 125.

2 Van BunepEN, “Recherches sur le Développement des Arachnactis,”
“Arch. Biol;;” xi; Lib:

470 G. HERBERT FOWLER.

tion, generally admitted to indicate an eight-mesenteried an-
cestor. Now this eight-mesenteried stage is unfortunately
reached in the Hexactiniz by various methods.! The moment
that we begin to apply homologies to these octomeral larvae we

Fie. A. KiGaeb:

The chief eight-rayed types of Hexactinian and Madreporarian larve. The
numbers indicate the order of appearance of the mesenteries; the dotted
outlines, the position of the future mesenteries which make up the total
twelve primaries.

i. Fig. A represents Sagartia, Actinia, Bunodes (Lacaze-Duthiers, corrobor-
ated by F. Dixon), an unknown larva, ? Bunodes (Boveri), and Cereactis
(Boveri).

ii. Fig. B holds for Adamsia (O. and R. Hertwig), and an unknown larva,
? Tealia (Boveri).

ili, A third type differs from that of Fig. A in transposing the order of appear-
ance of 2 and 4, and of 5 and 6 (not numbered in the figure) ; this occurs
in Rhodactis (McMurrich), Manicina (H. V. Wilson), and Cereactis
(Cerfontaine).

iv. A fourth variation is described in Haleampa (Faurot), which transposes
5 and 6 of type iii.

The four types therefore read—i. 3, 6,1, 5, 4,2; ti. 3,1,5 +5, 2,4; ii.
3; 0, 1, 0; 2, 45 av, 13,00, tn, eye

are met by a ring of difficult choices. If the twelve primary
mesenteries are homologous throughout the various genera and
species, there are at least three lines of descent in the group,

1 In the Cerianthew, according to van Beneden, the octomeral stage is
apparently attained by yet another path.

OCTINEON LINDAHLI. A71

three octomeral larve of phylogenetic value with different
muscular arrangements, which at first sight is unlikely.
That the eight mesenteries of one type are all homologous
with those of another is contradicted by a glance at the
diagrams given above; they have different muscle-relations,
and hold different positions in the twelve-rayed and adult
stages. An objection of less weight is that if there exist
that detailed homology which would follow from the value
assigned to them by McMurrich and Boveri, mesenteries
holding similar positions should appear in the same order
in both cases, but obviously do not. To refer the differences
between these two types to a “ Vereinfachung,”’ or to an
*‘ abbreviation” of development, is no explanation, and may
be made to cut equally well in two directions.

Again, there arise the cases of the Monaulee (Hertwig,
‘Report on the Actiniaria,’ Supplement; Chall. Rep. Zool.,
xxvi) and of the Holactiniz (Boveri, ‘‘ Das Genus Gyractis,
eine radial-symmetrische Actinien-form,” ‘ Zool. Jahrbiicher,’
Abth. Syst., vii, 241).

Fie. C. Fic. D.

Diagrams of the mesenterial relations of the Monaulee (C) and of the
Holactinie (D).

The essential difference between these and the Hexactiniz
lies in the fact that one pair only of ‘ directive’ mesenteries
is present in the first group, none (according to Boveri’s view)
in the second. Where in these can we look for homology
with the Hexactiniz? Is the single pair of directives of the

472 G. HERBERT FOWLER.

Monaulee homologous with the pair marked 3, 3 in Fig. A,
p- 470, or with the other? Are the two mesenteries opposite
to this pair to be regarded as a “ pair” of directives, or do
they form two halves of separate pairs? ‘This latter query
applies at any point to the Holactiniz.'

To crown the confusion, G. Y. and A. F. Dixon (‘ Proc. Roy.
Dublin Soe.’ [n. s.], vi) have described an abnormal Bunodes
with three pairs of mesenteries and three siphonoglyphs.

It seems almost impossible in the present state of our know-
ledge to deny that an eight-rayed ancestor is common to the
several groups of the Anthozoa: the Aleyonaria and Edwardsize
are permanently eight-rayed; the Madreporaria, Hexactinie,
and (without expressing an opinion on the points in dispute
between Boveriand van Beneden, we may add) the Cerianthee
have all yielded an eight-rayed resting stage in their ontogeny.
Nor is it perhaps a wanton use of the evidence to say that this
stage is the natural outcome of the earlier four-rayed condi-
tion, by further interradial specialisation in the first formed
four chambers. The evolution of a scyphistomoid ancestor
into the Lucernariz in one direction, the Anthozoa in a
second, the Scyphomeduse in a third, is very geuerally
accepted. The four mesenteries of the Lucernarie and
Scyphomeduse, like the more numerous mesenteries of the
Anthozoa, appear to have a threefold function: they carry (a)
digestive cells, (4) reproductive cells, (c) muscle cells; and it
is easy to conceive that, in a simple hydriform ancestor, a

1 T have cited these two groups merely to call attention to the absence of
evidence as to what constitutes homology between mesenteries, not because I
can accept them as groups of equal value with the Zoanthex, Cerianthez,
Edwardsiw, and Hexactinie. In a discussion of these points we may fairly
utilise evidence drawn from the Madreporaria, which anatomically agree so
closely with the Hexactinie. Is Lophohelia, which is devoid of directives
(Fowler, ‘Quart. Journ. Micr. Sci.,’ xxviii, 1), to be placed with Mussa and
Euphyllia (Bourne, ‘Quart. Journ. Micr. Sci.,’ xxviii, 21) in a group of
Monaulic Madreporaria, while Ampbihelia (Fowler, ‘Quart. Journ. Micr.
Sci.,’ xxviii, 413), the very next genus, of which the corallum, the mode of
growth, and budding agree almost exactly with those of Lophohelia, remains
with the Hexactinic corals ?

OCTINEON LINDAHLI. 473

special localisation of digestive cells on ridges (which may be
compared to a typhlosole) would lead to a concentration of
reproductive cells in their neighbourhood for a better nutri-
tion; and that the general musculature of the body-wall, pro-
bably originally both circular and longitudinal, like that of
Hydra, might, when carried out along the ridge, become
specialised on one side into retractor (longitudinal), and on
the other side into protractor (circular) muscles.

The alternative hypothesis—that the ridges or mesenteries
grew out for support like buttresses, and that the concentra-
tion of digestive and reproductive cells occurred on them
secondarily—does not seem so probable, and has neither
analogy nor observation in its favour; but the identical con-
centration of the three functions in Celenterate groups which
have evidently been long independent of one another (e. g.
Madreporaria and Alcyonaria), and the analogy of a typhlo-
solar increase of absorptive and secretive surface, occurring in
many divisions of the animal kingdom, are in favour of the
idea first suggested. The way in which one function may be
dropped and another retained is well seen in the Cerianthee,
where the contractile function has been taken up (? per-
manently retained) by the body-wall, and digestive alternate
with reproductive mesenteries. Any mesentery may become
specially utilised for a particular function; in Seriatopora and
Pocillopora (Fowler, ‘Quart. Journ. Micr. Sci.,’ xxvii, 1) the
six mesenteries specialised for digestion and reproduction are
precisely those six which in Madrepora pocillifera and
tubigera are arrested, are sterile and devoid of filaments!
(Fowler, ‘Quart. Journ. Micr. Sci.,’ xxvii, 1).

1 These two species of Madrepora were originally identified for me as
belonging to the species M. aspera and Durvillei. In the recent ‘Cata-
logue of the Madreporarian Corals in the British Museum,’ vol. i, ‘The
Genus Madrepora’ (Lond., 1893, 4to), the author, Mr. George Brook, whose
premature death has recently deprived zoology of a careful student of the
Anthozoa, has assigned them to the species quoted in the text.

In this case (Pocilloporide and Madreporide) I think we may safely com-
pare the twelve mesenteries of the polyps concerned, because of the orienta-
tion afforded by their axes when compared with the axes of the colony and

A7 4, G. HERBERT FOWLER.

Granting, then, the evolution of a four-rayed stage from a
hydriform ancestor, and its further advance into an eight-
rayed organism, from which branched the five great groups of
Anthozoa already mentioned, we may conclude, at any rate
until a good explanation is forthcoming of the apparent dis-
crepancy between the larval types shown in Figs. A, B, and C
(pp. 470-71), either (i) that, if the first eight mesenteries of the
various larve are homologous structures, and indicative of an
eight-rayed ancestor, this homology does not extend to their
musculature, and that mesenteries formed after the first eight
are not homologous with mesenteries of similar position in
adults whose larvee are of different types; or (ii) that, if the
first twelve mesenteries are homologous structures, they may
arise in any order, and this order is not of homologous or of
phylogenetic significance; or (ill) that slightly diverging types
have reconverged in our present group of Hexactinie. All three
conclusions appear to me to be almost equally difficult of accept-
ance, although the third becomes less improbable when the
facts are considered (1) that this group, at first sight so homo-
geneous, is being constantly shown to include Actiniaria which
do not conform to the hexameral type; (2) that the order of
development of the mesenteries has been efficiently studied in
only about eight genera. The improbable conclusion drawn
above may therefore be expressed in this way, that both from
embryology and from morphology comes evidence to show that
the group Hexactinie includes two or more groups, not clearly
distinguishable in the present state of our knowledge.

When dealing with questions of this kind, in which the evi-
dence is obviously incomplete, a writer can but express his
personal beliefs; but, while he cannot claim for them the
weight of admitted truths, he is allowed to apply them provi-
sionally in that capacity. On this score the beliefs which I
have stated—(1) that mesenteries are to be regarded as having
arisen primarily in connection with digestive cells, secondarily

its branches; both families have axial (superior) and abaxial (inferior) direc-
tives at the ends of the long axis of the stomodzum, an orientation similar
to that of the polyps of Alcyonium.

OCTINEON LINDAHLI. 475

with reproductive cells and concentrations of muscle-fibres ;
(2) that the eight-rayed larval stage is of phylogenetic value ;
(3) that any of these eight or of subsequently formed mesen-
teries may be specialised for the performance of one function,
digestive, reproductive, or muscular (whether protractor, re-
tractor, or depressor), and may drop one or both of the other
functions—may now be applied to Octineon.

At first sight this Actiniarian fits perfectly into Professor
MecMurrich’s scheme, taking a place immediately above the
Edwardsie, at the point where (inter alias) the lines of
descent of Hexactiniz and Zoanthez diverge—a stage at which
mesenteries 5 and 6 of Fig. A (p. 470) are throughout life less
developed than the remaining eight primary mesenteries (cf.
MecMurrich’s table, ‘Journ. Morph.,’ v, 150). Against this
view I would urge that the numerous mesenteries of lower
orders in Octineon are not incipient mesenteries, ‘ prophetic
germs’ of greater efficiency to come, but rather retrograde or
arrested mesenteries, which have (phylogenetically speaking)
lost their filaments, reproductive organs, and (most of their)
muscle-fibres, in correlation with the extraordinary muscular
development of ten primaries. The fact that the “ pairs” are
apparently not always of equal breadth (age) or length, and
that some of them die out to reappear at a lower level—that
they are, in fact, very irregularly developed—is an additional
argument in favour of this view. It can hardly be denied
that this degradation is true of mesenteries 5 (fig. 12) if this
figure be compared with an ordinary Hexactinian diagram,
and these are structurally identical with those of the lower
orders. Further, if the shifting (p. 467) of the eight retractor
muscles were to be carried a stage beyond its present condition,
the eight mesenteries which are at present connected with them
(1, 2, 4, 6, fig. 12) would be reduced to the same degraded
type. Lastly, if the suggestion made above as to the origin of
mesenteries he true, and they are all in the first instance
physiologically equivalent, the mere fact of enormously in-
creased efficiency in a few will surely lead either to the
gradual obliteration of the rest, or to their adaptation to new

4.76 G. HERBERT FOWLER.

functions. This second alternative has evidently affected me-
senteries 3 (fig. 12), but no others.

It is possible that the exaggerated development of retractors
has occurred in mesenteries 1, 2, 4, 6, in consequence of the
larva of Octineon being of the type figured as A (p. 470). Very
interesting are the special muscles of mesenteries 3, in which
the direction of the fibres is approximately at right angles to
those of the retractors; as stated above, they are probably to
be ranked as depressors,—muscles which, if present in other
Actiniaria, have not yet been distinguished, and therefore are
only slightly developed: their occurrence here, like the shifting
of the retractors, indicates the very great specialisation which
Octineon has undergone.

To the Zoanthid habit of incrustation of the external surface
of the mesoglea we can hardly attribute systematic import-
ance, although it has not yet been described as occurring out-
side the group. ;

The evidence seems, therefore, in favour of the view that
Octineon is the type of a new and highly specialised family,
descended from true Hexactinian ancestors.

OOTINEON LINDAHLI. 477

EXPLANATION OF PLATES 29 and 30,

Illustrating Dr. (G. He Fowler's) paper on. Octineon
Lindahli” (W. B. Carpenter).

PLATE 29.

These figures have been reproduced from the two quarto plates prepared
by Dr. W. B. Carpenter, but never published. Professor H. N. Moseley
wrote out an unrevised description of these plates, of which such part as
relates to the figures selected for reproduction is here printed. My own
additions are enclosed in square brackets.

I have been reluctantly compelled to differ from Professor Moseley in
interpreting his “ stomodeum ” as the upper part of the column, his “ ten-
tacular chamber ’”’ as the oral disc, and his “stomach” as the stomodeum.
Continuous series of microscopic sections (so far as I am aware, he was not
at work long enough upon the animal to have these prepared) and a com-
parison with an ordinary Actinia in complete retraction, have left no doubt in
my own mind as to the correctness of these interpretations, but I have
thought it only just to him to present his actual words, and to make my own
alterations in brackets.

Fig. 1.—A typical specimen, enlarged about [seven] diameters, viewed from
the oral face. The entire surface is thickly set with fine sand grains, frag-
ments of foraminiferous shells, spicules, &. At the summit of the central
visceral prominence is seen the opening of the invagination of the surface
leading to the mouth.

Fic. 2.—A closely similar specimen viewed from the aboral face, with the
adherent sand particles completely removed. ‘The finer radial striz mark the
courses of the mesenteries ; some of the broader radiating ridges correspond
with some of the eight large retractor muscles [and their associated mesen-
teries. Judging by the shading of the drawing, I have no doubt that
*‘aboral ” was only a slip of the pen for “oral ;” the aboral surface is very
nearly flat, but in the centre of this drawing is a cone with a central pit. ]

Figs. 3, 4, and 5.—Hxamples, with the adherent sand removed, showing
irregular varieties in [the] form assumed, produced by the indentation of the
margin and the outgrowth of lobes.

Fic. 6.—The invaginated stomodeum [oral dise, and inturned column]
removed from the body-cavity. At its base are seen [ten of] the evaginated
tentacles [lying in pouches of the oral disc] which there [surround ?] the
mouth, the entrance to the cesophagus,

478 G. HERBERT FOWLER.

Fic. 7.—View of the opposite side of the specimen shown in Fig. 6. The
inferior aperture of the digestive tract, the opening of the stomach, is here
seen [owing to the J-shaped curvature mentioned in the text]. From the
eight rings [ridges?] of tissue surrounding it radiate eight mesenterial
fringes; from the margin [of] the stomach, here inverted, six tentacles [in
their pouches] are seen to proceed. On the peripheral extremities of [those]
mesenterial filaments which lie uppermost in the figure, ovaries are seen to
be present. Two large retractor muscles, in a completely retracted condition,
lie just above the tentacles which are lowermost in the figure.

Fie. 8.—Base of the stomodeum [or rather, of the inturned part of the
column], showing its expansion, and the mode in which the tentacles [lying
in their pouches of the oral disc] proceed from its margin. The stomach
[stomodeum], lying beneath it, is hidden ; but the eight [mesenteries of the]
retractor muscles [which were torn away in the preceding figure] are seen
radiating from its margins, and some of the ovaries and mesenterial filaments
are seen in situ.

Fic. 9.—Specimen viewed from the aboral surface, with a disc of the
inferior body-wall removed from the central region, to expose the viscera. In
the centre is seen the slit-like pleated inferior opening of the stomach
[stomodszum] surrounded by the eight mesenterial filaments which radiate
from it. Beyond these, and from beneath their bases, pass outwards the
eight long and large retractor muscles [and their mesenteries]. The dorsal
and ventral pairs of these [are the directives, and] are placed exactly opposite
and in line with each other, and enclose each a single intermesenterial
chamber [formed by a?] pair of mesenteries only [while the spaces between
the other pairs of primary mesenteries contain a large number of mesenteries
of the lower orders]. Between the bases of the muscles are seen the
tentacles, eleven of which only appear. [A twelfth is probably present in
the sector to the left of the uppermost directives.] The eight mesenterial
filaments correspond in position with the eight retractors.

Fic. 10.—Much enlarged view of a but little compressed specimen laid
open so as to exhibit the essential internal structures (probably a combined,
more or less diagrammatic representation). The stomodzal tube [or
rather, the inturned part of the column,] is laid open, and it is seen that the
cuticle of the outer surface of the body, with its dense coating of sand and
shell particles, is continued to its base. Here it opens into a discoid
chamber, which may be called the tentacular chamber [formed by a folding of
the oral disc], since [the pouches for] the invaginated tentacles communicate
with it all round its periphery by open mouths. [In one of the pouches
pointing directly towards the observer, which has been cut across, the tentacle
can be seen lying as ina sheath.] In the middle of this chamber below, lies
the mouth, leading into the stomach cavity [stomodeum] below. The latter
is surrounded by the radiating retractor muscles, with the ovaries showing

OCTINEON LINDAHLI. 479

beneath them, and by the [sheaths of the] tentacles between the muscles.
Six retractor muscles are shown, and eight [sheaths for] tentacles. The
directive muscles are apparently the pair on the extreme right and those on
the extreme left of the figure [as there is only one entoccelic tentacle sheath
between them. It will be noticed, however, that in order to tally exactly
with Fig. 9 only one tentacle sheath should be pointing directly towards the
observer instead of two. The right-handmost of these is drawn as a tube,
but is perhaps the thick “ depressor ” muscle of mesentery 3 in Figs, 11 and
12. There are in this case only seven tentacle sheaths shown, the same
number as indicated by Fig. 9. None of the specimens in my possession
were nearly so spacious as the one here figured, but no doubt Dr. Carpenter
selected the least flattened and contracted specimens for dissection].

PLATE 30.

These figures are all from my own drawings, and, with the exceptions of
Figs. 12 and 23, have been outlined by the camera lucida. The histological
condition was so bad that I have represented the body layers diagrammatically
(except in Figs. 12 and 23) :—ectoderm blocked, mesogloea black, endoderm a

grey line.

a. Peripheral plate of mesentery. or. d. Oral dise.
b. Lateral plate of mesentery. ov. Ovary.
e. Central plate of mesentery. pe. d. Pedal disc.
d. Tendon. r.m. Retractor muscle.
dir, Directive mesentery. st. Stomodeum.
D. “Dorsal” aspect, in the Edwardsia | ¢e. Tentacle.
nomenclature. te. p. Tentacular pouch of oral disc.
ext. col, External part of the column. | VY. “Ventral” aspect, in the Hd-
int. col. Inturned part of the column. wardsia nomenclature.

Fic, 11.—Transverse section near the apex of the cone. Internally lies the
(broken) inturned part of the column; towards this radiate the peripheral
plates of the twelve primary mesenteries and a few of the lower orders ; from
it radiate the short central plates. The section has cut some of the tentacular
pouches, exhibiting in two cases the contained tentacles, and has also shaved
the upper part of two of the retractor muscles.

Fie. 12.—Diagram showing the general arrangement of the mesenteries, a
description of which will be found on p. 466, et seq.

Fig. 13.—Part of a section vertical to the animal, and at right angles to
the plane of the pedal disc. The inturned part of the column, distinguished
by its enormously thick mesoglcea, is cut nearly at a right angle, owing to its
flexure below ; in this region its sandy lining is replaced by ectoderm, thrown
by ridges of the mesoglea into six folds. It is seen to open into the tentacu-

VOL. 35, PART 3.—NEW SER. K K

480 G. HERBERT FOWLER.

lar chamber (H. N. M.) formed by the oral disc, from which radiate the
tentacular pouches. A few tentacles are cut at various angles. At the
bottom of the tentacular chamber is the opening into the stomodeum, and
through it into the ceelenteron. A few central plates of mesenteries project
from the upper surface of the inturned column.

Fics. 14—18 are taken from a series transverse to the animal, and repre-
sent longitudinal sections of the same retractor muscle, &c.,at various heights.
Compare p. 468 and Fig. 23.

Fries. 19—22 are taken from a series vertical to the plane of the pedal
disc, and are therefore transverse to the retractor muscle. They exhibit the
way in which it appears to be shifting in position away from the mesentery.

Fie, 23.—Diagrammatic vertical section of the animal, slightly less con-
tracted than Fig. 10; it shows both sides of a large mesentery with retractor
muscles. The position of the circular “mesodermal” muscle-fibres, which
in combination with the retractors achieve the inturning of the upper part of
the column, is indicated by dots.

STUDIES IN MAMMALIAN EMBRYOLOGY. 481

Studies in Mammalian Embryology.
III.—The Placentation of the Shrew (Sorex
vulgaris, L.).

By
A. A. W. Hubrecht, LL.D., C.M.Z.S.,

Professor of Zoology in the University of Utrecht.

With Plates 31 to 39.

A, INTRODUCTION.

As a sequel to my investigations on the placentation of the
hedgehog (‘ Quart. Journ. of Microscop. Science,’ vol. xxx,
part 3, 1889) I was very anxious to collect a similar series of
data with respect to different genera of Insectivora, in order
to determine to what extent the development of the placenta
might be said to agree in various representatives of this
archaic order of Mammalia. It will be seen from the follow-
ing pages—as also from what I hope to publish ere long about
Indian Insectivora—that this agreement is very small indeed.

The shrew was in the first place selected! as a type of com-
parison because the mole was then already being studied by
Professor Strahl, of Marburg, whose results with respect to
these questions have since been published in the ‘ Anatomische
Hefte’’ von F. Merkel und R. Bonnet, ii, 1892.

1 A preliminary communication was made by me to the Academy of
Sciences at Amsterdam, at its meeting of 27th September, 1890, in which
the chief results, which are fully described at this paper, have already been
briefly noticed (vide vol. viii, p. 79, of the ‘ Verslagen en Mededeelingen van
de Koninkl. Akademie van Wetenschappen te Amsterdam ’),

VOL. 35, PART 4,—NEW SER. vale

482 A. A. W. HUBRECHT.

Another paper on the placentation of the mole by one of
my pupils is in preparation, and will be published in a few
months hence. Weshall then be in the possession of sufficient
data concerning three European genera of Insectivora.

Although the collecting of a sufficient number of pregnant
shrews appeared at the outset to be rather an arduous under-
taking, still an organised search, continued through three con-
secutive summers, furnished me with the material required.
Kleinenberg’s fluid, in which the freshly extirpated uterus was
immersed in toto, proved to be in this case again—after
repeated comparative trials with Flemming’s mixture and
other reagents—the best and surest means of preservation.

The sections were cut after embedding in paraffin by means
of Caldwell’s or De Groot’s microtome; the series were
numbered and catalogued in the way indicated on p. 398 of
vol. xxx of this Journal.

In the explanation of the plates which belong to this article
the catalogue number of the respective section-series from
which the figures were taken is everywhere mentioned, and
thus a future comparison of these figures with the original
sections will always be possible. To such comparison I may
be allowed to invite any investigator, who, being occupied in a
similar line of research, desires to become acquainted by per-
sonal observation rather than by perusal of this paper with
the facts as they present themselves in Sorex. I must here
mention that not all the shrews whose uteri have served for
these investigations have been submitted to a separate specific
determination. Moreover for most of them such determina-
tion is at present no longer possible, as only a comparatively
small number (+ 100) of the total series of specimens was
preserved. Nevertheless the possibility cannot be denied that
in some few rare instances the uterus may have been taken
not from Sorex vulgaris, L., but from Sorex (Crossopus)
fodiens, Pall., which latter species is provided with one tooth
less in each upper jaw, but corresponds externally very closely
with the common shrew.

As no deviation from the normal succession of develop-

STUDIES IN MAMMALIAN EMBRYOLOGY. 483

mental phenomena has been noticed in my preparations
which would require a similar explanation, we may safely
conclude that either the phenomena of placentation are iden-
tical in Sorex vulgaris and Sorex fodiens, or that no
uteri of the latter species are among those that have been
investigated and figured by me.

Tn the illustrations of this paper I have endeavoured to give
the drawings on each plate, as far as possible, on scales of
enlargement that will allow of a more direct comparison of
the different figures among themselves. In figs. 1 to 16 the
general outlines (drawn with the camera) of the process of
placentation in the shrew are diagrammatically delineated. The
maternal tissues are there represented by red, the embryonic
tissues by black lines. The red numbers indicate the catalogue
number of the specimens; the small numbers (red or black)
have reference to the number of the figure in which the region
thus indicated is drawn with more histological detail under
higher power.

Leaving the histological questions to be fully discussed
further on, I will first, by the aid of these figs. 1 to 15, give a
general and succinct description of the principal facts that
present themselves concerning the growth of the shrew’s
blastocyst, and its more intimate connection with the mater-
nal tissues and the maternal circulation.

All these figures have been drawn with the camera from the
actual preparations with Zeiss’s apochromatic obj. 16 mm.,
oc. 1, tub. 160, distance between ocular and paper 24 cm.
They are thus strictly comparable also in regard to size.

To this paper no chapter is added in which the points of
agreement and of difference between the results here obtained
and those to which recent investigators (Duval, Strahl, Fleisch-
mann, Marius, Minot, Heinricius, Lisebrink, &c.) have
arrived with respect to other species of Mammalia, are dis-
cussed.

Nor are any general and comparative considerations with
respect to the theory of placentation here advanced. This
was done on purpose, because the material is already at hand

484 A. A. W. HUBRECHT.

which will enable me to study the placentation process of yet
five other genera of Mammalia hitherto not examined. It is
preferable to reserve the discussion on points of criticism and
on general conclusions, and to describe for the moment only
the facts and phenomena as they present themselves in every
one of the genera separately.

B. Tue Process oF PLACENTATION IN OUTLINES.

The following paragraph, preceding those in which the
detailed observations are recorded, is meant to be a recapitula-
tion of the general features by which the shrew’s placentation
is characterised. It precedes the full record of the details
instead of closing it, because the former will be better under-
stood and more easily grasped if certain general notions have
first been discussed.

The placentation of the shrew is brought about by processes
that take place in the maternal tissue, and by processes that
affect the blastocyst. At the outset these two sets of processes are
quite independent of each other; later on, when the blastocyst
has come to adhere against the maternal tissue, they are closely
related; still later the participation of the mother towards
the constitution of the ripe placenta is again reduced to the
maternal blood by which this organ is permeated.

The maternal processes are—

1°. Unexpected and somewhat peculiarly shaped local dis-
tensions of the uterine wall, accompanied by changes in the
distribution of glandular tissue, &c., in this wall.

2°. Considerable local proliferations of maternal uterine
epithelium.

The embryonic processes are—

1. Local changes in the outer wall of the blastocyst.

2. Special development of certain portions of the trophoblast
which finally constitute a syncytium, in which the allantois-
villi and the embryonic blood are in the closest contact with
maternal blood, the latter circulating in spaces of embryonic
issue without any endothelial lining.

STUDIES IN MAMMALIAN EMBRYOLOGY. 485

Beginning with the processes in the maternal tissue,
we notice that the distensions above referred to change the
aspect of the uterine horns, which were originally cylindrical
ducts, most considerably.

Not only in this sense, that spherical swellings indicate the
advance of pregnancy, as in other mammals, but the distensions
take a definite pear shape very soon after the embryo has wan-
dered from the oviduct into the body of the uterus. These
swellings are first connected by portions of great tenuity, which
only widen in the later stages of pregnancy.

This distension is undoubtedly independent of any direct
action of the blastocyst. It is a maternal phenomenon prepa-
ratory to the adhesion of the blastocyst. Simultaneously with
it the epithelial lining of the lumen of the distended portion
undergoes important changes.

These changes consist in a rapid increase of the epithelium
cells, which become more than one layer thick, and between
which vascular channels are enclosed, the final result being—

1°. A concave bell-shaped surface opposite the mesometrium,
on which numerous newly formed epithelial crypts open out,
gland openings being here and there interspersed between the
much more numerous mouths of the crypts.

2°. Lateral cushion-shaped surfaces, where no similar crypts
are present, and against which the blastocyst first adheres by
means of a zonary strip.

The embryonic processes are the following :—The blood-
vessels of the area vasculosa on the yolk-sac spread out against
this zonary strip, those of the allantois (soon ramified in digitate
villi) against the concave surface referred to sub. 1°; not,
however, before most important changes have taken place in
these two regions—changes that consist in the development
of a syncytial tissue of embryonic origin out of the outermost
layer of the blastocyst.

In this embryonic syncytium two regions—the counterparts
of the maternal proliferations above referred to—may be
distinguished :

1. A zonary syncytium in the region of the area vasculosa.

4.86 A. A. W. HUBRECHT.

This syncytium owes its origin to that portion of the surface
of the blastocyst which we will call the omphaloidean tro-
phoblast.

2. A bell-shaped syncytium opposite the mesometrium.
This syncytium owes its origin to that part of the outer wall
of the blastocyst which expands simultaneously with the
formation of the amnion (being, in fact, the epiblast of the
outer amnion fold), and which we will call the allantoidean
trophoblast.

_ Both syncytia enclose numerous cavities, into which maternal
blood penetrates.

The omphaloidean syncytium with the area vasculosa applied
against it is not indented by villi of the yolk-sac, such as are
found in the hedgehog. The area vasculosa is after some
time removed, the regions against which it has been applied
(proliferation and syncytium) being gradually but entirely
resorbed. A new layer of maternal uterus-epithelium applied
against the muscularis arises directly below the resorbed
portions.

When this has come about the bell-shaped syncytium oppo-
site the mesometrium remains alone in the field, and undergoes
a series of further transformations and complications, which
change it into the full-grown discoid placenta of the shrew.

These transformations can be summarised as follows:

(a) The allantoidean trophoblast is applied against the concave
maternal surface, and sends knob-like projections into the
mouths of the maternal crypts, the maternal epithelium being
destroyed wherever the trophoblast adheres. The projections
serve to fix the trophoblast very firmly against the maternal
proliferation. They donot enter the mouths of uterine glands ;
these are simply blocked by the trophoblast.

(6) The trophoblast undergoes a differentiation into an outer
layer, which assumes the syncytial character more fully, and
contains paler nuclei (plasmodiblast—van Beneden), and an
inner layer of which the nuclei stain more strongly (cytoblast—
v. Beneden).

(c) Internuclear blood-spaces are developed in the plasmodi-

STUDIES IN MAMMALIAN EMBRYOLOGY. 487

blast, and enter into communication with the maternal spaces
that have been laid bare after the disappearance of the maternal
epithelium.

(Z) The trophoblastic protuberances that have penetrated into
the crypts are hollowed out. Allantoidean villi enter into these
cavities.

(e) The allantois sends new villi against the cytoblast, which,
however, do not continue to grow centrifugally. The cytoblast
itself grows centripetally, the peripheral portions being gradually
transformed into plasmodiblast, the central portions at the same
time ensheathing the newly formed villi. The latter are thus,
while gaining in length, enclosed by an identical trophoblastic
matrix (in which maternal blood circulates), as are their earliest
predecessors.

(f) As the placentary region increases in breadth, space is
gained for the free development of these secondary villi. The
maternal proliferation at the same time flattens out to a super-
ficial covering of the growing placenta, and is finally reduced
to isolated nuclear remnants.

(7) In the final stage of the placenta the allantoidean villi are
no longerrecognisableas such, and the intervening trophoblastis
stretched to the utmost ; consequently there is only the thinnest
layer of plasmodiblast tissue to separate the maternal blood
fluid from the embryonic.

Trophoblastic lacunz containing maternal blood can be very
easily distinguished from those spaces in which embryonic blood
circulates by the fact of the much smaller size of the maternal
blood-corpuscles.

In the ripe placenta the most intricate intermixture of these
two sets of spaces is thus detected with great facility, whereas
the details of the genetic history of the placenta that are here
brought forward can by this important detail be easily traced
and tested.

The recapitulation here given shows that the placenta is
essentially an embryonic neo-formation, which is per-
meated by maternal blood that circulates in spaces devoid
- of endothelium. This embryonic neo-formation is preceded by a

488 A. A. W. HUBRECHT.

considerable proliferation of maternal epithelium, which, how-
ever, does not enter into the constitution of the ripe placenta,
but affords facilities of fixation and nutrition for the embryonic
neo-formation in its earliest stages.

The discoid placenta is, in the later stages of pregnancy, the
only connection between foetus and mother. The zonary con-
nection in the omphaloidean region is only temporary. Below
this another ring-shaped modification of trophoblast appears
at an early stage, and persists till the end as a thickened circular
membrane, forming an annular constriction round the allan-
toidean vessels that connect the foetus with the placenta.
There is reason to suppose that both this subdivision of the
trophoblast and that which is applied against the mesometrical
surface of the uterus-lumen play a part in the transport of
nutritive material to the inner cavity of the yolk-sac, where it
is being absorbed by specially modified hypoblast-cells.

C. DETAILS OF THE PROCESS OF PLACENTATION.
1. The Changes in the Uterine Tissues.

The various and successive changes by which the uterine
tissues are affected during pregnancy visibly influence the
outward aspect of the uterus. The position of the embryo and
placenta is not only marked by a swelling which increases in
size as pregnancy advances, but this swelling undergoes certain
unexpected changes of shape which we shall have to notice in
some detail. Several figures of the uterus in early stages of
pregnancy were already given in my paper on the development
of the germinal layers of Sorex (this Journal, vol. xxxi, 1890) ;
of these I will here refer to figs. 13, 14, 15, 42, 65, and 83.

The swellings are knob-like projections separated bythe more
tubular portions (fig. 47). With advancing pregnancy the
rounded knob becomes more ovoid, the tubular connecting
portions being, if possible, yet narrower (fig. 39). Of the
figures here alluded to, the following correspond to the figures
of transverse sections given in this paper:

STUDIES IN MAMMALIAN EMBRYOLOGY. 4.89
Fig. 83 (1890) is the uterus of which a section is figured in fig. 7 (1893).

», 65 (1890) corresponds to . 5 ; » 6 (1893).
» 42 (1890) 4s tees s ; : » 5 (1898).
» 15 (1890) 45 Ati : ; : » 4 (1893).
» 13 (1890) nate? : : : » 9 (1893).

On Pl. 34 of this paper further stages of pregnant uteri in
external view are figured. The numbers between crotchets
have reference to the catalogue number of the specimen ; and as
these numbers are also attached in red type to the transverse
sections of Pls. 31 and 32, it will be easy to compare these two
sets. The egg shape of the swelling (figs. 4—10) is seen to
be soon exchanged for a more pear-shaped one with a faint
constriction in the middle (figs. 11—13). The furthermost
projecting surface is the incipient placenta; the swelling
closer to the mesometrium is that of the inverted yolk-sac
and of the growing embryo itself. This latter swelling very
soon is seen to increase at a more rapid rate than the former ;
in fig. 38 it has become somewhat triangular, the placen-
tal area projecting in knob-like fashion. The connecting
portions are yet very narrow tubular ducts. As pregnancy
advances further these portions are finally also widened out,
and in the fully ripe uterus the embryos are no longer sepa-
rated by such formidable constrictions as in the early stages,
the placental area having at the same time become much more
flattened out (figs. 44, 45).

We now turn to the transverse sections, and the various
histological changes which accompany this transformation in
the outward shape of the uterus.

In the stage of figs. 1 and 2, where the embryos are yet
contained in the oviduct, and where any outwardly visible
local swelling is not detectable, we see the uterine muscularis
still in its full thickness, the outer layer of fibres being longi-
tudinal (fig. 16), the inner circular. Inside of the muscularis
the mucosa is of very varying thickness, as can be gathered at
a glance from the figs. 1,2, aud 16. Close to the mesometrium
a folded epithelium without any uterine glands opening out
into the lumen is present, and separated by an inconsiderable

490 A. A. W. HUBRECHT.

layer of connective tissue from the muscularis. The glands
are concentrated along the opposite surface, even more to the
right and to the left than exactly opposite the mesometrium,
where the mucosa is again somewhat thinner (fig. 16). In
this glandular portion the glands are very closely packed
together, and very tortuous ; they open out in that portion of
the lumen which in fig. 1 and fig. 16 forms a longitudinal
groove opposite the mesometrium. This narrow groove is thus
flanked by two cushion-shaped swellings of the mucosa. In
these thickenings more considerable blood-vessels are present
right and left (fig. 16, 6/.); smaller blood-vessels are seen
between the muscularis and mucosa, capillary ducts between
the glands.

It is very remarkable how in the next stage the transverse
section of this same region has undergone very considerable
changes, independently of any direct or active co-operation of
the blastocyst, which is as yet not adherent to the uterine
wall.

If we take the mesometrium as our starting-point, we notice
that in fig. 3 there are yet traces close to this mesometrium of
the two lateral recesses of the uterine lumen which were
visible in fig. 1, below the cushion-shaped swellings, carrying
glands and blood-vessels. Instead of the groove-shaped por-
tion of the uterus lumen that was found opposite the meso-
metrium in fig. 1, and that formed a 1 shape with those two
lateral recesses, we now find a wide bell-shaped space, between
which and the different parts of the uterine wall the relations,
more especially with respect to comparative thickness, have
become very different. Better than a detailed description, a
comparative glance at figs. 8 and 1 will explain this process.
A most considerable amount of stretching has taken place, the
antimesometrical part of the uterine wall has been reduced to
one half and even less of the thickness it had in the preceding
stage, and only the cushion-shaped regions yet fairly recog-
nisable as such have increased in thickness. The distribution
of the glands has assumed a very different aspect ; they are no
longer close together, but stretched over a wider area, and

STUDIES IN MAMMALIAN EMBRYOLOGY. 491

more numerous to the right and left than opposite to the
mesometrium. Their openings are in the same place as
before, but instead of the perpendicular groove of figs. 1
and 16, the region where they open out has become the
upper part of the concave upper wall of the bell-shaped
lumen.

The blood-vessels are found in corresponding situations as
before. The lateral swellings have chiefly increased by con-
nective-tissue proliferation, the muscularis having here also
become thinner (compare figs. 16 and 18), and the epithelium
being as yet only one cell-layer thick. The histological de-
tails are indicated in figs. 17, 18, and 66. In the last
figure it is seen that the epithelial lining also of the antimeso-
metrical concave surface is in this stage not more than one
cell thick.

It is this point which is the first to be modified. And this
modification is at the same time the most important change
that takes place in the maternal tissues preparatory to the re-
ception, fixation, and nutrition of the blastocyst. Moreover
it is a process which in other mammals has up to now not been
noticed, rather the contrary. The most recent trustworthy
observations on the placentation of mammals have brought
to light numerous instances amongst Insectivora, Rodentia,
Carnivora, and Cheiroptera, where the maternal epithelium
of the uterine lumen disappears in early stages of pregnancy.
I have myself described and figured this phenomenon in the
hedgehog. And thus it is certainly both an unexpected and
an important fact that in the shrew proliferation of this same
uterine epithelium takes place to a very considerable extent,
and that the cell material resulting from this proliferation is
of such high importance for the further attachment of the
embryo. Still we shall afterwards have to notice that the com-
plicated epithelial arrangement resulting from this proliferation
is not permanent, but that it disappears and is destroyed either
simultaneously or some time after the embryonic trophoblast
becomes attached to it, thus bringing about a final stage in
which comparison with the other mammals (where the uterine

492 A. A. W. HUBRECHT.

epithelium disappears without any special antecedent prolifera-
tion) is again possible.

This epithelial proliferation must now be described in detail.
When we compare fig. 4 with fig. 3 it is seen that the diameter
of the uterus and ofits lumen have not undergone any appreci-
able increase. Yet the epithelial proliferation has commenced
in fig. 4, and is already three to five cells thick in what will after-
wards be the placental region ; twelve to eighteen cells on the
two lateral cushion-shaped surfaces, where by this time (cf.
fig. 4) the blastocyst has commenced to adhere to the maternal
tissues. The difference between this and the foregoing stage
is still better seen if we compare fig. 19 with fig. 17, fig. 20
with fig. 18, whereas figs. 67—69 will allow us to discuss the
histological detail. The preparation figured in fig. 67 leaves
no doubt as to the proliferation being epithelial.

Here—and also in very many sections that were not figured
—the karyokinetic processes (well preserved and sharply de-
fined after staining with picro-carmine) leave no doubt about
the origin of the proliferated cell matter. It is the epithelial
cells lining the lumen that throw off new cells, which become
mixed up with connective-tissue elements that belong to the
layer situated between the epithelium and the muscularis.
However, the epithelial elements are much more numerous
than the connective-tissue ones. The proliferating process goes
on more rapidly on the lateral cushion-shaped surfaces; the
difference between the epithelial layer and the cells that have
originated out of it is soon effaced; these cells are themselves
rapidly multiplying, and thus the first distinction is about this
time created between the lateral maternal tissues against which
the yolk-sac adheres, and the bell-shaped maternal surface,
against which the allantois is going to be applied. Although
there is no sharp boundary line, still in fig. 20 and fig. 4 it
is easy enough to distinguish them in a general way.

Another distinguishing feature between these two portions
of the uterine surface is of great importance, viz. that in the
placental region the proliferated epithelial elements are very
soon seen to arrange themselves in a particular manner below

STUDIES IN MAMMALIAN EMBRYOLOGY. 493

the epithelial layer from which they have sprung. At com-
paratively regular distances the cells of the proliferation
arrange themselves in a peculiar radiating fashion, leaving a
central part without nuclei surrounded by an overcapping
layer of nuclei (fig. 69). In transverse sections this arrange-
ment of the proliferated cells could be termed fan-shaped, the
centre of the fan’s radii being situated somewhere in the ute-
rine epithelium.

In the following stages this arrangement becomes converted
into a functionally more important one. The centre of the
fan-shaped structure becomes an open crypt, the protoplasm
breaking up, and the peripheral nuclei forming the epithelial
lining of the crypt. The uterine epithelium breaks away from
under the crypt, and the inner lining of the crypt solders with
the surrounding epithelial surface at the lower border. All
this is figured in detail in figs. 70 and 71; whereas figs. 22
and 23 elucidate the same processes, and at the same time the
comparative thickness of the proliferation as compared with
the connective tissue and muscularis of the uterine wall. These
figures, compared with figs. 17 and 19, undeniably show that
the connective tissue has also increased in bulk. The open
crypts, secondary derivates of the epithelial proliferation, are
now spread over the concave surface where the placenta is
going to develop. Between the openings of these crypts the
mouths of the uterine glands are situated. It is not as diffi-
cult as it would perhaps seem to be to recognise a gland from
the newly formed crypts. Figs. 21 and 68 make this clear;
especially when the latter is compared to fig. 71, the well-defined
glandular epithelium is ever so much more distinct than the
epithelium of the crypt, which as yet is only sharply circum-
scribed where it passes into the surface layer (fig. 71) and
encloses a distinct lumen.

This lumen of the crypt becomes clearly circumscribed—
and with it the boundaries of the cells lining the crypts—in
the now following stages of figs. 24, 25, and 74.

The proliferation process has here reached its maximum de-
velopment ; its products, the epithelial crypts, are now ready

494, A. A. W. HUBRECHT.

to receive the processes of the trophoblast in the cavities of its
crypts. Between the epithelial proliferation here more fully
described maternal blood-vessels, supported by connective
tissue, have from the first taken their course. In figs, 24
and 25, but more especially in fig. 74, the relation of this
connective and vascular stroma to the epithelial tissue of the
cryptscan be clearly seen. There can be no doubt but that the
proliferated epithelial tissue is more massive than the san-
guiniferous strands penetrating between these epithelial crypts.
In comparison to the thickness of the wall of the uterus the
epithelial proliferation in figs. 24 and 25 is also seen to have
considerably increased in significance, and if we now compare
the outline figs. 3 and 4 with figs. 5—7, all of them corre-
sponding to the stages hitherto considered, we see that matters
have assumed a very different aspect, and that the lateral
cushion-shaped parts where the first omphaloidean attachment
of the blastocyst is brought about are no longer much thicker
than the antimesometrical concavity, but rather the contrary.
The epithelial proliferation, which at first took place at a
slower rate in this placentary region, has very rapidly overtaken
in its growth the lateral portions. The uterine glands, although
they are considerably flattened when the proliferation and
crypts have grown to the size of figs. 6 and 7, are always in the
possession of their duct, which takes its course towards the
surface parallel to the long axis of the crypts. In the lateral re-
gions of epithelial proliferation against which the blastocyst has
become attached there is no semblance, as we have already ob-
served (cf. fig. 7), of the formation of crypts ; here too, however,
delicate vascular tracts can be noticed between the proliferated
epithelial cells (fig. 20), these vascular spaces remaining in com-
munication with the vessels of the deeper connective tissue.
We have now obtained detailed information concerning various
changes which the maternal tissues undergo previous to and
simultaneously with the definite fixation of the blastocyst
against these tissues. When this fixation has come about a
new phase is entered upon, which is characterised by important
modifications. And whereas up to here we have noticed a

STUDIES IN MAMMALIAN EMBRYOLOGY. 495

marked progressive development in the maternal tissues, the
phase we now enter upon is one in which embryonic prolifera-
tions play the by far more considerable part. Proliferations of
the outer layer of trophoblast, more fully to be described in the
next paragraph, are henceforth seconded by vascular develop-
ment, first in the omphaloidean region (the vessels of the area
vasculosa), later on in the tissues of the allantois, which make
use of the roads that have been opened up by the trophoblast
in the future placental regions.

We may summarise the characteristic features of this new
phase as follows:

1. In the uterine tissue—

(a) The newly formed epithelial crypts are slowly but
gradually invaded by trophoblastic protuberances and ex-
crescences, which exercise both histolytical and vasifactive
functions.

(6) The lateral cushion-shaped proliferations of maternal
epithelium against which the trophoblast of the area vasculosa
is applied undergo a decided histolytical resorption, and finally
disappear.

2. In the blastocyst—

(a) The amnion is formed.

(6) The area vasculosa on the yolk-sac is completed.

(c) The allantois originates.

(d) The trophoblastic annulus and the trophoblastic pro-
tuberances noticed sub 1, a, make their appearance.

It will be seen from the above that the chief feature that is
novel to and characteristic of this phase is the intimate fusion
over a very extensive surface of embryonic and maternal histo-
logical elements, coupled with interesting and as yet only
imperfectly understood histolytical and histogenetical pro-
cesses. This makes it impossible to describe any further what
was set down for this paragraph, viz. ‘the changes in the
uterine tissues,” without making continual references to the
growth of the embryonic trophoblast. It is for that reason

496 A. A. W. HUBRECHT.

preferable to postpone the detailed description of the further
stages to the next paragraph.

In this one I will only indicate in a few sentences the general
outlines of the further changes that affect the maternal uterine
tissues down to the period of parturition.

The lateral cushion-shaped thickenings disappear with com-
parative rapidity, a new epithelial coating being rapidly formed
behind them, out of the confluence of what are in the beginning
(fig. 83) defects or fissures in the deeper regions of these
thickenings. The first appearance of this process of de-
hiscence in the deeper layers of the lateral portion is given in
outline in fig. 7, and figs. 8—12 furnish us with different
aspects of its further progress. It thus comes about that a
superficial portion of the maternal tissue against which the
area vasculosa is applied becomes more and more loosely
attached to the deeper lying portions, that strands of cells
(figs. 49 and 50, s.) connect them in some, but open spaces
separate them in more places, and that finally also the last
strands of tissue disappearing (cf. left half of fig. 12) a tongue-
shaped projection of tissue remains, which is only connected
with the rest of the uterine tissue superiorly, i.e. along the
margin of the placental region. From this moment onwards
the disappearance is even more rapid, and in fig. 13 these
tongue-shaped projections are no more present ; the portion of
the trophoblast adhering to them (omphaloidean trophoblast)
has become modified in a way which will be discussed in the
next paragraph, and at the same time the placentary region has
considerably increased. The important process of the forma-
tion of a new epithelial layer in the deeper portions, simulta-
neously with the process here referred to, can be better
gathered from those figures in which not only the outline, but
also the histological detail is given, viz. figs. 84 and 88. The
latter figure shows at the same time that this new epithelium
is thus disposed (with numerous folds) that it can allow of an
extraordinary amount of stretching, as will be necessary in the
subsequent stages of further advanced pregnancy.

From the moment the stage has been reached which is re-

STUDIES IN MAMMALIAN EMBRYOLOGY. 497

presented in fig. 13, the uterine tissue enveloping the embryo
is stretched and thinned out more considerably as far as the
non-placental region is concerned. But even in the placental
region the considerable local thickening brought about by the
formation of the placenta is, as we shall see in subsequent
paragraphs, not due to proliferation of maternal uterine, but of
embryonic trophoblastic tissue. Here, too, the actual uterine
tissue is flattened and stretched, its inner boundary line being
recognisable by the presence of deeply stained nuclear remains
of the blind ends of the epithelial crypts, of which the origin
and further development was noticed above, further blood-
vessels, and finally a few hardly recognisable gland remnants.
It deserves observation that the maximum of the decrease of
the true uterine tissue in the placental region is brought about
in stages long before parturition comes about (ef. fig. 14).

In the later stages of pregnancy the deeper regions again
somewhat increase in thickness in comparison to the bulk of
the placenta, preparatory to the severing of this organ and to
the restoration of the uterine surface after parturition.

As a detailed discussion of all these changes would fall out-
side of the scope of this paper, I will now only complete this
rapid sketch by special reference to some of the figures.

Figs. 18 and 20, when compared to fig. 16, bear testimony
to a very considerable increase of the uterine connective tissue
between the epithelium and the muscularis, partly preceding,
partly simultaneous with the epithelial proliferation. This
growth of connective tissue begins in the lateral cushion-shaped
regions ; somewhat later it is also noticed in the future placen-
tary region. In figs. 21—23 this is represented under lower,
in figs. 66, 72, and 73 under higher powers. A network of
blood-vessels is evidently being spun out in these regions for
the purpose of supplying the placenta in the later stages of
pregnancy. The comparative thickness of the connective-tissue
layer, which reaches its maximum in the stage of uterus No.
73, is soon encroached upon by the epithelial proliferation
from which it is separated by the fibrillar layer, which is very
clearly indicated in fig. 74, and which in the earlier stages (cf,

VOL, 35, PART 4,—NEW SER. MM

498 A. A. W. HUBRECHT.

figs. 68, 72, and 73) is seen to consist of an intermediate layer
of cells which are much closer together than the connective-
tissue cells further outside, and which are much more decidedly
fusiform than the proliferating epithelium cells situated below
them.

The reduction of this connective tissue has commenced in
figs. 24, 25, and 74; in the following stages (figs. 26—30) it
has reached its maximum, and if we compare one of them under
stronger powers (fig. 89, same stage as fig. 29) we see that the
maternal blood-vessels, which are situated outside the pro-
liferated region, and which supply the latter and the placentary
lacunz with blood, are directly enclosed by the muscularis.
The same can be noticed for the lateral regions in figs. 56, 57,
83, and 88a. Of compressed glands distinct traces can yet be
found, sometimes even (cf. fig. 30) with an unexpected dis-
tension among the proliferated epithelial cells.

The later increase in thickness of connective tissue between
the muscularis and the proliferated region (or rather the
remnants of it), which follows on the phases of compression or
suppression just described, can be noticed in figs. 31 and 82,
fig. 54 being a view of part of the latter figure under yet
stronger power. In fig. 54 we see between the muscular and
the connective-tissue elements peculiar corpuscles (c.) having
the aspect of dark granules round a lighter centre. In earlier
stages they are noticed in the blood-vessels, but as later on
they are not inside but outside those vessels, I have no sug-
gestion to offer as to their significance.

2. The Further Changes of these same Tissues in
connection with the Attachment of the Blasto-
cyst, and with Different Phases of its Later De-
velopment.

In the foregoing paragraph changes in the maternal tissue
have been described that occurred independently of any simul-
taneous embryonic growth that was in direct contact with the
maternal surface.

In this paragraph we shall have furtier to develop the history

STUDIES IN MAMMALIAN EMBRYOLOGY. 499

of these changes, and at the same time to trace the detail of
the very important interaction which henceforth becomes
established between maternal and embryonic growths—
(a) In the region of the yolk-sac.
(6) In the region where the allantois completes its
growth.

As far as the embryo is concerned, it is obvious that these
two regions are always very distinct. In the maternal tissue,
on the contrary, there is direct continuity between that portion
of it which is going to be in contact with the yolk-sac and that
which is preparing for the adhesion of the allantois, although
certain important differences have been noticed in the preceding
paragraph (formation of crypts, &c.).

The continuity ceases in the later stages of pregnancy, when
the yolk-sac is again removed from any contact with the
maternal tissue, because then the corresponding portion of the
maternal tissue is gradually resorbed and disappears. But
even before this final disappearance certain landmarks can be
noticed, by which it is possible to distinguish an allantoidean
from an omphaloidean region in the uterine wall. As early as
the stage of fig. 10, connective tissue is seen to penetrate in
wedge shape between the proliferated epithelium cells that
shall contribute towards the placenta and those that are in
contact with the yolk-sac. This is no full separation of the
two regions, but all the same a valuable indication of their
extension.

The actual contact between the outer layer of the blastocyst
and the maternal tissue is the first step of a new series of trans-
formations. It is a zonary region of the blastocyst that first
becomes attached to the maternal proliferation. This zonary re-
gion is equatorially situated with respect to the embryonic area;
the latter is, moreover, always facing the concavity of the uterus
lumen opposite the mesometrium. There is great probability,
as far as I can see, that the zonary region here alluded to is
already present before the attachment has yet come about.
The thickened trophoblast cells that were figured in vol. xxx1,
Pl. XXXVII, fig. 27, of this Journal, in my paper on the “ Ger-

500 A. A. W. HUBRECHT.

minal Layers of Sorex,” can hardly be anything else than the
predisposed zonary region of attachment. It is, however,
worthy of note that when the attachment comes about the
trophoblast cells that are applied against the maternal surface
are by no means more bulky, but rather excessively flattened.
This may be consequent upon a very rapid increase in surface
of the blastocyst.

Figs. 4 and 5 give us the two earliest phases of the attach-
ment of the blastocyst.

The fact that the two uteri, No. 52 and No. 73, contained
together no less than fifteen blastocysts, which have all been
sectionised, enables us to follow all the phases of this attach-
ment in detail. Three of the fifteen are not yet adherent to the
maternal surface; and of the remaining twelve, those that may
be said to represent the very earliest stage are as yet only
adherent by a small portion of the belt-shaped region; others
are already attached on one side; the majority, however, are
fixed all round, so that the fixation certainly comes about in a
very short space of time. In a few of the cases here mentioned
there was no visible change either in the maternal proliferated
cells or in the flattened embryonic trophoblast cells, against
the inner surface of which the still more flattened hypoblast
can be easily detected. These cases, no doubt, represent the
_ earliest incipient stages.

In the next stage there is a change both in certain of the
trophoblast and in certain of the maternal cells, in some cases
the maternal, in other the embryonic cells taking the lead.
The changes in the embryonic cells affect both the aspect of
the protoplasm and of the nucleus.

When examined in the preserved sections the cells are seen
to contain fine filaments which testify both to a small increase
in bulk of these flattened cells and to a more frothy arrange-
ment of their protoplasm. In some cases the nucleus of these
modified trophoblast cells is readily distinguishable, in others
it has undergone a transformation into smaller, deeply stained
granules, that (figs. 51 and 49) sometimes even become ex-
tremely minute.

STUDIES IN MAMMALIAN EMBRYOLOGY. | 501

The simultaneous change in the maternal proliferated epithe-
lium is this, that the sharply defined surface demarcation of the
layer tends to disappear, that the cell boundaries between the
cells of the superficial layers become less distinct, that the
nuclei stain more deeply, lose their sharp outline, and take the
aspect of spherical drops of nuclear matter in which the minute
details (nucleolus, granula, &c.) of ordinary nuclei can no
longer be detected. These changes, both in protoplasm and
nuclei, cause this part of the maternal tissue to be so exceed-
ingly like the adherent embryonic cell matter, that very soon
a distinction between the two becomes more difficult and even
impossible.

A most effectual adhesion of the two tissues has now come
about ; they may be said to be fused together. Especially in
the stage of fig. 5 (uterus, No. 73) the cementing process
between trophoblast and mucosa is very perfect, and in many
of the blastocysts secondary changes in the tissues of the belt-
shaped region of attachment have not yet commenced. Where
these are initiated they have the character of a granular de-
generation, as indicated in the lower part of fig. 51. In the
upper portion the fused parts have a yet more uniform aspect.

A more detailed and at the same time comparative inspec-
tion shows that this granular degeneration affects both the
maternal and the trophoblast cells. In how far it is preparatory
to vasifactive processes in this region cannot exactly be made
out, maternal blood penetrating into capillary spaces in this
zone of fusion independently of any preceding granular trans-
formation. :

Amongst the six blastocysts of uterus No. 73 there was one
yet partially protected by its zona pellucida; this is undoubtedly
abnormal for this stage, and a few other peculiarities in its de-
velopment seem to confirm this view. As such I may mention
partial and local proliferations and thickenings in the tropho-
blast instead of the flat application and parallel fusion against
the mucosa.

The stages represented by the uteri 45, 42, and 51 will be
discussed together. In all of them the fusion has become yet

502 A. A. W. HUBRECHT.

more intimate, and so has the confluence of protoplasm by
which a syncytial layer of embryonic derivation is created,
in which intercommunicating spaces for the maternal blood
are gradually evolved. The possibility of distinguishing ma-
ternal from embryonic derivates in this layer is yet further
diminished ; in the stage of figs. 7 and 49 (uterus No. 42) there
are, however, indications that secondarily a deeper layer of the
trophoblast becomes more sharply marked off against the
syncytial layer, and may thus be compared.to what Ed. van
Beneden! has called for the bat the cytoblast, i.e. a deeper
layer of the trophoblastic tissue in which the cellular character as
distinct from the syncytial is more prominent. Here in Sorex
it is only just indicated, and in many sections very imperfectly
visible, but certainly somewhat more distinct in later stages.

Altogether the phenomena here described in the ompha-
loidean region of Sorex do not testify of a very intense physio-
logical significance of this region. They have only an
evanescent existence for the short period that the area vasculosa
comes to extend along the belt-shaped zone of the original
attachment of the blastocyst. No omphaloidean villi (as in
Erinaceus) contribute to the increase of the surface or to a
more intimate contact between maternal and embryonic circu-
lation. As such the arrangement is perhaps more the here-
ditarily transmitted reminiscence of former higher perfection,
which has been gradually overruled by the allantoidean
arrangement.

The phase in which the interchange of nutritive materials
and of oxygen between the maternal blood and that of the area
vasculosa may be said to have reached its maximum is repre-
sented in fig. 10 (uterus No. 106).

In all the sections of this phase I find very marked self-
injections of numerous capillary blood-spaces that are imme-
diately contiguous with the above-mentioned cytoblastic layer,
and only separated by this from the now fully developed circu-
lation of the yolk-sac. These blood-spaces originated in the
syncytial layer, as was noticed above, and their lumen was

1 «Bulletin de l’Académie Royale de Belgique ’ (3), vol. xv, p. 351.

STUDIES IN MAMMALIAN EMBRYOLOGY. 503

brought into communication with the maternal capillaries
during the histolytic processes there described (cf. fig. 83).

As pregnancy advances, and as spaces that have originated
in these lateral proliferations by dehiscence (fig. 83) increase
in number and in size, the communication between the capil-
lary lacune in the syncytium and the afferent uterine blood-
vessels is brought about by vessels that take their course
through persisting strands of proliferated epithelial tissue that
keep up the connection between the deeper permanent and
the superficial epithelial tissue (s,s, figs. 9, 10, 50, 84).

The latter is, so to say, scaled off from the rest of the
mucosa. Even the connecting strands just noticed disappear
one by one, and in the stage of figs. 11 and 12 the maternal
blood-flow through this region, which reached its maximum in
the stage of fig. 10, has again become diminished, histological
and degenerative phenomena of the cells and the nuclei at the
same time increasing. There is a very marked difference
between the cells of this and the placental region. The com-
ponent cells of the latter are seen by the number of karyoki-
netic figures to be in rapid cell division ; those of the lateral
omphaloidean region show no signs of such activity ; growth
has here come to a standstill after the phase of fig. 10. In
many sections a sort of pseudo-karyokinesis was observed in
some of the component cells. I have figured this phenomenon
in fig. 84a. It may be gathered from these figures that the
nuclei rather testify towards a degenerative than towards a
reproductive process.

There is every reason to believe that in the process of final
resorption of the omphaloidean region of the mucosa an active
part is played by a special modification of the trophoblast, to
which we shall henceforth give the name of the tropho-
blastic annulus, and which is seen in what is evidently its
most active phase in figs. 10 and 84, i.e. simultaneously with
the maximum development of the omphaloidean circulation
and congestion. It is then seen to be a ring of trophoblastic
tissue of increased thickness as compared with the other por-
tions of the trophoblast. The component trophoblast cells are

504 A. A. W. HUBRECHT.

high and columnar, the nuclei situated in the inner half.
There is a very gradual passage of these modified cells to the
flattened ones which form the non-placental trophoblast below
the annulus. At the upper rim of the annulus the passage is
more abrupt. The ring, which is nowhere adherent to the
uterine surface but distinctly curved away from it, is there
attached to the region of fused maternal and embryonic ele-
ments, against which the area vasculosa is spread out. There
is, in fact, no interval between the cells of the omphaloidean
trophoblast and those of the trophoblastic annulus. The ter-
minal sinus of the area vasculosa is very close to the upper
rim of the trophoblastic annulus (fig. 88). From the point .
where the annulus meets the uterine wall there has been from
the first moment of its appearance (figs. 6—8, and 48—50) an
indication that cells properly belonging to the annulus pro-
liferate in a downward direction (an’ in these figs. and in
fig. 84), and are then applied against the tongue-shaped mater-
nal shred which has already been more than once alluded to.
This becomes more and more evident in later stages, and then
forms, as will be seen below, the permanent, also ring-shaped,
membranous connection between the trophoblastic annulus
and the outer circumference of the placenta.

In order to form an opinion as to the physiological signifi-
cance of the trophoblastic annulus we shall first have to describe
avery peculiar phenomenon, of which the first traces appear
about the time of the completion of the amnion, during the
congestion of the tissues consequent upon the gradual com-
pletion of the omphaloidean circulation. The histolytical pro-
cesses which at the time of this congestion are in full activity,
and which have partly contributed to prepare this circulation
of maternal blood outside the area vasculosa, are undoubtedly
connected with this phenomenon. It is an actual hemorrhage
which is invariably found to a lesser or greater extent just
outside the trophoblastic annulus, between it and the perma-
nent and the proliferated epithelium. As early as the stage of
uterus No. 51 (figs. 8, 9, and 50) there are traces of it. In
the last-named figure the extravasate is as yet insignificant ; in

STUDIES IN MAMMALIAN EMBRYOLOGY. 505

the stages of figs. 1O—12 it has become much more important ;
there. is a distinct blood-clot in which the separate blood-cor-
puscles can generally yet be traced in the sections, pressed
between the trophoblastic annulus and the uterine wall
(fig. 84). This blood-clot marks the lower end of what will
speedily become a tongue-shaped projection in the sections
(fig. 88), but which, in fact, is a thin cylindrical shred of de-
generating tissue, and which represents the remnants of the
proliferated omphaloidean region of the mucosa. Simultane-
ously with the resorption of these parts the placenta increases
in circumference. The trophoblastic annulus thus apparently
comes to be situated higher and higher up (figs. 11 and 12).
The area vasculosa is lifted off from the surface with which it
has entertained relations of exchange, but which is now no
longer available for that purpose.

Finally (fig. 18, uterus No. 180) the circumference of the
placentary region has so considerably increased that the tropho-
blastic annulus, as has already been indicated, is now connected
with the inner rim of the placenta by the same strip of tropho-
blast, and that could be noticed in its earliest stages on figs. 10
—12, 48, 50, and 84, and that has now grown out to a mem-
brane. As such it has more and more asserted itself, its first
phases being figured as noticed above, its final stages being
indicated in figs. 13—15, and more considerably enlarged in
figs. 33 and 34.

The last remnants of the omphaloidean trophoblast are visible
in the stage of fig. 13, though not indicated in that figure.
We find it in this stage as a mass of semi-resorbed cell-remains
mixed up with blood and nuclear detritus, and pressed in
between the outer surface of the trophoblastic annulus and
the outer rim of the placenta. In the stage of figs. 14 and 15
no further remains were present. The resorption has here
become final, and the allantoidean regions of mucosa and tro-
phoblast are now not only predominant, but entirely monopolise
the nutrition of the blastocyst.

We are now enabled more clearly to picture to ourselves
what part the trophoblastic annulus has to play in all this,

506 A, A. W. HUBRECHT.

And we can definitely affirm that this part is an active one
after inspection of such preparations as those of fig. 84. _ They
show us that the blood extravasate above alluded to, which is
such a constant and characteristic feature of the phase of
development that is here under discussion, is bodily absorbed
by the cells of the trophoblastic annulus. The blood-corpuscles
that have been set free between the uterine wall and the tro-
phoblastic annulus are seen to enter the cells of the latter toa
very considerable extent, and I have no doubt.that during life
this process takes place on an extensive scale. Whether the
cells of the trophoblastic annulus which thus absorb maternal
blood-corpuscles in a phagocytical way produce other matter
out of them, thanks to a special activity of their protoplasm,
must be left an open question. Still it is an undoubted fact
that about this time the cavity of the yolk-sac gradually comes
to contain matter which forms a characteristic coagulum in
this and the next phases.

This coagulum grows and increases, and is spread out against
the hypoblast that lines the mesometrical concavity of the yolk-
sac (fig. 56). It has a yellowish-green, glassy, and yet partly
granular appearance, sometimes with different shades of colour
disposed in parallel layers (fig. 57). It appears to be densest
in the immediate proximity of the hypoblast; towards the
lumen of the yolk-sac it grows less dense, and in the phases of
fig. 13 and following it fills the yolk-sac more or less com-
pletely, the area vasculosa and its modified hypoblast, of which
we shall come to speak by-and-by, being bathed by it.

In the early phases of development the cavity of the yolk-
sac does not contain similar coagulable matter. The first
appearance coincides (a) with the increased resorption and
final disappearance of the congested omphaloidean mucosa ; (0)
with the increase in size of the yolk-sac, which is then applied
with its whole surface against the mesometrical wall of the
uterus.

These particulars should be borne in mind when we are
going to speculate upon the origin and significance of the
coagulum. And if we then see that the trophoblastic annulus

STUDIES IN MAMMALIAN EMBRYOLOGY. 507

stands from the first in a particular relation to that ompha-
loidean mucosa, and to the disintegrated products into which
it gradually dissolves, and that the cells of the annulus actively
absorb the blood-corpuscles that are set free during this disin-
tegrating process, the hypothesis does not seem a strained one
which assumes the annulus to participate in the production of
the coagulable matter that is gradually accumulating inside
the yolk-sac.

Only one layer of hypoblast-cells coating the trophoblastic
annulus on the inside separates the protoplasm of the latter
from the yolk-cavity (figs. 50, 84, and 88). And the slight
increase in size of these hypoblast cells where they are applied
against the annulus (figs. 50 and 84) rather favours than con-
tradicts the supposition that they too aid in transporting
matter that is originally outside the yolk-sac towards the inside
of it, albeit with changed chemical and physical properties. __

The cells of the annulus merge very gradually, as was already
noticed, into those of the non-placental trophoblast. As this
is applied against the uterine epithelium, immediately behind
which numerous maternal blood-vessels convey the blood from
the mesometrium to the placenta (figs. 15a, 15 6, 56, 57,
11—13), it is neither a strained hypothesis to suppose that
here, too, certain substances transudate from the blood-vessels
through the uterine epithelium, are absorbed by the cells of the
non-placental trophoblast, and transported through this and
through the underlying hypoblast into the cavity of the yolk-
sac. Such a regular passage directed inwards through these
layers would go far to explain the very regular growth and
increase of the coagulable material all round the surface where
the non-placental trophoblast is in contact with the uterine
epithelium. Trophoblastic annulus and non-placental tropho-
blast, continuous anatomically, would then also physiologically
play a part to a certain extent analogous.

We must not forget that if the hypothesis of the derivation
of the coagulable matter here enunciated is not accepted,
another source must be indicated from whence this matter can
have been derived. And if we consider the other tissues by

508 A. A. W. HUBRECHT.

which the yolk-sac is bounded we find that only the area
vasculosa remains.

Now it is undeniable that the area vasculosa increases in
surface and its cells in bulk even after it has been delaminated
from the surface of the omphaloidean mucosa (figs. 11—13).
This considerable increase goes on in later stages of pregnancy.
Arborescent excrescences (figs. 15, 15 a, 65) contribute to ex-
tend its surface yet further, and we would thus be able to
argue that all this enlargement was at the same time the ex-
pression of a heightened activity, by which the coagulable
matter filling the yolk-sac is produced.

Such a conclusion would, however, to my idea, confuse
cause and effect. The unexpected ulterior development of the
vascular area on the yolk-sac, and the increase in bulk of the
hypoblast cells can hardly be explained as meant to contribute
to the production of special contents of the yolk-sac, which
would then be available for the growth of embryonic tissues
only on condition that the processes were reversed, and that
by the same channels that have first deposited the coagulable
matter it was next absorbed and conveyed to the embryo.

The explanation is ever so much more natural that it is the
presence of this coagulable matter inside the yolk-sac which
has brought about the increased ulterior growth of the vessels
of the yolk-sac.

The upper surface of the latter, in consequence of the pecu-
liar growth of embryo and placenta already noticed, has come
to be inverted into its own cavity, and over a considerable
surface the hypoblast cells beneath the area vasculosa can thus
be directly bathed by any fluid that is contained in the yolk- .
sac. If this fluid has nutritive properties, as in the case of
Sorex it may very reasonably be expected to have, then the
increased growth of the area vasculosa and the increased effi-
ciency of the apparatus (in casu of trophoblastic annulus and
non-placental trophoblast) by which this nutritive matter is
prepared and conveyed, is at the same time understood.

And so the disintegration of the omphaloidean portion of the
mucosa may be said to lead indirectly to the production in a

STUDIES IN MAMMALIAN EMBRYOLOGY. 509

second instance of food material for the growth of the embryo,
this material being at the same time partially derived from
other sources.

We have now to give a somewhat fuller account of the
changes in the area vasculosa that have been so often referred
to. The most evident change relates to the hypoblast cells of
this region, which become more bulky and which take a most
vivid green colouring, visible through the distended tissue of
the unopened living uterus. This colouring matter is extracted
by alcohol. It was not further chemically analysed, but after
the ordinary treatment of the specimens with hardening, stain-
ing, and embedding reagents, it can yet be distinguished as a
greenish-brown or reddish-brown colour. The increase in the
size of the hypoblast cells can already be detected when the
stage of fig. 12 isreached. From that of fig. 14 onwards it
is, however, most clearly marked, and at the same time another
process comes in the foreground in this region, viz. the forma-
tion of embryonic blood-corpuscles. Different stages in the
formation of those corpuscles are represented in figs. 583—63,
and I have no doubt that the preparations there figured
will well repay a conscientious study of this process. As the
dimensions of the vascular surface of the yolk-sac increase so
considerably in the period from fig. 13 onwards, new vessels
are being formed in all directions, and become visible as pro-
liferations that rise above the level of the surface (figs. 58 and
59) and gradually form a raised network with the pre-existing
larger vessels. The whole of this network is bathed, as was
noticed above, by the contents of the yolk-sac. In these
vessels the future lumen is filled with cells which in the first
instance are fixed and immoveable (fig. 58, extreme right), each
of them being granular and only gradually changing their
aspect—transition stages being present—in that of the ordinary
embryonic blood-corpuscles. At the same time the latter are
seen to become looser (fig. 58, middle), and are gradually
attracted into the general embryonic circulation.

As to the first origin of these cells, it can be noticed that
they arise out of larger polynuclear ones (figs. 59—63), of

510 A. A. W. HUBRECHT.

which in their turn the first origin will have to be settled.
For the present I will refrain from giving my own opinion on
the point whether these mother-cells are of direct hypoblastic
origin, or whether they are proliferations of the ever so much
thinner mesoblastic tissue which lines the massive hypoblast
of the area vasculosa.

Lately new researches on the origin of the blood and the
blood-vessels (C. K. Hoffmann, a.o.) have brought this ques-
tion again very much into the foreground, and rather than
here treating it incidentally I will limit myself to the pointing
out of the shrew’s yolk circulation as a favorable object for
the study of these problems.

Figs. 64 and 65, taken from the same section, show two
blood-vessels of the yolk-sac in a much later stage. The upper
one is a very much flattened space in a stretched portion of the
area vasculosa; the lower one is in the region close to the free
border of the placenta (cf. fig. 15a), where the surface of the
area vasculosa is thrown into very numerous folds, the free
space in the yolk-sac being in these later phases of pregnancy
more and more reduced. The coagulum is all the same pre-
sent in these later stages. The hypoblast cells are here seen to
have yet further increased in size.

An important peculiarity that should here be mentioned,
now that the blood-corpuscles of the embryo and their neo-
formation are being noticed, is this, that they are of such a
different size from those of the mother. Such is the case, not
only in the early (figs. 80—82, 85, 86, 89), but also in the later
stages of pregnancy, and offers a most valuable advantage for
recognising maternal from embryonic circulatory spaces. This
is especially important in the placental region, where the rela-
tive intermixture of these spaces is so extremely complicated,
and where quite normal self-injections are thus available,
showing the finest blood-spaces with the utmost clearness (figs.
52 and 53).

We have now finished the description of the later pheno-
mena in the omphaloidean region and the area vasculosa, in
that of the trophoblastic annulus, and of the non-placental

STUDIES IN MAMMALIAN EMBRYOLOGY. 511

trophoblast, as also of the participation of these different
regions in the processes of attachment and partly of nutrition
of the blastocyst. There now remains for our consideration
that yet more important portion of the trophoblast which takes
an active part in the formation of the placenta—the allantoi-
dean trophoblast.

Its first appearance as an independent layer is coincident
with the formation and completion of the amnion. From the
very first (cf. fig. 7) the amnion fold (that can primarily be
observed behind the embryo) has an ever so much thicker outer
than inner fold. This difference in thickness must be set down
to the account of one of the component cell-layers—the epi-
blast, whereas the somatic mesoblast is and remains exceed-
ingly thin. I have reason to believe (though I will reserve the
consideration of this hypothesis to a later publication) that the
growth of the amnion fold is not really a slow turning over of
a spread-out layer, but that from the very first—even as long
as they are spread out flat—the inner and the outer curve can
be said to be distinct.

In the earliest phase the thickened shield of epiblast is sharply
set off against the continuation of the epiblast external to it.
It seems to me very probable that from cells derived from
the first-named shield the epiblast entering into the inner fold
of the amnion is derived. The point of meeting between the
inner and the outer fold would then correspond to the rim of the
shield in the earliest phases. Inner and outer fold increase
simultaneously and at an equal rate. In Sorex they always
meet at a sharp angle ; they do not pass into each other along
a smooth curve. Be this as it may, the fact remains patent
that the outer fold of the amnion is composed of the somatic
mesoblast above referred to, and of a thickened layer of epi-
blast cells. This layer, as will be seen by reference to figs. 26,
27, and 75—79, is in many places two and sometimes more
cells thick: it is the allantoidean trophoblast. It cannot be
strictly defined topographically, as the allantoidean trophoblast
passes into the omphaloidean trophoblast quite gradually.
Still, the name is convenient for designating that portion of the

512 A. A. W. HUBRECHT.

trophoblast which forms a hemispherical overcapping of the
embryo, which is situated opposite the non-placental (equally
hemispherical) trophoblast, and which is separated from the
latter by two ring-shaped zones—the omphaloidean trophoblast
and the trophoblastic annulus.

The allantoidean trophoblast is a massive layer, and though
the cells may be less high individually than those of the tropho-
blastic annulus, still they seem to be yet more active physio-
logically speaking, and stain more deeply with picro-carmine.
Another peculiarity to be noticed on the allantoidean tropho-
blast even before the completion of the amnion (ef. figs. 8
and 9) is the presence of warts or projections of proliferating
cells. They arise in the first instance independently of corre-
sponding cavities on the maternal surface against which the
allantoidean trophoblast is going to be applied. But when this
application has come about, a trophoblastic projection is present
at whatever point we find the maternal surface indented, i. e. at
the mouth of every epithelial crypt.

The numerous crypt openings are thus in the stages of uteri
51 and 106 (figs. 8—10 and 75) already partially blocked by
- trophoblastic cell material, and the trophoblast cells continue
to proliferate in the first instance there where they have pene-
trated into the crypts. The trophoblast applied against the
concave surface between the crypt openings is also in prolife-
ration ; later on (figs. 13 and 94) this becomes yet more marked.

When once the crypt openings are filled by massive knobs of
proliferated trophoblast, the latter are seen to become hollowed
out very rapidly, a phenomenon which goes apace with the
further development of the allantois. The latter makes its first
appearance while the amnion is being completed (fig. 8). After
this has come about, and while the omphaloidean circulation
reigns as yet supreme (fig. 10), the allantois rapidly extends
against the concave surface when the trophoblast has just be-
come adherent against the mouths of the epithelial crypts that
have arisen in this region in preparation of the processes which |
are now going to follow. In these processes both this region
of the trophoblast and the allantois play the principal parts.

STUDIES IN MAMMALIAN EMBRYOLOGY. 513

Before describing those processes in detail it will be well
once more to summarise them. Superficially it would seem as if
the phenomena could be thus characterised :—the trophoblastic
knobs become inserted in the crypts and vascularised by the
allantois, and by further growth and division of all these parts
the full-grown placenta comes into existence. Nothing is, how-
ever, further from the truth. It can be easily understood that
the number of crypts is limited when once the trophoblast has
come to be applied against the latter. Thus new crypts and
knobs cannot possibly arise in the same way as the original
ones after the adhesion of the trophoblast against the maternal
surface.

Neither do I find traces, when once the trophoblastic pro-
tuberances have become firmly fastened in the crypts and have
received allantoidean villi in their subsequent cavities, of any
penetration of trophoblastic tissue into the proliferated
maternal tissue between the already existent maternal
cry pts.

The positive interpretation of the facts which I desire to sub-
stitute for the insufficient hypothetical suggestions just brought
forward is the following :—After the maternal crypts have
received the plugs of trophoblast in their cavities a destruction
of the maternal epithelium as far as it is in contact with the
trophoblastic plugs follows.

Trophoblastic proliferation and histolysis in the surrounding
~ maternal tissue going hand in hand, the firm adhesion of the
blastocyst in this placental region is very soon brought about.

And itis to obtain this degree of very firm connec-
tion that the crypts and protuberances have to a great
extent served. The trophoblast has, so to say, become
anchored by its wart-like protuberances in the maternal pro-
liferation, and it is now going to prepare the placenta by its
own activity. For this the firm adhesion between trophoblast
and mucosa which has now been established is of the highest
importance. It is henceforth possible for the trophoblastic
tissue to expand into a syncytium of considerable size, and to
tap the maternal circulation without any danger of unservice-

VOL. 35, PART 4,—NEW SER. NN

514 A. A. W. HUBRECHT.

able extravasates. The trophoblastic knobs may yet penetrate
somewhat further into the crypts, but it is not in this penetra-
tion that the chief feature of the placental development is
found. That chief feature is the increase in thickness of the
trophoblast, and the decrease in thickness of the maternal
crypt region. From this latter blood passes into the
former; no further maternal contribution, nor indeed any
penetration of maternal growths between the trophoblast (or
the allantois villi that are enclosed therein), can anywhere be
noticed.

The trophoblast is spun out against a maternal surface spe-
cially prepared for its firm adhesion. A glance at figs. 1O—15
will further explain this. In these figures the maternal crypts
represented by the red dotted lines are not represented in
their real number. The fact is, they are very closely pressed
together in the early stages (cf. figs. 24 and 74), and only those
crypts were inserted in the drawings with the camera (figs. 6
—18) which happened to possess a more or less distended
lumen.

The depth of the zone occupied by the crypts is seen to re-
main about stationary between the stages of the figs. 11—18.
On the contrary, the depth of the zone in which the tropho-
blastic protuberances are situated has very considerably in-
creased,

It follows from this that when once the firm adhesion is
brought about there is no very active further penetration of
the trophoblast into the crypts, nor any extension of the
cryptal and intercryptal proliferation downwards between the
trophoblastic tissue. Great activity is, however, displayed ;
first, in the trophoblast between the villi, where intercommuni-
cating spaces are being evolved, that enter into communication
with the maternal circulation ; second, in the trophoblastic
layer that covers the inner concavity of the developing placenta.

The latter layer attains to a most considerable thickness
(cf. figs. 13, 30, and 94), preparing new points of insertion
for secondary and tertiary allantoic villi (figs. 13, 29, 30, 31,
91, 94), which in their turn {grow out to the length of the

STUDIES IN MAMMALIAN EMBRYOLOGY. 515

first formed as the placenta yet further increases in size and
in thickness.

Simultaneously with this growth of secondary and tertiary
allantoidean villi, the thick layer of trophoblast in which their
insertion took place furnishes the material for the blood-
cavities by which these extending villi are being surrounded.
That same trophoblast layer has in the further stages (figs. 14
and 15) no longer such a considerable thickness, and we can
now understand how there can be no doubt that all the tissues
that contribute to the formation of the part of the placenta
which is marked in black are of embryonic origin.

During the considerable increase in superficial extension of
the placenta the cryptal region has at the outset not remained
quite stationary, though its increase is in no comparison what-
ever to that of the trophoblastic region. Its growth is more
adapted to the obvious fact that it has to spread over an ever
enlarging area. This leads to a decrease in thickness that is
very obvious when we compare the figs. 13, 14, and 15 with
each other. And whereas in the first-named figure the crypts
are yet very clearly observable, this is much less the case in
figs. 14 and 15, the whole of the cryptal and intercryptal
tissues being here only represented by a layer of deeply stain-
ing nuclear matter and nuclear detritus.

After having given this summary description of the important
modifications that go on in the placentary region, we shall have
to analyse in further details the processes that cause them, and
to furnish the confirmation of the interpretation here given.
We must then go back to the stage of uterus Nos. 42 and 45
(figs. 24, 25, and 74).

We here have the mucosa before us of the placental region
in the phase of its highest development, anteriorly to the
adhesion of the allantoidean trophoblast against it. The epi-
thelial crypts, that have originated by the proliferation above

described (p. 493), arein possession of a very distinct epithelial
lining, passing continuously into that of the uterus lumen. They
are closely pressed together, and some of them bifurcate towards
the outer circumference, the blind ends of the crypts being

516 A. A. W. HUBRECHT.

thus considerably more numerous than the mouths. Between
the high and massive epithelium of two neighbouring crypts
there is everywhere a core of connective tissue with capillaries,
the latter with a flattened endothelium. Especially close to the
surface of the mucosa many of these blood-spaces are disposed
parallel to that surface, some of them immediately below the
epithelium. .

In fig. 74 it can be easily seen that the character of the
tissue between the crypts is not that of ordinary connective
tissue, nor that the endothelium referred to is as yet very
much flattened. The tissue is in a state of most active pro-
liferation, and the resemblance to the proliferating epithelium
is very close. In the stages of figs. 26, 28, 29, 30, the epithe-
lial character comes yet more into the foreground, because the
tissue between the crypts is now reduced to actual capillary
vessels, taking their course strictly radially. In this stage
the endothelium is really flattened (fig. 82), and the radial
capillaries are on all sides supported by cryptal epithelial
cells.

The moment the adhesion of the trophoblast has come about
the maternal epithelium disappears. Whether it partly dis-
integrates in its place, or whether it is actively absorbed by the
outer layer of trophoblast cells, cannot be decided with abso-
lute certainty ; perhaps both processes contribute, and the
phenomenon is readily comparable to what was noticed in the
omphaloidean region and represented in fig. 83.

In fig. 75 the maternal epithelium is yet seen to the right
of the trophoblastic knob; to the left of this it is disintegrat-
ing between the darker stained outer trophoblast layer and
the sanguiniferous deeper layer. In the same way the mater-
nal cryptal epithelium disappears wherever a trophoblastic
knob penetrates into a crypt; in this case the preparations
countenance the view that active destruction and absorption of
the maternal cells by the trophoblastic ones takes place (fig. 80).
As a rule the nuclei belonging to the trophoblast are smaller
than those of the maternal proliferation, although this is of
course no rigorous means of distinction (cf. figs. 82 and 89).

STUDIES IN MAMMALIAN EMBRYOLOGY. 7

However this may be, a stage is soon reached—and figs. 11, 28,
and 80 are good representations of it—in which the tropho-
blast forms a continuous layer over the maternal surface, and
over the mouths of the crypts, acting as a sort of pseudo-
epithelium.

This trophoblast layer has at the same time commenced ano-
ther transformation, which is of the highest importance for
the correct interpretation of the further phenomena of placen-
tation. Of this transformation the earliest appearance of the
allantoidean trophoblast—even before it is as yet applied
against maternal surfaces—has given evidence already. We
there notice in the free trophoblast an evident tendency to
differentiate into two layers, the inner one of these generally
staining more intensely (figs. 75—78). This duplicity is also
marked in the trophoblastic knobs (figs. 27, 27 a, 78, and 79).
We may safely infer that this subdivision of the trophoblast is
the same as that which was noticed by van Beneden for the
bat, and afterwards by Masius, Duval, and others for the
rabbit and other rodents. Van Beneden applied the names of
“‘cytoblast ” and ‘ plasmodiblast” to these closely contiguous
subdivisions of the trophoblast.

We will follow his example, and henceforth designate in the
allantoidean trophoblast of Sorex the inner layer as the cyto-
blastic, the outer as the plasmodiblastic one. Both of them
rapidly increase in extent and thickness as the trophoblast
continues to spread over the maternal surface. Superficial
inspection of preparations, as those of figs. 28 and 80, would
lead us to the conclusion that the real extent of the tropho-
blast was limited to the darkly stained layer and knobs there
indicated. Still these only represent the deeper cytoblast.
The fact is that the plasmodiblast, which is the outer layer,
commences to be fused intimately with the maternal tissue in
the stage of figs. 27, 27 a, and 79, and is busily engaged in
developing lacunary and intercommunicating spaces just out-
side the limits of the darker cytoblast layer (fig. 79). These
future blood-spaces are spread out in the plane of the layer,
and thus come to be directly contiguous to the actual blood-

518 A. A. W. HUBRECHT.

vessels that are already present in the maternal proliferation,
which, as we have seen on p. 516, are also very markedly spread
out horizontally below the layer of epithelium, which has
come to disappear now that the trophoblast has taken its
place. Out of this immediate contiguity an actual communi-
cation of the maternal blood-spaces with those in the plasmo-
diblast is very soon developed, as can be clearly seen in fig. 80.
The plasmodiblast does not henceforth develop independently
of the cytoblast ; on the contrary, it is continually being added
to by cells or cell sheets from the darker stained cytoblastic
layer getting more detached and travelling inwards. This is
clearly indicated in fig. 93, which has reference to the uterus
No. 85, i.e. one stage later than (fig. 12) uterus No. 3 (fig.
11), from which fig. 80 is taken. In fig. 93 the detachment of
cytoblastic elements from the more deeply stained layer that
arrange themselves into parallel superposed layers of plasmodi-
blast, between which open spaces develop, in which blood
gradually penetrates, is very distinctly visible, as it is also in
numerous preparations of this and the earlier stages that have
not here been figured.

As further growth goes on, these horizontal blood tracts
assume a more vertical course, new horizontal ones developing
below them (cf. fig. 89).

Also in fig. 87 the penetration of blood in trophoblastic
spaces is indicated,—this time in a section that cuts the pla-
centa circularly (cf. fig. 118).

Thus, through the action of the massive trophoblast, not
only a firm attachment of the blastocyst in the placentary
region is brought about, but between the trophoblastic knobs
which have caused that firm attachment a sanguiniferous plas-
modiblast is very early established. Allantoidean villi pene-
trating into the trophoblastic knobs that become hollowed out
as they lengthen centripetally are thus bathed by maternal
blood circulating in embryonic spaces.

There is thus neither necessity nor even possibility for any
further penetration of proliferating maternal tissue between
the villi. That space is filled by the allantoidean trophoblast,

STUDIES IN MAMMALIAN EMBRYOLOGY. 519

composed of cytoblast and plasmodiblast, which are in con-
tinuous growth and in rapid increase.

Of the mother only the blood and the blood-corpuscles pene-
trate into the numerous and intricate lacune of this region.
Both the external cap of maternal tissue and the internal core
of trophoblastic tissue (with allantoidean villi embedded in it)
out of which the placentary region is composed, show the traces
of their respective growth in numerous karyokinetic figures.
In the stage of fig. 11 these are yet numerous in the maternal
crypt region; in that of fig. 13 they are already rare or absent,
and the increase of this layer may be said to have ceased. On
the contrary, the number of karyokinetic figures in the cyto-
blast is in this stage very striking. Of direct nuclear division,
without karyokinetic intermediate stages, examples are found
in the plasmodiblast.

The glands, as far as they persist in the placentary region
(sometimes with unexpected local dilatations), are not invaded
by trophoblastic knobs or villi; their openings towards the
uterine lumen are closed by the trophoblast, and they play no
part in the fixation of the blastocyst, or in the facilitation of
the intercourse between the maternal and the embryonic cir-
culation. In the latest stages of pregnancy all vestiges of the
glands have disappeared in the region between the villi.

And so we see that there is in the placentary region a
gradual substitution of maternal tissue by embryonic tissues,
corresponding in extent to what is indicated in figs. LO—15
by the extension of the red and the black divisions.

Always with this restriction, that in the region which is in-
dicated by black lines maternal blood penetrates into the
trophoblastic tissue that fills up all the white space between
the black lines.

Thus the maternal blood is transported from the maternal
arteries into channels that are wholly built up of embryonic
material, returning back along similar spaces to the superficial
(red) covering of the placental region and to the maternal
veins.

The phenomena of growth and development of the tropho-

520 A. A. W. HUBRECHT.

blastic tissue that surrounds the allantoidean villi (and through
the intervention of which the latter are bathed by maternal
blood) are thus identical with the further phenomena of growth
and development of the placenta. The latter does not contain
maternal elements other than blood. Trophoblastic tissue is
the material out of which the placenta is built up.

It is soaked with maternal blood, and allantoidean vessels
with their ramifications have been carried into it by the villi.
In the ripe placenta the villi are no longer recognisable as
such, more or less in the same way as I have formerly described
and figured this for the hedgehog (1. c., figs. 56 and 57).

And whereas in the growing placenta of Sorex a maternal
(red) and an embryonic (black) part can be distinguished, in
the full-grown one the maternal portion is no longer present
as such, but forms an insignificant and interrupted sheet of
nuclear remains between the main mass of the placenta and
the muscularis with the afferent and efferent vessels. It is in
the plane of this sheet that the severing of the placenta at birth
is effected. In the shrew it is actually shed asin the hedgehog,
and not resorbed in loco as in the mole. .

We must now attend to a few details connected with this
process of development. Starting from the figs. 81, 89, 93,
82, and 94, which have already been referred to, we must
repeat that in these stages the fusion of trophoblast and
maternal epithelial proliferation has become most intimate,
that it would be difficult to draw any sharp line of demarcation
in figs. 81, 82, and 89, but that, all the same, the difference
between plasmodiblast and maternal crypt tissue and between
trophoblastic blood-spaces and maternal capillaries is unmis-
takable. The thickness of the maternal layer in comparison
to the embryonic one is about 1:1 in figs. 12 and 89. The
active increase of the maternal crypt tissue by further pro-
liferation has now nearly come to a standstill, and henceforth
the growth of the placenta means the increase of the tropho-
blastic strands and of the villi between them. The increase of
the plasmodiblast at the cost of new layers of cytoblast was
already noticed above. But as pregnancy advances certain

STUDIES IN MAMMALIAN EMBRYOLOGY. 521

other processes of growth of these parts come under observation.
It is no longer in the lamellar fashion (of fig. 93) that the
cytoblastic elements come to be transformed into plasmodi-
blastic tissue, at least not generally. A phenomenon more
frequently observed in the later stages, when the uterine
swellings have reached the size of fig. 38 (cf. fig. 31) and
more, is represented under higher powers in figs. 90—92.
The separation between cytoblast and plasmodiblast is less
distinct. In the region immediately surrounding the villi
this was already noticed in comparing the two stages of figs. 81
and 82 that follow so closely upon each other. But whereas
in the stage of fig. 82 the distinction between cytoblast and
plasmodiblast had become difficult between the villi, fig. 94
(belonging to the same stage as fig. 82) shows that it was
yet very easy towards the lower surface of the placenta, where
the trophoblastic proliferation is most active.

Later on, however, as the comparison of the last-named
figure (94) with those just mentioned (90—92) shows, the line
of demarcation becomes less marked. There is yet a certain
difference in the staining, and many of the cytoblast nuclei are
recognisable by their more copious absorption of picro-car-
minate. But one glance at the figs. 90—92 will convince us of
the very gradual transition between cytoblast and plasmo-
diblast. Also with respect to the fact that hardly any more
cell boundaries can be distinguished between the cytoblast
nuclei; so that not only the plasmodiblast, but also the
cytoblast, should then be considered as a syncytium.

Now the new process of transformation of cytoblastic into
plasmodiblastic—i. e. into sanguiniferous—tissue consists in
the appearance of nests of nuclei marked off by a comparatively
very distinct line of demarcation (fig. 91). These nests gra-
dually separate themselves from the surface, and are apparently
carried inwards (fig. 90), where for some time they persist
(fig. 92). After that they gradually develop cavities, into which
maternal blood-corpuscles are seen to penetrate, and which
thus become part of the general circulatory apparatus of the
plasmodiblastic syncytium. Fig. 91 clearly shows, however,

Dae A. A. W. HUBRECHT.

that such blood-spaces not only arise in those nests of nuclei,
but also between them. The whole of the set of blood-spaces
in the syncytial tissue here referred to is thus gradually
developed by processes that, though closely comparable, are
not yet identical in the earlier and in the later phases of
pregnancy. The final point to which they lead up, and which
we find represented in the fully ripe placenta (figs. 32 and
52—54,), is this, that the lower surface of the placenta where
the vessels of the allantois are applied against it very much
resembles the phase of fig. 90, but that the rest of the tropho-
blastic tissue is so exceedingly attenuated between the allan-
toidean villi and their ramifications that by the stretching of
the syncytium only a very thin partition of trophoblast, with
nuclei regularly distributed in it (figs. 52, 53, and 86), sepa-
rates the maternal blood from the embryonic.

The latter, circulating in the allantoidean villi, is surrounded
by the tissue of the villus, which in the earlier stages is indeed
substantial (figs. 85 and 86). There is to each villus a central
blood-space (figs. 85 and 30) and numerous peripheral ramifi-
cations immediately under the surface (figs. 81, 82, 85, and 86).
But as the placenta increases in size, the trophoblast in tenuity,
and the villi with their ramifications in length, this primarily
more massive tissue is also extraordinarily stretched, and
finally not more thau the thickness of one cell ensheaths the
embryonic corpuscles in the villi.

I have no doubt that in the fully ripe placenta even this
covering disappears as far as the finer ramifications are con-
cerned, and that there only the thin trophoblastic partition
above referred to separates the maternal from the embryonic

blood.

We may conclude from the foregoing that the passage of
blood from the maternal vessels into the embryonic trophoblast
takes place in the shrew at a later period of the development
than in the hedgehog. For this passage communications must
necessarily originate (p. 35) between the maternal blood-vessels,
and the embryonic lacunary spaces which are intended for the

STUDIES IN MAMMALIAN EMBRYOLOGY. 523

reception of that blood. The genesis of such communications
is undoubtedly facilitated in the shrew by the fact that not
only the trophoblast is composed of young and newly formed
cells, but that the same holds good for the maternal prolifera-
tion with the crypts. The spaces which in both are destined
for the transport of maternal blood fuse, so to say, in statu
nascenti, and this helps to explain how in the shrew the
process of fusion can only be traced with so much difficulty,
and how the boundary lines between maternal and embryonic
proliferations become so very soon untraceable.

The final reduction of the maternal cryptal tissue to a layer
of nuclear nests interspersed between the placentary tissue and
the muscularis (see figs. 32 and 54) need not be discussed in
detail. The reduction is a gradual one, and partly figured in
fig. 86, where the maternal tissue between this cryptal prolifera-
tion and the muscularis has not yet developed anew—rearranged
itself preparatory to parturition, as it has already done in
fig. 54.

The more deeply stained maternal nuclear nests just referred
to are thrown off with the placenta. The way in which the
maternal surface regenerates after parturition will not be here
discussed, but will be reserved for another paper, in which I
will then include comparative considerations with respect to
other genera of Insectivora (Hrinaceus, Talpa, and Tupaja).

524 A. A. W. HUBRECHT.

EXPLANATION OF PLATES 31 to 39,

Illustrating Professor A. A. W. Hubrecht’s paper ‘‘ Studies
in Mammalian Embryology. IIJ.—The Placentation of
the Shrew (Sorex vulgaris).

List of Abbreviations.

a. T, Allantoidean trophoblast. o. 7. Omphaloidean trophoblast. ¢r. az.
Trophoblastic annulus. az’. Embryonic cells which grow downwards from
the upper rim of the trophoblastic annulus and adhere against the maternal
tissue. zp. 7. Non-placental trophoblast. 4. Trophoblastic knobs which
become hollowed out and into which allantoidean villi penetrate. U. Lumen
of the uterus. w. #. Uterine epithelium. g/. Uterine glands. M. Meso-
metrium. ep. Epithelial crypts in the proliferating portion of the maternal
mucosa. §. Strands of epithelial tissue between the superficial and deeper
maternal layers in the omphaloidean region. 4/. Blood-vessels. dm. Longi-
tudinal, and em. Circular fibres of the muscularis. y. Yolk-sac. a.v. Area
vasculosa. Z#.c. Extra-embryonic celoma. am. Amnion. pr.am. Proamnion.
all. Allantois. vz. Allantoidean villus. g.c. Green coagulum in the yolk-
sac. mes. Mesoblast. s.m. Somatic mesoblast. Ay. Hypoblast. az. Nuclear
nests in the syncytium (incipient blood-spaces).

PLATES 31 and 32.

Outline sketches of median transverse sections through uterus and embryo
in successive stages of pregnancy of Sorex vulgaris. All the figures
drawn with the camera. xX 27. Embryonic tissue in black, maternal
tissue in red outlines. The red numbers in brackets refer to the catalogue |
number of the specimen. The small numerals connected with dotted lines
refer to figure-numbers on the subsequent plates in which the region thus
indicated is represented as seen under stronger powers. Cf. Figs. 36—47
for the outward aspect of the uterus in the stages from that of Figs. 8—15.

Figs. 1 and 2.—The blastocyst has not yet arrived in the cavity of the
uterus, but is still contained in the oviduct.

Fig. 1.—Utr. Mus. Cat. n Sorex 124 7,37. 45.
Fig. 2.— “1 5A 1102,27r.16s.

Fries. 3 and 4.—Sudden and peculiar extension of the uterus lumen and of
the wall, the latter having decreased in thickness considerably opposite the
mesometrium. The blastocyst has arrived in the uterus.

Fig. 3.—Utr. Mus. Cat. n* Sorex 24,5 r. 21s.
Fig. 4.— or aA » 026,37. 268.

STUDIES IN MAMMALIAN EMBRYOLOGY. 525

Fies. 5 and 6.—Proliferation of the maternal epithelium having extended
over the surface opposite the mesometrium, the uterine wall has here again
thickened. The blastocyst has become attached to the maternal surface by
means of the omphaloidean trophoblast.

Fig. 5.—Utr. Mus. Cat. n* Sorex 73 f, 27.19 s.
Fig. 6.— x x 45¢c,3r.14s.

Fic. 7.—First appearance of amnion, trophoblastic annulus and area vascu-
losa. Dehiscence in lateral maternal tissue, which has commenced in the
stage of Fig. 5, has here made rapid progress.

Fig. 7.—Utr. Mus. Cat. n° Sorex 42 ¢,37.11s.

Fies. 8 and 9.—Proamnion and amnion simultaneously in process of forma-
tion. Secondary maternal epithelial crypts at their maximum of independent
development. First appearance of trophoblastic knobs on the allantoidean
trophoblast. Area vasculosa does not as yet extend down to trophoblastic
annulus.

Fig. 8.—Utr. Mus. Cat.n* Sorex 519,37. 29s.
Fig. 9.— 5S op 51d,2,47r.2s.

Fic. 10.—Amnion closed and completed; allantoidean trophoblast partly
applied against concavity of future placentary region; trophoblastic knobs
just entering epithelial crypts. First appearance of allantois as free knob
projecting in extra-embryonic ceelom. Area vasculosa completed down to
trophoblastic annulus. Active participation of the latter in resorbtion of
hemorrhagic extravasates consequent on disintegration of lateral tracts of
maternal tissue.

Fig. 10.—Utr. Mus. Cat. n* Sorex 106 e,37.9s.,and1062,17.9s.

Figs. 11 and 12.—The trophoblastic excrescences have penetrated into the
secondary epithelial crypts and have in their turn become hollowed out.
Villi of the allantois penetrating into these trophoblastic cavities. Epithelium
of maternal surface and crypts has disappeared wherever embryonic trophoblast
is in contact with it. In Fig. 11 the area vasculosa commences to loosen its
hold upon the lateral maternal surface; in Fig. 12 it is hardly any longer
adherent. The actual distance between the trophoblastic annulus and the
placentary region has at the same time diminished. The embryo is being
sunk into the yolk-sac, thus pushing before it the area vasculosa, which is
thereby turned inside out. The differentiation of the allantoidean trophoblast
into a cytoblastic and a plasmodiblastic portion has commenced. In Fig. 12
the last traces of the ensheathing of the allantoidean villi by distinct cytoblast
are yet visible as a thin black line.

Fig. 11.—Utr. Mus. Cat. n* Sorex 36,2,17.9s.
Fig. 12.— a 35 85 9,3 7.1 5s.

Fies. 11 a@ and 11 4.—Two sections along a plane perpendicular to all the
foregoing. Same stage as that from which Fig. 11 was taken. The planes of
sections accompanied by the corresponding numbers are indicated in Fig. 11
by lines and numerals, In Fig 11 a only secondary crypts of maternal epithe-

526 A. A. W. HUBRECHT.

lium and uterine glands, g/. (flattened against the outer circumference), are cut,
the former being indicated by dotted lines. In Fig. 11 4 trophoblastic pro-
tuberances, which have already penetrated into the more centrally situated
crypts, are indicated by thin black lines. Other dotted lines inside of these
refer to allantoidean villi filling up the cavity of the hollow trophoblastic

protuberances.
Fig. 11 a.—Utr. Mus. Cat. n* Sorex 3 a, 2,5 7.19 s.
Fig. 11 6.— s 55 By Oe (eee leh

Fic. 18.—The visible contrast between the maternal secondary crypts and
the trophoblastic ingrowth, ensheathing allantoidean villi, has considerably
diminished, making it more and more difficult to distinguish the maternal
from the embryonic histological elements in the placentary region. Of the
crypts, however, the peripheral blind ends are yet preserved and sufficiently
distinct. In the trophoblast we notice a region of special activity on the
concave placental surface where the cytoblastic thickening is particularly con-
siderable, secondary allantoidean villi that are rapidly forming being ensheathed
by it. The increase in size of the placenta is accompanied by the total dis-
appearance of the lateral maternal tissue against which the area vasculosa has
been applied. The whole of the area vasculosa (which has not stopped
growing, but, on the contrary, remains on the increase) is henceforth situated
below instead of above the trophoblastic annulus. The attachment between
the latter and the border of the placenta has become a circular membrane.
The head of the embryo is still enclosed in a proamnion, the cavity of the yolk-
sac is being considerably encroached upon by the embryo, and the inverted
area vasculosa is sinking down into it.

Fig. 13.—Utr. Mus. Cat. n° Sorex 130 4,3 7.15.

Fie. 14.—There is a considerable increase in the size of the placenta, but
the relation of the parts has remained very much the same to what it was in
Fig. 13, The layer of proliferating trophoblast is no longer so exceptionally
thick on the concave placental surface. The trophoblast (and the allantoic
villi embedded in it) occupies seven eighths of the thickness of the uterine
wall in the placental region. The remnants of the cryptal region have been
still more considerably reduced. The trophoblastic annulus has undergone
no further modification. The area vasculosa has extended considerably.

Fig. 14.—Utr. Mus. Cat. n° Sorex 61 a, 27.17 s.

Figs. 15 and 15 a.—The border region of the placenta just before birth.
The actual thickness has hardly increased when compared to Fig. 14, but the
ramification of the allantoic villi and the intervening strands of trophoblastic
tissue is ever so much more complicated. The maternal epithelial prolifera-
tion has now finally become reduced to mere nuclear remnants, staining more
deeply with carmine. The thickness of the maternal tissue outside these
nuclear elements has somewhat increased when compared to Fig. 14. In
Fig. 15 @ peculiar foldings of the wall of the yolk-sac are indicated.

Fig. 15.—Utr. Mus. Cat. n* Sorex 100,37. 10s.
Fig. 15 a,— 35 .. 100,1 7.8 s,

STUDIES IN MAMMALIAN EMBRYOLOGY. 527

Fie. 15 4.—The wall of the uterus (thick red line), the non-placental
trophoblast and the yolk-sac in the vicinity of the mesometrium, drawn on
the same scale and from a similar preparation as Figs. 15 and 15 a.

Fig. 15 6.—Utr. Mus. Cat. n° Sorex 90a,17.8s,

PLATE 33.
(All these figures x 100.)

Fie. 16.—Part of a transverse section (cf. Fig. 1) through a uterus in
which the blastocysts are as yet contained in the oviducts. The coiled uterine —
glands are seen to be massed together in the antimesometrical regions. The
uterine lumen is more or less |-shaped. Opposite the mesometrium the
uterine epithelium is much higher now than in the stages of Figs. 17 and 19.

Utr. Mus. Cat. n° Sorex 124 f,3 7.3 s.

Fie. 17.—Part of a section through the wall of the uterus opposite the
mesometrium, in a later stage of pregnancy (cf. Fig. 3 and Fig. 66). The
epithelium is flattened consequent upon the stretching of the wall and the
distension of the lumen.

Utr. Mus, Cat.n® Sorex2a,57.18s.

Fic. 18.—Ibid., through the region where the uterine wall of the same
specimen is thickest (cf, Fig. 3). The epithelium has as yet not entered
upon any proliferating process.

Utr. Mus. Cat, n* Sorex 2@,57. 25s.

Fic. 19.—The same as Fig. 17, but of a somewhat later stage (cf. Fig. 4).
Proliferation of the epithelium has commenced (cf. Figs. 67—69).
Utr. Mus. Cat. n° Sorex 52 ¢e,37. 24s.

Fic, 20.—The same as Fig. 18, but in a stage in which the proliferation of
the lateral uterine epithelium is already considerable (cf. Fig. 4). This pro-
liferation differs in character from that of Figs. 67—71, there being no
formation of crypts. The omphaloidean trophoblast will very soon become
adherent against it. Blood circulates in capillary spaces between these pro-
liferated epithelium cells.

Utr. Mus. Cat. n® Sorex 52¢,3 7, 24s.

Fies. 21—26.—Later stages of a portion of the uterine wall opposite the
mesometrium (cf. Figs. 4—9). The secondary epithelial crypts which originate
by the peculiar proliferation of the uterine epithelium, that commenced in
Fig. 19, gradually attain their full development in Figs. 25 and 26 (ef. Fig.
74). In the latter figure many of the crypts are not indicated because they
are flattened and pressed together. Maternal blood-vessels are everywhere
present between the crypts. In Fig. 26 the amnion is nearly complete, and

528 A. A. W. HUBRECHT.

the allantoidean trophoblast is beginning to be applied against the uterine

epithelium.
Fig. 21.—Utr. Mus. Cat. n* Sorex 52%, 3 7. 22 s.
Fig. 22.— : 2 73. f, 21. 20 s.
Fig. 23.— 2 A 73.b,2,1 7.16 s.
Fig. 24,— = oD 450,387.17 s.
Fig. 25.— p 55 42¢,37.8s.
Fig. 26.— on PA 51f, 47.15 s.

Fies. 27 and 27 a.—The same, with the allantoidean trophoblast applied
against the uterine wall and the trophoblastic protuberances, £., fitting into
the secondary epithelial crypts, ev. The specimen from which these figures
-were taken having been differently preserved from all the others, and the
histological details of the proliferated epithelium being less distinct, this
region is indicated by a blank space in Fig. 27. There are unmistakable
indications of the separation of this trophoblast into a cytoblastic and a plas-
modiblastic portion (cf. Fig. 79).

Fig. 27.—Utr. Mus. Cat. n° Sorex 106 f,3 7. 12s.
Fig. 27 a.— A; + 106 f, 37.45.

Fic. 28.—The same, in a yet later stage. The trophoblastic protuberances,
k., adhere firmly in the epithelial crypts, ey., the protuberances being partly
hollowed out and filled out by incipient allantoic villi, v7. Plasmodiblast
extends between the darker zone of the cytoblast and the protuberances.

Utr. Mus. Cat. n°> Sorex 36,3 7.17 s.

PLATE 34.
Figs. 29—34 x 100. Fig. 35 x 13. Figs. 36—47 natural size.

Fic. 29.—Fragment of a section through a yet later stage (cf. Fig. 12).
The allantoidean trophoblast has already considerably developed; the cyto-
blastic portion of it has been more deeply stained by the picro-carminate.
P. Tissue belonging to the maternal epithelial proliferation. 7. Tissue be-
longing to the trophoblast (cytoblast + plasmodiblast). The sanguiniferous
villi of the allantois that penetrate into this have on purpose been entirely
omitted in this figure. White spaces in this figure thus indicate where they
are situated. Only the lower boundary line of the allantois is indicated,
with a few excrescences adhering against the trophoblast and representing
incipient new villi (cf. Fig. 30). Comparison with Figs. 12, 81, and 89 will
further elucidate this preparation. The difference should be noticed between
the extent and situation of the area vasculosa in this and in the next figure
(cf. also Figs. 12 and 13).

Utr. Mus. Cat. n* Sorex 859,2 7.175.

Fie. 30.—The same, in a later stage (cf. Figs. 13, 82, and 94). The
cytoblastic trophoblast is only here and there visible as a distinct layer out-
side the allantoic villi; generally it cannot be readily distinguished from the

STUDIES IN MAMMALIAN EMBRYOLOGY. 529

intervening trophoblast (plasmodiblast) surrounding the blood-spaces (ef. Figs.
18, 82, and 94). On the concave free surface of the placenta, however, the
cytoblastic trophoblast has attained to a maximum of thickness. It is here,
moreover, more or less honeycombed, incipient new allantoidean villi entering
into those recesses and being there attached and subsequently vascularised. In
this figure the connective tissue of the allantoic villi has been indicated, but the
spaces in these villi through which the embryonic blood-corpuscles circulate
have been left open (white). Two full-grown double villi, both of them
bifurcating, are here indicated, and three or four incipient ones. The tropho-
blastic layer, az’, being the continuation of the trophoblastic annulus (cf. Figs.
9—13, 50, 84, and Figs. 33 and 34), is closely applied against the allantois,
but does not in any way fuse with it. A uterine gland, with partly distended,
partly flattened lumen, is visible in this preparation. P. and 7. as in
Fig. 29.
Utr. Mus. Cat. n° Sorex 130 4,37.10s.

Fie. 31.—A yet further stage of placentation. The trophoblastic region
in which the allantoidean villi are embedded occupies about five sixths of the
thickness of the wall; the epithelial proliferation and gland-remains about
one sixth. Here, too, the embryonic blood-corpuscles are not indicated, and
the parts of the villi where they circulate remain white. The wide allan-
toidean vessel which supplies the three to four villi here represented is only
indicated in outline, as is also the lower boundary line of the allantois. The
maternal proliferation is yet more reduced, and of the crypts no distinct
remains persist (cf. Figs. 14 and 86).

Utr. Mus. Cat. n* Sorex 6l1a,27r. 3s.

Fic. 32.—A fragment of a section through the ripe placenta, perpendicular
to the surface, with low power (cf. Figs. 15 and 15a). The spaces in which
embryonic and maternal blood-corpuscles circulate, although interwoven ina
most complex manner, can be distinguished (a) by the size of the blood-
corpuscles, which are very much smaller if maternal, (4) by the fact that in the
syncytium the maternal corpuscles circulate in spaces without any trace of
endothelium, whereas round the embryonic corpuscles the traces of the
endothelium of the allantoidean vessels (which in this figure are again repre-
sented by white spaces) are often preserved as flattened nuclei (cf. Figs.
52—54),

Utr. Mus. Cat. n* Sorex 100,27.12s.

Fic. 33.—The trophoblastic annulus, terminal portion of area vasculosa,
and downward membranous continuation by which the trophoblastic annulus
is connected with the placental border, in the stage of the uterus No. 77 (cf.
Figs. 85 and 44). To the right of the membrane, az’, a transversely cut
allantoidean vessel is indicated (compare with Fig. 14).

Utr. Mus. Cat. n* Sorex 77 4a,3 7.45.

Fie. 34,—The same for a yet later stage, viz. that of uterus No. 100. The
VOL. 35, PART 4.—NEW SER. 00

530 A. A. W. HUBRECHT.

area vasculosa is here not indicated; only its point of attachment is marked
by a thin line.
Utr. Mus. Cat, n* Sorex 100,27.1s.

Fic. 35.— View of the inside of the uterus (in a stage corresponding to Fig.
44, which is somewhat further advanced than Fig. 14) after the mesometrical
half has been dissected away. The view is thus directed towards the inner
surface of the placenta. Three considerable allantoidean vessels, pressed in
between the amnion (which is not indicated in the figure) and the area vascu-
losa turned inside out (cf. Fig. 14), are seen to take their course towards a
central ring-shaped opening, through which they disappear. This ring is
the upper constriction of the trophoblastic annulus. The area vasculosa is
attached to it immediately below this rim. There is in the figure an irregular
lighter centre, and a darker outer circumference. The latter is continued on
the (here absent) inner surface of the mesometrical half. This darker colour
is caused by the special green pigment of the hypoblast cells of the yolk-sac.

Prep. Sorex n*™ 77.

Fics. 36—47.—Different phases of pregnant uteri of Sorex vulgaris
following on the earlier stages that were represented in vol. xxxi of this
Journal, Pl. XXXVI (and following) figs. 13, 15, 43, 65, and 83. The
figures are natural size, drawn after the spirit specimens; the catalogue
number between crotchets.

Fig. 36.—Uterus, No. 101 (one of the swellings; cf. Figs. 60, 61,90—92).

Fig. 37.— Uterus, No. 106. a—A. The individual swellings, each of them
containing an embryo (cf. Figs. 10, 27, 76, 79, 84).

Fig. 38.—Uterus, No. 61 (one of the swellings; cf. Figs. 14, 31, 58).

Fig. 39.—Uterus, No. 130 (ef. 13, 30, 82, 85, 94).

Fig. 40.—Uterus, No. 85 (ef. Figs. 12, 29, 81, 88, 89, 93).

Fig. 41.—Uterus, No. 90 (one of the swellings).

Fig. 42.—Uterus, No. 26 (one of the swellings; cf. Figs. 57, 59, 62).

Fig. 43.—Uterus, No. 51 (cf. Figs. §, 9, 26, 50, 55, 75, 77, 78).

Fig. 44.—Uterus, No. 77 (one of the swellings; cf. Figs. 33, 35).

Fig. 45.— Uterus, No. 100 (one of the swellings; cf. Figs. 15, 32, 52—54).

Fig. 46.—Uterus, No. 80 (one of the swellings; cf. Fig. 86).

Fig. 47.—Uterus, No. 3 (cf. Figs. 11, 28, 56, 80, 87).

PLATE 35.
Figs. 52—54 enlarged x 200. Figs. 48—5], 55—65 enlarged x 260.

Fic. 48.—First appearance of the trophoblastic annulus in a stage corre-
sponding to Fig. 6 (though not from the same preparation). About a dozen
cells in each section take part in the formation of the annulus. These cells

STUDIES IN MAMMALIAN EMBRYOLOGY. 531

are as yet only somewhat enlarged. They are uot the same enlarged tropho-
blast cells that were figured for Sorex on Pl. XXXVII, fig. 27, vol. xxxi of
this Journal. The latter have become applied against the maternal tissue,
and have given origin to the omphaloidean trophoblast. The shred of
omphaloidean trophoblast (0. 7.) that is here figured gives ample indication
of active histolysis going on in this portion of the trophoblast.. In the
preparation the shred here figured is not applied against any maternal tissue,
and its trophoblastic derivation is thus all the more indubitable. It is im-
portant to determine this, as it will be seen in Figs. 49—51 how eminently
difficult is the unravelling of maternal and trophoblastic tissue in the ompha-
loidean regions as soon as these two have become contiguous.
Utr. Mus. Cat. n* Sorex 45 c,2 7. 28 s.

Fic. 49.—In this figure the trophoblastic annulus is rather less than more
distinct than in the former (cf. Fig. 7). The omphaloidean trophoblast, is
undergoing a granular transformation simultaneously with degenerative
phenomena in the maternal tissues.

Utr. Mus. Cat. n* Sorex 42 e, 27.14.

Fie. 50.—The maternal tissue in partly degenerative resorption (cf. Fig. 8),
in consequence of the fusion with trophoblastic tissue. To the right strands
of epithelial tissue (s.) keep up the connection between the superficial and
the deeper maternal layers not indicated in this figure. These latter hecome
transformed into a new uterine epithelium (cf. Figs. 84 and 88). The fusion
between omphaloidean trophoblast and maternal tissue is very complete in
this preparation just above the upper rim of the trophoblastic annulus. The
hypoblastic nuclei are distinct. There are traces of blood-extravasate between
the trophoblastic annulus and the maternal tissue (cf. Fig. 84).

Utr. Mus. Cat. n° Sorex 510,27.17s.

Fig. 51.—A yet earlier phase in the attachment of omphaloidean trophoblast
against the lateral maternal epithelial proliferation (cf. Fig.5). A considerable
portion of the trophoblast has undergone a granular metamorphosis. Above
this there is another region where the distinction between embryonic and
maternal proliferated tissue is indeed impossible. In this stage a tropho-
blastic annulus is not yet even indicated.

Utr. Mus. Cat. n® Sorex 73 ¢,3 7.13 s.

Fies. 52—54.—Three regions of the placenta that was represented in
Fig. 32, more considerably enlarged. Fig. 52 from the lower, Fig. 53 from
the middle, Fig. 54 from the upper portion. The difference in size between
embryonic and maternal blood-corpuscles (the former being by far the larger)
renders distinction between maternal and embryonic blood-spaces very easy.
The preparation may be considered as a very perfect self-injection. The
maternal blood circulates in spaces delimitated on all sides by trophoblast that
is spread out to the utmost degree of tenuity. The embryonic blood circulates
in similar spaces, nowhere communicating with the former, and showing here

532 A. A. W. HUBRECHT.

and there flattened remnants of the endothelium of the vessels of the allan-
toidean villi. In Fig. 52 the embryonic blood-spaces are wider than in Fig.
53. In Fig. 54 no blood-corpuscles are indicated, whereas the remnants of the
secondary epithelial crypts are visible as nuclear conglomerates. C., peculiar
corpuscles in the deeper maternal layers.

Utr. Mus. Cat. n° Sorex 100, 27.12 s.

Fies. 55—57.—Three figures of different phases of development of the non-
placental trophoblast aud of the extenuated uterine wall. Fig. 55 taken
from the phase of Figs. 8 and 9. The lower wall of the blastocyst is in this
phase comparatively thick, the hypoblast even much more so than the non-
placental trophoblast. In Fig. 56 (corresponding to Fig. 11) the same layers
have become more considerably attenuated, and so has the uterine wall, the
epithelium of which is consequently no longer thrown into any folds. In
Fig. 56 a thin layer of a greenish granular coagulum is applied against the
inner surface of the yolk hypoblast. In Fig. 57 this has considerably increased
in thickness, and can be traced in the lower half of the yolk-sac all round.
The greenish coagulum has parallel surfaces. To the left of gc. in Fig. 57 a
further granular precipitate is indicated, which in this preparation fills part of
the crescentic yolk-sac. In comparing Fig. 57 with 56 we see that the
further extenuation of the muscular layers of the uterus is especially marked.

Fig. 55.—Utr. Mus. Cat. n* Sorex 51¢,47.13s.
Fig. 56.— mn 55 36,47r, 458.
Fig. 67.— op as 26a,37.10s.

Fies. 58—65.—Hight different regions and different stages of development
of that surface of the yolk-sac, which becomes inverted from the stage of
Fig. 12 onward. This is the surface, stretching between the trophoblastic
annulus and the embryo, on which the area vasculosa develops. After the
inversion the vascular region increases in surface, and so does the number of
the vessels. The hypoblast cells increase in size, proliferate freely, and
acquire a deep green tinge. Neoformation of blood-corpuscles is very actively
going on in this region.

Fig. 58.—Utr. Mus. Cat. n* Sorex 61 a,27.10s.

Fig. 59.— + of 26a,37.15s.
Fig. 60.— as oh 101 a, 3,3 7. 15s.
Fig. 61.— x ss 101 a, 3,2 7.12 s.
Fig. 62.— - e 26 a,27.10s.
Fig. 63.— 5 5 92 a,2,17.10s.
Fig. 64.— 53 y 100,17.8s.

Fig. 65.— a Bs 100,17.8s. ‘

STUDIES IN MAMMALIAN EMBRYOLOGY. 533

PLATE 36.
All the figures in this Plate x 480.

Fic. 66.—Part of the uterine wall opposite the mesometrium, in the phase
of Fig. 8 (cf. Fig. 17). The stretching of the uterine wall is very consider-
able ; the blastocyst has not yet become attached.

Utr. Mus. Cat. n* Sorex 24,51. 23 s.

Fie. 67.—Part of the same, one stage further (cf. Figs. 4 and 19). Pro-
liferation of the uterine epithelium has commenced. Karyokinetic figures
demonstrate the participation of the epithelium cells.

Utr. Mus. Cat. n° Sorex 526,47. 20s.

Fie. 68.—Section through another swelling of the same uterus, showing
the mouth of one of the uterine glands (cf. Figs. 4 and 21).
Utr. Mus. Cat. n™ Sorex 527,37: 21s.

Fie. 69.—Section through the same swelling from which Fig. 67 was taken,
somewhat further from the mesometrium (cf. Fig. 4). The proliferated
epithelium cells have commenced to arrange themselves in a radial fashion.
Uterine epithelium as yet continuous. A few cells in the border region,
between the epithelial proliferation and the connective tissue of the uterine
wall, are seen to become spindle-shaped in Figs. 68 and 69. This is later on
further accentuated into a special fibriform layer (cf. Figs. 72—74, and also
Figs, 22—25).

Utr. Mus. Cat. n®* Sorex 520,47.15s.

Fie. 70.—Section through a further stage of development. The prolifera-
tion forms already a thicker layer, the uterine epithelium breaks away from
under the radially arranged groups of cells, which are going to be the secon-
dary epithelial crypts.

Utr. Mus. Cat. n° Sorex 73f, 3 7. 22 s.

Fic. 71.—Another section through another swelling of the same uterus,
The passage of the uterine epithelium into that of the crypt is already more
complete, but quite different from the mouth of a gland.

Utr. Mus. Cat. n* Sorex 73¢,37,. 14s.

Fic. 72.—The boundary region between the epithelial proliferation and the
connective tissue of the uterine mucosa. The blood-vessels in the latter are
more spacious than the capillary ducts in the former. Of the uterine glands
one is partly cut along the lumen, the other only tangentially through the wall.

Utr. Mus. Cat. n* Sorex 73 6,2,37, 8s.

Fic. 73.—The same, in a somewhat later stage. The proliferated epithe-
lium cells are more closely packed, and between them and the connective
tissue there is a layer of fusiform cells, first noticed in Fig. 69. This section
was somewhat too tangential to show the secondary crypts well (cf. Fig. 25,
which has reference to the same uterus).

Utr. Mus. Cat. n° Sorex 42 ¢,27.13 s.

534 A. A. W. HUBRECHT.

PLATE 37.

Fig. 80 x 200. All the other figures x 260. Both the maternal and the
embryonic blood-corpuscles are coloured red on this Plate.

Fic. 74.—Part of a section through the future placentary region when
the proliferation of the maternal epithelium, which precedes the adhesion
of the blastocyst, has attained its maximum of development. The crypts
reach through the whole thickness of the proliferation (cf. Figs. 6 and 24).

Utr. Mus. Cat. n®* Sorex 45¢,37.17s.

Fic. 75.—One of the proliferating knobs of allantoidean trophoblast
penetrating into the mouth of a secondary epithelial crypt in the future
placentary region. Large nuclei reveal other epithelial crypts tangentially
cut (cf. Figs. 8 and 9).

Utr. Mus. Cat. n* Sorex 51¢e27.8s.

Fie. 76—79.—Four different portions of the allantoidean trophoblast
before its adhesion against the maternal tissue. In all—but most especially
in Fig. 79—the differentiation of cytoblast.and plasmodiblast has commenced.

In Fig 76 the amnion is just being completed.

In Figs. 78 and 79 distinct and massive trophoblastic knobs are re-
presented.

Fig. 76.—Utr. Mus. Cat. n* Sorex 106 ¢,27.19s.

Fig. 77.— 5 es 5la,27. 24s.
Fig. 78.— 55 > 51 ¢c,5 7. 20s.
Fig. 79.— 5 i 106 f, 47.15 s.

Fie. 80.—A similar section to that of Fig. 28 (ef. also Fig. 11), more con-
siderably enlarged. The allantoidean trophoblast and its protuberances have
become fused with the maternal tissue, and have further proliferated under
partial destruction of the latter. The trophoblastic knobs are being hollowed
out, allantoidean villi penetrating into the cavities thus originating. The
differentiation of plasmodiblast between the protuberances has commenced.

Utr. Mus. Cat. n* Sorex 36,2,17.17s.

Fie. 81.—Part of the placentary region of Fig. 12, more considerably
enlarged. The upper half of the figure represents maternal tissue (secon-
dary epithelial crypts, with blood capillaries between them), the lower half
trophoblastic tissue. Below this is the allantois, with a distended blood-
vessel and two villi. The trophoblast is subdivided in a more deeply stained
cytoblast (which is the superficial layer, and which visibly ensheathes the
villi) and an intervening mass of cells (plasmodiblast) in which the maternal
blood circulates. In this figure these two have been torn asunder in the
lower part of the figure; in more normal circumstances they firmly adhere
together. The exact boundary line between the “ plasmodiblast ” and the
maternal tissue cannot be distinctly indicated; it takes its course somewhere

STUDIES IN MAMMALIAN EMBRYOLOGY. 535

between the tops of the villi. With Fig. 81, Figs. 29, 89, and 93 should be
directly compared, as they are all preparations from the same uterus.
Utr. Mus. Cat. n* Sorex 85 f, 37.13 s.

Fic. 82.—A section through a corresponding region in a Jater stage (cf.
Fig. 18). The plasmodiblast between the villi is being attenuated as the
number and the length of the villi increase. The more deeply staining layer
of cytoblast surrounding the villi is no longer distinct; its elements have
assimilated with the intervening plasmodiblast. The passage of blood from
a maternal vessel, with distinct endothelium, into the trophoblastic blood-
spaces can be distinctly traced in this and the neighbouring sections. The
cytoblastic layer of the trophoblast, if it is no longer distinct round the
villi, is all the more massive on the free concave surface of the placenta,
which is, however, not represented in this figure (cf. Figs. 30 and 94, taken
from the same uterus).

Utr. Mus. Cat. n* Sorex 180 a,27.138 s.

PLATE 38.
Figures 83—85 x 260. Fig. 84a x 60. Fig. 86 x 480.

Fie. 83.—Fragment of a section through the region of the omphaloidean
trophoblast and the adjoining maternal tissue, in an early stage (cf. Fig. 6).
One of the first phenomena of dehiscence in the deeper proliferated maternal
layers is visible as a solution of continuity in the middle of the lower half
of the figure. The omphaloidean trophoblast is considerably thickened—
without as yet being adherent to maternal tissue—in the upper half of the
figure; in the lower half adhesion has come about, and at the same time
considerable histolytical transformations have commenced. Maternal uterine
epithelium is only intact in the very topmost portion of the figure. Maternal
blood is seen to pass from the blood-spaces of the maternal proliferation
into the syncytial tissue, in the region where maternal and trophoblastic
elements have fused together. Remnants of compressed uterine glands are
seen between the muscularis and the epithelial proliferation.

Utr. Mus. Cat. n* Sorex 45¢,57.65.

Fic. 84.—Section through the uterine wall in the region of the tropho-
blastic annulus (cf. Fig. 10). There is a considerable blood-extravasate
between the annulus and the tissues that are in process of resorption.
Blood-corpuscles are being actively absorbed into the protoplasm of the cells
of the trophoblastic annulus. In the region of the omphaloidean trophoblast
there is a very intimate fusion between embryonic and maternal tissue. In
the deeper layers of the proliferated maternal epithelium dehiscence is
actively going on, and a fresh layer of uterine epithelium is developing below
the tissues that are being resorbed. From the upper rim of the trophoblastic

536 A, A. W. HUBRECHT.

annulus a cell-layer (az’.) descends, that is parallel to the annulus but adherent
to the maternal tissue. The hypoblast adhering against the trophoblastic
annulus is cut tangentially in the upper part of the figure. Its continuation
along the omphaloidean trophoblast, missing in this section, is again present
in the neighbouring sections. aa
Utr. Mus. Cat. n* Sorex 106 a,37.15s.

Fig. 84.a@.—Pseudo-karyokinetic and other phases of the nuclei of Fig. 84

in the region where histolysis is most actively going on.

Fie. 85.—An allantoidean villus in an early stage ; massive, with a central
blood-vessel and peripheral blood-spaces communicating with that central
one (cf. Figs. 30, 81, 82, and 86).

Utr. Mus. Cat. n* Sorex 180 4,2 7.13 s.

Fic. 86.—-The tops-of two other allantoidean villi, in a later stage of
pregnancy (the embryonic blood-corpuscles are here represented as red discs,
the maternal as dots). The tissue of the villus is less compact, the blood-
spaces in it are arborescent and partly intracellular. Between the two villi,
as well as right and left of them, strands of trophoblastic tissue are pictured,
carrying spaces with the so-much smaller maternal blood-corpuscles. It has
here already become considerably more difficult to distinguish between the
tissue of the villus and of the intervening trophoblast than it was in Figs.
81 and 82. On the other hand, it is here not yet quite so difficult as in the
still later phases of pregnancy that are represented in Figs. 52 and 53. There
is as yet hardly any other tissue between the muscularis and the maternal
epithelial proliferation, as far as it is yet preserved (cf. Figs. 13, 14, 15,
and 54).

Utr. Mus. Cat. n* Sorex 804a,37.558.

PLATE 39.
Figs. 87, 90—92 x 480. Fig. 88 x 100. Figs. 89, 93, 94 x 260.

Fig. 87.—A portion of Fig. 11 4, more considerably enlarged. In the free
space in the middle of the figure an allantoidean villus is going to penetrate.
The layer of darker stained nuclei surrounding that space is the deeper
(cytoblastic) layer of the trophoblast. Blood already circulates in tropho-
blastic spaces.

Utr. Mus. Cat. n* Sorex 3 a, 2,5 7. 22s.

Fic. 88.—The region of the trophoblastic annulus of Fig. 12, under higher
power. Lateral maternal proliferated tissue already considerably resorbed.
New epithelium that has originated behind this (cf. Fig. 84) is further
developed and conspicuously folded.

Utr. Mus. Cat. n* Sorex 85 a,27.14s,

Fie. 89.—Another part of Fig. 12, more considerably enlarged, further to

demonstrate the fusion between the proliferated maternal and trophoblastic

STUDIES IN MAMMALIAN EMBRYOLOGY. 537

tissues. Blood circulates in the maternal proliferation in capillaries that
generally have a distinct endothelium. In the trophoblast it circulates in
intra- and intercellular cavities. Two allantoidean villi are here only indi-
cated by their embryonic blood-corpuscles. Between these villi the cyto-
blastic and plasmodiblastic portion of the trophoblast is situated, the two
having become artificially separated from each other as in Fig. 81. It is
impossible to distinguish between those cells that have a maternal origin and
those that are derived from the embryo, otherwise than genetically.
Utr. Mus. Cat. n° Sorex 85 f,3 7. 15s.

Figs. 90—92.—Three portions of the allantoidean trophoblast taken from
one and the same series of sections of uterus No. 101 (Fig. 36). These
sections indicate how the allantoidean trophoblast proliferates and is trans-
formed into intercommunicating blood cavities, nuclear nests breaking away
’ from the outer layers of the syncytium, and arranging themselves into blood
channels by processes of stretching and hollowing out. In principle it is this
same process by which, in the earlier phases such as that of Figs. 79, 80, and
93, the trophoblast has developed and become the peculiar sanguiniferous
interstitial tissue between the villi.

Fig. 90.—Utr. Mus. Cat. n* Sorex 101 4,1,37.5s.
Fig. 91.— BS op 101 a4, 1,17. 10s.
Fig, 92.— 53 % 101 a@,1,27.10s.

Fic. 93.—The same in an earlier stage (uterus No. 85, Fig. 40; ef. also
Figs. 12, 29, $1, and 89). The darker layers of cytoblast have the somewhat
more faintly stained plasmodiblast between them. The latter is in this stage
seen to be split off from the former in parallel layers that develop blood-
spaces between them (cf. Figs. 29, 30, and 89). The cytoblast at the top of
the figure belongs to a cavity in which an allantoidean villus is situated (cf.
Fig. 29, which represents the same stage).

Utr. Mus. Cat. n°. Sorex 85 a, 27, 23 s.

Fig. 94.—The same, one stage later (cf. Figs. 18, 30, and 82) at the
period that the concave layer of cytoblast has attained its maximum of
development. Here, too, the newly-formed blood-spaces are primarily
arranged in a more lamelliform way close to the inner layer, whereas higher
up between the villi this vasifactive plasmodiblast is already more like that of
Figs. 90—92. Indentations on the lower surface for the reception of secondary
or tertiary allantoidean villi.

Utr. Mus. Cat. n* Sorex 180 4,3 7.105.

'
dis a NR

ca ati 5

; rn bs . ce
‘ ne ey ay eae ae pag este a We
wiwhy .4 me if iv -_ vD,

; ing
oO

: ow Loot Re oe Za ety’ isa se

es)

v2

MINUTE ANATOMY OF LIMNOCODIUM. 539

Some Further Contributions to our Knowledge
of the Minute Anatomy of Limnocodium.

By

R. VT. Ginther, B.A.,
Lecturer of Magdalen College, Oxford.

With Plate 40.

Wui te engaged in a study of the anatomy of Limnocnida
tanganjice I was naturally rather desirous of making a com-
parison between the structural features of Limnocnida and
those of the first-described fresh-water medusa, Limnoco-
dium Sowerbii, in order to attempt to ascertain if the two
forms possessed any similar modifications which might possibly
be brought into relation with their life in fresh water.

For the furtherance of this object Professor Ray Lankester
very kindly placed all such preserved material of Limnocodium
as existed in the Department of Comparative Anatomy of
the Oxford Museum at my disposal. I take this opportunity
of thanking Professor Ray Lankester for this and all other
assistance he afforded me during the progress of my work,
which was carried on in the new Laboratory of Comparative
Anatomy at Oxford. I also wish to express my deep sense of
gratitude to the President and Fellows of Magdalen College
for enabling me to continue my studies in Oxford by prolong-
ing my Demyship at that College.

Hitherto I have, unfortunately, not had an opportunity of
examining any fresh specimens of Limnocodium. The material
which was best suited for histological examination had been
killed in osmic acid some years previously, but was neverthe-

540 R. T. GUNTHER.

less well enough preserved for the elucidation of many struc-
tural points. All my observations are based on sections cut
by the ordinary paraffin method and stained in various ways,
but those which had been coloured with Kleinenberg’s hzema-
toxylin gave as satisfactory results as any. Notwithstanding
the age of the material, I have been able to confirm most of
the observations of Allman and Lankester (2 and 4), and to
add some further details regarding the structure of the ten-
tacles, the sense-organs, and the male reproductive organs.

The general proportions of the various parts of the body as
seen in meridional section are shown in Pl. 40, fig. 1. In the
section of the individual there figured the mouth of the bell is
considerably contracted, so that the manubrium is enclosed in
the subumbrellar cavity ; but in the living condition, when the
animal is floating in the water, the bell has a much more
flattened shape, and the manubrium then projects considerably
beyond the margin of the umbrella. The mesoglea of the
umbrella is fairly uniform in thickness throughout, being only
very slightly thicker in the centre than near the circular canal.
In the manubrium, however, the mesoglea is not so evenly
distributed, but is thickest at the distal end of that organ, the
proximal end being almost completely destitute of any gela-
tinous middle layer whatever. This distribution of the meso-
gloea may be in relation to the great extensibility of the
manubrium. It is probable that the extension of that organ
is chiefly effected by the elongation of its proximal end, and
that a well-developed muscular layer exists between the limit-
ing epithelia of this region.

In none of the preceding papers are the figures of the
arrangement of the organs situated at the margin of the
umbrella satisfactory, while that of the “diagrammatic
meridional section” of Allman (4, p. 182) is erroneous and
misleading. The relations of the organs situated round the
periphery of the umbrella are exhibited in transverse section
in Pl. 40, figs. 2 and 3. Fig. 2 is a section passing along a
radial canal (7. c.) and through the base of a radial tentacle
(te.), while fig. 3 is taken several sections further on, passing

MINUTE ANATOMY OF LIMNOCODIUM. 541

through the base of another tentacle, not a radial one. The
circular canal (c¢c. ¢.) as seen in cross-section is roughly
triangular in shape. The epithelia of the radial canals and
the endoderm lamella join the epithelia of the circular canal
at the apical angle, while the velum and tentacles arise near
the interior and exterior basal angles of the circular canal
respectively. On the basal side, i.e. the side between the
attachment of the velum and that of the tentacles, the ecto-
derm is much thickened and modified to form the “ nettle-
ring” (ne¢.), and between the nettle-ring and the point of
attachment of the velum is the nerve-ring (n. 7.).

Tentacles.—The tentacles are usually carried turned back
over the aboral surface of the umbrella, and, like those of
many of the Trachymeduse, are adnate to the margin of the
umbrella for a short distance. The tentacle roots are not
entirely surrounded by mesoglea as are those of Cunina, but
only lie in a furrow on the umbrella margin. This attach-
ment of the tentacles to the umbrella is doubtless connected
with their upright carriage, and is very similar to the condition
obtaining in Limnocnida; but in Limnocodium the embedded
part or “root” of the tentacle consists of endoderm only,
whereas in Limnocnida the embedded tentacle root is en-
sheathed with ectoderm.

The structure of the tentacles has been described both by
Professor Allman and by Professor Lankester. With regard
to the question of the presence or absence of an axial cavity
and of the condition of the endoderm, the account by Professor
Allman (4, p. 133) is as follows :—‘‘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 character-
istic of the solid tentacles in the Trachomeduse and Narco-
meduse.”’

Professor Ray Lankester, in an addendum to his second
paper (8), says, ‘‘ Endoderm-cells consist of a dense, highly
refringent substance, which is somewhat wrinkled by the action
of the reagent;”’ and further, ‘‘In some cases a small amount

542 R. T. GUNTHER.

of granular cell-substance may be seen radiating from the
nucleus, but the whole cell body otherwise has been meta-
morphosed into a homogeneous cartilaginoid substance. There
is no continuous lumen, although the cells are disposed in a
single series around the axis of the tentacle, and leave, on
shrinking, a small space where their adaxial surfaces should
come into contact. This potential lumen appears not to be
continuous even in the specimens treated with reagents, and in
living specimens it has no existence.”

In all the individuals which I have hitherto examined the
larger and older tentacles were always hollow throughout their
length (Pl. 40, figs. 2 and 5), and it is only to the younger
and smaller ones that the above-quoted descriptions of Allman
and Lankester can apply. Moreover, in a considerable number
of sections of tentacles examined, the lumen of the tentacles
was found to be directly continuous with the lumen of the
ring canal, and the endodermic lining of the tentacle was
directly continuous with that of the ring canal as seen in
section in Pl. 40, fig. 2.

From this it appears that the tentacles of Limnocodium are,
morphologically speaking, hollow tentacles, though it is
quite possible that under certain circumstances they often
contract to such an extent that the lumen vanishes. An
indication of this contractile power is afforded by the exist-
ence of a powerful circular muscular coat at the bases of
the ectodermal cells. In obliquely cut sections such as the
one figured in Pl. 40, fig. 5a, these circular muscles may be
seen at the two ends of the section as transverse lines (circ. m.).

The endodermal lining of the tentacles consists of very large
clearish cells very similar to those of the tentacles of Limnoc-
nida. Their contents are very probably of a gelatinous nature,
which gives the tentacles a certain amount of firmness.

Nervous System.—The nerve-ring lies on the inner or
subumbrellar side of the circular canal at the attachment of
the velum. It is seen in transverse section in Pl. 40, figs. 2
and 38. The nerve-fibres composing the ring, as in other
medusz, are divided into two bundles separated by the sup-

MINUTE ANATOMY OF LIMNOCODIUM. 543

porting lamella of the velum. Of the two divisions of the
nerve, the outer division on the side next the nettle-ring is
the most strongly developed of the two. Its fibres are per-
fectly well demarcated from the surrounding cells, and are
always easy of observation. They seem, moreover, to be en-
closed in a sheath or neurilemma. On the other hand, the
fibres which constitute the inner bundle of the nerve-ring
(“unterer Nervenring” of Hertwig) are more difficult of certain
demonstration, since they are not distinctly separate from the
bases of the ectoderm-cells of the region.

Sense-organs.—Concerning the velar sense-organs, mar-
ginal bodies, or refringent bulbs, I have nothing to add to
Professor Ray Lankester’s description. I can only confirm
his account of these most remarkable organs in every particu-
lar. They consist of a small almost spherical multicellular
refringent body, which in the fully developed organ is attached
by a thin stalk near the nerve-ring, and is suspended in an
elongated sac which is embedded in the thickness of the meso-
gloea of the velum. The general relations of the organ are
shown in PI. 40, fig. 3, while in fig. 6 are several views of
sections of refringent bulbs carefully drawn with the camera
lucida. The refringent bulbs consist of cells of two kinds.
The more peripheral ones—the “ cortical cells” of Lankester
—are thin, and often so much stretched over the more central or
“ medullary cells” that they are difficult of observation. These
cortical cells are of ectodermal origin, and are continuous
with the lining of the sac in which the entire organ is enclosed.
The medullary cells impart the highly refringent appearance
to the bulbs, and are of endodermal origin, being budded off
from the lining of the circular canal, and subsequently become
completely enclosed by ectoderm. In the young condition
these medullary cells have a granular appearance (Pl. 40,
fig. 6, a, med. c.) like the cells lining the circular canal, and
only become clear and refringent as the bulb approaches
maturity. Even in fully developed bulbs a few of the medul-
lary cells near the point of attachment of the bulb still retain
a certain granular character (fig. 6, d, e).

544, R. T. GUNTHER.

In his original paper Professor Ray Lankester drew attention
to the unique nature of this peculiar organ, and pointed out its
relation to the endodermal sense-organs of other craspedote
medusz, such as the Trachomeduse.

The interest attaching to this organ is now all the greater
because another medusa (Limnocnida tanganjice) has
been discovered with similar and similarly situated sense-
organs, of which the axial cells are also endodermal. The
new medusa, moreover, is also an inhabitant of fresh water.
In Limnocnida the organ is identical in every important
respect, the chief differences being that the sacs are not pro-
longed into the velum, and that the refringent bulbs, as a
rule, consist of fewer cells, but in both Limnocodium and
Limnocnida no otolithic concretion is formed.

Endoderm.—The endoderm of the gastric cavity has
already been the object of a very thorough investigation by
Professor Ray Lankester in his paper on the “ Intra-cellular
Digestion of Limnocodium” (8). Proceeding from the mouth
towards the stomach, three regions have been distinguished
by Professor Lankester, all differing in regard to the nature
of their epithelial lining. The endoderm of the first region
nearest the mouth is composed of more or less granular cubical
cells, the nuclei of which are situated near the bases of the
cells (Pl. 40, fig. 9). The epithelium of the second region
(fig. 10) is very much higher; the cells composing it are pos-
sibly ciliated, though no cilia could be observed in the preserved
material. The nucleus is situated near the middle of the cell,
and separates the protoplasm of the inner haif, which is fairly
clear, from the very granular protoplasm of the outer half.
The epithelium lining the third region (fig. 11) or stomach
proper is composed of very large cells, and it is in this region
that intra-cellular digestion occurs. Among the large vacuo-
lated digestive cells are numerous goblet gland-cells, and here
and there a fragment of food may be seen which is undergoing
intra-cellular digestion, as in PI. 40, fig. 11, 2.

Fig. 12 illustrates the very abrupt transition between the
large-celled digestive epithelium of the third region and the

MINUTE ANATOMY OF LIMNOCODIUM. 545

very small and cubical cells which coat the subumbrellar wall
of the stomach. Similar cells are found lining the radial
canals and part of the circular canal. The genital sacs are
lined with a taller epithelium, a description of which is given
below.

The circular canal is lined with a low epithelium, similar to
the one lining the radial canals, except on the outer side.
There, between the point of attachment of the velum and the
origin of the tentacles, all round the inside of the thickened
patch of ectoderm which forms the nettle-ring, the cells of the
circular canal become much thicker, and in some places the
cell outlines are not well defined (Pl. 40, fig. 2, end.). The
function of the modified cells is not at all clear at present ;
but it is noteworthy that in Limnocnida a modified mass of
cells occurs in exactly the same position, but is very much
more largely developed.

Reproductive Organs.—As the material at my disposal
consisted solely of male individuals in various stages of
maturity, I have not been able to examine any females.

The reproductive organs in the male consist of four sac-like
outgrowths on the subumbrellar aspect of the four radial
canals. The distal walls of these sacs are very thick, being
chiefly composed of testicular tissue, as shown in section (Pl. 40,
fig. 7). The lumen of each of the sacs is a ventral diverti-
culum of the lumen of the radial canal, and is lined by a con-
tinuation of the endoderm of the radial canal. The cells of
this endodermal lining of the gonads are columnar, and rather
taller than the cells of the ordinary epithelium of the radial
canal. The nucleus of each cell is roundish, with a well-
marked nucleolus, and is situated near the base of the cell.
The protoplasm is highly granular, and near the free margin
of many of the cells it contains an ovoid mass which stains
deeply, and is probably the product of some secretory activity
of the cell (Pl. 40, fig. 8, end.).

The mesoglea is exceedingly reduced where the testicular
tissue is thickest, but it is fairly well developed all round the
stalk of the sac (fig. 7, m. s.), to which it imparts some rigidity.

VOL. 35, PART 4,—NEW SER. PP

546 Rk. T. GUNTHER.

Immediately to the outside of the mesoglea the great bulk
of tissue is chiefly composed of developing spermatozoa. As
in the spermarium of Oceania, described by O. and R. Hertwig
(1, p. 27), three different tissue zones may be distinguished.
Proceeding from within outwards (Pl. 40, fig. 8), the first or
basal layer of ectoderm consists chiefly of large round nuclei
with but little protoplasm proportionately ; secondly, there is
a thick layer of spermatozoa in various stages of development ;
and thirdly, there is an epithelial covering over all, the cells of
which send down processes in among the bundles of spermato-
zoa, and also seem to be in connection with long fusiform cells
penetrating between the spermatozoa in the second layer.

In a single favorable section through the gonad of a male
Limnocodium of a certain degree of maturity all the various
stages of developing spermatozoa can be observed ; consequently
Limnocodium is a far more favorable object for the examination
of the process of spermatogenesis than the majority of
Wydroids, in which, as a rule, a complete series of stages of
developing spermatozoa are not found in the same gonophore.
In most Hydroids in which spermatogenesis has been studied
the developing spermatozoa of one gonophore all progress at
about the same rate, and so in a mature bud the younger
stages do not occur, and vice versa. In fact, it is of rare
occurrence that more than two different stages of developing
spermatozoa occur in the same bud. In Limnocodium, on the
contrary, all stages are often present.

The following stages in the development of the spermatozoa
may be distinguished :

1. The sperm mother-cells (Pl. 40, fig. 8, a), situated next
the endoderm. These are characterised by their large nuclei,
which stain but slightly. They more or less correspond to the
Hertwigs’ first layer in their description of the spermarium of
Oceania (1). Each of the nuclei of these cells has a well-
marked nucleolus. Eventually they divide by karyokinesis,
and give rise to—

2. The daughter spermatoblasts (fig. 8,8). The nuclei of —
these cells are relatively much less than half the size of the

MINUTE ANATOMY OF LIMNOCODIUM. 547

nuclei of the sperm mother-cells of the first layer, but their
chromatin is in a more compact condition and stains more
deeply. The nuclear matter of these cells now apparently
undergoes a sort of condensation, as indicated both by dimi-
nution in size as compared with the size of the cell, and
also by greater intensity of tingibility (fig. 8,y). At this stage
the cells apparently undergo a second division, but whether
by karyokinesis or not could not be ascertained, owing to
their extreme minuteness. In either case the result is—

3. Anumber of cells (fig. 8, 6) with very small, deeply stain-
ing nuclei, which, by the drawing out of their protoplasm into
the tail, give rise to—

4, The spermatozoa themselves (fig. 8,<). The spermatozoa
‘are of the ordinary hydroid type, with well-marked heads and
relatively short tails. They probably escape to the exterior by
the dehiscence of the outer epithelium of the spermaria.

At stage 3 the cells seem to become segregated into groups.

The process of spermatogenesis, as described above, is
totally at variance with the views of André de Varenne (5)
regarding the condition of the nucleus of the sperm-cells. De
Varenne makes the following statement:—‘‘Dans toute la
durée du développement des spermatozoides, en prenant la
cellule mére dés son début, le noyau n’a pas changé.”” Limno-
codium certainly affords us a most effective refutation of any
such view, if any further objection to the view was needed after
the extensive researches of Thallwitz (6) on hydroid sperma-
togenesis.

Conclusion.

In conclusion, I am afraid that the foregoing observations
do not shed very much light upon the question of the systematic
position and genetic affinities of Limnocodium. It is seemingly
a case in which an increase of knowledge is correlated with an
increase of difficulties. All attempts to find a resting-place for
Limnocodium in the system of Haeckel have been unsatisfac-
tory. Of the four sub-orders into which the Medusz are
divided by Haeckel, neither the Anthomeduse nor the Narco-
medusz can receive Limnocodium on account of the position

548 R. T. GUNTHER.

of the gonads. Hence the Leptomeduse and the Tracho-
medusze are the only two sub-orders which can be considered.
Allman (4) enrolled Limnocodium among the Leptomedusz
on the erroneous assumption that the sense-organs were entirely
derived from the ectoderm of the velum. This, however, was
shown not to be the case by Ray Lankester (2), who proved their
partial endodermic origin, and accordingly transferred the me-
dusa to the ranks of the Trachomedusz. This change he, more-
over, supported by pointing out that Limnocodium resembles
the Trachomedusz in certain other respects. On the other hand,
it has been urged, and I think justly, that Limnocodium has a
fixed hydroid stage, a thing which is quite unknown among
Trachomeduse, and certainly absent in some of them. Inmy
own opinion, to include Limnocodium among the Tracho-.
meduse in the present state of our knowledge cannot but
render that group an unnatural one, or more unnatural than
it is at present. On the other hand, it cannot be denied that
Limnocodium has reached a Trachomedusan grade of develop-
ment in the possession of sense-organs with an endodermal
axis. Limnocodium, then, is a medusa descended from Lepto-
medusan ancestors, which has developed sense-organs with an
endodermal axis independently of the Trachomedusz.

List oF AUTHORS REFERRED TO IN TEXT.

1. O. and R. Hertwic.—‘Organismus der Medusen,’ Jena, 1878.

2. E. Ray Lanxester.—‘ Limnocodium Sowerbii,’ ‘Quart. Journ.
Micr. Sci.,’ xx, 1880.

8. E. Ray Lanxester.—“ On the Intra-cellular Digestion and Endoderm of
Limnocodium,” ‘Quart. Journ. Micr. Sci.,’ xxi, 1881.

4. G. J. Autman.—* Limnocodium victoria,” ‘Journ. Linn. Soc.,’ xv,
1881.

5. A. DE VARENNE.—‘ Recherches sur les polypes hydraires,’ Paris, 1882.

6. J. Tuatitwitz.—“ Uber die Entwickelung der mannlichen Keimzellen bei
den Hydroideen,” ‘ Jen. Zeitschr. f. Nat.,’ xviii, 1885.

7. G. H. Fowier.—“ Hydroid Phase of Limnocodium Sowerbyi,”
‘Quart. Journ, Micr. Sci.,’ xxx, 1890,

MINUTE ANATOMY OF LIMNOCODIUM. 549

8. R. T. GintHER.—“ The Fresh-water Medusa of Lake Tanganyika,” ‘ Ann.
Mag. Nat. Hist.,’ ser. 6, xi, 1893.

A complete bibliography of Limnocodium Sowerbyi up to 1890 will
oe found in No. 7 in the above list.

EXPLANATION OF PLATE 40,

Illustrating Mr. R. T. Ginther’s paper, “Some Further
Contributions to our Knowledge of the Minute Anatomy
of Limnocodium.”

Reference Letters.

e.c. Circular canal. circ. m. Circular muscle. ort. c. Cortical cells of
sense-organ. ect. Kctoderm. exzd. Endoderm. m. Mouth. med. c. Medul-
lary cells of sense-organ. mz. Manubrium. ms. Mesoglea. net. Nettle-ring.
m.r. Nerve-ring. 7. c. Radial canal. s. 0. Velar sense-organ. ¢e. Tentacle.
v. Velum.

Fig. 1.—Meridional section through an entire Limnocodium, in which the
velum is strongly contracted. The ectoderm (ect.) is represented by a single
line; the mesogloea (ms.) is shaded; the endoderm (ezd.) lining the gastric
cavity is represented by a double contour. x 15.

Fig. 2.—Radial section of margin of umbrella, showing the circular canal
(c. c.) with a radial canal (7. c.), cut longitudinally, opening into it. -The
basal portion of a radial tentacle (¢e.) and of the velum (v.) are also re-
presented.

Fig. 3.—A similar section to Fig. 2, but not passing through a radial canal
and tentacle. In this section a velar sense-organ (s. 0.) in its chamber has
been cut through.

Fie. 4.—Longitudinal vertical section of a portion of the velum, showing
several of the chambers of the velar sense-organs cut across. Note the thin
pavement epithelium lining these chambers.

Figs. 5a, 6, c.—Transverse sections of tentacles, showing lumen. 5a is
cut diagonally, and shows the circular muscles (circ. m.) at the two ends as
transverse stripes.

Fie. 6.—Five drawings of sections of velar sense bulbs. a@ is a longitu-
dinal section of quite a young stage which has only just become separated
from the endoderm. 4 and c are transverse, @ and ¢ longitudinal sections of
mature sense bulbs. e shows the stalk of attachment.

550 R. T. GUNTHER.

Fig. 7.—Radial section through an entire spermarium (sp.) and through a
portion of the radial canal on which the spermarium is situated.

Fie. 8.—Section through a portion of the wall of the spermarium of
another individual. a—e are various stages in the development of the
spermatozoa. ;

Fic. 9.—Endodermic lining of the oral end of the manubrium (Region I).
Fic. 10.—Endoderm of the middle or region II of the manubrium.

Fic. 11.—Section through the wall of Region III of the gastro-vascular
cavity. Note the much vacuolated nature of the ordinary digestive cells of
this region, and also the numerous goblet gland cells (g..c.), each filled with
numerous granules of secreted matter. wz is a foreign particle undergoing
intra-cellular digestion.

Fie. 12.—Section through an upper corner of the gastro-vascular cavity,
showing the junction between the endoderm of Region ILI, figured in Fig. 11,
and the small-celled endoderm of the subumbrellar roof of the gastro-
vascular cavity.

NOTE ON THE MESENTERIES OF ACTINIANS. aia |

Note on the Mesenteries of Actinians.
By

A. Fraser Dixon,
Royal College of Science, Dublin.

In a paper published in the last number of the ‘ Quarterly
Journal of Microscopical Science’ (193, January, 1894) on
**Octineon Lindahli,” the author, Dr. G. H. Fowler, gives
a figure—woodcut Fig. A,—and as my name is given in con-
nection with it, perhaps I may be allowed to make a statement
regarding it. The figure “represents Sagartia, Actinia,
Bunodes (Lacaze-Duthiers, corroborated by F. Dixon).” Some
time ago Professor Haddon, who was then writing the first
part of his ‘‘ Revision of British Actiniz,’ asked me to cut
sections of young specimens belonging to the genera whose
development Lacaze-Duthiers had described. He asked me to
do this because the Hertwigs, from their observations on
Adamsia diaphana, assumed that Lacaze-Duthiers had
made a mistake in determining which were the first eight
mesenteries to arise in Sagartia, Actinia, and Bunodes. They
assumed that in these genera, just as in Adamsia diaphana,
the most “ ventral’ of the dotted mesenteries in Fig. A of Dr.
Fowler was developed so early as to form one of the eight
mesenteries in the stage with eight mesenteries. If this were
so the eight-mesentery stage in Sagartia, Actinia, and Bunodes
could not correspond to the permanent condition in Edwardsia,
or to the condition described first by Professor Haddon for the
larva of Halcampa, because in these no “ lateral” mesentery is
present with its muscle plate pointing “ dorsalwards.” The
Hertwigs’ figure for Adamsia is given by Dr. Fowler in Fig. B,
and this represents what they assumed to be the arrangement
also in young specimens of Sagartia, Actinia, and Bunodes.

552 A. FRASER DIXON.

The sections which I made at this time showed that the
Hertwigs’ assumption was incorrect, and that Lacaze-Duthiers
was right in his determination of the first eight mesenteries.
They showed that the two mesenteries assumed by the
Hertwigs to be formed very early—in fact, to be the first
formed—were not present at all in the eight-mesentery stage,
but only arise as this stage is passing into the twelve-mesen-
tery stage. The sections confirmed Lacaze-Duthiers that
the four mesenteries dotted in Dr. Fowler’s Fig. A were the
latest developed. Unfortunately I was not able to obtain
specimens younger than these with eight mesenteries, and so
made out nothing regarding the order of appearance among
these eight. Professor Haddon, however, in his “ Revision”
(‘Sci. Trans. Roy. Dub. Soc.,’ vol. iv, series 2, p. 350) states,
“ Tn these representative species of three different families of
Actiniz the development of the mesenteries is similar in all,
both as regards the order of their appearance and the dis-
position of their muscles, and they are also identical with those
of the larva of Halcampa.”” This from my observations is too
wide a statement, and it would only have been safe to say “ in
these three species the disposition of the first eight mesenteries
is similar to that found in Edwardsia and in the larva of Hal-
campa, the arrangement of the muscle plates also correspond-
ing; further, the arrangement of the next four mesenteries
takes place in positions similar to those noted for Halcampa.”

That this was really the point on which Professor Haddon
wished to insist, and that he was not thinking of the order in
formation of the mesenteries of the eight or ‘‘ Edwardsia stage,”
is, I think, certain to anyone who reads his paper on “ The
newly hatched Larva of Euphyllia,” read before the Royal
Dublin Society in March, 1890. In this last paper, on p. 133,
Professor Haddon states that the order of development among
these first eight mesenteries in the forms studied by Lacaze-
Duthiers requires re-investigation, and that in Euphyllia, at all
events, he has a priori reason for believing that the order of
Lacaze-Duthiers is not present, but that that described by
Wilson for Manicina obtains. Hence it is almost certain that

NOTE ON THE MESENTERIES OF ACTINIANS. 553

the too wide statement made in the “ Revision of British
Actiniz ” would have been corrected by Professor Haddon if
he had had an opportunity of seeing the proofs of it; un-
fortunately, however, he started for Torres Straits immediately
after it was read.

It thus is evident that the type represented by Sagartia,
Actinia, and Bunodes is separated from the type represented
by Manicina (Wilson), Ariactis (McMurrich), and perhaps
also Euphyllia (Haddon), merely by observations of Lacaze-
Duthiers which were made without sections.

I can only add my regret that this mistake should have
caused so much confusion.

-

he
‘

o P » Tis j es : 71) ee
to. OR Ge ai: pee

iv

é

y

‘7 “a rin Neh Hig 4s
!

=

- —-
«4 <
£ a thin” ,
ames.
‘ 7 |
pps “ elt ne
mn.
é if ‘
Paes ft) '
7 - ,
‘a - ms ot 1 eae (|
‘ 3 3
A H
eg
iS
<
HM
a

INDEX TO VOL. 385,

NEW SERIES.

Actinians, mesenteries of, by A. F.
Dixon, 551
Amphioxus, collection of, from Torres
Straits, by Arthur Wil-
ley, 361
> pharyngeal bars of, by
Benham, 97
Apus and Branchipus, reproductive
elements of, by J. HE. 8S. Moore, 259

Benham on the structure of the
pharyngeal bars of Amphioxus, 97

Bipinnarie from the English Channel,
by W. Garstang, 451

Branchipus and Apus, reproductive
elements of, by J. HE. 8. Moore,
259

Buchanan on Eupolyodontes Cor-
nishii, a Polynoid with branchiz,
433

Ciona, perivisceral cavity of, by A.
H. L. Newstead, 119

Dendy, studies on the comparative
anatomy of sponges, No. V, 159
Distichopora violacea, Hickson,

129
Dixon, A. F., on the mesenteries of
Actinians, 551

Eupolyodontes, a Polynoid with bran-
chise, by F. Buchanan, 433

Fowler on Octineon Lindahli,
461

Frog, orientation of the egg of, by
Morgan and Tsuda, 373

Garstang, W., on Bipinnarie from
the English Channel, 451

Gobius capito, development of the
head in, by H. B. Pollard, 335

Goodrich on the fossil Mammalia of
the Stonesfield slate, 407

Giinther, R. T., contributions fo our
knowledge of the minute anatomy
of Limnocodium, 539

Hickson on the early stages of the
development of Distichopora, and
on the fragmentation of the nucleus,
129

Hubrecht on the placentation of the
shrew (Sorex vulgaris), 481

Kirkaldy, J. W., on the head-kidney
of Myxine, 353

Limnocodium, contributions to our
knowledge of the minute anatomy
of, by R. T. Giinther, 539

Limulus, brain and sense-organs of,
by W. Patten, 1

Mammalia, fossil, from the Stonesfield
slate, by E. 8. Goodrich, 407

556

Mesenteries of Actinians, by A. F.
Dixon, 551

Moore, J. E. §., on some points in
the origin of the reproductive ele-
ments in Apus and Branchipus,
259

Morgan and Tsuda on the orientation
of the frog’s egg, 373

Myxine, head-kidney of, by J. W.
Kirkaldy, 353

Newstead on the perivisceral cavity
of Ciona, 119

Nucleus, fragmentation of, by Hick-
son, 129

Octineon Lindahli, anundescribed
Anthozoon of novel structure, by
G. H. Fowler, 461

Patten on the brain and sense-organs
of Limulus, 1

Peripatus from Dominica, by E. C.
Pollard, 285

INDEX.

Pollard, E. C., on the Peripatus of
Dominica, 285
» H. B., on the development

of the head in Gobius capito,
335

Polynoid with branchie, by F. Bu-
chanan, 433 i

Protochordata, studies on the, by
Arthur Willey, No. II, 295

Shrew, the placentation of, by Hu-
brecht, 481

Sorex vulgaris, the placentation
of, by Hubrecht, 481

Sponges, studies on the comparative
anatomy of, No. V, by Arthur
Dendy, 159

Tsuda and Morgan on the orientation
of the frog’s egg, 378

Willey on a collection of Amphioxus
from Torres Straits, 361

Willey’s studies on the Protochor-
data, No. II, 295

PRINTED BY ADLARD AND SON,
BARTHOLOMEW CLOSE, E.C., AND 20, HANOVER SQUARE, W.

aa ma

ne i

ny 4 ns

vA a eh :
i ian ie ty
J
Ah 4
hs wil Wil ry
7 a
! ay)

W va j una hy)

Py

iy my Ki ty

i |
‘Wis
Wan

ih |
in
i

Hp LE | wii ; Bt Lb PR il We L ely 4 ih }
i fea a ve i 1) ALN i I iif a ‘;

——
=o

~~

a

Patatadehieleiee
MUTE
hit

——

me erg er ae Pae TA OPO PEO ELA WL CT TO PO Rea