VOLS9
[PES
rg Z 4 |
Pibrary of the THuseum
OF
COMPARATIVE ZOOLOGY, |
AT HARVARD COLLEGE, CAMBRIDGE, MASS.
Founded by private sudscription, tn 1861.
Deposited by ALEX. AGASSIZ.
No.
FF?
fin
QUARTERLY JOURNAL
OF
MICROSCOPICAL SULENCE.
EDITED BY
E. 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 ; Corresponding Member of the Imperial
Academy of Sciences of St. Petersburg, and of the Academy of
Sciences of Philadelphia; Foreign Member of the
Royal Bohemian Society of Sciences.
WITH THE CO-OPERATION OF
ADAM SEDGWICK, M.A., F.R.S.,
Fellow and Lecturer of Trinity College, Cambridge ;
AND
W. F. R. WELDON, M.A., F.B.S.,
Jodrell Professor of Zoology and Comparative Anatomy in University College, London;
late Fellow of St. John’s College, Cambridge.
VOLUME 39.—NeEw Series.
With Xithographic Plates and Engrabings on Wood.
LONDON:
J. & A. CHURCHILL, 7, GREAT MARLBOROUGH STREET.
~ 1897,
CONTENTS.
CONTENTS OF No. 153, N.S., MAY, 1896.
MEMOIRS : PAGE
The Blood of Magelona. By W. Buaxtanp Benuaw, D.Se.Lond.,
Hon.M.A.Oxon., Aldrichian Demonstrator in Comparative Ana-
tomy in the University of Oxford. (With Plate 1) : : i
Fission in Nemertines. By W. Buaxtanp Brnuam, D.Sc.Lond.,
Hon.M.A.Oxon., Aldrichian Demonstrator in Comparative Ana-
tomy in the University of Oxford. (With Plates 2 and 3) eae)
Studies on the Nervous System of Crustacea. By Epear J. ALLEN,
B.Se., Director of the Plymouth Laboratory of the Marine Biolo-
gical Association. (With Plate 4) : : : cae
Notes on Oligochetes, with the Description of a New Species.
By Evwin 8. Goopricu, B.A., Assistant to the Linacre Pro-
fessor of Comparative Anatomy, Oxford. (With Plates5and6) 51
On the Development of Lichenopora verrucaria, Fabr. By
Sipney F, Harmer, M.A., B.S8c., Fellow of King’s College,
Cambridge; Superintendent of the University Museum of
Zoology. (With Plates 7—10, and two figures in text) nee
CONTENTS OF No. 154, N.S., AUGUST, 1896.
MEMOIRS:
Letters from New Guinea on Nautilus and some other Organisms.
By Artuur Witty, D.Sc. : : : . 145
The Brain of a Foetal Ornithorhynchus. Part I.—The Fore-brain.
By G, Exxior Smitu, M.D., Ch.M., Demonstrator of Anatomy,
University of Sydney, N.S.W. (With Plate 11) 3 5 eH
On Arhynchus hemignathi, a New Genus of Acanthocephala.
By Artuur HE. Surerezy, Fellow and Tutor of Christ’s College,
Cambridge, and University Lecturer in the Advanced Morpho-
logy of the Invertebrata. (With Plate 12) 5 : . 207
iv CONTENTS.
PAGE
Zoological Observations in the South Pacific. By ARTHUR WILLEY,
D.Se.Lond., Balfour Student of the ome of Cambridge.
(With Plate 13) ‘ - é : . 219
Chlamydomyxa montana, n. sp., one of the Protozoa Gymno-
myxa. By E. Ray Lanxester, M.A., LL.D., F.R.S., Linacre
Professor in the University of Oxford. (With Plates 14 and 15) 233
CONTENTS OF No. 155, N.S., NOVEMBER, 1896.
MEMOIRS :
The Constitution and Development of the Society of Termites: Ob-
servations on their Habits; with Appendices on the Parasitic
Protozoa of Termitide, and on the Embiide. By Professor B.
GrassI in collaboration with Dr. A. Sanpias. oe Plates 16
—20) . : ; : : . 245
On Ctenoplana. By ArtHuR Witter, D.Sc. (With Plate 21) . 323
An Attempt to deduce the Vertebrate Eyes from the Skin. By
H. M. Bernarp, M.A.Cantab., F.L.S. (With Plate 22) . 343
The Reproduction and Metamorphosis of the Common Eel (An-
guilla vulgaris). By Professor Grassi ‘ : s ovk
CONTENTS OF No. 156, N.S., JANUARY, 1897.
MEMOIR:
Changes in the Cell-organs of Drosera rotundifolia, pro-
duced by Feeding with Egg-albumen. By Lity Huts, Physio-
logical Laboratory, Oxford. (With Plates 23 and 24) . . 387
Observations upon the Development and Succession of the Teeth in
Perameles ; together with a Contribution to the Discussion of
the Homologies of the Teeth in Marsupial Animals. By J. T.
Witson, M.B., Professor of Anatomy, and J. P. Hirt, F.LS.,
Demonstrator of Biology in the University of ces New
South Wales. With Plates 25—32) : F . 427
TITLE anpD INDEX.
JUN Le 1896
The Blood of Magelona.
By
W. Blaxland Benham, D.Sc.Lond., Hon. M.A.Oxon.,
Aldrichian Demonstrator in Comparative Anatomy in the University of Oxford.
With Plate 1.
THe small Annelid first discovered and briefly described by
Fr. Miller! in 1858 has more recently received more elaborate
treatment at the hands of Professor W. M‘Intosh.? It presents
several peculiarities deserving of still more detailed research ;
and, at Professor Lankester’s request, Dr. M‘Intosh has kindly
sent to me from time to time living specimens, obtained at St.
Andrews, which have served as the starting-point of my own
observations.
Amongst these peculiarities, the most striking and astonishing
is afforded by the contents of the blood-vessels ; and I propose
at present to limit my remarks to an enumeration and descrip-
tion of the observations and experiments which I have made
on this fluid.
In order to obtain a greater quantity of quite fresh material
I spent a week at the Marine Laboratory at St. Andrews
during the summer of 1894. Professor M‘Intosh was good
enough to send me a list of low tides—at which time only can
Magelona be obtained in any considerable number. My
very hearty thanks are due to Dr. M‘Intosh for his permission
to make use of the laboratory and appliances for the purpose
1 Fr. Miller, “ Hiniges iiber d. Annelidenfauna d. Insel 8. Catharina,” &c.,
‘ Arch. f. Naturgesch.,’ 1858, p. 211.
2 W. C. M‘Intosh, “ Beitrage zur Anat. von Magelona,” ‘Zeit f. wiss.
Zool.,’ 31, 1878.
VoL. 39, PART 1.—NEW SER. A
2 W. BLAXLAND BENHAM.
of my study, and for his advice and kinduess in aid of my work
there.
Magelona lives in the sand at and below ordinary low-water
mark, and the lower the tide the more abundant are the worms.
My sojourn at St. Andrews was timed so as to cover par-
ticularly low tides occurring between August 30th and Septem-
ber 4th. I found the worms limited to a comparatively small
area on the sands near the harbour, close to the Laboratory,
though they occur more sparingly over a larger area. The
work involved in their capture is rather heavy, as they occur
at a depth from eight to sixteen inches below the surface of the
sand. I was fortunately able to obtain the help of the laboratory
attendant, A. W. Brown, who accompanied me in my search.
The process was as follows:—Having selected a spot which he
believed from experience to be likely, Brown dug as deeply into
the sand as possible. As he raised the spade-load the sand na-
turally broke across, and if Magelona was present the worms
would be seen stretching across the gaps thus formed in the
spadeful of sand. The worms are rather brittle, and consider-
able care had to be exercised in raising the mass of sand, and
further separating it, so as to liberate the worms. Sometimes
such a spadeful would produce only one or two worms, at other
times considerable numbers might be obtained; but the work
of sorting through such masses of wet sand was no light task.
We usually spent a couple of hours in this manner till the tide
drove us back from the Magelona area.
The body of Magelona papillicornis presents two
regions,—a short “ thorax,” with thick, muscular wall; and an
“abdomen,” through whose thin walls the gut and its sandy
contents can beseen. The general appearance then of the worm
is sand-colour, for there is no pigment in the skin or body
wall, but the very short thorax is of a faint madder- pink tint,
or in some cases dull white, according, as later observations of
captive specimens showed, to the condition of oxidation of the
blood. When the worms, separated more or less from the
sand, are placed in clean aérated sea water the colour of the
thorax soon becomes a deeper madder-rose colour; as the water
THE BLOOD OF MAGELONA. 3
becomes foul the tint of the worm becomes fainter and fainter,
till the thorax is white. I first observed this change while
working upon worms sent from St. Andrews to me at Oxford.
The worms had been put into clean sea water on their arrival,
and had been put aside overnight. Next morning they were
no longer madder-rose, but white. On pointing this out to
Professor Lankester, he at once suggested aérating the water
by squirting air into it; this I did, and soon the madder-rose
tint reappeared. It was this small but extremely interesting
change in colour which called our attention to the blood, for it
was evident that we had to deal with a respiratory pigment of
a kind unusual in the Chetopoda.
The fact that the tint of the worm is due to contained blood,
and not to any pigment in the skin, is readily recognised by
the unaided eye when the alternating processes of eversion
and retraction of the ‘ proboscis” or “introvert” is watched.
So long as the introvert is at rest within the body the thorax
is coloured; when eversion takes place the tint becomes quite
faint (Pl. 1, fig. 2),—in fact, frequently the thorax becomes
white. The blood in the thorax is contained in greatly dilated
vessels, which block up and obliterate nearly the whole of the
ceelom (fig. 3); the introvert is a hollow sac traversed by thin
bundles of retractor muscles, the cavity of the sac being con-
tinuous with the dilated vessels, so that on eversion nearly the
whole of the blood in the thorax is driven into the introvert—
this flow of blood is, of course, the cause of the eversion,—and
the thorax is more or less completely deprived of its colour.
The fact that the abdomen is not tinted by the blood to any great
degree is due to the small size of the blood-vessels in this region.
The arrangement of these blood-vessels has been minutely
described by M‘Intosh, and need not detain us at present.
But the remarks of this author, so far as they relate to the
blood itself, may here be quoted :—‘‘ The blood is a coagulable
pale rose-red fluid, containing numerous corpuscles. On
being shed these group themselves in different clumps. The
size of the spherules is nearly constant, although variations
may occur. They exhibit molecular movement, and their
4, W. BLAXLAND BENHAM.
contour is changed by pressure. Many are ovoid, others
circular or irregular. If studied within the blood-vessels the
corpuscles exhibit a refringent body or nucleus-like central
structure. In addition to these spherules various other elements
exist, which may perhaps be explained as a development of
spherules inside cells. Here and there a coagulum may be
seen between the corpuscles. Treated with strong acetic acid,
very noticeable changes occur. The whole field is now covered
with a granular débris (Pl. 38, fig. 9), and in place of the
corpuscles we see masses of granular cells of larger size. The
acid probably dissolves the envelope of the corpuscles or
alters their walls, so that the (probably fatty) contents run
together to form larger masses.”
It will be seen later that this description, though true to a
certain extent, does not exhaust the subject. I may mention
that nearly the whole of my own observations had been made
before I had acquainted myself with the above account of the
blood given by M‘Intosh,—that is, in reading his paper some
time back I had not paid special attention to his statements
as to the blood itself.
1. Spectroscopic Examination.—A fter observing the re-
markable change of colour on oxidation I wished to examine
the blood spectroscopically, and to this end I sought for some
reagent which would extract the pigment; but though I was
unsuccessful in my attempts, yet these experiments are, I
believe, worthy of record, as the action upon the blood of
some of the reagents is sufficiently peculiar. These will be
described below.
Having been unsuccessful in extracting the colouring matter,
I examined the blood itself by means of Zeiss’s spectroscope.
For this purpose I compressed a worm with its introvert everted,
so as to give a thin film of pink fluid, and examined it through
the micro-spectroscope; but no absorption bands were visible.
I then had short pieces of glass tubing cut and fixed to glass
slips, so that I could fill them to a greater or less extent, and
get a column of fluid of different depths. The process of
obtaining sufficient blood to fill even a tube one eighth of an
THE BLOOD OF MAGELONA. 5
inch deep was tedious. ‘The method I adopted was as follows :—
A number of living worms were taken, one by one, laid on
blotting-paper, so as to dry them as much as possible; then,
while holding them just above the stout piece of tube, I
snipped the everted introvert or the thorax itself; the worms
were then left head downwards over the tubing, and the blood
gradually flowed out of the cut end down the side of the tube
and collected at the bottom. A considerable number of worms
had to be thus treated, and as the blood tends to coagulate the
process occupied considerable time ; nevertheless I was able to
collect what appeared to be a sufficient quantity of blood for
spectroscopic examination. The result, however, was nil.
I could obtain no absorption bands, though I used both bright
sunlight and gaslight, and took all precautions to keep the
light out of the apparatus except that which passed through
the blood. I also employed solutions of blood in salt solu-
tions and in water. These solutions were colourless, but I
thought it possible that even then the colouring matter might
give some result; but I was equally unsuccessful.
I believe, then, that we must conclude that the colouring
matter of the blood of Magelona causes no absorption bands
when a beam of light passes through it, and is spectroscopically
analysed.
2. The Histology of the Blood.—The fact that all the
reagents which I employed had greater or less action on the
blood rendered difficult the ascertainment of the real structural
condition of the fluid. The only way in which the true con-
ditions could be studied was by compressing the worm and
examining the fluid as it flows along the vessels ; for even the
mere shedding of the blood, without the addition of reagents,
might be supposed to produce certain alterations in its con-
dition, though from what follows I do not thiuk this to be the
case. The blood thus examined in situ is seen to consist of
very small madder-rose coloured globules, varying in size within
certain small limits. These globules choke the vessels, and
there appears to be very little plasma. The appearance pre-
sented in sections of worms fixed in Hermann’s solution is
6 W. RLAXLAND BENHAM.
very similar to that exhibited by the living subject. Further,
these corpuscles or globules do not flow freely in the vessels as
the walls contract, but adhere together in clumps, and some--
times a mass or clump will be separated by some considerable
space from neighbouring masses, appearing to indicate a colour-
less plasma. This massing of the globules also occurs in the
freshly shed blood, as will be seen hereafter, but its occurrence in
the blood-vessels is extremely peculiar, and M‘Intosh expresses
the phenomenon by speaking of the blood as “ coagulable ”—
a term which perhaps is scarcely true in its original sense, —
for there is no evidence of the separation of “fibrin” or
anything of that nature; the globules or corpuscles rather
adhere together.
The general shape of the corpuscles is spherical ; in size they
average 0°002 mm. in diameter in the living condition; they
are not amceboid; they are homogeneous, and rather oily in
appearance. They are not nucleated, as the addition of re-
agents will show; and I did not observe the refringent body
within them mentioned by M‘Intosh. Nuclei, however, do exist
in the blood, though usually in an isolated condition, and are not
recognisable with certainty in the living worm ; but in stained
sections, as well as in preparations treated with osmic acid, fol-
lowed by picro-carmine, such nuclei are more or less abundant
(fig. 4), and though usually quite deprived of any protoplasmic
envelope, yet in some cases I noted a granular mass surrounding
the nucleus, without any boundary or recognisable limit, and I
believe this to be the fluid plasma of the blood, coagulated by the
reagents. In a preparation of the blood mounted in glycerine,
in which the globules have undergone a considerable amount of
fusion, so that the globules are of very varied sizes, I note several
instances of what appear to be nucleated corpuscles. In these
cases the nucleus, of the usual size, is deeply stained, and lies at
one side of an unstained “ globule” (figs. 8, 9); sometimes, in
fact, the outline of the nucleus seems to project slightly beyond
the general outline of the globule, but this may be due to the
greater refractive index of the former. These globules, of which
I give some figures (fig. 6), vary in size, but are much larger
THE BLOOD OF MAGELONA. a
than the unaltered corpuscle as seen in absolutely fresh blood,
and I suggest that two or more normal globules have fused
together and involved the nucleus. The isolated nuclei may
occur singly or in groups of three or four (fig. 7).
The size of the nucleus, as seen in sections, is rather greater
than that of the globules themselves (Pl. 1, fig. 4). The chro-
matin granules vary considerably in number and arrangement ;
‘sometimes there is but one such granule, at other times several.
In the glycerine preparation the former condition is more
frequent, and the chromosome is then rod-like (fig. 7, a, 0, c).
A. The readiest way to obtain the blood as free as possible
from other matters is to dry a worm by placing it on blotting-
paper, place it ona dry slide, and pierce with a sharp needle the
proboscis, which is pretty sure to be everted. The blood now
oozes out slowly, but does not flow freely, the globules adhere
together in masses, and the masses are connected by narrower
bridges of corpuscles, giving rise to a coarse network. This
mass changes its shape and arrangement, as neighbouring
nodes or clumps run together by the shortening of the bridges ;
further, very delicate threads (really due to strings of globules)
are seen (¢.) traversing the meshes, and owing to the continued
stretching of the narrowing bridges isolated clumps of cor-
puscles may come to occupy the meshes as these threads
snap asunder. This change of arrangement may be due to
evaporation of the plasm and to surface tension. The appear-
ance presented by such a drop of blood shed in this way,
and examined without a cover-slip, is shown in fig. 10.
The whole mass is pink in colour, highly refringent and oily-
looking; no outlines of the separate corpuscles are visible
in the larger masses, though in the smaller groups they may
be recognised.
When such a drop of blood is covered (fig. 11) the outlines of
the corpuscles become evident; and though the majority are
similar in size, here and there larger ones are met with, and it
is quite possible that the latter are nuclei. The change in
arrangement of the bridges and meshes is seen to continue. The
appearance presented by this mass of corpuscles is strikingly
8 W. BLAXLAND BENHAM.
similar to the “soapy foams” manufactured by Butschli’s
method; the “alveolar layer” is formed by regular arrange-
ment of the corpuscles along the margin of the mass, and the
slow movement recalls that described by him.
B. The blood when freshly drawn by rupture of the “ intro-
vert ” in sea water (after removal of as much of the sea water
as possible from the slide), and at once covered, presented the
following appearance. Instead of a mass of similar small glo-
bules, these were seen to be of various sizes,—most of them
much larger than the corpuscles seen within the vessels, being
as much as 0:04 mm. in diameter (fig. 15).
From later experiments, carried out with greater care so
that no water should be present, I am led to believe that the
result here obtained was due to the presence of a certain
amount of sea water on the slide; this causes the globules to
fuse with one another, hence the varied sizes of the globules.
3. Appearances presented by the Blood on the
Addition of Various Reagents.
1. Normal salt solution, passed under the cover-slip of
such a preparation as described above (A.), causes the globules
to separate from the masses if there has been no great amount
of pressure. If the latter has occurred, only the corpuscles
lying at the edge of the mass separate. At first the colour is
unchanged, but soon it disappears.
This disappearance of the colour was much more evident
when the tubes used for spectroscopic analysis containing a
little salt solution, the worms cut as above described, and
the blood allowed to flow into the salt solution. I hoped to
get a pink solution, but the colour very soon changed and
became colourless. This may be due to deoxidation rather
than to any destructive power of the salt; but whatever the
cause, the result was a disappearance of colour.
2. Distilled Water.—When water is added to blood
already treated with salt solution, the following extraordinary
series of phenomena ensues:—As the water displaces the
salt solution the outlines of the globules disappear: this
THE BLOOD OF MAGELONA. 9
takes place naturally at the edge of the masses first, and then
the mass presents the appearance of a colourless, oily-looking
drop, in which more or less numerous rounded granules are
embedded (fig. 12). These granules are the globules, and the
longer the water acts the fewer distinct globules remain, till
finally the mass of globules is replaced by a homogeneous, oily-
looking drop (fig. 13). Meanwhile the movement already de-
scribed in the fresh blood is going on, and is aided by the current
of water as it is drawn under the cover-slip ; the narrow bridges
uniting the masses become drawn out into longer and thinner
threads—the globules of which they were composed have fused,
and the threads, like the masses at the nodes of the network,
become homogeneous. These threads may thin out till they
break, the two ends are drawn into the masses at the nodes,
and the latter become rounded off. In such a way the network
may be resolved into a number of larger and smaller droplets,
clear, colourless, and without any indication of the globules of
which they were originally composed.
If a group of globules or corpuscles not forming part of a
network be examined, it can be readily seen that the neighbour-
ing globules do fuse with one another, forming larger
globules: these larger ones fuse with others, and thus, as a
greater and greater number of globules become involved, a
great oily-looking droplet is formed, in which all trace of the
original globules has disappeared. In this reaction there is no
swelling up of the original globule—neighbouring globules
fuse with one another (fig. 14).
The addition of normal salt solution to such a mass, so as
to replace the water, did not lead to the reconstitution of
separate globules, so that there can be no mistake about the
actual fusion.
Such a droplet is shown in fig. 16; after salt solution has
been added, it lost its circular shape, and became irregular
there being a marked movement—almost amoeboid in general
appearance; and small irregular outgrowths comparable to
pseudopodia became formed, and some of these separated from
the mass as pear-shaped, and later circular droplets (z.). It
10 W. BLAXLAND BENHAM.
may be suggested that these are the original globules, but such
is not the case ; they are very much smaller, and vary more in
size. Further, only a few such separations occurred, even
after prolonged action of the salt solution.
When water is added to freshly drawn blood the same
series of reactions occur (fig. 14). The colour of the corpuscles
is quickly changed by the water, the masses becoming colour-
less. I also “ bled” worms into a few drops of water in a
tube, in the hope of obtaining a solution of colouring matter.
The reaction of the corpuscles or globules to water indicates
that these coloured globules are not provided with any mem-
brane or envelope. They appear to be droplets of coloured
material, and their general appearance suggested some oily
substance.
Chloroform.—Employing freshly shed blood on a dry
slide, and running in chloroform as soon as possible, I observed
the gradual disappearance of the corpuscles. Each corpuscle
as the chloroform reached it first lost its colour, then rapidly
became smaller and smaller, till nothing was left but a few
very minute granules. Amongst these a number of small
ovoid, highly refringent bodies, being a faint olive-green in
tint, make their appearance.
Here, too, I have no doubt as to the fact that the corpuscles
are dissolved : there is no mere “ diminution in size,’ such as
is observable in the case of mammalian corpuscles treated
with chloroform ; there is an absolute disappearance of the
globules of Magelona, the only trace of them that remains
being a few granules.
Ether produced effects similar to those produced by chloro-
form. Both these reactions seem to point to fat or oil of some
kind.
Absolute alcohol gave the same reaction. I added nitric
acid (10 per cent.) to the small olive-green refringent bodies,
and found that they are insoluble in the acid. But both these
and the smaller granules were dissolved in strong nitric acid.
Alcohol (70 per cent.) had similar action.
In all the above cases the colour of the corpuscle disappeared.
THE BLOOD OF MAGELONA. 11
Osmic acid coagulates the substance of the corpuscles, but
does not otherwise change them. They only become distinctly
brown after some minutes, and do not give the dark brown or
black colour characteristic of fat.
Nitric acid (25 per cent.) and Hydrochloric acid
(20 per cent.) cause the corpuscles to become granular,
though they do not actually disappear. The colour is changed.
Potash (80 per cent.) has a similar action.
Glacial acetic acid dissolves the globules, leaving minute
granules, though the action is very slow.
Further, picro-carmine stains the globules, as also does
eosin: they retain their homogeneous appearance, however.
These tests are unfortunately inadequate for the identifi-
cation of the substance of which these globules are composed ;
but from the ether, chloroform, and alcohol reactions, I think
we may conclude that it is one of the “ fats ” which are soluble
in these reagents, and moreover, give rise to crystalline bodies
on evaporation. And amongst these fatty bodies, lecithin
appears to give a similar reaction with water, in that it swells
up in a peculiar way, forming droplets and threads, as I have
described above.
4. General Remarks.
From these observations it will be seen that the blood of
Magelona is totally different in structure from that of any
other Chetopod, in that it consists mainly of very small
madder-rose coloured, non-nucleated globules, embedded (rather
than floating) in a very small amount of colourless plasma:
amongst the corpuscles occur isolated nuclei. It was originally
demonstrated by Professor Lankester that nuclei occur in the
red fluid of the common earthworm, and this observation has
been extended to sundry other Annelids by various observers.
In these cases, as in Magelona, the nucleus is surrounded by
very little, if any, protoplasm, and floats freely in the perfectly
liquid plasma, which is coloured red by hemoglobin, or ina few
cases green by chlorocruorin or chlorochromin ; while in some
Oligochzetes the plasma is colourless. No other structural
12 W. BLAXLAND BENHAM.
element has been described in the fluid contained in the blood-
vessels of any Annelid. But a few worms are known in which
the blood-vessels are absent, whilst the fluid in the celom is
coloured ; and in these cases (Glycera, Capitellide, Poly-
cirrus) the hemoglobin exists within the corpuscles, the plasma
of the coelomic fluid being colourless. The coloured corpuscles
of this fluid in Glycera and Capitellide are, however,
entirely unlike the coloured globules in the blood of Magelona.
In the case of Capitella I took the opportunity of examin-
ing the corpuscles during my stay at St. Andrews, as I was
able to get abundant material. My observations are not new;
Claparéde and Eisig have already sufficiently demonstrated
the character and constitution of these corpuscles. Lankester!
was the first to show by spectroscopic examination that they
contain hemoglobin. The corpuscles when drawn fresh from
the worm are circular, somewhat flattened discs, yellowish red in
colour. A distinct membrane is visible, and granules of various
small sizes are present. Stained with picro-carmine, a small
round nucleus is seen in each corpuscle. When fresh cor-
puscles are treated with salt solution no effect is noticeable.
On the addition of distilled water the hemoglobin is dissolved
out and the nucleus is rendered distinct, and the corpuscle
itself becomes slightly smaller. This is no doubt due to os-
mosis, the water passing into the corpuscle and causing it to
become spherical: it thus loses in diameter in one direction as
it gains in the other,—a process exactly like that occurring
when Vertebrate corpuscles are similarly treated.
Prolonged action of water causes the outlime of the cor-
puscles to become less and less distinct, till it is almost
impossible to recognise it, the nucleus and refringent bodies
alone remaining. This reaction is quite different from what
we have seen to occur in the caseof Mageiona, but resembles
the corresponding reaction with human blood.
Chloroform dissolves the hemoglobin, the corpuscles become
granular, and the outline becomes gradually less distinct, but
the membrane remains for a long time. But there is no
1 * Proc. Roy, Soc.,’ xxi, 1873, p. 70.
THE BLOOD OF MAGELONA. 13
marked change in size; there is nothing approaching the
gradual though rapid disappearance of the globules seen in
the case of Magelona.
Ultimately, though after a considerable time, the corpuscles
appear to become dissolved, breaking up into granules.
In Glycera, too, the corpuscles, as Claparéde, Lankester,
and others have shown, are nucleated—the nucleus being oval
and the corpuscle larger than that of Capitella. I was unable
to obtain Glycera in any quantity, —in fact, I only came across
two small specimens while collecting Magelona, and was unable
to make many observations on them. But previous authors
have compared these corpuscles with those of Vertebrates.
The so-called “corpuscles” or coloured globules of Ma-
gelona thus differ from the coloured corpuscles observed in
other Annelids, not only in position, viz. within blood-vessels
instead of in the ccelom, but also in structure and in their
behaviour to chemicals, and I believe that they are different
in constitution.
These coloured globules in Magelona are to be compared
with the coloured plasma of the ordinary Cheetopod blood,
rather than with the coloured corpuscles of Capitella,Glycera,
and Polycirrus. These globules, in fact, though recognisable
as separate elements, do adhere together in the blood-vessels to
form a fluid of the consistency of thick oil. If we suppose
that a little more water were present in their composition we
may imagine the globules to fuse with one another, and so form
a more freely flowing plasma—such as is obtainable by ex-
periment. This plasma would then be quite comparable with
that of the blood of ordinary Annelids. But if we imagine the
individual globules to have a firmer envelope—more resistent
to reagents than it is in fact—we should have more perfectly
defined corpuscles, comparable to the hematids in the blood of
mammals. These adhere together, just as those of Magelona
do, on being shed from the vessels.
The globules of Magelona stand, as it were, midway between
the coloured liquid plasma of Annelids generally and the
coloured corpuscles of mammalian blood. The latter are
14, W. BLAXLAND BENHAM.
known not to be cells, but to be parts of cells, produced in
many cases within cells by the modification of the protoplasm,
its coloration and subsequent separation into globules, which
are then set free by the breaking up of the parent cell.
With regard to the development of the globulesin Magelona
I have not yet come to any definite conclusion, though I
expect to find that a similar origin of the corpuscles obtains.!
The formation of the plasma in ordinary Annelids is practically
on the same lines, but the fluid, in place of separating as
globules or corpuscles, is discharged en masse (see Lankester’s
observations “On the Connective and Vasifactive Tissues of
the Medicinal Leech,” ‘Q. J. Micr. Sci.,’? vol. xx, 1880,
p- 307).
5. The Colouring Matter of the Blood.—We have
already seen that no absorption bands were obtainable by means
of the spectroscope, nor have I been able to separate any pig-
ment by the use of various reagents. But from mere tint alone,
and its change of colour on deoxidation, it is evident that the
pigment is quite different from any hitherto recorded amongst
the Chetopods; it is most evidently not hemoglobin. The
other colouring matters known in the Chetopods are the greens
of the Sabellids and Chlorhzmids, known as chlorocruorin.
In no other Chetopod do we find the madder-rose colour
characteristic of Magelona; but in Sipunculids a similar blood-
pigment exists, presenting a similar loss of colour on deoxida-
tion. This colouring matter, as seen in Sipunculus nudus,
closely resembles that of Magelona. Its peculiarities were
first described by Lankester,” and subsequently by Krukenberg,’
who gave to it the name ‘‘ Hemerythrin.”
The blood-pigment in the Sipunculids occurs in the nucleated
circular disc-shaped corpuscles of the fluid contained in the
1 Miss Buchanan describes some of the early phenomena of the formation
of the blood of Magelona (‘ Rept. Brit. Assoc.,’ 1895, p. 469), in which she
suggests that the “corpuscles” arise from subdivision of multinucleate cor-
puscles derived, apparently, from a special blood-forming organ in the dorsal
vessel.
2 Lankester, ‘ Proc. Roy. Soc.,’ xxi, p. 80.
3 Krukenberg, ‘ Vergl. Physiol. Stud.,’ lst reihe, 3te Abth, p. 82.
THE BLOOD OF MAGELONA. te
body-cavity. In Phoronis the pigment is hemoglobin, as
Lankester showed by its spectrum, and is not hemerythrin,
as Krukenberg assumes. In Phascolosoma elongatum
Schwalbe describes the tint as light rose or faint greyish red
(matt graurothlich), which on oxidation grows darker, and finally
becomes “ burgunderrothe.” This description of the colour
scarcely agrees with that of Magelona, nor does that given
by Ehlers and Keferstein for other members of the group—
Sipunculus and Priapulus,—for they term the colour of
the body-fluid ‘ wine-red;” but in Sipunculus nudus at
any rate, as I can testify from my own observations, the tint
is very close to that seen in Magelona.
6. Relation of Vessels to Celom.—The fact that in
the thorax the ventral vessel dilates to such an enormous extent
as to nearly obliterate the ccelom is a very interesting and im-
portant piece of evidence in favour of Professor Lankester’s
suggestions (‘ Quart. Journ. Mic. Sci., xxxiv, 1893) as to the
manner in which the heart of Arthropods, with its series of
ostia for the entrance of blood, has been derived from a Cheeto-
pod vascular system, owing to the enlargement of the afferent
vessels or veins, till the latter fuse with one another to form a
great “ pericardial blood sinus,’’ to the exclusion of the ceelom.
Here, in Magelona, we have a great dilatation of the ventral
vessel and lower part of the “ afferent” vessel leading upwards
to the dorsal vessel. This great lateral dilatation (J. ext.) com-
municates with the ventral vessel just behind the septum,
between consecutive segments, by a comparatively narrow
opening, as shown in the figure. If this dilatation extended
further upwards to the dorsal vessel we should have a condition
closely approaching that of the Arthropod vascular system.
This ventral vessel of Annelids no doubt corresponds to the
great ventral blood sinus of Astacus, &c., whence the blood
makes its way up to the heart, either directly or by way of
the gills.
I hope, in a later contribution, to give a further account of
the vascular system of this interesting worm, Magelona.
16 W. BLAXLAND BENHAM.
EXPLANATION OF PLATE 1,
Illustrating Dr. W. Blaxland Benham’s Paper on ‘‘ The
Blood of Magelona.”
EXPLANATION OF FIGURES.
Fie. 1.—Side view of the anterior region of Magelona, with the “‘pro-
boscis” withdrawn. The thorax is coloured by the blood contained in its
vessels.
Fic. 2.—Side view of the anterior region of Magelona, with “ proboscis”’
(a) everted. The thorax is now almost colourless, the blood being driven
into the proboscis.
Fie. 3.—A transverse section through the thorax of Magelona, showing
the great dilatation of the ventral blood-trunk and its extension into the
“lateral chamber” of the body. On the right side one of the dorso-ventral
muscles is represented separating the lateral extension of the vessel from the
median portion. This dilated vessel comes to occupy almost the entire ccelom.
The epidermis (ep.) is somewhat diagrammatically represented; the remainder
is drawn with the camera. D.v. Dorsal blood-vessel, with thick muscular
coat. V.v. The dilated ventral vessel, with thin nucleated wall. Jat. ext.
and Z. ext. Lateral extension of the ventral vessel. /.v. Vessel from ventral
to dorsal vessel. Ca. Celom. wn. Nerve-cord. G. The gut. cére. Circular
coat of muscles of the body-wall. ¢.m. Longitudinal muscles. dor.v.m.
Dorso-ventral muscle. 0d/. Oblique transverse muscle. (Camera Zeiss B 2,)
Fic. 4.—A group of blood-globules. (x 550.) Camera drawing of
contents of a blood-vessel from a section stained in borax carmine. The size
of the corpuscles is practically unaltered by the preservatives. . Nucleus.
Fic. 5.—Two globules from same section, greatly magnified. The con-
tents have been coagulated by reagents.
Fic. 6.—A group of globules (x 550) from a preparation of blood killed
with Fleming’s solution, stained in picro-carmine, and mounted in glycerine.
The globules have ‘‘fused” with one another, so that the field is filled by
larger globules of various sizes. A coagulated plasma (p/.) is represented.
A nucleus (z.) shows the relative size of the globules. (Camera drawing.)
. Fic. 7.—Some isolated nuclei, showing varying numbers of chromatic
masses. From same preparation.
Fie, 8.—An example of nuclei apparently contained within large globules
(from picro-carmine, glycerine preparation). #. Nucleus. pl. Plasma.
Camera. xX 550.
THE BLOOD OF MAGELONA. ily?
Fie. 9.—Other examples of the same more highly magnified. In a the
nucleus, overlying the globule, projects beyond its margin.
Fic. 10.—Freshly drawn blood, without reagent, uncovered (Zeiss, F 2),
The globules do not freely separate from one another, but adhere together,
forming an irregular mass or network. c. A few isolated globules. ¢. De-
licate thread-like bridges of corpuscles. #’. The same broken across by the
movement of the mass, resulting in the isolation of some globules.
Fic. 11.—A portion of the same preparation covered; the same magnificas
tion (Zeiss, F. 2). The irregular mass is now seen to consist of masses of
corpuscles or globules, forming a network with irregular meshes.
Fig. 12.—A portion of a similar drop of blood in salt solution, to which
water has been added. The threads of the network have become greatly
elongated and more delicate. The margins (a) of the masses have become
homogeneous, owing to the fusion of neighbouring globules; these still exist
unaltered in the centre of the nodes (4).
Fic. 13.—Further action of water (lower magnification).
Fic. 14.—A small drop of blood after addition of water, showing the
fusion of a large number of globules to form larger and larger clear masses.
Fie. 15.—Another drop of blood, with a trace of sea water. (Camera
drawing. xX 550.) Various sized drops have been formed by fusion of
globules.
Fic. 16.—Action of salt solution after water. (a) Mass in water, action
commencing. (4) A mass of globules in water have fused to form a round
mass. (c) Salt solution added; change of shape and movement of the mass;
protrusion of processes. (d) The changes are more marked, and small drop-
lets (w) are separating from the ends of processes. (This figure is drawn
on a slightly larger scale than the other two.) In the mass there is a hole (7)
which in @ has made its way to margin.
VOL. 39, PART 1,—NEW SER. B
ged. : Ns (ie Sa AP Ain Mb
7h. >» SA Fe oy =k eae; coy SX r
any so A
-
abate. Ole WP Sate Sine at
: ‘ ewes iv glk ale plat tira
o£ § ug e ie} b a est 5 veo VS Bre
a Day
aa y ; le
At hb 0 7] 708 Cee ae
;
= alt Yaer ee Ire a ee Ais: .
= ii yra ¥4 = FS" ia Wale ‘ : i BAT ae al Sie i
CR orf aaah eee
oie ss130 Sry ge os. + a » EAS tcp ie oe © +i
f at t tv = way a om ae utp? A) Sol ; wile -
x ton © ts, x = oe “
fee yf aa SVK a) cit =i fra) +aig@yrr an |
rarze (© gd tp ee wR » UDO aan
: : ert & 19 hana
res 7 6 > 1a i. ; ‘ !
4 De,
= e rs
7)
. way hi
Fd a
% rt hes i" ‘aa¢ + % =
a — ’ *
= : y i '
2 ee
“| f 4 ¥ fia,
22° i + te Ae The r Spy ser 6S ‘ 4 '
7 ~ {
45751 Scat oT i’ f
tH SIRR eis Swth sh ihe wien’ aie
oo FMW UE CLT SEER aie
s
é
_ 4 » et
F ee a AY Ear Fi 3
> ‘ —_
G \ ..
¥ fe
FISSION IN NEMERTINES. 19
Fission in Nemertines.
By
W. Blaxland Benham, D.S¢.Lond., Hon. M.A.Oxon.,
Aldrichian Demonstrator in Comparative Anatomy in the University
of Oxford.
With Plates 2 and 3.
Ir is a well-known fact that many Nemertines break up into
pieces when irritated; but although this statement is current
in text-books as well as in special memoirs the matter has
received very little attention, and I find no account of any
internal changes which may take place in the tissues of the
body previous to the process of fragmentation. Indeed, it
appears that the statement is made rather vaguely, and its
application is scarcely so general as might be inferred. There
is no doubt but that Carinella, Lineus, and other elongated
and comparatively slender forms do fragment, and that these
fragments can remain alive for a considerable time, as Dalyell,
and later M‘Intosh have recorded ; and further, it appears that
the anterior end of such a fragmented Lineus may produce a
new posterior end, and M‘Intosh figures various stages in the
formation of a head to amore posterior piece. As the specimen
which he kept under observation lived in very unfavorable
conditions of food, &c., he did not observe the completion of
the head even after some months, but remarks that in a state
of nature such a regeneration most probably takes place fre-
quently.1. Further, we are left somewhat in doubt as to
whether a Nemertine does of its own accord, and independently
of irritation, normally break up into pieces.
1 M‘Intosh, ‘Mar. Brit, Annelids: Nemertea,’ p. 125,
20 W. BLAXLAND BENHAM.
My attention was called to this matter in 1894 on examining
a small Carinella from St. Andrews that happened to be
amongst a number of specimens of Magelona, which Pro-
fessor M‘Intosh had most kindly sent to me for the purpose
of certain work on the latter worm on which I was engaged.
This Carinella presented, in the hinder region of its body,
two conspicuous constrictions at nearly equal distances from
one another, A B, and each piece thus marked out was further
constricted, though much less deeply, C D, into two nearly
equal portions (Pl. 2, fig. 1).° So striking was this apparent
“segmentation ”’ of the body, that I preserved the worm for
sectionising. It was killed in corrosive sublimate to which
1 per cent. acetic acid was added.
Later on in the year, while working in the Marine Labora-
tory at St. Andrews, with a view to collect and observe living
Magelona, I came across two more specimens of this same
Nemertine, which I found living in the sand with the Poly-
cheete below the ordinary low-water mark. Both these worms
were about the same size as the one I had obtained previously,
and both presented similar phenomena of ‘ segmentation.”
One of these was unfortunately mislaid; the other I killed,
stained, and mounted entire.
The Carinella to which these observations refer is a small
worm about two inches (55 mm.) in length when in moderate
extension. It is pure white, without eye-spots, and without
pigment except for a girdle of a faint yellowish-brown colour,
about a quarter of the worm’s length from the anterior end
(fig. 1, 6). This is very faint even in life, and is due to pig-
ment in the epidermis, Further, the proboscis is coloured
faint blood red for a short region just in front of this pigmented
band (p).
The body is nearly cylindrical in section; the head is pointed,
and slightly marked off from the body by a lateral vertical
furrow on each side at the level of the mouth; the head is not
noticeably wider than the body (fig. 2).
The posterior end of the worm, as longitudinal sections
showed, is not complete—a portion has already been separated,
FISSION IN NEMERTINES. 21
so that I cannot state the true length; but at first sight this
end does appear entire and rounded off (PI. 2, fig. 1).
In transverse sections (fig. 10) probably the most noticeable
feature is the thinness of the circular muscles; so thin is this
layer that it cannot be distinguished in transverse section
through the middle of the body until longitudinal sections
are called in to help:! the basement membrane is also thin.
The two nerve-cords lie between the basement membrane
and the circular muscles, and this position at once shows that
the worm belongs to the genus Carinella; but as to its
specific identity I cannot be so sure. I hesitate to confer a
new name on the worm, as it bears a certain likeness to
C. linearis, Montagu, which is described by M‘Intosh; and
one of his two woodcuts illustrating this species (p. 206) also
shows constrictions of the hinder end, recalling at once the
phenomenon presented by the present worm, although the
“ segments ” are not so nearly equal in size. But this point
would not be of any moment, for in the stained entire worm
(Specimen I1) there is less equality in the size of these “ seg-
ments” than in the one I first noted and drew when alive.
There is, however, a statement in M‘Intosh’s description
which renders an identification doubtful, for he states that
C. linearis is “ flattened,” and this is certainly not the case
with my worm at any rate—transverse sections do not exhibit
any flattening. M‘Intosh states that “the inner muscle of the
body-wall shows a marked tendency to separate in the dorsum.”
If this means that the longitudinal coat is traversed by a
band of tissue in middle line my specimen agrees with his.
Nevertheless no great weight can be placed on this fact, for
Birger has shown that this occurs in probably all species of
Carinella. But from the general description, and from
quotations from Montagu’s MS., it appears at any rate
probable that the present worm is C. linearis, for it agrees
in habitat, and in the fact that it ‘‘ secretes a tenacious slime
from its body, which, collecting sand, readily forms a covering
1 The thickness of the circular muscle-coat has been intentionally exag-
gerated in the figure.
22 W. BLAXLAND BENHAM.
like that of Sabella.” It agrees more closely with this species
than with any more recently described, such as C. albida,
Biirger, so that I must leave the matter of species open.
The hardened specimen was sectionised as follows :—(1) The
anterior end was’cut sagittally (vertical longitudinal); (2) the
hinder end was cut horizontally, though rather obliquely
to the horizontal plane; (3) the middle region of this body
was transversely sectionised. The stain used was carmalum,
though some sections were stained with picro-carmine.
With regard to the posterior “segmented” region of the
body, one is struck at once by the fact that at the level of each
of the four constrictions there is a transversely arranged
row of small nuclei traversing the longitudinal muscles. They
are readily seen even with a low power (fig. 3), and are more
evident at the deeper constrictions a, B, than at the shallower
ones ¢, D.
What appears to be the commencement of the phenomenon
is illustrated by the sections across the constriction c (fig. 4).
At this point the tissues are as yet perfect—there is no
rupture. The section figured passes through the lateral line
of the body, on one side cutting along the nerve (NV); at the
opposite side the nerve is notinvolved. The nuclei constituting
the transverse row are not as yet as definitely arranged as they
will be; nevertheless they are more numerous along the level
of the constrictions than elsewhere, and take a deeper tint in
the stain. It may here be mentioned that the accompanying
figures are drawn with the camera, and so far as was possible
every nucleus is put in its true position, except in the case of the
epidermis and intestinal epithelium where these are involved.
The nuclei which are thus conspicuous appear to belong to
the connective tissue in which the longitudinal muscles are
wrapped ; they are smaller and more deeply stained than those
of the muscles. Although the impression given by the section
fig. 4, through the region c, appears to indicate an irregular
transverse line, yet at a further stage this line becomes very
much more distinctly transverse, and, moreover, it is in reality
double; the nuclei are arranged into two rows close to one
FISSION IN NEMERTINES. 23
another, as is shown in fig. 5, through the region p. Here,
however, we have a further phenomenon: the double row of
connective-tissue nuclei having been established, the longi-
tudinal muscles break across between the two rows of nuclei.
This rupture appears to start just below the epidermis, and
then travels inwards. But already the epidermis exhibits a
furrow even at an earlier stage, as at c (fig. 4), so that it is not
merely the result of a contraction of the muscles.
The muscles, once ruptured, naturally contract, and leave a
space which appears in sections to be partially occupied by a
coagulum (2), though it may bea pre-formed material—part of
the connective tissue. This appears to be the more likely, as is
seen at a later stage in the process, as in fig. 6, which passes
through the middle of the region a. Here the rupture of the
muscle has gone a step further, and is extending towards the
opposite side; this figure also illustrates another point—viz.
that the rupture may commence at one point of a plane and
extend in all directions in a radial direction, so that while it
is complete at one side of the body it may not have yet com-
menced on the opposite side.
In this figure, for instance, the muscles have shrunk con-
siderably at the left side; yet, though the rupture is extending
towards the right, the muscles are still entire below the epi-
dermis of the right side. By the rupture of the muscles the
linear arrangement of the connective-tissue nuclei is to some
extent destroyed.
Finally, the epidermis, which during this process has become
thinner, as well as furrowed at the plane of constriction, gives
way; but the basement membrane still appears to resist the
rupture, for in fig. 7 there is a distinct membrane (dm.) left
after shrinkage of the epidermis. We must consider this
basement membrane, as Hubrecht and others have suggested,
as a firm skeletal tissue, and its resistance to the process here
described is only what would be expected.
When the rupture is complete, or even before it has travellod
all the way round the worm, the circular muscles come into play
(right side of fig. 8), and drawing together the margins of the
24 W. BLAXLAND BENHAM.
wound formed by the rupture, form a boundary to the portion
separated as well as to the anterior region. The gonads are thus
held in place, and do not project through the wound, and hence
the rounded end of the body looks like the true anal end. During
the above process the intestine has, of course, become nipped.
After the epidermis has been ruptured one notices the
surfaces of the wound to be covered with a flat epithelium
(fig. 8, ep’). Whence comes this? Is it derived from the
connective tissue? It looks as if the row of nuclei before
mentioned had flattened out, and so given rise to the flat nuclei
of this membrane.
This appears to be the general history of the process of
fission. But we are no nearer to the answer—how do the
muscles become ruptured? why do they all give way at this
particular plane? Further, is it merely a rupture of the
fibrille, or is there a degeneration of these fibrillz over a cer-
tain small area, viz. between the two rows of nuclei? Have
these nuclei, or rather the cells belonging to these nuclei, any
part to play in this rupture or degeneration, or are they merely
a preparation for the new membrane which forms the ends of
the fragments of the body ?
At present I am unable to answer these questions with
certainty. It appears to me, after a careful study of my pre-
parations, that the muscle-fibrils do rupture; at any rate, if
there is a degeneration of tissue, it occurs over only a very
minute distance: the ends of the ruptured fibrils appear clean-
cut (fig. 6), and I see no sign of any modification in their
substance which would point to a degeneration.
With regard to the second part of the query, viz. the action
of the row of nuclei, I would suggest that they do play an
active part in the process. The muscle-fibrils are wrapped
together by connective tissue, which has the form of a network,
through the meshes of which the fibrillz pass (fig. 9). At the
point of future rupture this network is denser and the meshes
smaller, and it may be that the cell-substance becomes actively
contractile, and really nips the muscle-fibrils in two. On
the other hand, we must not overlook the possibility of the
FISSION IN NEMBERTINES. 25
occurrence of some kind of solvent action—the fibrils may be
“eaten through,” as it were; but I know of no means of
deciding the matter.
The process, whatever its details, seems to be different from
that which occurs in Planarians, such as Microstoma, which
reproduce asexually by fission.
In his account of the process, van Wagner (‘ Zool. Jahr-
buch,’ iv, 1891) expressly states that there are no nuclei in
the “ septa,’ which make their appearance along the lines of
the future separation of the individual into zooids. Each septum
is represented in v. Wagner’s illustrations as being of some
homogeneous tissue, passing from the subepidermic muscles
to the muscularis of the intestine. He gives no details as to
how the “ parenchyma” separates, which bear any resemblance
to those detailed in the process of rupture of the muscles seen
inthis Nemertine. Further, there is a great deal of ‘‘ regene-
ration ”’—i.e. formation of new brain, new pharynx, &c.—in
Microstoma before separation takes place. Of this there is
no trace in Carinella. I need not enter here upon the vexed
question as to the distinction between “ division” and “ gem-
mation.” In Carinella there seems no doubt that the process
is “ division.”
A second part of the problem relates to the meaning of the
whole process in regard to the life of the animal. It is
generally assumed, and apparently rightly assumed, that the
process of rupture will be of advantage to the Nemertine, in
that each of the pieces thus formed will be able to give rise to
gonads, as M‘Intosh observed; in most cases, however, there
does not seem to be any direct relation between rupture and
genital maturity. But in the present instance there is such a
relation, for in both the specimens the genital elements are
only present in the hinder part of the worm, which
is being separated off from the rest.
In the specimen I (fig. 1), which is a female, the gonads are
present in each of the four “segments,” and no trace of ova
exists in front of the line B, while behind the line A the ova are
larger and more numerous than in front of it.
26 W. BLAXLAND BENHAM.
The specimen II, which was mounted entire, is a male, and
the spermatozoa exist only in the hindermost part of the body.
It would appear, therefore, that the gonads make their appear-
ance simultaneously over a certain stretch of the body, com-
mencing at the hindermost region; that as they ripen this
region begins to be constricted off from the rest of the body.
A second tract now commences to ripen, and in its turn becomes
marked off from the non-sexual anterior region of the body ;
and so the process goes on up to the point of origin of the
proboscis.
Further, each such region may become restricted into two
(or more?) pieces ; and it will be very interesting to ascertain
whether this segmentation goes any further than in the speci-
mens to hand. It is possible that the segmentation of this sexual
region of the body, which is already beginning in our specimens,
may ultimately become so far repeated that each “‘ segment ”
will contain only a single pair of ovaries. In that case we
should have a strong case in support of Hubrecht’s theory of
segmentation, as elaborated in his ‘ Report on the Nemertines
of the Challenger Expedition.’
However that may be, we have in this Carinella what
appears to be an undoubted preparation for the spontaneous
subdivision of the sexual region of the body into a number of
isolated portions, and this subdivision is due to the develop-
ment of the gonads, and is thus directly related to the pro-
pagation of the species.
We have yet to learn whether each of the pieces so isolated
will produce a new head; but it appears to me a needless
assumption that this would happen. As each “segment”
drops off, the tissues after decomposition might set free the
generative products; but since genital ducts (fig. 3, g.d.) are
already forming, and are in the hindmost segments nearly
completed, it appears that the segments live an independent
life for some time after separation. This fact, therefore, mili-
tates against the idea of any greatly extended “‘ segmentation.”
Anatomical.—The Nemertine which forms the subject of
the present contribution exhibits one or two points of interest
FISSION IN NEMERTINES. 27
in its histology, which differs from that described by Birger!
for Carinella and its immediate allies.
The lateral nerve-stems do not exhibit the characters
described and figured by that author, who finds an inner and an
outer neurilemma separated by a considerable space. In this
lie numerous ganglion-cells, chiefly collected above and below
the “punktsubstanz,’ which is surrounded by the inner
neurilemma, and to which the processes of the ganglion-cells
pass. In the present worm the ganglion-cells are very few, at
any rate towards the middle of the worm, though they are
more numerous anteriorly. But these cells lie not only above
and below, but at the side of the nerve (fig. 11). Further, I
do not find distinct inner and outer neurilemma. The nerve-
cord is surrounded by a distinct but delicate membrane, with
flattened nuclei (m’ m’”); this appears to correspond to Biirger’s
outer neurilemma. The ganglion-cells are situated in a net-
work of tissue, resembling, in general, the substance of an
ordinary Invertebrate nerve-cord, and the meshes of this net-
work are occupied by fine dots—the “ punktsubstanz ”’ (p.).
Lying near the centre of the nerve-cord is a distinct tube,
with well-marked wall, against which lie here and there nuclei
resembling those of the ganglion-cells. This tube (ch.) usually
appears empty, but in some sections a faint coagulum with
very minute scarcely visible granules init. There is no net-
work, such as occurs in a nerve-cord, and I take this tube to
be a “ giant fibre” or “‘ neurochord,” such as Biirger finds in
Nemertine nerve-trunks generally.
In Carinella, according to that author, there are none,
whereas in the Schizonemerti he finds several of them. It
might be suggested that what I regard as the wall of a tube or
neurochord is, in reality, Biirger’s “ inner neurilemma;” but
the contents of this tube are certainly not like the structure
seen in an ordinary Invertebrate nerve-cord. My sections
were stained in picro-carmine, and others in alum carmine.
I did not employ hematoxylin, and it is possible that a
1¢ Unters. ub. d. Anat. u. Histol. d. Nemertinen,” ‘Zeit. fiir wiss.
Zool.,’ 50.
28 W. BLAXLAND BENHAM.
renewed investigation with other preservatives and stains may
lead to a different interpretation of this “tube ;” but, so far
as my preparations go, | cannot think that this is an “inner
neurilemma,” or that the contents form the “ punktsubstanz.”
The muscles of the body-wall are chiefly interesting for the
great thickness of the longitudinal coat, and the very feeble
development of the circular coat in the greater part of the
body, though anteriorly this coat becomes rather thicker.
The ‘inner coat of circular muscles,” which has so often
been described for the Paleonemertini, has been identified
by Birger as the muscle of the rhynchocele (proboscis sac)
and gut, which, owing to the slight development of the paren-
chyma, comes to lie immediately within the Jongitudinal coat.
This explanation brings the muscular system of the Palzone-
mertini into agreement with the general scheme of Nemertine
musculature.
This inner coat (r/.c.) is in the present worm very thin indeed,
but can easily be recognised in longitudinal sections. I have
exaggerated its thickness in the figure of a transverse section,
This inner coat can be traced as a sheath round the rhynchoceele
and intestine (éné. c.), and a second coat surrounding the rhyn-
choceele alone.
Birger describes a decussation of this inner coat in the
median dorsal line. I am unable to satisfy myself whether
this exists in the present worm; but traversing the longi-
tudinal muscles is a vertical strand of tissue (fig. 10, a.) stain-
ing like the basement tissue and continuous with it dorsally,
and with a similar tissue surrounding the rhynchoceele ;
whether the circular muscles accompany this strand—as I
believe they do—I was unable to determine. The intestine, as
I have just stated, is partially—i.e. on the ventral side—
surrounded by a layer of circular muscles; in addition, its
dorsal wall is formed by a fairly thick layer of longitudinal
fibres (int. lg.), lying between the intestinal epithelium and
the wall of the rhynchoceele.
The proboscis presents throughout the greater part of its
extent a glandular epithelium, surrounded by circular muscles,
FISSION IN NEMERTINES. 29
and then by a thick coat of longitudinal muscles. But the
epithelium changes its character for a short space near its
anterior end; the gland-cells, so conspicuous elsewhere, no
longer catch the eye in section, if indeed they are present ;
but the cells are long, narrow, with flat nuclei at the sides, and
close to the free end there is a series of round nuclei arranged
in a well-defined row.
In longitudinal section this region occurs between 2 and
3mm. from the anterior end; whether it corresponds to the
blood-red tract on the proboscis, noticeable in the living worm,
I am not certain.
In all the Palzonemertini the parenchyma is but very
feebly developed, but in the present worm, at any rate behind
the middle of the body, it is practically absent. I can detect
no space between the wall of the gut-muscle, and body-wall,
i.e. between the inner circular muscles and the longitudinal
muscles, except the longitudinal blood-vessel on each side and
a nearly homogeneous tissue (y.) passing up from it to the inner
wall of the intestine; this tissue resembles the ordinary con-
nective tissue, which occurs in some sections in sheets quite
like this little piece. This is all that represents the “ paren-
chyma”’ posteriorly.
There is one point in which this Carinella does not agree
with Biirger’s statement as to the arrangement of the gonads.
He states that in the Palzeonemertini there is usually more
than a single pair of gonads in a transverse plane, so that a
transverse section through this region would exhibit more than
two gonads and ducts. ‘This is not the case with the present
Carinella. The gonads are strictly paired; there is one,
and only one, on each side,—each with its duct formed in the
usual way. ach ovary when nearly ripe consists of three or
four large ova, surrounded by a common membrane of flat
cells, which is continuous with a heap of indifferent rounded
cells on the outer side of the ripe cells (fig. 12). Amongst this
mass of rounded cells one can distinguish one or more cells,
still small, but distinctly larger than the rest, with a distinct
germinal vesicle (ov.); these are evidently young ova.
30 W. BLAXLAND BENHAM.
EXPLANATION OF PLATES 2 and 3,
Illustrating Dr. W. Blaxland Benham’s Paper on “ Fission in
Nemertines.”’
Fie, 1.—View of Carinella (No. 1) immediately after death, showing the
“segmentation ”’ of the hinder part of the body. (x 6.) A, B, GC, D, indi-
cate the constrictions in order of depth, 4 being deepest. The drawing also
shows the general appearance of the worm. pp is the red region of the
proboscis. The body is coloured orange in the region just in front of 4.
Fic. 2.—Ventral view of the anterior end, to show the shape of the head.
M. Mouth. &. Aperture of the rhynchodeum. JZ. Ciliated furrows.
Fie. 3.—Camera drawing of the hinder end of the worm. A, B, C, D.
The constricted regions, each marked out internally by a double row of
nuclei. The section is nearly horizontal. ov. Ovaries, fully developed in the
hinder “‘segments.” ov’. Less developed ovarian sacs in anterior regions.
g.d. Genital duct. zt. Intestine. ep. Epidermis. /y. Longitudinal muscle.
Mag. Zeiss AA 2.
Fic. 4.—The region C enlarged. The section is at a different level from
that of the preceding figure, as it cuts the nerve-cord on the left, and does
not involve the intestine. ep. Epidermis. 4.m. Basement tissue. circ.
Circular muscles. J/g. Longitudinal muscles. 2. The transverse row of
nuclei. WV. Lateral nerve-stem. Mag. Zeiss B 4.
Fic. 5.—Horizontal section through the region D, to show the next stage in
the process of fission. The longitudinal muscles have ruptured, contracted,
and left a cavity just below the nerve-cord. 2. A coagulum.
Fic. 6.—The next stage in the process, illustrated by section through the
region 4. The ruptured muscles are more numerous, and the cavity (sp.)
therefore more extensive, and is traversed by connective tissue (c¢.¢.). Mag.
Zeiss B 4.
Fic. 7.—Another portion of the same region, to show the coagulum (z.),
the broken ends of the longitudinal muscle-fibres, and the unruptured base-
ment membrane (4.m.). 4v. is the lateral blood-vessel.
Fie. 8.—Portion of A (fig. 3) at a different level, enlarged. The rupture
has extended nearly right across, and the ends of the two pieces are bending
inwards owing to the contraction of the circular coat of muscles. This figure
also shows the flattened epithelium (ep’.) which covers the broken ends of the
muscles.
FISSION IN NEMERTINES. 31
Fic. 9.—A somewhat diagrammatised view of the muscle-fibres and con-
nective tissue in the region of one of the “septa.” m. Muscle-fibres.
c.f, Connective-tissue network, through the meshes of which the fibrils pass:
the network is denser in the region of the “septum” than elsewhere. 7!
Nucleus of ‘‘ septum.”
Fic. 10.—Transverse section of the worm (mag. B 2,cam.). ep. Epi-
dermis. 4.¢. Basement tissue. 4v. Blood-vessel. circ. Circular muscles
of body-wall. 7g. Longitudinal muscles. zz¢. Intestine. izé.c. Circular muscle
of intestine. zz¢./y. Longitudinal muscle of intestine. Pr. Proboscis. P.ep.
Outer epithelium of proboscis. RA. Rhynchocel. R.ep. Epithelium of
rhynchocel. rh.c. Its coat of circular muscles. WV. Lateral nerve-stem.
x. Streak of tissue resembling connective tissue interrupting the longitudinal
muscles dorsal (? decussation of r/.c.). y. ‘Connective tissue” near blood-
vessel.
Fic. 11.—View of section across lateral nerve-stem. 4.¢. Subepidermic
basement tissue. circ. Circular coat of muscles. ch. Neurochord, with
distinct membrane and (?) coagulum within. m!. Outer membrane (? neuri-
lemma). #', Inner membrane, separated at one point by a space (sp.), probably
due to shrinkage. gy. Nuclei of nerve-cells. .s. Nuclei of sheath. p.
Punktsubstanz in the meshes of nerve-cord. D.V. indicate the dorsal and
ventral surfaces of the nerve-cord.
Fie. 12.—A portion of a section across the region A to show the ripe
ovary. ov. Ovary. O. A ripe ovum. w. Wall of ovary. s. “Septum.”
b.m. Basement tissue exposed after the rupture of the epidermis.
a Roti Phir smear ey
tea mins Emr eSiiabieand sty eel bea enbtar sat aul
an een nlp ind te Oa Sre mitt 7h Age Ty ning we
¢ - omuiiy SAT Sst @ De petals Wank AY oad) Vbe—é a iad th
‘ae, cro wollte aut) “ abhagas HD Wipe, gee 1: ow"
nae ern
7 neck uy . ead a ; Cu), ure ny ny ity " vebvrewtcheme
- Z “of ee re ca 2) 2 oe a} th io% af inaif {ov eped® a)
‘| i viru? vases Pyllinl Jo a t iia oo ~
7 Viv ( atratyl 1 aa. ch Link yeaa an cont
sl i a - . i so ios wil awh
é ’ > ¢ =
: Vip 4 F ; A ri A dat OOS ia
2 i lint iy nde os f rr
€ iY 7 . :
tw Se cl" ae ;
= al he sles} See le. al Tea
; ~-"e
? yo» urea}
> . e} * ‘wit
, : - ae. a
t j if 0 ts, Py “nts
, \ ree ifs xe
; ey olf ho OO ae
tt SPL Ripetecdl
7 nae we aes Pe
aS
i? ;
A
4 .
, .
f ’
. Qghmaeny
-— AY Oe
a
STUDIES ON THE NERVOUS SYSTEM OF CRUSTACEA. 33
Studies on the Nervous System of Crustacea.
By
Edgar J. Allen, B.Sc.,
Director of the Plymouth Laboratory of the Marine Biological Association.
With Plate 4.
IV. Further Observations on the Nerve Elements of the
Embryonic Lobster.
In the first part of these studies! an account was given of
certain nerve elements which had been demonstrated by the
action of dilute solutions of methylene blue upon the ganglia
of the embryonic lobster (Homarus vulgaris). The elements
then described arose, in most cases, from cells which lay in the
anterior ganglia of the thorax. In the present paper a number
of additional elements occurring in these ganglia will be
noticed, together with those found in the posterior thoracic and
in the abdominal ganglia.
The method of investigation has been the same as that
described in my former communication, the results in this
case, however, having been obtained principally from embryos
which were near the point of hatching. Several of the new
elements differ in essential points from any of those previously
described, and are of interest in throwing additional light upon
the manner in which different portions of the central nervous
system, or different movements of the body, are co-ordinated
one with another.
1 This Journal, vol. 36, pt. 4, 1894; a preliminary notice occurs in ‘ Pro-
ceed. Roy. Soc.,’ vol. lv.
VOL. 39, PART 1.—NEW SER. C
34 EDGAR J. ALLEN.
The observations will be dealt with in the following order:
I. Elements arising from cells situated in the THoracic
GANGLIA.
(2) Elements of types the same as or similar to those
previously described.
(6) Elements belonging to types not previously described.
I]. Elements arising from cells situated in the ABDOMINAL
GANGLIA.
III. Elements arising from cells outside the central nervous
system.
I. Tuoracic GANGLIA.
(a) ELEMENTS SIMILAR TO THOSE PREVIOUSLY DESCRIBED.
B, Tuorax xt! (fig. 1).—The cell lies in the anterior
portion of the lateral mass of ganglion-cells of Thorax x1,
and gives off a fine fibre, which passes into the neuropile
( punkt-substance”’), and there breaks up into three main
branches. Of these branches one passes forwards into the
next ganglion in front, another backwards to the ganglion
behind, whilst the third curves round the central ganglion-
cells, and, keeping close to these, runs forward towards the
brain on the opposite side of the cord to that on which its
cell lies.
By actual observation this fibre has been traced as far for-
wards as Th. viii, but there can be no doubt that it behaves
in the same way as B, Th. v, and B, Th. vii (Pt. I, pl. 35,
fig. 3),? and runs forward to the brain. A comparison of the
three elements makes it evident that they belong to one and
the same system, B, Th. vitt, and B, Th. x1, especially resem-
1 The denomination of the elements to be adopted here will be a continua-
tion of the system used in the first paper. On comparing fig. 1 of this paper
with fig. 1 on Pl. 35 of Part I (this Journal, vol. 36), it will be seen that in
the present case the lateral mass of ganglion-cells in each ganglion (shaded
blue) is more completely divided into an anterior and a posterior portion by
the neuropile (left white in the figure) and the fibres of the anterior nerve-root.
The present diagram represents more nearly the appearance presented by very
late embryos, whilst the former figure represents that found in earlier ones.
2 This Journal, vol. 36.
STUDIES ON THE NERVOUS SYSTEM OF CRUSTACEA. 35
bling one another in details of arrangement. Also, in a pre-
paration of an earlier embryo, in which the fibres B, Th. v,
and B, Th. vir1, were completely stained, a third fibre was
observed ending with them in the brain, running close to them
along the cord, but proceeding backwards beyond Th. vir.
Its destination at that time remained undetermined, but there
can be little doubt that it was the fibre of this element, B, Th. xt.
It will be observed, too, that Th. xr is exactly the ganglion
in which we should expect to find another element of this class.
Such elements have already been described (see Pt. I, pl. 36,
fig. 3) in Th. 11 (with branches to Th. 1 and Th. 111), in Th. v
(with branches to Th. 1v and Th. vi), in Th. vitr (with
branches to Th. vir and Th. 1x), and now we find a similar
one in Th. xi (with branches to Th. x and Abd. 1). In this
way each ganglion of the thorax must be influenced by these
elements, which end in a particular region of the brain.
The series of elements of which the element B, Th. x1, is a
type resembles in many respects the series of elements de-
scribed by Retzius,' in Amphioxus lanceolatus, the cells of
which give rise to the giant fibres in the nerve-cord of that
animal. These fibres, after leaving the cell and crossing to
the opposite side of the cord, are described as running forwards
for some distance, but their ultimate fate was not determined.
A knowledge of the nature and position of their endings would
be of great interest.
PropaBLe Motor Exements or Anterior Roots.—The
elements E, G, and H (fig. 1. Compare also Pt. I, pl. 35,
fig. 1), the fibres of which leave the ganglia by the anterior
nerve-roots, and which have already been described for the
anterior ganglia, occur also in Th. 1x, x, and xt.
The elements E (4) and H (e) (fig. 1) resemble in general
features the element E, but differ from it in the number and
mode of branching of the fibres which they give off to the
neuropile. Their characteristic appearance and the situation of
the cells may be seen from the figure. They occur upon both
sides in the ganglia from Th. vii1 to Th. x1.
1 Retzius, ‘ Biol. Untersuch. Neue Folge,’ ii, 1891.
36 EDGAR J. ALLEN.
The element F (0) (fig. 1), which occurs in Th. x and x1,
resembles the element F (Pt. I, pl. 35, fig. 1), and may be the
representative of that element in these two ganglia.
The element P, Th. vi (fig. 1), presents a striking appearance
when stained, and differs in important respects from the
elements previously described. ‘The cell lies in the anterior
portion of the lateral mass of ganglion-cells. The fibre curves
backwards to the neuropile, and after giving off what, from its
relatively small diameter, may be regarded as a subsidiary
branch, divides into two main branches, one of which passes
immediately out of the ganglion through the anterior nerve-
root, whilst the other runs across as a stout transverse fibre to
the opposite side, where it turns back again forming a loop
with itself, and was traced as far as the centre of the ganglion.
When the element has stained upon both sides the two trans-
verse branches lie close together and appear as one stout fibre.
The subsidiary branch, which leaves the fibre before it bifur-
cates, takes the somewhat complicated curved course repre-
sented in the figure, and was traced to the neuropile of the
opposite side. As, however, neither this nor the main branch
was observed to break up into finer branches, it seems pro-
bable that the staining was in all cases somewhat incomplete.
Similar elements have not been found in any other ganglion.
ProspaBLeE Motor ELemeEnts or Posterior Roots.—The
cells of the elements just described all send out fibres through
the anterior nerve-root of the ganglion in which the cell is
situated. In the two following cases, however, the fibre leaves
by the posterior root.
In Th. vist, an element (fig. 1, K, Th. vir) similar to K,
Th. 111 (Pt. I, pl. 35, fig. 1), has stained. The cell is situated
in the central mass of ganglion cells, the fibre decussates with
its fellow of the opposite side and passes out at the posterior
root of the ganglion. This element has been found only in
Then and Th. vars.
In Th. x1 is an element which most nearly resembles the
element J of the anterior ganglia. This is denominated J (3)
(fig. 1). The cell is situated at the posterior end of the central
STUDIES ON THE NERVOUS SYSTEM OF CRUSTACKA. 37
mass of ganglion cells of the ganglion, and is relatively large.
From the cell the fibre passes forwards and slightly outwards
to the centre of the neuropile, where after turning downwards
it bifurcates, sending one branch out of the ganglion through
the posterior root, whilst the other runs as a transverse fibre
to the neuropile of the opposite side. The element has not
stained in any other ganglion of the thorax, but as will be
seen later, what is probably the same element occurs in the
abdominal ganglia.
The element O in Th. vit (fig. 1) also sends its fibre through
the posterior nerve-root. The cell lies in the anterior portion
of the central mass of ganglion cells and pursues the course
indicated in the figure, giving off branches to the neuropile
on the side only on which the cell lies. The element occurs
on both sides of the ganglion Th. vir, but corresponding
elements have never stained in any other ganglion.
ELEMENTS BELONGING TO NEW TYPES.—The motor elements
referred to in Part I, and those already noticed in the present
paper, are all characterised by the fact that the fibre leaves
the central nervous system through one of the roots of that
ganglion in which the cell attached to it is situated. The
portion of the element which lies within the central nervous
system is therefore entirely confined to one ganglion. In the
cases now to be described, whilst the cell lies in one ganglion
the fibre passes out of the cord by the nerve-root of some
other ganglion.
The element Q, Th. vir, is an instance of this class. The
cell lies in the anterior portion of the central mass of ganglion
cells of Th. vir, gives off a fibre which runs outwards and
then upwards to Th. v1, where it passes out by the posterior
root of the ganglion. The fibre gives off a stout arborescent
branch in Th. vir, and a straight transverse branch in Th. v1,
which passes across to the opposite side of that ganglion.
This pair of elements has only stained in Th. vi.
Three pairs of elements having many of the characteristics
of the above, but differing in detail, are found in Th. tv, v,
38 EDGAR J. ALLEN.
and vi. The element R, Th. v1, shows their principal cha-
racters. The cell in this case lies in the posterior portion of
the central mass of ganglion cells of Th. v1.1 It is somewhat
smaller than the cell of the element Q, and its fibre runs
forwards for some distance before turning outwards. After
proceeding in the outward direction, the fibre again turns
forwards and runs into Th. v, leaving the central nervous
system by the posterior root of that ganglion. Two principal
branches are given off during its course, one arborescing in
the neuropile of Th. vi, the other in that of Th. v.
Another pair of elements, whose cells lie in one ganglion
whilst the fibres pass out through the posterior roots of the
ganglion in front, is that lettered S in Th. x1 (fig. 1). In this
case, however, the elements of the opposite sides decussate,
giving rise to the characteristic figure shown (fig. 1). The
cell of this element is small, and lies on the central surface of
the ganglion. The element occurs also in the abdominal
ganglia, and will be subsequently described in more detail. It
has not stained in any other ganglion of the thorax, except
Th. x1.
ELEMENTS HAVING TWO OR MORE BRANCHES WHICH PASS OUT
OF THE CENTRAL NERVOUS SYSTEM BY THE NERVE-ROOTS OF
DIFFERENT GANGLIA.—T wo pairs of elements of this kind have
stained in the thorax, the cells of one pair being found in
Th. vir, those of the other in Th. vi1rt. In the latter case
(Fig. 1, T, Th virz) the cell les in the anterior portion of the
lateral mass of ganglion cells, near the point at which the
anterior nerve-root leaves the ganglion. The cell gives off a
moderately fine fibre, which very soon bifurcates, one branch
passing immediately out of the ganglion through the anterior
nerve-root, whilst the other runs forwards along the ganglionic
cord. The forward branch, keeping close to the lateral masses
1 As was mentioned in Part I (see this Journal, vol. 36, p. 465) the central
masses of ganglion-cells of the embryo, shaded blue in fig. 1, divide at a later
stage into an anterior and a posterior portion, the two portions belonging to
two adjacent ganglia. The line of demarcation is seen in Th, IX (fig. 1).
STUDIES ON THE NERVOUS SYSTEM OF CRUSTACEA. 39
of ganglion cells, pursues a perfectly straight course until it
reaches Th. 111, where it gives off a branch, which passes out
through the posterior root of that ganglion. After giving off
this branch the fibre continues to Th. 11, where it turns and
leaves the ganglion through the posterior root. In one or two
preparations, another branch appeared to be given off from the
fibre in Th. tv, and to pass through the posterior root of that
ganglion, but I was never able to make myself quite sure of
this point.
From the preceding description and the figure (fig. 1), it will
be seen therefore that the element T, Th. viir supplies fibres
to three (possibly four) nerve-roots of different ganglia, namely,
the anterior nerve-root of Th. vir1, the posterior root of Th. rv
(probable), the posterior root of Th. 111, and the posterior root
of Th. 11, and that all these fibres have their origin in a single
cell.
A corresponding element occurs also in Th. vir (fig. 1, T,
Th. vit). The cell is similarly situated to that of Th. vii1, and
a fibre passes almost immediately to the anterior root of
Th. vir. A second fibre pursues a straight course forwards,
keeping close to the corresponding fibre of Th. vii1, until it
reaches Th. 111, where it gives off a branch to the posterior
nerve-root of that ganglion. The fibre then continues to run.
forwards, but its ultimate destination has not been satisfactorily
determined. It appears to turn inwards in Th. 11.
Mention may here be made of a fibre which stains in almost
every preparation from the earliest stages onwards. It is
inserted in fig. 1, T (a). The fibre appears to run longitudinally
through the ganglionic cord from the level of Th. 11 to the
posterior region of the brain, and to give off branches to two
nerve-roots. One branch passes out through the posterior
root of Th. 11, whilst the second leaves the brain by a small
nerve which arises immediately anterior to the nerve which
supplies Antenna 11. A few small fibres are given off to the
neuropile of the brain at the angle which the latter branch
makes with the longitudinal fibre.
No cell has ever been seen to stain in connection with the
40 EDGAR J. ALLEN.
fibre just described, and whether the originating cell lies inside
or outside the central nervous system remains unknown. It is
possible that we are here dealing with an element similar to
the elements T in Th. vir and vii, the cell of which however
has never taken up the methylene blue.
Elements in which one ganglion cell gives rise to two or
more fibres passing out of the ganglion by different nerve-roots
have been described by Retzius in Aulastomum gulo.! In
that case the two fibres traced passed out by the two roots
of the ganglion in which the cell lay, whilst a third, whose
ultimate fate was not determined, passed into the general mass
of longitudinal fibres of the ganglionic chain.
Misce,tnanzous Exements.—In Th. iv a system of
elements occurs, which offers certain difficulties in resolving
it into its component parts. It is found to stain either com-
pletely or incompletely in embryos at almost all stages of
development. Its position and relations to the neighbouring
ganglia are shown in fig. 1 (U, Th. iv), whilst the details will
be best seen in fig. 4. As may be gathered from the latter
figure only two cells, belonging to corresponding elements of
the opposite sides, have stained, but it appears to me to be most
probable that the system contains several sets of elements, the
cells of some of which have never taken up the methylene blue.
The only clue which it has been possible to obtain as to the
course pursued by the individual elements is the fact that in a
number of preparations of late embryos (near hatching) the
portion of the system represented in fig. 5 has been alone
stained, and may therefore represent a single element. If this
be the case, the course of the element may be described as
follows (cf. fig. 1, U., figs. 4 and 5): Starting from a cell in
the anterior portion of the lateral mass of ganglion cells of Th.
iv, the fibre runs near the dorsal surface of the ganglion with
an almost straight course inwards towards its centre, where it
gives off a pair of branches (figs. 4 and 5, c.) which run down-
1 Retzius, ‘ Biol. Untersuch. Neue Folge,’ ii, ‘ Zur Kenntniss des centralen
nervensystems der Wiirmer.’
STUDIES ON THE NERVOUS SYSTEM OF CRUSTACEA. Al
wards and break up into tufts of fibres near the ventral surface.
The original fibre continues its course to a point a little beyond
the centre of the dorsal surface of the ganglion, where it
bifurcates, one branch turning upwards and curving outwards
to the posterior root of Th. 111 (figs. 4 and 5, a), whilst the
second branch continues in a transverse direction for some
distance, and then turns suddenly forwards and runs into the
neuropile of Th. 111 (figs. 4 and 5,6). Its fate here is a little
uncertain, but it often presents an appearance which suggests
that it ends in a tuft of fine branches, as in fig. 4, b. (on the
right hand side of the figure). It should be mentioned that
there is one difficulty in regarding the portion represented in
fig. 5 asa single element, namely, that two central branches
(c) are stained, whereas there appear to be only two when the
complete system (fig. 4) is stained.
Returning to fig. 4, it will be seen that the whole system
there represented contains, in the first place, two elements
similar to that in fig. 5, and lying upon opposite sides of the
ganglion. In addition to this pair of elements, there is a fibre
upon each side (fig. 4, d.), which appears to start from the
point where the transverse fibre turns forwards to form the
branch 6. This fibre (d) runs backwards for a short distance
and then turns outwards to the posterior root of Th.1v. It is,
however, probably not simply a branch of the element already
described, but has a transverse portion of its own running
parallel and close to the transverse portion of that element.
In some preparations it is clear that the main transverse fibre
of the whole system is of a composite nature. This is indicated
on the right-hand side of fig. 4.
A third fibre belonging to the system (fig. 4, e) appears to
spring from near the centre of the transverse fibre, to curve
forwards, and finally to pass out at the posterior nerve-root of
Th. 111. It seems to be impossible to determine from embryos
whether the fibres d and e are independent elements whose
cells have never stained, or merely branches of the other
elements of the system. It is to be hoped that the study of
young adults may throw light upon this point,
42 EDGAR J. ALLEN.
Two other fibres in the same region often stain, and are
shown in fig. 4, fandg. The fibre f appears to enter by the
posterior nerve-root of Th. 111, to pass transversely across the
ganglion, and leave by the posterior root of the opposite side,
the straight course being broken by a slight indentation at the
middle line. It is not unlikely that there are in reality two
elements which decussate in the middle line, and are there
connected with cells which have remained unstained. The
fibre g has a similar relation to the posterior roots of Th. rv.
Traces of a system of elements similar to that just described
occur in Th. 111.
A pair of elements exists in each ganglion from Th. vi to
Th. x, which unfortunately seldom stains, but when stained
presents several points of interest. The appearance is generally
that shown in fig. 1, Th. rx, W., but in one preparation the
element stained on one side only of the ganglion, and fortu-
nately showed the position of the cell. This element, which
actually occurred in Th. vr is inserted in fig. 1 for the sake of
clearness in Th. x. (W.). Since in another preparation in
which the element was stained, a cell just commenced to take
up a similar position, I feel little doubt that this figure (fig. 1,
Th. x., W.) represents the true condition of the element. Its
course may then he described as follows :—Starting from a
cell, which lies in the anterior portion of the central mass of
ganglion cells, the fibre passes first backwards, decussates with
its fellow of the opposite side, and then turns outwards to the
neuropile, upon entering which it gives off a small tuft of fine
branches. After giving off this tuft, the fibre continues to run
transversely through the neuropile to about the level of the
lateral mass of ganglion cells, where it bends suddenly forwards,
giving off a short richly arborescent branch at the angle, and
then runs to the next ganglion in front, in the neuropile of
which it ends in another tuft of fine branches. No branch
has stained which passes out through either of the nerve-roots.
If the above description represents in reality the entire course
of the element, it must serve to put different parts of two
adjacent ganglia into communication with each other.
STUDIES ON THE NERVOUS SYSTEM OF CRUSTACEA, 43
II. ABDOMINAL GANGLIA.
Mertuop or Preparation.—Staining of elements in the
abdominal ganglia can be observed in two ways. In the case
of embryos at an early or medium stage of development which
have been prepared, as described in Part I, for the staining of
elements in the thorax and in which the abdomen is allowed
to remain undisturbed, fibres which have taken up the methylene
blue in the thorax often continue to absorb the colouring
matter in the abdomen, and the cells with which they are con-
nected are thus brought to light. The best results for the
abdominal ganglia can, however, be obtained by special pre-
paration of embryos which are very near the hatching point.
In such embryos the abdominal ganglia may be dissected out
from the surrounding tissue by careful manipulation with
needles. Special care must be taken not to injure or stretch
the ganglia and their continuity with the ganglia of the thorax
should be maintained. Ifthe embryos, thus prepared, be placed
with the dorsal surface uppermost in very dilute methylene
blue (1: 100,000 may be used to commence with, and the
strength gradually increased), satisfactory staining of many of
the elements of the abdomen will soon take place.
A variation of this method, which often gives good results,
consists in removing, or even simply tearing, with needles the
cuticle on the ventral surface of the abdomen (after having
first turned the abdomen backwards, and caused it to lie in a
line with the thorax), and allowing the embryo to lie in the
methylene blue with the ventral surface uppermost. In this
case, as soon as the staining is thought to be satisfactory, the
ganglia must be dissected out before they are examined. A
cover-glass may be placed upon them, and the elements rapidly
drawn, or the preparation may be fixed with ammonium picrate,
and mounted in glycerine diluted with an equal volume of
saturated solution of the picrate. Such preparations, however,
only retain their full colour for a comparatively short time.
For practical purposes a large number of fresh preparations,
examined with a cover-glass, have been found more instructive
44, EDGAR J. ALLEN.
and satisfactory. The same elements are seen again and again
with such frequency, that little doubt can remain as to the
accuracy of an observation.
The elements of the abdomen are similar in kind to those
described for the thorax, and will be considered under the
same general headings.
ELEMENTS OF WHICH BOTH CELL AND FIBRE LIE ENTIRELY IN
THE Gane ionic Cua1n.—In each of the abdominal ganglia,
from Abd. 11 to Abd. v1, a pair of elements similar to that
represented in fig. 6, B (2) has been found to exist. Each
element has its origin in a large cell situated in the anterior
lateral portion of the ganglion near its ventral surface. From
this cell a fibre arises which runs inwards and backwards, de-
cussating with its fellow of the opposite side at the middle line
and subsequently giving rise to two fibres, one of which runs
backwards, whilst the other runs forwards towards the brain.
The backward branch breaks up into fine fibres in the neuro-
pile of the ganglion in which the cell lies. The forward branch
runs as a longitudinal fibre along the ganglionic cord and in
all probability enters the brain. Unfortunately, owing to
difficulties of technique, it has never been possible to trace a
single longitudinal fibre of one of these elements through its
entire course, but a consideration of the whole evidence leaves
little doubt as to their entrance into the brain. One of the
most satisfactory direct observations was made upon an embryo
near the point of hatching, in which the abdominal ganglia
had been dissected out, as described above. In such prepara-
tions the pair of elements under consideration most frequently
stains in Ab. 11, and the fibres can generally be traced forwards
through Th. x and 1x. In one preparation, however, where
there was a wound at the level of Th. 111, the fibres of the
elements upon both sides were distinctly visible as far forwards
as this point.
But it is in embryos at an earlier stage, in which the
abdomen has been left uninjured, that these elements most
frequently appear. In order to make the evidence clear it is
STUDIES ON THE NERVOUS SYSTEM OF ORUSTACEA. 495
necessary to again draw attention to the method of preparation
adopted for such embryos, when it is desired to obtain staining
of the longitudinal fibres. The yolk is removed with needles,
and the embryo placed with the exposed thoracic ganglia upper-
most in the dilute methylene blue, the abdomen being allowed
to remain turned in underneath the thorax. The ganglionic
cord is then cut across, generally at tlie level of the cesophagus,
and the methylene blue enters the fibres at the wound. In
preparations so made the elements under consideration (B (4),
figs. 6 and 7) have frequently taken up the blue, and can be
well seen on turning the embryo over and lifting back the
abdomen. The longitudinal fibres of the posterior ganglia
can be traced through the anterior ones and show no sign of
terminating. They are clearly continued forward into the
thorax, although they cannot be individually followed on
account of the flexure of the abdomen. On turniug the
embryo back again so that the thoracic ganglia can be ex-
amined, all the longitudinal fibres coming from the abdomen
are seen to commence at the wound which has been made,
and when this wound lies at the level of the cesophagus it is
obvious that they must be fibres which enter the brain. No
endings of longitudinal fibres coming from the abdomen have
ever been observed in the thorax, and preparations of the
kind just described have been so frequently made that there
can be practically no doubt that the fibres in question enter
the brain.
These elements (B (0), figs. 6 and 7) are evidently the same
as those seen by Retzius in the abdominal ganglia of the adult
Astacus. (See ‘ Biol. Untersuch.’? Neue Folge I, pl. xi, fig. 1;
pl. ix, fig. 4.)
A pair of elements of this kind, as has been already stated,
occurs in Abd. vi, and is represented in fig. 7. In this ganglion
a second pair of elements, whose cells are situated in the
posterior portion of the ganglion, also frequently stains. The
fibres from these cells (fig. 7, B (c)) have been traced as far
forwards as the abdominal flexure, in preparations in which
the only wound in the thorax has been at the level of the
46 EDGAR J. ALLEN.
cesophagus, and they probably therefore enter the brain. The
fact that two pairs of elements in Abd. vi send fibres to the
brain, whilst only one pair has been found in the other abdominal
ganglia is in accordance with the known composite nature of
this ganglion.
PropasLe Moror Evements.—Of elements consisting of a
cell in a ganglion and a fibre passing out of the cord by one
of the nerve-roots, two principal kinds have been found in
the abdomen, as in the thorax, namely (1) those in which the
fibre passes out through one of the roots of the ganglion in
which the cell lies, the whole of the element within the central
nervous system being confined to one ganglion; and (2) those
in which the fibre passes out through a nerve-root of some
ganglion other than that in which the cell lies.
These elements are represented in figs. 2and 3; fig. 2 giving
the appearance presented when the elements upon both sides
of a ganglion are stained ; fig. 3, that when they are stained
upon one side only, and indicating therefore the course of the
individual elements. Each of the elements to be described
has been found in all the ganglia from Abd. 1 to Abd. v.
ELEMENTS CONFINED TO ONE GaNnGLION.—The element a,
shown in figs. 2 and 3, takes its origin in a cell, which lies
near the centre of the ganglion and at its ventral surface.
The fibre passes first upwards and outwards, turns inwards,
and after a short course divides into two branches, one of which
passes out of the ganglion by the posterior nerve-roots, whilst
the other runs across to the opposite side, keeping close to the
corresponding branch from the element of that side (fig. 2)
and subsequently breaks up on the opposite side of the gang-
lion. The fibre gives off numerous fine branches during its
course through the ganglion, which have not been represented
in the somewhat diagrammatic figs. 2 and 3, but which may
be seen in fig. 8, drawn from a preparation fixed with am-
monium picrate (fig. 8, a; Abd. 1 and 111).
Element 6 (figs. 2, 3 and 8) is similar in most respects to
STUDIES ON THE NERVOUS SYSTEM OF CRUSTACEA. 47
the element just described. It differs chiefly in the position
of the cell, which lies in the posterior lateral portion of the
ganglion. The fibre curves first upwards and forwards, and
then downwards, when it divides into two branches similar to
those of element a.
The element J (4), which has already been described as
occurring in the last ganglion of the thorax (fig. 1, Th. x1) is
found also in the ganglion of the abdomen (Abd. 1-v), and is
shown in figs. 2 and 3. The cell lies near the middle of the
posterior end of the ganglion. The fibre runs at first forwards
and outwards, and then turns inwards and backwards, giving
off a small branch to the neuropile at the angle. Subsequently
it divides into two branches, one of which passes out at the
posterior root, whilst the other runs over to the opposite side
of the ganglion, where it breaks up in the neuropile.
The element D (figs. 2 and 3) arises from a comparatively
small cell, which lies in the anterior half of the ganglion near
the middle line. The fibre runs backwards and slightly out-
wards, decussates with its fellow of the opposite side forming
the characteristic figure represented in fig. 2, gives off a small
arborescent branch to the neuropile, and then runs backwards
to the posterior root of the ganglion through which it passes.
ELEMENTS NOT CONFINED TO ONE GaNnGLion.—Two pairs of
elements have taken up the stain in each of the abdominal
ganglia, whose fibres pass out through one of the roots of the
ganglion immediately anterior to that in which the cell lies.
The element e (figs. 2 and 3) has the cell situated near the
middle line at the anterior end of the ganglion. From the
cell, which is small, the fibre runs outwards and backwards,
and then, turning sharply inwards, runs as a transverse fibre
close to its fellow of the opposite side. On reaching the other
side it turns forwards, giving off an arborescent branch at the
angle, runs into the next ganglion in front, and there passes
out at the posterior root. When the elements upon opposite
sides are stained it is generally impossible to distinguish the
two transverse fibres, which lie so close together in this and
48 EDGAR J. ALLEN.
in the other similar cases in the abdominal ganglia, that they
appear as one.
The element S (figs. 2 and 3) is similar to the element S
already described in the thorax (Th. x1, fig. 1). The cell,
which is very small, lies near the centre of the ganglion and
at the ventral surface. The fibre runs first upwards, back-
wards, and outwards, and then curves forwards and inwards.
After running for some distance in this direction it decussates
with its fellow of the opposite side, the two fibres lying for a
short distance close together. The fibre then turns forwards
and outwards, gives off an arborescent branch to the neuropile,
and then proceeds forwards in a straight line to the ganglion
in front, where it passes out by the middle root of the three
which spring from that ganglion.
From the foregoing description of the motor elements found
in the abdomen, it will be noticed that they, in nearly every
case, supply fibres to the posterior nerve-roots of the ganglia,
whilst the greater number of those described for the thorax
in this and the previous paper send their fibres to the anterior
nerve-roots, a few only supplying the posterior roots. The
probable reason for this difference is not difficult to find.
The anterior roots chiefly supply the appendages, which are
well developed in the thorax of the embryo, whilst in the
abdomen they are wanting. The posterior roots, on the other
hand, supply the muscles of the body itself, which are well
developed in the abdomen, but less so in the thorax owing to
the considerable space occupied by the still unabsorbed yolk.
III. ELements Arisinc FRoM CELLS OUTSIDE THE
CENTRAL Nervous SystEmM.
In Part I a number of elements were described, arising from
cells which lay in the ectoderm of the dorsal surface of the
abdomen. These cells give off fibres which enter the abdo-
minal ganglia and there bifurcate in the Y-shaped figure,
which is characteristic of sensory nerve-fibres in all divisions
of the animal kingdom, sending one branch forwards and the
STUDIES ON THE NERVOUS SYSTEM OF ORUSTACEA. 49
other backwards along the ganglionic cord. Numerous ele-
ments of this kind have stained in embryos near the point of
hatching, and it has been possible to trace the fibres inside
the central nervous system for considerably greater distances
than was done before.
In the thorax, elements having the characteristic Y-shaped
bifurcation enter the various ganglia by the posterior roots.
One such fibre is represented in fig. 1 at M, Th. x. From this
ganglion (Th. x) it has been traced forwards by direct ob-
servations as far as Th. 1, but there was no indication of a
definite ending there, the blue colour becoming gradually less
distinct and suggesting that the true termination had not been
reached.! From Th. x1 the element has also been traced as
far as Th. 1.
In the abdomen, elements of this kind enter by the middle
root of the three which belong to each ganglion. From Abd.
1 the forward branch has been actually traced as far as Th. 11,
but gave no indication of a definite ending there.
From the three ganglia already mentioned the forward fibre
has been directly observed to pass through at least nine or ten
ganglia. Cases in which the fibre could be traced from these
and neighbouring ganglia through five, six, and seven ganglia
occurred in a large proportion of the numerous preparations of
late embryos that were made. The fibres from spindle-shaped
cells lying in the telson can, in embryos of moderate age, be
seen to enter the last abdominal ganglion and then to run
forward through the anterior ganglia. Such fibres have been
traced as far as Abd. 1, but could not be followed further on
account of the flexure of the abdomen.
The fact mentioned in Part I, that elements of the kind
1 If an element stains at all frequently it is generally possible to form an
opinion as to whether the actual termination has been reached, or whether
the staining is incomplete. In the former case the terminal portion is, gene-
rally, at least as deeply stained as any other part of the element, and it may
continue to take on a stronger colour after there has been a considerable
fading of the other parts. In the case of incomplete staining, on the other
hand, the colour becomes more and more faint towards the end of the fibre
until it finally disappears.
VoL. 39, PART 1.—NEW SER. D
50 EDGAR J. ALLEN.
under consideration frequently stain in the abdomen, when
the longitudinal fibres commence to take up the methylene
blue at a wound made at the level of the esophagus, from
considerations of a similar nature to those already adduced in
the case of the abdominal elements B (4), leaves little doubt
that the fibres actually go to the brain. The observations
now recorded, although they do not directly demonstrate the
point, render it still more probable.
With regard to the second branch given off by these ele-
ments after entering a ganglion, which is directed backwards
along the ganglionic cord, no definite termination has been
found. It has never been seen to pass through more than
three ganglia, and can generally only be followed through two.
This, however, is probably due to incomplete staining, and the
entire course of this branch remains yet to be determined.
EXPLANATION OF PLATE 4.
Illustrating Mr. Edgar J. Allen’s “ Studies on the Nervous
System of Crustacea.”
Fic. 1.—Brain and thoracic ganglia of Homarus embryo. es, (so-
phagus. ¢r.dr. Transverse bridge behind cesophagus. com. Csophageal
commissure. azt.11. Ganglion of Antennaul. Th.1—1x. Thoracic gan-
glia. For individual elements see text. Somewhat diagrammatic.
Fic. 2.—Three abdominal ganglia of Homarus embryo. Motor elements
inserted upon both sides. For individual elements see text. Somewhat
diagrammatic.
Fic. 3.—Ditto. Motor elements inserted upon one side only.
Fics. 4 and 5.—System of nerve elements in Th. m1 and Th. Iv of Ho-
marus embryo.
Fic. 6.—Abdominal ganglion of Homarus embryo. B (4). Element
sending fibre to brain.
Fie. 7.—Sixth abdominal ganglion of Homarus embryo. B (4) and
B (c). Two elements sending fibres to brain.
Fic. 8.—Second and third abdominal ganglia of Homarus embryo.
Camera drawing from preparation preserved in ammonium picrate. a. 0.
motor elements.
NOTES ON OLIGOCHATES. 51
Notes on Oligochetes, with the Description of
a New Species.
By
Edwin 8. Goodrich, B.A.,
Assistant to the Linacre Professor of Comparative Anatomy, Oxfoad.
With Plates 5 and 6.
Tue following observations were chiefly made on some
worms sent to me by Mr. Damon from a garden at Weymouth.
I have to thank Professor Ray Lankester for much help during
my researches.
On the Structure of Enchytreus hortensis, n. sp.
At first sight this little Enchytreid presents no very dis-
tinctive characters, It is, when full grown, about 15 mm. in
length, and milky white in colour, the anterior end being
sometimes yellowish. The chetz (fig. 18) have a straight
shaft and a hooked inner end, but this hook is not always so
pronounced or so sharp as in the cheta figured. The bundles
contain from three to four chet, generally three in the
dorsal and four in the ventral bundle. There are no dorsal
pores. A small dorsal head-pore is situated between the pro-
stomium and first segment. The brain (fig. 1, 67.) is larger
behind than in front ; its posterior margin is slightly indented
or rounded. The dorsal blood-vessel arises from the wall of
the intestine about the 17th segment. Three pairs of septal
glands (fig. 1, sept. g/.) are present; two sets of bulky glands
in front, and a third pair of smaller glands behind, each of
which is subdivided into two lobes. The well-developed sali-
52 EDWIN S. GOODRICH.
vary glands (fig. 1, sa/. gl.) are unbranched and of considerable
length. The nephridium (fig. 2) is compact, with a portion of
the canal slightly coiled in the pre-septal region. The funnel
of the sperm-duct is of moderate length and very thick-walled.
The spermatheca has a distinct duct beset with glandular cells
near its external aperture, and possesses a posterior blind
sac (fig. 23, post. sac). Three chief varieties of corpuscles are
found in the celom ; one of these bears a peculiar product in
the shape of a coiled thread, to be described in detail further
on. In a favourable light the cuticle can be seen in most
specimens to be covered throughout with fine hair-like pro
cesses (fig. 8,i.p.), similar to those described in Vermiculus(8).
This species is probably distributed all over England, as I
have found it near Oxford, London, and Weymouth.
On the Nephridium of Enchytreus hortensis.
I have thought it worth while to give in detail the results
of observations carried on for some years on the nephridia,
since not only does a great deal remain to be described in
these interesting organs, notwithstanding the large number of
authors who have written on the subject, but also on account
of the recent publication by M. Bolsius of descriptions totally
at variance with the observations of other naturalists and my
own (4).
As M. Bolsius gives an account of the literature, I need
only say that former writers (Claparéde, Eisen, Vejdovsky,
Michaelsen, Ude, &c.) all agree in describing the nephridium
of Enchytreeids as consisting of a more or less compact mass
of cells pierced by a continuous ciliated convoluted canal
leading from the nephrostome to the exterior. Bolsius, on
the contrary, declares that the lumen of the canal forms a
complex network “ des canalicules anastomosés dans toutes les
directions ;” and, moreover, that it is totally devoid of cilia:
‘Les canalicules et le canal collecteur ont été dessinés par
nous sans cils vibratiles : c’est que nous n’en avons pas trouvés,
ou plutdt c’est que nous en avons constaté l’absence compléte
dans nos nombreuses préparations.” Since the first of these
NOTES ON OLIGOCHATES. ba
statements has been included in the excellent monograph Mr.
Beddard has just published (2), it seems all the more important
that it should be criticised before it becomes generally
accepted.
Let us now examine the structure of the nephridium in En-
chytrzus hortensis as seen in the living animal. The organ
(fig. 2) consists of an oval main body, flattened from side to
side, and narrowing in front to a neck, which piercing the
septum ends in an open funnel (fig. 2, neph. st.). Behind the
nephridium passes into a lobe leading to the body-wall. The
protoplasm of the nephridial cells, whose boundaries I have
been unable to distinguish, is very granular, especially in the
anterior region, which is frequently of a brownish tinge. The
funnel (figs. 8—5) has a protruding upper lip (up. J.) and
a truncated lower lip (/./.). At the extremity and along the
edges of the upper lip are set numerous very long extremely
fine waving cilia (ezt. cil., figs. 4 and 5), while from the inner
surface springs a large bunch of powerful undulating cilia (fl.).
As far as I have been able to determine, cilia are not present
on the lower lip; but rising from its middle region is a peculiar
sharp protoplasmic process (m. p., figs. 4 and 5), sometimes
branched and ragged.
Already in Nereis (7) and Vermiculus (8) I have drawn
attention to the two kinds of cilia attached to the lip of the
nephrostome, which indeed have not escaped the notice of
previous authors, but the exact disposition and differences
of which have not heen sufficiently insisted upon. Whilst the
inner beat rapidly, simultaneously, and rhythmically, forming
an undulating bunch! or “ flame,” which, no doubt, propels
liquid down the lumen of the canal; the long external cilia,
on the other hand, do not beat in unison: each cilium seems
to move independently with trembling motion, sweeping down
towards the protruding process of the lower lip. These cilia
seem rather to guard the mouth of the funnel and select
1 The seat of the motion seems to be at the base of the cilia, the wave
passing along them as an undulation passes along a cord shaken at one ex-
tremity.
54 EDWIN S. GOODRICH.
particles from the celomic fluid than to propel it down the
canal,
Passing backwards from the nephrostome is the nephridial
canal (neph. c., figs. 2, 3, and 5). I may say at once that,
although I have not been able to follow it throughout its whole
extent with certainty, owing to the opaque and granular nature
of the post-septal region and the small dimensions of the
tortuous lumen, I believe the main duct to be continuous and
single (i.e. without anastomosing branches), except in that
region which leads directly to the exterior, where the lumen
occasionally undergoes secondary subdivision (figs. 2 and 7).
Although, therefore, I cannot claim to be able to prove the
unbranched character of the canal, yet amongst the number of
Enchytreid nephridia I have had under observation during
the last three or four years I have never come across a single
undoubted case of such anastomosing. The course of the canal
drawn in fig. 2 does not profess to be accurately represented in
all its details, nevertheless I have been able to observe that at
least the main features of the canal’s course, if not the less
important of its convolutions, are constant, not only in the
different nephridia of the same worm, but also in the nephridia
of different individuals of the same species. For the first part
of its course it has a very narrow lumen, and is not ciliated.
It then widens slightly, and undergoes several complicated
twists in the more posterior and dorsal region of the main body
of the nephridium. It is along this second region of the canal
that here and there we find a slight oval enlargement of the
lumen, in which is situated a bunch or “ flame” of cilia (ce. a.,
fig. 2), similar to the “flame” in the nephrostome, and con-
stantly driving the liquid in one direction, namely, to the
exterior. As far as I can make out, there are no more than
1 Anyone examining the living worm could hardly miss seeing these cilia.
It is, of course, just possible that the worms observed by M. Bolsius were of
exceptional structure, but I am more inclined to think that he was misled,
both with regard to the absence of cilia and the anastomosing of the canal by
employing the method of sections alone.
NOTES ON OLIGOCHMTES. 55
five such ciliated regions in each nephridium.' Their essential
similarity to the ‘ciliated ampullz ” found in the nephridium
of Vermiculus (8) scarcely needs pointing out. From the
second region the canal passes to a third, where the lumen is
often wider still, and is entirely devoid of cilia. The course of
this, the terminal region of the canal, lies along the postero-
dorsal edge of the main body, and from thence passes through
the posterior limb to the external pore. It will be noticed that
the blind sacculations of the canal (mentioned by Michaelsen,
9a) are here very large (fig. 2, d/. b7.), and frequently a sub-
division of the lumen takes place, forming a more or less com-
plicated network (figs. 2 and 7), which must not, however, be
confused with the supposed anastomosing branches between
the different regions of the canal. Finally the lumen
expands to form a vesicle (e. v.), which opens to the exterior
by a small transversely elongated pore in front of the ventral
chete.
Sections only confirm these observations. No limits are
distinguishable between the cells. The funnel has two nuclei
in its upper lip near the point of insertion of the cilia (fig. 3),
and one at the base. Small deeply staining nuclei are seen
in the main body (n., fig. 3), and larger nuclei are distributed
apparently in more intimate connection with the second and
third region of the nephridial canal (n., fig. 6).
On the Celomic Corpuscles of Enchytrzus
hortensis.
Most characteristic of this worm are the celomic corpuscles ;
these are of three kinds. Floating in the ccelomic fluid, or
creeping on the surfaces exposed to it, may be seen a small
number of ameeboid corpuscles of very variable character.
Some (fig. 28, a.) are more or less spherical with short- pointed
processes ; their protoplasm is pretty clear, and contains, as a
rule, a few highly refringent round granules, apparently of a
1 M. Bolsius (4) justly criticises previous authors for describing the canal
as ciliated, and yet not figuring any cilia. M. Roule (10) alone does so, and
represents them erroneously along the entire length of the canal.
56 EDWIN 8S. GOODRICH.
fatty nature. Others are formed of more opaque and finely
granular protoplasm, prolonged into fine branching processes
of great length (fig. 28,.): corpuscles intermediate in character
between these two forms also occur.
The second variety is very characteristic of the Enchy-
treide in general, and has been described by almost every
writer on this group of the Oligocheta. Much larger than
the amceboid form (nearly twice the size), these corpuscles
are of regular and constant oval shape, somewhat flattened,
thicker in the middle than at the periphery, and frequently
bluntly pointed at one or both ends. The body of the cell is
made up of large spherical, clear, and refringent granules,
embedded in a delicate meshwork of protoplasm (fig. 15, a).
The nucleus is round or slightly oval, in which case its long
axis is invariably situated at right angles to that of the cor-
puscle ; staining reveals a dark nucleolus lying in the less
deeply staining nucleus, n (fig. 12, a). These corpuscles are
found either freely floating in the ccelom, or attached to the
body-wall or septa by means of a short stalk, apparently of
cuticular nature. In the latter case the corpuscle is always
fixed by its more pointed extremity. The remains of the stalk
of attachment may generally be detected on those corpuscles
which float freely (st., fig. 15, a). Although my observations
as to the origin of these corpuscles are unfortunately neither
complete nor conclusive, yet I am strongly inclined to believe
that they arise from the coelomic epithelium lining the body-
wall and septa. Certain it is that smaller, flatter, and younger-
looking corpuscles of similar structure may sometimes be seen
attached to the epithelium. The fact that similar apparently
young corpuscles may occasionally be seen floating freely in
the fluid does not seem to me fatal to the view that normally
they remain attached to their place of origin until they have
reached their complete development. Never have I found
intermediate stages between these oval corpuscles and the
ameeboid variety, or the granular cells of the modified epithe-
lium covering the gut, from which they differ markedly, as
already noticed by Vejdovsky (13). In the youngest worms I
NOTES ON OLIGOCHATES. ay)
have been able to examine, the oval corpuscles are relatively
about as abundant as in the adult; but the free ones appear
to be more rounded and flattened, and with fewer granules,
thus resembling what I have taken for the younger stages in
the adult. As no definite signs of extensive disintegration of
the corpuscles are apparent, and as I have never observed any
case of their multiplication by division, it would seem that the
life of each individual corpuscle must be of considerable
length, and that they are but rarely renewed.
The third and most interesting kind of ceelomic corpuscle
should probably be reckoned as a variety of that just described.
The body of the cell, though smaller, is formed of a similar
meshwork, enclosing granules of identical appearance and pro-
perties ; the nucleus also resembles that of the ordinary oval
corpuscle (n., fig. 12), but lying on one side is a colourless
refringent body of peculiar structure. This body, in the
fully developed form, is of the shape of a thick disc or of a
truncated cone, thicker at the edge than in the centre (thr.,
figs. 12—15), the flat surface being next to the nucleus.
When the celomic corpuscles are squeezed out of the worm
under a cover-glass, and they come into contact with some
strange fluid, such as distilled water, or even salt solution, it
becomes apparent that the refringent body is formed of a
long thread of transparent homogeneous substance,
closely coiled like a rope (fig. 13). Surrounded by a thin layer
of protoplasm the thread-coil is obviously an endoplastic pro-
duct of the ccelomic corpuscle. At first I took these strange
bodies for parasites, but subsequent observations have proved
this view to be false.
As already mentioned, immediately the corpuscle comes
into contact with some foreign liquid, such as distilled water,
weak acids, or alkalies, the body of the cell begins to dis-
integrate, the coil of thread swells and gradually unwinds,
forming a tangle of loops, amongst which I have never de-
tected free ends (fig. 13). The thread appears to be thicker
in some places than in others, or perhaps it would be more
correct to say that the coils stick together in certain regions,
58 EDWIN S. GOODRICH.
Like the ordinary oval cells, the thread-corpuscles are found
both floating and attached by a stalk to the body-wall or septa.
In the latter case the thread-coil is invariably situated on
that side of the nucleus which is far from the point of attach-
ment. Looking in the adult worm for possible stages in the
development of thread-corpuscles, smaller cells can be found
in which the coil is smaller relatively to the granular cell
body (figs. 10, 11,d, e, f, and 12,6). Such small and appa-
rently young cells are scarce, and generally attached. Turning
now to very young worms, we find in them the thread-
corpuscles scarcely less abundant than in the adult relatively
to the whole number of ccelomic corpuscles. They are, how-
ever, as a rule, flatter and more oval in shape (fig. 9), with
more loosely packed granules and smaller thread-coils. In-
deed, these refringent bodies differ from those of the adult,
not so much in the circumference of the disc as in its thickness
(cp. figs. 9, d, and 138, a, both side views). Similar shallow
discs occasionally occur in the adult (fig. 10, a). It will be
seen, moreover, that in what I have taken to be the younger
stage of development the disc presents its edge to the nucleus
(figs. 9 and 10) ; whilst in the later stages the nucleus is lodged
on its flat surface (figs. 12, c, and 13). Whether the disc has
during growth turned round or merely changed shape is a
doubtful point. These facts lead to the conclusion that the
thread-corpuscles originate from the coelomic epithelium lining
the body-wall or septa, passing through stages essentially
similar to those of the young corpuscles of the ordinary oval
type, and are not formed by the development of a thread
within full-grown oval corpuscles. The thread itself would
appear to be formed at the expense of the granules, cp. (figs.
9,10, and 13); but this appearance may be deceptive, and
merely due to the increase in size of the thread alone. The
chemical nature of the substance composing the thread is
discussed in detail below; it differs widely in its chemical
properties from that of which the granules are formed, and
from chitin.
On the function or fate of the thread in the ccelomic cor-
NOTES ON OLIGOCHATES. 59
puscle of this Enchytreid I am sorry to say that my observa-
tions throw no light. It occurs in immature and mature
worms at all seasons of the year. It has been searched for in
vain, either discharged into the coelomic fluid or amongst the
tissues of the body. Nor have empty cells been found which
could have discharged the thread. It might be suggested that
this thread is an excretory product ; however, as far as I have
been able to make out, it is never got rid of,—indeed, it would
seem to be most ill adapted for passing through the compli-
cated coils of the nephridial canal. Possibly it may be a waste
product of harmless nature, which is allowed to remain and
accumulate throughout life, but here also facts are wanting to
enable us to come to any definite conclusion.
Amongst the other Enchytreeids or the Oligochetes in gene-
ral no products in the celomic corpuscles which could be
compared to these threads seem to have been hitherto observed.
Amongst the Polychetes, on the other hand, there have been
described in Ophelia radiata ameeboid celomic corpuscles
containing a chitinous rod! of unknown function, first noticed
by Claparéde (5), and recently studied in great detail by Dr.
Shaepi (11). The latter adopts the view urged by Dr. Hisig (6),
that these rods of chitin are of the nature of an excretory
product.?
In the epidermis of many fish and amphibia cells have been
described which secrete a thread-like substance.
Perhaps the thread-cells of Myxine, described by Johannes
Miller, Blomfield (8), and others, resemble more closely the
thread-corpuscles of Enchytrzus hortensis than any other
cell hitherto known, especially in the earlier stages of their
development. But here, again, the product appears to be of
different chemical nature, and the cells in their origin and
situation are so different that they scarcely help us to solve
the problem.
1 T have found similar but longer chitinous rods in Ophelia limacina.
2 As the thread in Enchytreus hortensis is certainly not of chitin, a
comparison with Ophelia (even supposing the excretory nature of the rod had
been proved) would not warrant our coming to a similar conclusion,
60 EDWIN S. GOODRICH.
On the Spermatheca of Enchytreids.
In 1885 (9) Dr. Michaelsen first made known the remarkable
discovery that the spermatheca in the Enchytreidz, which lies
in the 5th segment, communicates directly with the alimen-
tary canal. More recently Mr. Beddard (1) has shown that
the same thing occurs in Sutroa, and he discusses the subject
in the monograph already mentioned (2, pp. 127-8). As such
a case of the direct communication of the cavity of the ceso-
phagus with the exterior by means of a sac invaginated from
or derived from the epidermis‘is unknown in any other group
of Invertebrates, and is of great morphological interest, I give
a series of longitudinal and transverse sections through these
organs. Fig. 27 represents a longitudinal horizontal section
through the 5th segment of Marionia enchytreoides
St. Loup.!
As the section is slightly oblique it shows on one side the
opening of the spermatheca to the exterior (sp. op.), and on the
other the opening into the alimentary canal (éné. op.). In
Pachydrilus, and forms closely allied to it, such as Marionia,
the spermatheca is simple, and opens into the cesophagus at
its hinder extremity. The spermatheca figured contains no
spermatozoa, showing that the internal opening is formed
before the spermatozoa are received, and the point at which
the fusion has taken place between the two organs can be
detected, since the lining of the csophagus is ciliated (cs.),
while that of the spermatheca is not (sp.).
In Enchytreus hortensis the spermatheca fuses with
and opens into the esophagus at a point more nearly opposite
its external opening, sending backwards a large blind sac in
which are lodged the bulk of the spermatozoa.
It will be seen that the communication between the sperma-
theca and the alimentary canal is very wide (figs. 21 and 25,
and the diagram, fig. 23); the external spermathecal pore
remains of normal size (sp. op.). The spermatozoa (spiz., figs.
1 Professor Marion kindly sent me a number of these worms from Mar-
seilles.
NOTES ON OLIGOCHATES. 61
20—25) arrange themselves round the walls of the sperma-
theca, with their long heads deeply sunk in or between the
lining cells, and their tails hanging into the cavity. Many
bundles of spermatozoa also lie loose in the cavity, and may
get carried into the alimentary canal (Michaelsen, 9).
On the Celomic Corpuscles of Pachydrilus, sp. (?).
It is not without hesitation that I put on record the following
very incomplete observations on a small species of Pachydrilus,
of which I unfortunately only found two specimens amongst
the Enchytreids sent to me by Mr. Damon from Weymouth.
On examining the worm under the microscope, the coelom
was seen to contain a number of remarkably long and slender
cells, attached to the body wall by one extremity (fig. 16). The
longest of these cells reached nearly across a segment, being
abont ten times as long as an ordinary celomic corpuscle
(cp., a., and c., fig. 17). Near the middle was an oval clear
region, indicating the position of the nucleus (n., fig. 17), whilst
the body of the cell is formed of granular protoplasm, in which
the granules could be observed circulating. No thick or
definite wall was apparent covering the somewhat irregular
surface, and the whole cell appeared to be very flexible, being
twisted about backwards and forwards with the motion of the
coelomic fluid. An expanded sucker-like foot or base ( ft.)
served to fix it to the body-wall. In stained sections the
nucleus appears small and rounded as in the ordinary oval
corpuscles.
Although it may seem extremely probable that these strangely
shaped cells were parasitic Protozoa, yet the fact that there
were present cells representing in appearance every stage
between the elongated form and the ordinary ceelomic corpuscle
(fig. 16) suggests that the former are derived from the latter.
62 EDWIN S8. GOODRICH.
Notes on the Action of Chemical Reagents on the
Celomic Corpuscles of Enchytreus hortensis
and Vermiculus pilosus.}
In a former paper I described the characteristic coelomic
corpuscles of Vermiculus pilosus (8, fig. 28, pl. 28); they
are spherical cells, filled with large oval refringent granules.
If these be watched under the high power of the microscope
in distilled water, or even in normal salt solution, the large
white granules will be seen to dissolve, whilst minute and
yellower-looking granules come into view.? Under the in-
fluence of weak alcohol, a solution of iodine, very weak acids,
or alkalies the same thing occurs. With osmic acid (1 per
cent.) the dissolution of the large granules takes place with
almost explosive rapidity, the small granules in this, as in the
other cases, remaining sharply defined. Whether the smaller
granules are hidden by the larger, and only become visible
when these have been dissolved, or whether they remain as an
actual residue from the large granules, is a point difficult to
settle. I feel convinced from appearances observed that the
large granules really include the smaller.
It is difficult, if not impossible, to permanently fix the large
granules. Sometimes they could be fixed for a time by a
sudden drenching with absolute alcohol, and they would then
be unaffected by ether.
The small highly refringent granules, which remain when
the large have disappeared, are insoluble in water, iodine solu-
tion, alcohol, and ether. They are soluble in a solution of
caustic potash, the mineral acids, in oxalic acid, and with
difficulty in acetic acid.
1 T am indebted to Professor Gotch for many useful hints while conducting
these experiments.
2 The difficulty of conducting chemical experiments on small bodies which
have to be watched under the high power is very great, and often renders the
result uncertain. I have, therefore, always repeated the experiments many
times over, and used strong solutions of the reagents (unless otherwise
stated).
NOTES ON OLIGOCHATES. 63
Turning now to the oval corpuscles of Enchytreus hor-
tensis, we find here also white granules soluble in distilled
water, in potash, lime water, acetic and tannic acid, with diffi-
culty (perhaps not entirely) in oxalic acid, and also with diffi-
culty in weak mineral acids, but easily in strong. When
treated with osmic acid (2 per cent.) these granules are gene-
rally reduced to smaller and more refringent bodies. This
also takes place under the influence of alcohol, strongly recall-
ing the much more pronounced action in the large granules in
Vermiculus. The white granules are insoluble in ether, and
weak mineral acids are unaffected by silver nitrate and by
iodine. They do not answer to the xanthoproteic test, and
are not doubly refractile.
It would seem possible that we have in these granules,
more firmly combined, the same two substances which are so
easily separable in the corpuscles of Vermiculus.
The Thread.—The coiled thread in the celomic corpuscles
of Enchytreus hortensis is insoluble in distilled water (hot
or cold) and salt solution. Iodine solution shrivels it slightly,
and stains it yellowish brown. In strong alcohol the thread
contracts to an irregular mass, which is not farther acted upon
by ether. In strong potash (30 per cent. sol. of KOH) the
thread is not dissolved in the cold; but as soon as the cell
has been destroyed it uncoils, forming very characteristic
U-shaped loops, the two limbs of the U being closely applied
(fig. 14). When heated in strong potash portions of the
thread are dissolved or broken up into minute globules, a
highly refringent thick U-shaped loop remaining. On boiling
this portion appears to be dissolved.! A solution of silver
nitrate, oxalic and picric acids produces no effect. Glacial
acetic acid, on the contrary, has a most marked influence.
Under its action the thread swells up and becomes transparent,
then melts down to an irregular coagulated mass. On further
action this refringent mass swells up again, and begins vio-
1 It is almost impossible to make quite certain as to whether the thread is
really dissolved in boiling (in this and other cases), as the operation cannot
be conducted under the microscope.
64 EDWIN S. GOODRICH.
lently to “boil up,” and hecomes dissolved, leaving only one
(or two) minute yellowish and very refringent coagulated
granules, which undergo no further change. Strong HCl and
H,SO, have at first somewhat the same action as acetic acid,
swelling up the thread to a transparent viscid mass, which,
however, is not dissolved in the cold. Strong HNO, acts in
the same way, the swollen thread being further reduced to a
small granular mass, which, on heating, is broken up into
globules, and finally dissolved on boiling (?). Osmic acid
shrivels the thread a little, and stains it a pale greyish tint.
With tannic acid it becomes reduced to an irregular shrivelled
mass. The thread is insoluble in lime water,—in fact, the
addition of lime water to the celomic fluid is one of the
best methods .of showing up the threads in a mass of cells!
(figs. 15, 16).
The xanthoproteic test gives no result as with the white
granules. The thread does not stain readily with eosin; but,
on the other hand, it stains readily with methyl blue, and
especially with methyl green. In an acidified solution of
Victoria blue it stains dark purple, and again dark blue in an
alcoholic solution of cyanin.
The chief results are shown below in tabular form. The
sign + means that the substance is soluble in the reagent,
the sign — that it is not obviously soluble without heating.
3 : | | |
ducts in the celomic] :| S |x $4 sHESM Whe Re I Lol ie 5
me meaertclee $3 3 a a3 z8 ac ea sig = = zd 3
Ael| 4 ROB Re asel/Os|<4s/m m = Ssis
Of Vermiculus— | |
1. Large white | |
granules . J + |+—/—| + |... | oe | + | + [+] + 4! + [+
2. Small residual | |
granules . .| — | — |—| + | «| «. | + | + J+) + J+] = J=
Of Enchytreus— | | |
3. White granules) + | — /—| + | + | +/+] + /+! + [4/4—J—
par- |
tially,
4. Thread. . J — | — |J—| —| - | —/] —] + ‘| a i a
1 The thread does not give rise to double refraction.
NOTES ON OLIGOCHATES. 65
From this table it will be seen that we have at least four
distinct endoplastic products and perhaps three distinct
chemical substances in the ccelomic corpuscles of Vermiculus
and Enchytrzeus, which are probably of an albuminoid nature.!
That none of the products are mucin is proved by their
solubility in acetic acid ; that none of them are chitin is proved
by the first three being soluble in caustic potash and the
fourth in acetic acid. That they are neither of a fatty nor of
an amyloid nature is shown by the action of ether, osmic acid,
and iodine (which does not stain them dark).
Notes on the Cuticle and Chete of Oligochetes.
It has long been known that the cuticle of the Oligocheta is
not made of chitin (Timm, 12; LHisig, 6), since it is readily
soluble in caustic potash; but few, if any, observations seem
to have been made with a view to acquiring a more definite
knowledge of the character of the substance of which it is
formed. Below are given the results of some experiments
performed on the cuticle of the common earthworm, Lum-
bricus herculeus. The worms were killed in 30 per cent.
alcohol, and placed for a time in water until the cuticle could
be easily peeled off. The cuticle, having been thoroughly
washed in salt solution and distilled water, was then experi-
mented upon in test-tubes. It is insoluble in alcohol and
ether; soluble in hot distilled water (before boiling-point is
reached), in solutions of ammonia or weak potash, in lime
water, in weak mineral acids, in oxalic acid, and with some
difficulty in acetic acid.” Cuticle dissolved in distilled water,
boiled and filtered, gives a slightly viscid solution of whitish
opalescent colour, which does not thicken on cooling. On
1 In a recent paper on “The Process of Secretion in the Skin of the
Common Eel,” Professor E. Waymouth Reid comes to the conclusion that
the thread-like secrections from the epidermis are formed of an albuminoid
possibly allied to keratin (‘ Phil. Trans.,’ v., 185, B. 1894). The substance
presents no great resemblance to that found in Enchytreus.
2 The same results were obtained with the cuticle of Vermiculus and
Enchytreeus.
VOL. 39, PART 1,—NEW SER. E
66 EDWIN S. GOODRICH.
evaporation the solution leaves a film of transparent substance
without crystallisation. Boiled with caustic potash and cupric
sulphate we get no reduction. ‘The solution answers to a
certain exteut to the proteid tests, turning straw-coloured
with the xanthoproteic test, and pale lilac with the biuret
test. Boiling with Millon’s reagent produces no more than a
pinkish hue. Tannic acid in the solution acidified with acetic
acid gives a white precipitate.’ Sodium sulphate and po-
tassium ferrocyanide give no precipitate. If we treat the
solution with absolute alcohol we get a white precipitate, and
the filtered liquid answers no more to the above tests.
The Chete.—It is generally assumed that the cheetz are
chitinous, but this seems to be by no means always the case. The
cheetee of Vermiculus and Enchytreeus are insoluble in distilled
water (hot or cold), alcohol, ether, caustic potash (even boiling),
acetic and oxalic acids, hydrochloric acid (even after prolonged
boiling for some minutes). In strong sulphuric acid they
become swollen and soft, finally dissolving on heating.
The cheetz of the earthworm are insoluble in distilled water,
alcohol, ether, caustic potash, lime water, acetic and oxalic
acids; soluble in strong hot sulphuric acid. Boiled in strong
caustic potash the chete become somewhat swollen and broken
up, but do not dissolve. If they be now washed and placed in
concentrated hydrochloric acid the proximal region (that part
which lies in the body-wall) is rapidly dissolved, partially in
the largest cheetz, and almost entirely in the smaller. If the
acid be now boiled the resisting distal region is soon dissolved,
leaving only a thin colourless layer which formed a sort of
cap over the cheta. This shell appears to be insoluble in
hydrochloric acid like the entire cheta in Vermiculus. The
chetz of Vermiculus, Enchytreeus, and Lumbricus become
orange with xanthoproteic test, and crimson with Millon’s
test.
Below are set forth in tabular form the chief of these
results :
1 It behaves, in fact, somewhat like a peptone.
NOTES ON OLIGOCHMTES. 67
wa ee ee eee
3S mld 2 : ae <
Ssh1[4/ 8 |/eele8/35/2] = | £8. | 2.
Sas | 8 | 2 1es|8e (se) 5 don | Se
AFS | 4)/ RA /OS/AeEl/das\|H) Aas As
1. Cuticle of Lum-
IDM WI Q All Ge —;/—}] +] +] + [+l + Pale Pale
2. Cheete of Vermi-; yellow. | pink.
culus and En-'
chytreus). =. .| = —-!—-/]-]-—-]-— —-| — | Orange.| Red.
3. Chete of Lum-
WINGER > o « of = ~!|—|—|{—]-— j\-| +. | Orange. | Red.
with
| insol.
resi-
due.
From this table it would again appear that we have three
albuminoid substances distinct from each other, and from those
found in the coelomic corpuscles. The chete of the small
worms (Vermiculus and Enchytreus) are formed of a substance
which is not chitin, since it is insolublein HCl. The chetz
of Lumbricus, on the contrary, are probably chiefly composed
of chitin, or some substance closely allied to it, since they are
insoluble in caustic potash, and partially soluble in HCl. On
the other hand, so far as the solubilities show, the cuticle
appears to be formed of a substance closely allied neither to
chitin nor to mucin.
List oF WORKS REFERRED TO.
1. Bepparp, F. E.—‘‘ A Contribution to the Anatomy of Sutroa,” ‘Trans.
Roy. Soc. Edinb.,’ vol. xxxvii, 1892.
2. Bepparp, F. E.—‘ Monograph of the Order Oligocheta,’ Oxford, 1895.
3. Biomrretp, J. E.—‘The Thread-cells and Epidermis of Myxine,”
‘Quart. Journ. Mier. Sci.,’ vol. xxii, 1882.
4. Bousius, H.—* L’Organe segmentaire d’un Enchytreide,” ‘Mem. d.
Acad. Pont. d. Nuovei Lincei,’ vol. ix, 1893.
5. CuarparkpE, E.—‘ Annélides, Chétopodes du Golfe de Naples,’ Geneva,
1868.
6. E1stc, H.—‘ Die Capitelliden des Golfes von Neapel,’ Monog., Berlin,
1887,
68 EDWIN 8S. GOODRICH.
7. Goopricu, E. §8.—‘On a New Organ in the Lycoridea, &c.,” ‘ Quart.
Journ. Micros. Sci.,’ vol. xxxiv, 1893.
8. Goopricn, EH. 8.—“On the Structure of Vermiculus pilosus,”
‘Quart. Journ. Micros. Sci.,’ vol. xxxvii, 1895.
9. Micuartsen, W.— Untersuchung iiber Enchytreus MObii,” ‘In.
Diss.,’ Kiel, 1886; and ‘ Vorlaufige Mittheilung,” ‘ Zool. Anz.,’ 1885.
9a. MicuaEtsen, W.—“ Enchytreeiden-Studien,” ‘Arch. f. Mikros. Anat.,
Bd. xxx, 1888.
10. Route, L.—“ Etudes sur le développement des Annélides,” ‘Ann. Sci.
Nat.’ (7), vol. vii, 1889.
11. SHarprri, Th.—‘‘ Das Chloragogen von Ophelia radiata,” ‘Jen. Zeit.
f. Naturwiss.,’ Bd. xxviii.
12. Timm, R.—‘* Beobachtungen an Phreoryctes, &.,” ‘Arb. Zool.-Zoot.,’
Wurzburg, vol. vi.
13. Vespovsky, F.—‘“‘Beitrige z. vergl. Morph. d. Anneliden,” ‘Monog.
d. Enchytreiden,’ Prag, 1879.
EXPLANATION OF PLATES 5 and 6,
Illustrating Mr. Edwin S. Goodrich’s “ Notes on Oligochetes,
with the Description of a New Species.”
List oF REFERENCE LETTERS.
él. bv. Blind branch of the nephridial canal. 7. Brain. 3.0. Body-wall.
e.a@. Ciliated enlargement of nephridial canal. circ. m. Circular muscles.
cel. Celom. cel. corp. Celomic corpuscle. cal. epith. Colomic epithelium.
d. bl. v. Dorsal blood-vessel. epid. Epidermis. e.v. End vesicle. ect. cil.
External cilia. ~ Point at which the spermatheca has fused with the ceso-
phagus. jl. Flame-like bunch of cilia. /. Expanded foot whereby the cell is
fixed to the body-wall. g/. Spermathecal gland. 4. p. Hair-like processes.
7. e. Internal hooked end. iz¢. op. Internal opening of the spermatheca into
the cesophagus. 7,7. Lower lip. Jong. m. Longitudinal muscles. m. p.
Median process. 2. Nucleus. z.c. Ventral nerve-cord. xeph.c. Nephridial
canal. xeph.p. Nephridiopore. zeph.st. Nephridiostome. @s. (sophagus.
op. sp.d. Opening of spermathecal duct into the main chamber of the sperma-
theca. post. sac. Posterior sac of the spermatheca. s. Septum. — sal. gl.
Salivary gland. sept. g/. Septal gland. sp. Spermatheca. sp. d. Sperma-
thecal duct. sp.op. Spermathecal opening. spéz. Spermatozoa. st. Stalk of
attachment. hr. Thread in the celomic corpuscle. w.p/. Upper lip. v. d2.v.
Ventral blood-vessel.
All the figures, except figs. 16, 17, and 28, refer to Enchytreus hor-
tensis, n. sp.
NOTES ON OLIGOCHATES. 69
Fic. 1.—Enlarged view of the anterior region, showing the shape of the
brain, salivary glands, and septal glands.
Fic. 2.—Enlarged view of the nephridium, drawn from the living.
Fie. 3.—Section through part of the nephridium and nephrostome. Z.
apoch. 4 mm., oc. 4 c. cam.
Fic. 4.—Enlarged ventral view of the nephrostome, from the living.
Fic. 5.—Hularged side view of the same.
Fic. 6.—Transverse section through the posterior region of the nephridium.
Z., D, oc. 4, cam.
Fic. 7.—Enlarged view of a portion of the last region of the nephridial
canal, from the living.
Fic. 8.—Optical section through the body-wall. Z., D, oc. 3.
Fic. 9.—Thread-corpuscles from a young worm (fresh). a, 0,¢. Front
views. d. Side view. Z., D, oc. 4.
Fic. 10.—Young (?) thread-corpuscles from an adult worm. a. Side view.
b. Front view. Much enlarged (fresh).
Fic. 11.—Oval (a, 4, c) and thread-corpuscles (d, e, f/). Flemming and
carm.-alum. Z., D, oc. 4, cam.
Fig. 12.—Oval (a) and thread-corpuscles (4, ¢). Corrosive and bor. car-
mine. (The thread has been acted upon by the reagents leaving the cavity it
formerly filled.) Z., F, oc. 2, cam.
Fic. 13.—Series of figures indicating the action of water on the thread-
corpuscle (a). Much enlarged.
Fic. 14.—The thread after the action of caustic potash. Enlarged.
Figs. 15 and 16.—An oval corpuscle before (a) and after (a’), and a thread-
corpuscle before (4) and after ('), the action of lime water,
Fic. 17.—Enlarged view of a portion of Pachydrilus, sp. (?), showing the
long cells in the ceelom; from the living.
Fie. 17a.—Section through an elongated cell from the ccelom of Pachy-
drilus, sp. (?). Corrosive, bor. carmine. Z., D, oc. 4, cam.
Fie. 18.—Cheta. Z., D, oc. 4, cam.
Figs. 19, 20, 21, and 22.—Series of transverse sections through the sperma-
theca, taken from before backwards. Z., B, oc. 3, cam.
Fig. 23.—Diagram showing the communication of the spermathece with
the exterior and the cesophagus ; in longitudinal section.
Fries. 24, 25, and 26.—Series of longitudinal horizontal sections taken from
above downwards, showing the relations of the spermatheca. Z., B, oc. 4
comp., cam.
Fig, 27.—Longitudinal horizontal section through the fifth segment of
Marionia enchytreoides. Z., B, oc. 4, cam.
Fic. 28.—Ameeboid corpuscles (fresh). Much enlarged.
ere Te
ss
jieth
i. me ‘ f hee
A os ek
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 71
On the Development of Lichenopora
verrucaria, Fabr.
By
Sidney F. Harmer, M.A., B.S8c.,
Fellow of King’s College, Cambridge ; Superintendent of the University
Museum of Zoology.
With Plates 7—10, and two figures in text.
I HAVE formerly had occasion to describe certain very re-
markable phenomena in the development of Crisia (6), and I
then ventured to suggest that embryonic fission would be found
to be characteristic of the whole group of Cyclostomatous
Polyzoa. The development of Crisia being, to the best of my
belief, without an exact parallel in the entire animal kingdom,
it was important to test my former results by the study of some
other Cyclostome. A chance discovery of large numbers of the
colonies of Lichenopora verrucaria, Fabr., in all stages of
development, enabled me not only to confirm the main fact of
the occurrence of embryonic fission, but to discover certain
previously unsuspected phenomena in the life-history of that
species. An abstract of my results has been communicated to
the Royal Society (8).
It would not be easy to find two genera of Cyclostomata
which are more unlike than Crisia and Lichenopora;
and the occurrence of a fundamentally similar mode of develop-
ment in the two forms may be taken as giving good reason to
believe that we are really dealing with a normal characteristic
of the group. Crisia is an erect, branching form, whose
calcareous branches are interrupted at intervals by chitinous
joints. Development here takes place in the interior of a
72 SIDNEY F. HARMER.
pear-shaped ovicell, which replaces a zocecium at a greater or
smaller distance from a joint, according to the species. Lichen-
opora, on the contrary, is a plano-convex disc, usually
closely attached by its flat surface to some foreign object ; its
zocecia project from its upper side, and between them is a
calcareous lamina or “crust,” which forms the roof of a large
compound ovicell. The development of this ovicell commences
with the early stages in the growth of the colony. Whilst a
Crisia-colony may produce a profusion of ovicells at the
breeding season, or may be entirely devoid of these structures,
the ovicell of Lichenopora is a single, complicated struc-
ture, whose growth is intimately connected with the develop-
ment of the external features of the colony. In dealing with
this question I am obliged to limit myself to L. verrucaria,
the only species of which I have obtained an ample supply of
material. It may, however, be pointed out that, if this species
does not materially differ from other species of the same genus,
it is the character of the ovicell which has been taken as the
distinguishing feature of the genus or even of the family to
which it belongs.
According to Hincks (9, p. 471) the zocecia of the Licheno-
poride are ‘ disposed in more or less distinct series, which
radiate from a free central area.’’ ‘Thearea here referred to is,
in L. verrucaria, the roof of the ovicell, and the definition
applies only to moderately advanced colonies (Pl. 7, fig. 7),
in which the ovicell possesses a calcareous roof. But when
development has reached a certain stage, every colony possesses
an ovicell, and this is in marked contrast to most of the
Cyclostomata, in which ovicells are not present in a very large
proportion of the colonies which may be examined.
A still more noteworthy feature remains to be pointed out.
The internal processes which precede the external appearance
of the ovicell commence almost with the beginning of colonial
life. The individual formed by the metamorphosis of the free
larva gives rise, almost at the same time, to two new zoccia.
One of these two commonly becomes fertile, and forms the
starting-point of the series of stages by which the ovicell
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 73
becomes stocked with numerous larve. The other zocecia of
the colony remain, for the most part, unfertile, although many
of them produce spermatozoa. It is clear that they have the
function, not only ef feeding themselves, but also of providing
nutritive material at the expense of which the ovicell and its
brood of larvee may develop. The restriction of the female
reproductive function to a single individual, or at least to a
very small number of individuals of the colony, is by no means
in accordance with what is known of other colonial animals.
The remarkable nature of the phenomenon will be most clearly
realised when it is understood that the fertile zocecium is
usually one of the two “ blastozoites,” which are first formed
by budding from the “‘ oozoite.”’ }
My material was collected in Norway at Lervik, in Stordo,
and at Godosund, off the north coast of Tysnaes6, during the
end of June and the beginning of July. At that time of the
year there is no difficulty in collecting an indefinite number of
colonies from the earliest stages immediately after metamor-
phosis to the fully developed condition. The entire colonies
were preserved for the most part in corrosive sublimate, to
which, in some cases, a few drops of nitric acid and of acetic
acid were added. ‘The internal details were studied prin-
cipally from sections, which were prepared for me by my wife,
but to some extent in entire colonies stained without decal-
cification. The external features were investigated in dry
mounts and in Canada-balsam preparations, whether stained
or unstained.
Owing to its discoidal or flattened form, the colony will
obviously tend to rest on one of its flattened surfaces when it
is mounted. In order to examine certain features which can
only be made out when the colony is looked at edgeways, it is
necessary to make some special arrangement. This has been
successfully accomplished by folding a piece of black paper into
* These terms, first introduced by Lacaze-Duthiers, have been appro-
priately employed by Prouho (18) in describing the colonies of Polyzoa. ‘The
oozoite is the individual developed from an egg, i. e. the metamorphosed
larva or primary zoccium of the colony. The blastozoite is an individual
which has been formed as a bud.
74 SIDNEY F. HARMER.
a series of ridges whose cross section has the form of MM. The
iichenopora is then mounted (dry) on the slopes of the
ridges ; and this allows of the examination of one edge of the
colony. An improvement on this method, whereby each
colony could be examined in a series of positions, was effected
by taking a short length (about 18 mm.) of brass tubing,
through the axis of which a long needle was passed, the
interval between the needle and the tubing being then filled
up with sealing-wax. The outer surface of the brass was
covered with black paper, on which numerous colonies were
mounted. The needle, to which a suitable head is attached,
serves as an axis permitting the revolution of the brass cylinder
under the microscope.
The largest colony, among my preparations, has a maximum
diameter of 5°47 mm. ‘The older colonies are invariably
altered in appearance by the occurrence of secondary thicken-
ings of the surface.
Literature.—A large number of descriptions and figures
of the ovicells of Cyclostomata are scattered through the litera-
ture of the Polyzoa, and some of these are alluded to below.
The only paper which needs special notice here is the second
of Smitt’s admirable series of memoirs dealing with the Scan-
dinavian Polyzoa (20). Smitt has thoroughly understood the
general growth of the colony, and to a large extent that of the
ovicell of L. verrucaria, the particular species we are con-
sidering. But as Smitt was not concerned with the embryonic
development, he paid no attention to the earliest processes in
the formation of the ovicell. As, moreover, I cannot agree
with all the statements of this observer, a complete account of
the development of the entire colony must be given. In further
justification of this course, I may mention that although it is
perfectly obvious to any one who has made an independent
study of L. verrucaria that Smitt had grasped many of the
important facts, the figures which he has given are insufficient
to explain his meaning to anyone who is not well acquainted
with the species.
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 795
Smitt has described correctly the general growth of the
colony, and he has clearly appreciated the arrangement of the
alveoli (see below) or interstitial spaces. He has the merit
of having pointed out that these structures give rise to the
ovicell, and he calls attention to the important fact that they
do not reach the basal lamina of the colony. He has further
stated quite correctly that the central area of the colony, which
is at the same time an important part of the ovicell, owes its
character as a space devoid of zoccia to the fact that the area
originates by the divergence from one another of the most
centrally placed zocecia. Smitt has also given some account
of the secondary thickening of the colony.
The size of the largest colonies (exceeding 8 mm.) recorded
by Smitt is somewhat greater than that of the largest colonies
I have found. I have not been able to discover any colony in
which the ovicell had as many trumpet-shaped openings (one to
eight) as are described by Smitt. Ido not think that so large a
number would ever occur in a young colony, in which the
number seems to be invariably one or two, and in this respect
I find myself in agreement with Levinsen (12, p. 28) ; but I
think that it is perfectly possible that a larger number of
openings are developed in colonies in which a considerable
number of primary embryos (cf. p. 133) are formed as a second
generation. I have obtained no actual evidence that this is
the case, and I have some reason for believing that the larve
of the second generation make use of the old apertures of the
ovicells. Ridley has suggested (19, p. 452) that Smitt has
mistaken zocecia in which incomplete transverse septa have
been developed for the apertures of ovicells.
I do not think that Smitt is right when he describes (p. 478)
the secondary thickening by the formation of the so-called
“cancelli”’ (cf. my figs. 8 and 9) as giving rise to two or three
ovicells, arranged in superposed layers. I do not believe that
there is ever more than a single space which is occupied by
embryos, although that space can fuse with other spaces by the
absorption of the calcareous septa which are at first present.
Smitt describes (p. 476) the first septum formed in the young
76 SIDNEY F. HARMER.
colony at the outset of its growth as bending to the left. He
appears to have missed the curious fact that the growth is
to the left in some colonies, and to the right in others. He has,
I think, not quite appreciated the way in which the proximal
part of the colony becomes covered by the growing edge. His
fig. 7 (pl. x) omits an essential part of the arrangement,—that
is to say, the actual growing edge shown in my own figs. 3—5.
Smitt describes the colour of the ovicell as inclining to blue
(“ till fargen dragande at blatt’). I suspect that this was due
to the use of a glass which was not quite achromatic. The
ovicell. is, according to my observations, perfectly white.
Finally, I must call attention to the small, scattered tubes
with an even, round mouth which Smitt describes as some-
times occurring in the colony. These he compares with the
small tubes of Diastopora hyalina (= D. obelia, Johnst.),
and suggests that they may have something to do with the
production of the male generative organs. ‘This suggestion is
clearly wrong, since the testes are developed inside ordinary
zocecia. I have not observed these small tubes as a part of the
Lichenopora, but I think it possible that Smitt may be
referring to the Infusorian Folliculina, whose tubes are
commonly found growing on the Lichenopora.
External Features of the Colony.—Fig. 7 represents a ma-
ture specimen, with a diameter of 1:92 mm., in which secondary
thickenings have hardly commenced. The colony is nearly
circular, and it is surrounded by the delicate calcareous
‘basal lamina,” which in this particular case is upturned,
although more usually it is closely adnate to the seaweed.
The arrangement of the zocecia may be understood by ima-
gining a number of the quills of quill pens to be arranged
in an obliquely vertical position, radiating from a common
centre. Those near the centre will approach the vertical
position, while those nearer the margin are more nearly
horizontal. It must further be supposed that the nib of
each pen is uppermost, or nearer the centre, so that the orifice
of the tube is completely concealed when the colony is looked
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 77
at from above, except in those individuals which stand nearly
vertically. The cylindrical zocecium is, in fact, obliquely cut
off in the way which has been indicated by the above com-
parison, and this will be better understood by referring to
fig. 6. The “nib” of the zocecium is commonly in the form
of a single spike, but it may be split, so that two, three, or
even more spikes occur. ‘These are very delicate structures in
young zocecia, but they are readily broken off, or lost in other
ways, in old zoecia. In young colonies the zocecia usually
possess more than one spike each. At the growing margins of
old specimens the occurrence of a single spike is much more
common—if, indeed, it is not the rule.
Between the more centrally placed zocecia stretches a calca-
reous lamina (fig. 7)—the roof of the ovicell. This lamina is
not even, but is obviously composed of a number of convex
portions, separated from one another by shallow grooves. The
ovicell opens to the exterior by means of a trumpet-shaped
aperture, within whose expanded mouth is a straight edge,
turned towards the centre of the colony. The section of the
trumpet is at this level semicircular. The straight edge cor-
responds to the level at which the trumpet springs from the
roof of the ovicell. On the opposite side, however, the wall
of the trumpet is longer, so that it becomes continuous with
the roof of the ovicell at a somewhat lower level. Its passage
into the roof is at the same time more gradual and less angular
than on the side which faces the centre of the colony.
This arrangement of the aperture of the ovicell may be
regarded as characteristic of L. verrucaria. The straight
edge may correspond with the valve which I have described
(5) in the ovicell of Crisia, and its function may be to restrain
the too rapid birth of the larve.
The early stages in the growth of a Cyclostome colony have
been well described by Barrois (2 and 8), and my results agree
with his so far as the general character of the growth is con-
cerned. It has not, however, been previously noticed that the
study of the earliest stages of the colony is essential for the
proper understanding of the ovicell of Lichenopora.
78 SIDNEY F. HARMER.
Fig. 1 (Pl. 7) gives two views of an extremely young
colony of L. verrucaria. The colony is attached by means
of a circular disc, whose diameter is ‘16 mm. The disc is
formed, as Barrois has shown, by the calcification of the body-
wall of the fixed larva. It gives off a calcareous tube which
lies nearly horizontally, and ends in an open! mouth, whose
diameter is ‘18 mm. On looking down into this mouth (Fig. 1, 8),
it is seen that the cavity is divided by a septum which appears
triradiate in end view. The septum does not in any part
reach the margin of the aperture of the tube, while proximally
it passes into the lower wall of the tube in such a way that the
cavity of the disc is continuous with the part marked 1, and is
entirely cut off from 2 and 3.
A curious point may now be noticed. About. half the
colonies of L. verrucaria may be described as “ right-
handed,” and about half as “left-handed,” and this will
become intelligible from a consideration of the woodcut (Fig. 1).
Fic. 1.—Lichenopora verrucaria; diagram to explain the difference
between “right-handed” and “left-handed” colonies. The growing
margin of the colonies is omitted. I, IJ, and ILI are supposed to be
seen from above, as in PI. 7, fig. 2. (For explanation, see text.)
1 The mouth appears open (as in later stages) because it is uncalcified.
The delicate layer of cells which stretches across it, and can readily be
demonstrated in sections, disappears in the dried colony.
ON THE DEVELOPMENT OF LICHENOPORA VERRUOARIA. 79
I shows the end view of a young colony, similar to fig. 1,
Plate 7. The space or zoccium A is uppermost, while B
and c are nearer to the seaweed. Assuming it to be possible
for the whole tube shown in I to roll over, it is obvious that it
can do so in the direction shown by either of the two arrows.
Should it roll to the left, we shall have the condition shown in
II, where a and c have come to the upper surface, while B alone
is left in contact with the seaweed. III shows the result of
the rotation of I to the right.
Returning to Pl. 7, fig. 1, it will be understood that the
young colony is attached by the flat surface of its plano-convex
disc, as well as by the flattened wall of the zoccia 2 and 3.
This precludes any actual rotation of the parts already deve-
loped, but it does not preclude an alteration in the relative
position of 1, 2, and 3, as the colony increases in length. This
it does in such a way as to assume the position which would
be arrived at if it at first grew quite symmetrically, but subse-
quently experienced a rotation of the common mouth of the
tube, while leaving its proximal end unaltered. In other
words, the zoecium 1 may either grow to the right or to the
left, as shown in IV and V of the woodcut, the other two
zocecia experiencing a corresponding change of position. A
colony in which 1, as seen under the compound. microscope,
has grown to the right, will be described as “ right-handed ”
(IV), while a ‘ left-handed” colony has the appearance of V
in the woodcut.
It is obvious that a cross-section of either IV or V, imme-
diately above the constriction which marks the passage of 1
into the disc, will have the appearance of I; while a section
taken more distally will show the first zocecium (1) in the
position of a in IT or III, according as we are dealing with a
*‘ right-handed ” or a “left-handed” colony. It is clear that
the zocecium 3 in IV corresponds with B in II; while 3 in V
corresponds with cin IIT; and the object of giving different
symbols to these zowcia in 1V and V is to call attention to the
fact that the individual properties of the zoccia in the older
colony depend on the direction in which the primary zocecium
80 SIDNEY F. HARMER.
grows at the commencement of the formation of the colony.
The whole colony acquires a certain symmetry with regard to
3, the axial zocecium, and in studying the characteristics of 2
and 3 it matters not which of these corresponds with B and
which with cin I of the diagram. The necessity for a close
attention to these zocecia will become apparent later.
The distinction between right- and left-handed colonies per-
sists throughout life, although it ceases to be obvious in colo-
nies of more thana certain age, when looked at in the usual
position, from above. The difference can readily be appreciated
in young colonies, in which the early stages of the embryonic
development are taking place. This will be seen by comparing
fig. 3 (left-handed) with fig. 4 (right-handed).
The two kinds of colonies seem to occur in about equal
numbers. Of sixteen chosen at random as a test case, eight
were right-handed and eight were left-handed ; and this result
agrees with the other observations I have made on this point.
The actual position of the three zoccia can be seen from
fig. 2, which is an end view of a young colony which was
proved to be left-handed by examining it in another position.
1 is the primary zoccium, which in future will be designated
by the symbol z!, the symbols z? and z> being employed for
the zocecia 2 and 3 respectively in either 1V or V of the wood-
cut (p. 78). The tubular end of 2? has in fig. 2 reached a
higher stage of development than that of z). zis in contact
with the surface of attachment, while z! and z? are on the
free surface of the colony. The position of these three zoccia
is also shown in fig. 3, a somewhat later stage seen from above.
The colony may be described as having the form of a flattened
funnel, whose narrow end originates in the disc. The funnel
is attached by one of its flattened sides ; and its lip, which is
closely attached to the seaweed on this side, here projects
considerably further than on the opposite side. The wide end
of the funnel is, in fact, obliquely truncated.
2° has acquired an axial position, which it retains through-
out the later development. z! and z’ are, at their upper ends,
arranged symmetrically with regard to z°, although the wood-
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 81
cut on p. 78 (IV and V) shows that 2? and 2 are at their
bases symmetrical with regard to z!, and are clearly a pair of
sister zocecia produced by the primary zocecium of the colony.
Figs. 2 and 14 illustrate further the manner in which the
number of zocecia becomes increased. This happens by the
forking of a pre-existing septum ; and the zoccia are at first
bounded externally by the common growing rim of the colony.
The way in which the zocecium is completed externally is shown
in fig. 14, which represents a part of the growing margin of an
old Lichenopora. No sooner is the outer wall completed
than the zocecium commences to grow upwards as a free tube,
although its inner side has before then commenced its upward
growth. ‘The obliquely truncated character of the adult orifice
is thus a marked feature even at the first formation of a
zocecium.
Fig. 14 shows that the septa do not reach the actual rim of
the colony, and that the outer wall of the zocecium always
grows up from within the edge. Since this happens in the
entire circuit of the mouth of the funnel-shaped colony, part
of the rim capable of forming new zoccia is left all round, and
this is the proximate cause of the assumption of the completely
discoidal shape which is so characteristic of the genus Lichen-
opora. The growing edge is well seen in figs. 38 and 4, from
which it is evident that that part of the rim which is situated
on the proximal side of the first three zoccia grows at first
much less energetically than the opposite part.
We may here notice an important generic difference between
Lichenopora and such a genus as Tubulipora. In the
latter, z! and z? are formed in such a way that the proximal
part of the rim of the young funnel-shaped colony is used up
in forming their outer walls. There is thus no edge capable
of giving rise to new zocecia between z! and z? on the one
hand and the disc of fixation on the other hand. The colony
thus assumes a fan-like shape instead of the discoidal form
which is assumed by Lichenopora. This distinction has
already been pointed out by Smitt (20, p. 476).
Fig. 4 represents a right-handed colony of Lichenopora
vou. 39, PART 1.—NEW SER. F
82 SIDNEY F. HARMER.
which has a maximum diameter of ‘59 mm. The number of
zoccia has increased, and it will be noticed that the zocecium
which we have called z? still maintains its axial position in
the colony. The zoccium between z? and z! may be termed
z‘, while that between z® and z? will be called z*. It will be
unnecessary to give special symbols to the other zoccia.
Growth has, so far, proceeded almost exclusively on the
distal side of the first three zocecia; but the characteristic
features of the adult Lichenopora are soon acquired by the
growth of that part of the margin of the colony which inter-
venes between them and the disc. In fig. 2 the diverging
zocecia z! and z* are separated by a quadrangular area which
is the outer wall of a pyramidal space situated between 2},
z°, 2°, and the outer wall of the funnel-shaped colony. In
fig. 4 this space has become divided into two, and at a later
stage two or three zoccia will be found between the distal
ends of z! and z?, and directed in such a way that they radiate
from z® towards the disc of attachment. The position of these
zocecia is shown in figs. 5 and 12. Their development has
taken place simultaneously with the extension of the proximal
part of the rim of the funnel, which grows in such a way
as to cover that part of the funnel which intervenes between
the disc and the open mouth. ‘The rim or basal lamina here
grows horizontally, keeping in close contact with the wall of
the funnel, and finally covering it and the disc completely.
Having done this, it comes in contact with the seaweed on
the proximal side of the disc of attachment ; and the latter is
thus completely covered. Growth has not, however, been
confined to this region. The colony has been increasing hori-
zontally in every direction, z® forming a centre approximately
equidistant from all parts of the margin.
The general nature of these processes will be understood
by comparing figs. 5 and 7. In the fully formed discoidal
Lichenopora no trace can be seen from the upper side of the
proximal part of the funnel ; but it is merely necessary to turn
the colony over to find that that part persists throughout life,
without any increase of size, A back view of the adult colony,
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARTA. 83
showing the disc and the proximal part of the funnel, has been
well figured by Smitt (20, pl. xi, fig. 6). It is obvious that
the increase of the diameter of the colony is accompanied by
an increase in the number of zocecia which are present.
The divergence of the zocecia from the centre of the discoidal
colony results in the occurrence of a central space, round which
the zocecia are arranged; and this space is the commencement
of the ovicell. Two points must be here expressly noticed :
firstly, that the ovicell is not restricted to certain colonies, but
is developed in all colonies which reach the proper stage of
growth; and secondly, that the description of the young ovicell
must be given with a word of warning. The account of the
growth of the colony has been taken from entire specimens,
and largely from dried specimens. In the latter at least, we
should not expect to find much trace of any structures which
are not calcareous; and even in dealing with Canada-balsam
preparations of entire colonies, it is extremely difficult to see
much of the delicate, uncalcified membranes which really
exist. But it must here be noticed that the study of sections
shows emphatically that the various spaces of the colony are
roofed in by living membranes. Thus the spaces at the growing
edge shown in fig. 14 are really closed, and the same is true
of the interzocecial spaces which give rise to the ovicell. Even
the orifices of the zocecia are not widely-open tubes, but are closed
during life by a membranous diaphragm, which can be widened
to allow of the protrusion of the tentacles. Ofall this, nothing
appears in the dried colony ; but it must be clearly understood
that the cavity of a young ovicell is morphologically a body-
cavity, and is not an external space converted into an ovicell
by the formation of a calcareous roof.
The development of the ovicell can be most easily described
after the structure of the complete ovicell is understood.
Fig. 6 represents a thick free-hand section showing the whole
length of a single radius of a colony which has not been decal-
cified. ‘The zoccia diverge from the centre of the colony.
Each zoecium reaches the attached basal lamina, a short part
of which it forms. The older zocecia are very much longer
84. SIDNEY F. HARMER.
than the younger ones. The roof of the ovicell connects the
diverging zocecia, aud the cavity of the ovicell is clearly, in the
central part of the colony, a space which completely surrounds
a part of the zocecia. The roof of the ovicell is not perfectly
even, but is formed of convex portions, shown in surface view
in fig. 7.
The ovicell increases in size during the growth of the colony ;
and the nature of this increase may be understood from fig. 14.
The marginal spaces, marked out by the vertical septa which
start from the basal lamina, are all destined to give rise to
zocecia, which are the only spaces which extend down to the
base. The zoccium is first clearly indicated by the appearance
of one or more of the spikes which occur on the upper margin
of the orifice of all the zocecia. At about this stage (fig. 14, 4),
one of the septa forming the side walls of the zocecium gives
off a branch which runs more or less transversely. ‘The cavity
of the young zocecium is in this way cut off from the cavity on
the distal side of it. Up to this time the zocecium has been
a short horizontally-placed tube, whose lower wall is part of
the basal lamina, and whose free upper wall is parallel to the
base. No sooner is the cavity of the zocecium completed on its
distal side than it begins to alter its direction of growth. The
woodcut Fig. 2 is a diagrammatic representation of the margin
of acolony. The thick lines represent two young zocecia and
ji
8 e Se = : — C - - ‘
Fic. 2.—Diagram of the growth of the margin of the colony of Licheno-
poraverrucaria. (For explanation, see text.)
part of the basal lamina as they would be seen in a radial
section, and the dotted lines represent the condition of the
corresponding parts after a certain amount of growth has taken
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 89
place; acu Bis the basal lamina. The zocecium whose point
is marked p has its upper end free, even in the earlier con-
dition, while that marked £ is still incomplete on its distal
side (cf. fig. 14). It is soon completed on that side by an
obliquely vertical septum, which starts at Fr; and the pro-
longation of this septum forms the lower wall of the zocecium.
By the time that the margin of the colony has extended to BI,
the two radial septa, which at first formed the lateral parts of
the zocecium £, have grown in length and in height; and they
are supposed to have meanwhile given rise to the beginning of
a new zocecium, G, which has reached the condition of & in
the earlier phase.
In examining the woodcut in the condition indicated by the
dotted lines, it is clear that a space occurs between the two
older zocecia, and a much younger one between the two younger
zoecia. Smitt (20) alludes to a space of this nature as an
“ interstitialrum ;” but it will be convenient to substitute the
term “ alveolus.”
The zocecia may be arranged in radial series, as shown in
the woodcut; or their position may be for the most part
alternate, as in fig. 14. The upper sides of the zocecia are
marked with longitudinal ridges, which have been omitted
from most of the figures for the sake of simplicity, but are
shown in figs. 8 and 9. The projecting spikes of the zocecia
are commonly continuous with these ridges, some of which
become exaggerated at their lower ends to form lateral walls
for the alveoli. These are thus pyramidal spaces, pointed at
their lower ends, and at first only closed by membrane above.
They increase in length by the growth of the zocecia, and by
the extension of the vertical septa on sides where they are not
bounded by zocecia. The septa do not, however, grow so high as
the zocecia; but after they have reached a certain height, each
alveolus commences to become roofed by the calcification of the
membrane which is already present in that situation. The
calcification of the roof of each alveolus begins at the margin
and gradually advances to the centre, in the same way that an
opening is gradually closed by the contraction of a sphincter
86 SIDNEY F. HARMER.
muscle. An alveolus which is incompletely roofed in is
shown in fig. 6.
The original extent of the alveoli is indicated by the grooves,
which correspond with the original septa, on the surface of the
ovicell (cf. figs. 7 and 14). It is obvious that the ovicell is
gradually increased in size by the addition of fresh alveoli all
round its margin. The vertical septa are absorbed after the
alveolus has become part of the ovicell, and a continuous space
thus results, through which the more centrally placed zocecia
pass as free pillars. Fig. 6 illustrates this point, and the con-
formation of that part of the roof of the ovicell which lies to
the left of the right-hand zocecium indicates the loss of the
vertical septa which were at first present.
We have still to consider the earliest development of the
ovicell, which, to the best of my belief, has not been studied in
detail by any previous observer. Fig. 12, though by no means
an early stage, will serve to introduce us to a phenomenon
which is of the highest importance for the understanding of
the history of the colony. This phenomenon may be spoken
of as the occlusion of the fertile zoecium.
In fig. 12, all parts of the ovicell which have received a
complete roof are shaded. The preparation is a transparent
one, made in Canada balsam, so that the walls of the zowcia
can be seen through the roof of the ovicell. Parts which are
seen through some other part of the colony are indicated in
dotted lines. The colony is left-handed—z! and z? project in
the ordinary way beyond the roof of the ovicell, while z?, the
occluded zocecium, is completely closed by the roof, and would
probably not have been visible at all in a dried, opaque pre-
paration. In the immediate neighbourhood of the occluded
zocecium arises the trumpet-shaped aperture of the ovicell. The
study of decalcified sections is necessary in order to understand
the significance of these facts, but it must for the present be
assumed that we are justified in alluding to z? in this colony as
the “fertile? zocecium—i.e. as the zocecium in which the
embryo is first developed.
The roofing in of the ovicell commences with the
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 87
occlusion of the fertile zowcium. This process is not
easy to see in a colony which is resting in its position of most
stable equilibrium on its flat surface; but it can be seen
without difficulty in a colony of suitable age which has been
mounted on a rotating cylinder in the manner described on
p- 74. No general rule can be laid down with regard to the
size of the entire colony in its relation to the closure of the
ovicell. The smallest colony in which I have found this
process actually beginning measured ‘77 mm. in diameter,
while the largest colony in the same condition as regards its
ovicell was 1:13 mm. across. In nearly all the cases I have
actually observed at this stage, 2? was the fertile zocecium,
although this rule is by no means invariable. A reference to
the woodcut on p. 78 will show that the orifice of 2? may be
described as having an axial and an abaxial side. The occlu-
sion of the zocecium commences with the formation of a calca-
reous film, starting from the abaxial edge of the zocecium.
This film grows inwards as a cap to the zocecium (fig. 18) ;
but while it is closely fitted on to, and indeed arises from the
abaxial edge of the orifice, it grows over the axial side of the
edge as an arch which does not fuse with the edge. The film
is the beginning of the roof of the ovicell; and it is clear
from fig. 13 that a passage leading from z? into the cavity of
the ovicell is left between the calcareous film and the axial
side of the edge of the orifice of z2. Through this passage
the embryo passes from the fertile zowcium into
the ovicell. It should perhaps be pointed out that this
statement is to some extent a matter of inference. I cannot
claim to have first found this passage in a particular colony,
and then to have observed the embryo on its way through the
passage. But I have been able to show that the embryo
actually passes from the fertile zocecium into the ovicell
through an aperture close below the roof of the latter; and it
may fairly be concluded that the passage whose formation we
have just considered is really formed for this purpose.
No sooner has the fertile zocecium been occluded than the
formation of the orifice of the ovicell is commenced. This is
88 SIDNEY F. HARMER.
seen in an early stage in fig. 5, where the trumpet-shaped
mouth is not yet formed. Transparent preparations (fig. 12)
show that the trumpet is developed in the immediate neigh-
bourhood of the occluded zocwcium. The appearance of the
fully formed trumpet is shown in fig. 7.
The occlusion of a fertile zoecium may or may not result in
the obliteration of all signs of its presence. In fig. 13 the
principal spike of z? is still visible externally, in spite of the
occlusion of that part of the orifice. In other cases, a careful
examination of a colony may result in the detection of the
ends of two or three spikes in the immediate neighbourhood
of the aperture of the ovicell, which clearly belong to a zoecium,
and from their relation to the trumpet must have belonged to
the occluded fertile zoccium. What is probably part of the
orifice of z*? is seen as a zigzag line below the aperture of
the ovicell in fig. 7.
The study of sections shows certainly that z? is not neces-
sarily the fertile zoecium, although the cases in which either
this or z> is fertile very largely preponderate over all other
cases put together. The study of entire colonies amply con-
firms this, if we may for the present assume that the conjunc-
tion of an occluded zocecium with a trumpet-shaped aperture
belonging to the ovicell is a strong reason for believing the
zoecium to have been fertile. In many cases where the open-
ing of the ovicell is formed in connection with z®, it occupies a
very characteristic position with regard to that zoecium, being
placed immediately on the distal side of the original orifice of
2°, and with its trumpet-like mouth directed towards the disc
from which the colony started its growth. The opening of
the ovicell less commonly has a similar relation to 2”.
It will be remembered that some few of the oldest zoccia of
each colony form a group which diverge from the centre of
the discoidal colony. These zocecia are connected with one
another by vertical septa which do not necessarily reach the
roof of the ovicell. In other words, a part of the roof which
starts at one vertical septum does not always become connected
with the adjacent septum, but may form an arch extending
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 89
completely above that septum to one beyond it. In this way
is formed what may be called a compound alveolus; i.e. an
alveolus which at its formation is partially subdivided by one
or more vertical septa which do not reach the roof of the ovi-
cell. The closure of the central part of the ovicell commonly
and probably usually takes place by the formation of one or
more compound alveoli, whereas the extension of the ovicell
in a mature colony appears to take place for the most part by
the formation of simple alveoli.
The extent of the central area of the colony, devoid of
zocecia, depends in the first instance on the extent to which
the older zocecia diverge from one another; but it may be in-
creased by the occlusion of zoccia. This process, which can
be best made out in transparent preparations, has nothing to
do with the occlusion of the fertile zoecium. It is highly
probable that the occlusion follows the loss of the polypide.1
The process does not appear to take place to any great extent
in young and healthy colonies of L. verrucaria, but in old
colonies a considerable number of the more centrally placed
zocecia may be occluded. A calcareous lid may in these cases
be added in the position of the original orifice (Pl. 7, fig. 9),
and it may or may not be formed in an oblique position corre-
sponding with the original condition of the orifice. This is
very different from the state of an occluded fertile zocecium, in
which the closure takes place at the level of the roof of the
ovicell.
All the older colonies of Lichenopora differ from those
which have so far been described by the occurrence of a very
striking process of secondary thickening of the surface. Since
old colonies are further remarkable for producing new broods
of embryos (cf. p. 132), which are developed in a different way
from the first brood, the idea suggests itself that the second-
arily thickened colonies are the result of a second year’s
growth. I have no actual evidence in favour of this sugges-
tion, but the fact that secondary broods of larve are produced
1 Cf. my former paper on Crisia (5, p. 142).
90 SIDNEY F. HARMER.
in old colonies at the same time of year as primary broods from
young colonies is in favour of that view.
In fig. 14 places will be noticed, as at B, where the neigh-
bouring alveoli do not meet accurately. These places are the
** cancelli”” which are so commonly met with in accounts of
the genus Lichenopera. The cancelli do not at first develop
to any great extent; but this condition is altered in the old
colonies (figs. 8, 9).
Fig. 9 shows three zocecia of an old colony. One of these
zocecia has been occluded by a calcareous cap at a considerable
height above the level of the ovicell. This is quite similar to
the structures which are well known to occur in profusion
over the more centrally placed zocecia of certain species of
Diastopora.
The alveoli which are shown in fig. 9 are separated from
one another by shallow grooves. At these regions, and at the
junction of a zocecium with the roof of the ovicell, vertical
septa are rising up. Fig. 8 shows the appearance of a colony
after this secondary thickening has been proceeding for a longer
period. The system of vertical septa, which in the earlier
stage started only at the intersection of the alveoli with one
another or with the zoccia, has now extended so as to cover
the entire surface of the ovicell. The porous roof of the latter
can be seen at the bottom of the open cancelli, which have
for the most part no calcareous roof. The base of the right-
hand zoccium is surrounded by very large cancelli, the walls
of which (as in fig. 9) are continuous with the longitudinal
ridges of the outside of the zocecium. These cancelli are
obviously encroaching on the base of the zocecium, which can
be seen at the bottom of the spaces. Those marked ‘cancelli”’
are beginning to acquire a calcareous roof. The process of
roofing in is much further advanced at the base of the left-hand
zocecium, where one of the cancelli is completely closed. In
this way the bases of the zocecia of old colonies commonly
become surrounded by a set of blister-like swellings, which
have been formed above the original roof of the ovicell. These
are presumably the structures which Smitt regards as forming
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 91
a new ovicell superposed on the old one (cf. p. 75). Judging
by the analogy of the walls of the alveoli, I think it possible
that the floor of these blister-like spaces may ultimately be
absorbed, and the ovicell thereby enlarged.
It is hardly possible to examine figs. 8 and 9 without feeling
convinced that there must be a layer of living tissue outside
the roof of the ovicell. Although I have not obtained complete
evidence with regard to the morphological character of this
layer, its presence can be demonstrated in vertical sections of
stained colonies which have not been decalcified. In pre-
parations of this nature a film of stained nucleated tissue can
be seen, passing over the outer side of the calcareous parts of
the upper surface of the colony, and thus covering the upper
side of the roof of the ovicell, and extending up the outside of
the zocecial tubes to the orifices of the latter, where the tissue
passes round to join the part of the body-wall which is invagi-
nated into the orifice of each zowcium. The ridges on the
outer side of the zocecia are as a matter of fact undeveloped
septa, which can continue their growth under certain circum-
stances. It is thus natural to find that the outer wall as well
as the inner wall of the young and active zocecia should be
covered with living tissue.
Where the protoplasmic layer passes over the convex surface
of an alveolus, or up the uniform surface of a zocecium, it is
usually tightly stretched over the adjacent calcareous matter.
But in places where it crosses a groove between two alveoli,
the layer usually appears as a tangent of the two curved surfaces.
Hence one sees in sections a more or less triangular space,
roofed only by a delicate membrane, at the junction of two
alveoli. It is obvious that this is in accord with the arrange-
ments shown in fig. 9.
The calcareous part of the upper surface of a Lichenopora
is thus not the outermost layer; and it becomes a question
how far it corresponds with the “ectocyst”? of some other
Polyzoa. In the Entoprocta, the Phylactolemata, and the
Ctenostomata, in which calcification does not usually occur,
the outermost layer is a mere cuticle; and it is to this layer
92 SIDNEY F. HARMER.
that Allman (1, pp. 8, 13) gave the name of “ ectocyst.”
There is, however, a good deal of reason for believing that the
arrangements are not quite so simple in calcareous Polyzoa.
Ostroumoff (14) announced in 1885 that the use of silver
nitrate demonstrated the evidence of an epithelial layer external
to the calcareous skeleton in several genera of calcareous
Polyzoa; and later (15, p. 12) expressed himself to the effect
that in calcareous forms, such as Lepralia, the calcareous
matter separated the ectoderm into two layers, one of which is
outside the skeleton. Vigelius (22, sep., pp. 3, 4, pl. vi, fig. 24)
describes and figures an extra-skeletal layer in an encrusting
Cheilostome, and in a later paper (23) states that the skeleton
is probably usually formed inside the ectoderm-cells. Pergens
(17) finds that the skeleton of the calcareous forms examined
by him is overlaid by a cuticle merely.
The relation of the calcareous skeleton to the parts of the
living body-wall appears to need further elucidation; but the
condition found in Lichenopora makes it probable that, in
that genus at least, there is more than mere ectoderm outside
the calcareous layers of the upper surface of the colony. The
cavities of the cancelli are probably morphologically parts of
the body-cavity; and as such are no doubt surrounded by more
than ectoderm.
It appears to me extremely likely that the relations of the
calcareous septa in Lichenopora may not be very different
from those of the septa of a Zoantharian Coral, as shown by
Koch’s well-known researches. The outer wall of the young
funnel-shaped colony (fig. 3) is immediately derived from the
body-wall of the metamorphosed larva; and it is of course a
young condition of the basal lamina. All the septa which
divide the zocecia, the alveoli, and the cancelli from one another
are ultimately derived from the basal lamina; the first forma-
tion of a young septum at the growing edge always appearing
as an outgrowth from that layer (fig. 14), and other septa being
always formed from pre-existing septa.
I have obtained no evidence of the existence of soft tissues
on the lower side of the basal lamina, and I do not think it
ON THE DEVELOPMENT OF LICHENOPORA VERRUOARIA. 93
likely that any such exist. The view which seems to me to be
suggested by my observations and by those of previous workers
is that the basal lamina is the outermost layer of the body-
wall; that the body-cavity it encloses has been repeatedly
subdivided by the ingrowth of a series of branching and
anastomosing ridges or septa from the body-wall. The whole
upper surface of a Lichenopora colony thus corresponds with
what would be seen in an end view of a growing point of
Crisia. But whilst in the latter the growing edge is sub-
divided by a very simple set of septa, which result in the
formation of only two series of zocecia, that of Lichenopora
is divided by an immensely complicated series of septa. The
parallel is not, however, quite accurate ; since in Crisia, as in
Tubulipora (cf. p. 81), the septa reach the growing edge of
the colony, and the mature zoccia are thus completely outside
the “ growing-point,” because part of the growing edge is used
up in toto in the formation of each zocecium. The morphology
of the Lichenopora colony can be most easily understood by
considering a young colony like fig. 3. It will then be seen
that while the lateral body-wall is complete from the beginning,
the mouth of the expanded funnel is closed by a living body-
wall. While this is constantly encroached on and altered by
the formation of septa, whether those giving rise to new
zocecia or those which form alveoli cr cancelli, the body-wall
is never completely calcified in this region. For this uncalcified
body-wall, which closes the mouth of the funnel-shaped colony
of Lichenopora, of the growing points of Crisia, &c., the
name “terminal membrane” may be suggested. It is this
layer which gives rise to the polypide-buds, and is invaginated
into the orifice of every zocecium.
If the septa are really, as there seems every reason to believe,
derivatives of the basal lamina, there is no reason for not re-
garding them as ingrowths of a true calcified ectocyst; and
they may accordingly be compared, in a general way, with the
septa of a Coral (as already pointed out), or with the chiti-
nous ingrowths which form the endoskeleton of a crayfish.
94, SIDNEY F. HARMER.
Internal Structure.—The general appearance of a_ section
through a colony in which the production of the first brood
of embryos is at its height will be understood from fig. 11,
representing a section, parallel to the flat surface, of a colony
in about the same stage as that shown in fig. 12.1 The section
has a maximum diameter of °80 mm. The ovicell may be ex-
plained by comparing its contents with an Ameba with a
series of blunt pseudopodia extending in the intervals be-
tween the zocecia. The “ pseudopodia”’ contain a considerable
number of secondary embryos, and the nucleus may be repre-
sented by the “fertile brown body.”
This structure, although an inert body which plays no direct
part in the development, is in fact the centre about which all
the most important phenomena in the development of the ovi-
cell take place. The “ pseudopodia”’ form a complex structure
for which the name “embryophore”’ may be suggested, and
they lie in a cavity which may be regarded as the body-cavity
of the ovicell.
The zoccia can be identified, in a good and well orientated
series of sections, by an examination of the proximal end of
the colony, which, it will be remembered, retains throughout
life the form assumed by the young Lichenopora. Hence
it may be confidently stated that z is a section of the pri-
mary individual of the colony, and that z? and z® represent
respectively the second and third zowcia, as defined in the
earlier part of this paper.
z} possesses a polypide, whose tentacles, enclosed in their
tentacle-sheath, are seen in the figure; z? is quite empty
basally, but contains some embryos distally, as shown by the
figure; z° contains a brown body basally, but it has no polypide,
and it opens at the level of the section into the ovicell. Two
of the zoccia contain brown bodies, that of A not being visible
in the section figured.
The fertile brown body is in the neighbourhood of z?, near
whose upper end is developed the trumpet-shaped aperture of
1 The secondary embryos are more developed in fig. 11 than they would
have been in fig. 12.
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 95
the ovicell (not seen in the section figured). There is no
aperture in connection with z’, or indeed with any of the other
zocecia. The colony has every appearance of being perfectly
healthy. The sections of its basal part show that nearly
every zoecium has a functional polypide. Only one of these
polypides has any trace of a testis, which consists of a small
mass of degenerating spermatozoa.
The occurrence of an ordinary brown body in z’? makes it
improbable, as will be shown later, that this was the fertile
zocecilum ; and it is in the highest degree probable that the
primary embryo of this colony was produced by 2%. This is
indicated by the position of the apertnre of the ovicell and by
that of the fertile brown body, as well as by the occurrence of
embryos within the cavity of z2. The connexion between z°®
and the ovicell probably implies that z*? had been occluded
(cf. p. 89).
My preparations include a practically continuous set of stages
- between fig. 11 and the commencement of the development ;
and my results are based on the examination of complete
series of sections through some three hundred colonies, in
eighty-nine of which no embryos can be detected, while in
seven more, degeneration had set in after the commencement
of the development. The large number of fertile colonies
examined enables me, to a large extent, to eliminate abnorma-
lities in the development.
An examination of the cases in which normal embryonic
development was not found will give some useful information.
The negative result in at least twenty cases was probably due
to the fact that the colonies were too young. A careful com-
parison of my measurements of entire colonies and of sections
shows that the commencement of the embryonic development
is to be looked for in colonies which measure about *40—
‘43 mm. in total length, and about ‘25 mm. in transverse
diameter, i.e. in colonies which are at about the stage of
fig. 2.
In the next place I am able to confirm the suggestion I
formerly made (6, p. 212) to the effect that the normal develop-
96 SIDNEY F. HARMER.
ment of the ovicell is dependent on the activity of the poly-
pides. On decalcifying a number of colonies of Lichenopora,
a considerable number of all ages and sizes will be found to
be practically empty. In studying the embryonic development,
time is simply wasted if preparations are made of colonies in
which the polypides are not for the most part fully active and
functional, or at least in which the polypides have not recently
been functional. A recently degenerated polypide may be
difficult to distinguish, in a preparation of an entire colony,
from a functional polypide; whereas it is quite easy to recog-
nise the small, compact, brown body which indicates that
the histolysis of the polypide took place at some more distant
period.
The conditions under which the embryos are nourished
appear to be very different in different Cyclostomes. In
Crisia ramosa I have shown that the embryos are contained
in a highly protoplasmic reticulum, which, it can hardly be
doubted, serves for the transference of nutriment to the de-
veloping larve. In C. eburnea the reticulum is reduced to
a minimum, and the path of the nutrient substances is pro-
bably somewhat different from that inC. ramosa. Analogous
phenomena probably occur in the genus Lichenopora; and
L. verrucaria may be compared with C. ebur nea, in which
the reticulum is but slightly developed. Although I have not
yet been able to obtain a supply of material for the proper
examination of other species of Lichenopora, I have obtained
a series of sections through a single fertile colony of L. hispida,
Flem. The ovicell contains a comparatively small number
of young secondary embryos, which are embedded in a large
solid nucleated mass which probably corresponds, in func-
tion at least, with the “ follicle” of C. ramosa (cf. 6, Pl.
xxii, fig. 6).
My examination of this single series of sections of L.
hispida enables me to state that the embryonic development
of that species probably differs to a very considerable extent
from that of L. verrucaria. Another curious difference
between the two species may here be pointed out. While in
ON THE DEVELOPMENT OF LIOHENOPORA VERRUCARIA. 97
L. verrucaria the brown bodies formed by the degeneration
of the polypides are absorbed from time to time, so that a
zocecium hardly ever contains more than one or at most two
brown bodies, those of L. hispida (again judging from my
single series of sections) accumulate at the basal end of the
zocecium, a considerable part of which is tightly packed by a
mass of brown bodies.
It seems clear that considerable differences in detail occur in
the embryonic development of different Cyclostomes; but I
believe that it will be found that the phenomena are funda-
mentally the same throughout the group.
Whatever may be the character of the arrangement by which
the embryos are directly nourished, there can be no doubt that
it is the polypides which ultimately supply the material at the
expense of which they grow. In a very large proportion of
the cases in which no embryo could be discovered, the poly-
pides were not functional in all or some of the zowcia. A
similar result is obtained by the examination of the degenera-
tion of the embryo in fertile colonies. Leaving out of account
doubtful cases, I have fifteen colonies in which degeneration of
the embryo has clearly occurred. In eleven of these cases all
the polypides of the colony had degenerated, this being pro-
bably the proximate cause of the degeneration of the embryo.
In some cases new polypide buds were being developed, and
the colony would obviously have survived, and would probably
have developed a new embryo from one of the younger poly-
pides. One of the other four cases probably points in the
same direction. The fertile zocecium contains two brown
bodies, a polypide, and an embryo which commenced to dege-
nerate in the ‘ suspensor stage ” (cf. fig. 24, a normal embryo
at this stage), The remains of the embryo are between the
two brown bodies, and it is highly probable that the distal
brown body represents the polypide which originally supported
the suspensor. The abnormal degeneration of this polypide,
at an earlier period, probably resulted in the degeneration of
the embryo, which is still just recognisable, although a new
polypide has grown up in the fertile zocecium.
VOL. 39, PART 1.—NEW SER. G
98 SIDNEY F. HARMER.
A much larger number of degenerating embryos might have
been obtained, if the colonies of which sections were to be cut
had been taken at random, instead of having been chosen with
some regard to the condition of their polypides.
I am unable to suggest any definite cause for the degenera-
tion in most cases. A Protozoon, a species of Folliculina,
is commonly found with the basal end of its tube embedded in
the Lichenopora, its free end projecting from the surface of
the colony. These organisms can often be recognised in the
sections of the colonies, and in many cases they are found in
degenerating specimens. It is quite possible that the Foll1-
culina may in these cases be the cause of degeneration ; but
it is equally possible that it is the effect, and that the Protozoa
settle down in parts of the colony which have lost their zocecia.
They are found, not uncommonly, in parts of the colony in
which their presence does not seem to have had any effect on
the adjacent zocecia.
Degeneration may start at any stage of the embryonic de-
velopment. I have observed it most commonly in colonies in
which the embryo had developed up to the “ suspensor stage ; ”
but I have also found it commencing at later stages, as, for
instance, at the beginning of embryonic fission, or even later.
In describing the development of the structures found within
the ovicell of Lichenopora verrucaria, it is desirable to
keep the account of the first brood of larve entirely separate
from that of later broods. The appearance of an ovicell con-
taining larve belonging to the first brood has already been
considered (fig. 11); and this is the condition which is usually
found in colonies which have a diameter of 1 mm. or there-
abouts. The origin of all these complicated struc-
tures is to be looked for in colonies which consist
of only three or four zoecia.
For descriptive purposes it will be convenient to classify
the embryos in a certain number of stages; and it will be
seen that the form of the entire colony has a distinct relation
to these stages, although wide variations occur in the period
at which any particular stage of development is passed through.
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 99
The following stages are naturally marked out by the phe-
nomena of the development.
Stage A.—Formation of the definitive egg (figs. 15—18).
The entire colony in which this stage occurs measures about
‘45 mm. in length, and about ‘25 in transverse diameter
(between figs. 2 and 8).
Stage B.—Formation of the follicle and suspensor (figs.
19—22). The colony does not differ materially in size from
those in which Stage A occurs.
Stage C.—‘Suspensor stage” (figs. 10, 23, 24), the
embryo being supported by a functional polypide. Colony in
the condition of figs. 3 and4; usually about ‘50—-60 mm. long,
and about °35—'55 mm. broad.
Stage D.—The polypide has degenerated, and the embryo
is supported by a brown body found near the distal end of the
fertile zocecium (figs. 27, 28). Colony as in Stage C.
Stage E.—Disappearance of the “ suspensor”’ and enlarge-
ment of the embryo (figs. 29, 30).
Stage F.—Commencement of embryonic fission (figs. 31—
35). Diameter of colony about ‘60—:80 mm.
Stage G.—Fully formed ovicell (figs. 7, 11, 12).
The measurements given here and elsewhere, so far as they
refer to sections, must be taken as approximate. The plane
of the sections may not be such as to permit of a satisfactory
measurement being taken at all, while in the case of an
oblique series the whole length or breadth of a colony may
not be seen in any one section ; and the necessity of estimating
the length obviously introduces a source of error. The delicate
calcareous lamina which forms the edge of the entire colony
is usually not distinct in sections. Lastly, there is consider-
able variation in the size of the colony and in the time at
which development commences.
In spite of these drawbacks, I believe that it will be useful
to record the measurements of the colony, wherever possible,
as a guide to future investigators. The measurements given
in the sequel refer in all cases, unless the contrary is stated,
to the sections, and not to the colony before decalcification,
100 SIDNEY F. HARMER.
During my study of the development of L. verrucaria, I
have been much struck with the frequency with which the
zocecia z” and z® form the starting point of the. ovicell. This
result was first obtained by a study of sections; but, as has
already been pointed out, it was afterwards amply verified by
an investigation of the entire colonies. Some statistics will
bring out the striking nature of this phenomenon. In 108
colonies belonging to Stages B to G, or being fertile colonies
in which degeneration of the embryo was taking place, I have
been able to determine the fertile zocecium. In the cases of
the older colonies (Stages E to G), it is easier to recognise the
fertile zocecium when it happens to be z? or z? than when it is
a younger zocecium ; and the older stages have consequently
been omitted in the second column of the table.
Table showing the Frequency of the Occurrence of
z* or z° as the Fertile Zowcium.
Stages B—G
and degenerating Stages B—D.
fertile colonies.
No. of cases in which z? is fertile ‘ ‘ 44 j 99
2 rs Z . ‘ - Si ‘ 15
29 y) 2° or 23 ee) . . 8 ° 5
55 R 5 Pizaee : : 15 ; 1
“4 3 Piz us : ; 8 : 6
re a a younger zocecium is
fertile . : 3 6 ‘ 5
- s two zoccia are fertile . 6 : 3
“A a the fertile zocecium could
be determined . 5 108 ‘ 75
Total number of colonies investigated . : 175 : 95
As it is by no means easy to determine the fertile zocecium
in unfavorably orientated sections, there is obviously room
for error in the above results; but as I have only entered
cases in which I have felt confident that I could really deter-
mine which zoccium was fertile, I believe that the table is
substantially accurate. In estimating the bearing of these
numbers it must be noticed that in a considerable proportion
of the cases in which the fertile zocecium is entered as ?2? or
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 101
?28, the doubt usually arises in consequence of the difficulty of
distinguishing between z? and z® in the sections. Thus in
most cases where a fertile zoceclum was marked ?2z?, it was
certainly z° if it was not 2’.
Another line in the table refers to cases in which the fertile
zocecium was either z? or z®, but it was not certain which of
these two was fertile.
We thus obtain the result that in the colonies of all ages
up to Stage G the fertile zocecium was either z? or z> in
96 cases, while a younger zocecium was fertile in only 6
cases in which a single embryo was present. The correspond-
ing numbers for Stages B to D only are 67 and 5.
Even if a liberal deduction is made from the 96 of the first
column or the 67 of the second, on the ground of uncertainty
or possible errors of observation, we are still entitled to con-
sider that the fertile zocecium which produces the first brood
of embryos in L. verrucaria is normally one of the two zocecia
which are first formed by budding after the larva has meta-
morphosed itself into the primary zocecium of the colony. Two
of the six cases alluded to in the table as colonies in which
two zocecia are fertile, have embryos in both z? and z3. In two
more cases the fertile zocecia were probably z? and z?, In the
fifth case z?, and in the sixth case z® was fertile in addition to
a younger zocecium than either of these two.
The table further brings out the fact that z? is more often
fertile than z?. An examination of young colonies which have
been stained and mounted whole, without decalcification, shows
that z is really older than z’, although there is no evidence of
this in the later stages. Thus a colony ‘82 mm. long by
‘15 mm. broad had a mature polypide in z', and a small
polypide-bud in z*. Another, °35 mm. by ‘24 mm., had poly-
pides in z! and z?; while a third, -43 mm. by ‘27 mm., was
developing for the first time a polypide-bud in 2’, in addition
to having functional polypides in the older zoccia.
z” is thus clearly older than z’, and we accordingly have the
following striking result. ‘There is a strong tendency in the
development of Lichenopora verrucaria for the ovicell to
102 SIDNEY F. HARMER.
owe its origin to the oldest zocecium which is formed asa bud
and does not result from the metamorphosis of the larva; or,
in other words, the ovicell tends to be formed from the first
blastozoite of the colony. I have found no case in which the
ovicell owes its origin to z'; but it must not therefore be con-
cluded that the primary individual differs from the others in
being devoid of generative organs. z! commonly has a well-
developed testis, while I have on several occasions observed
structures in z! which I consider to be egg-cells.
It will be seen from the figures of the entire colonies that the
determination of z®? and z? is easy enough in sections in many
cases in which the orientation is good. It is more difficult
when, as commonly happens, the disc from which the colony
originates is injured or lost when the Lichenopora is removed
from the seaweed. The determination of the fertile zocecium
is greatly facilitated by noticing the position of the tentacles of
the polypides.
The woodcut on p. 78 shows that z! may be considered to
have an inner (axial) and an outer side. During the retracted
condition of the polypide the tentacles lie close to the inner
side of the zocecium, while the cecum of the stomach is situated
more externally. The orientation of the polypide of z’ is like
that of the polypide of z'; but the polypide of z* faces in the
opposite direction, its tentacles lying, like those of z!, close to
the inner border of the zoccium. ‘This appears to be true of
all the polypides developed in any zocecium. After a given
polypide has degenerated into the condition of a brown body,
the newly formed polypide is found to have the same orienta-
tion of its tentacles as its predecessor had. The younger
zocecia have a corresponding arrangement. In any colony the
polypides are normally orientated during retraction in such a
way that the tentacles of a polypide are nearer to the centre of
the circle enclosed by the growing margin of the colony than
is the stomach. This point is, however, not brought out by
any of the figures which illustrate this paper. It may be noted
that the number of the tentacles is constantly eight.
A further point which deserves special notice is the character
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA, 103
of the “ brown bodies.” Barrois (2 and 3) has shown that the
brown body formed by the degenerating parts of the larva
during metamorphosis has disappeared completely by the time
that the first polypide is mature. He has not, however, de-
scribed the way in which this disappearance takes place.
Ostroumoff (16, p. 185) states that the cecum of the primary
polypide is formed ‘“ at the cost of the brown body;” but he
does not enter into further details.
I can confirm Barrois’ statements with regard to the dis-
appearance of the degenerated remains of the larva, since I
have found that the primary zocecium, when very young, pos-
sesses a mature polypide, but has no trace of a brown body.
I have found this condition lasting till the stage is reached
when z” has a mature polypide, and z’ possesses either a bud or
amore or less mature polypide. At a later stage the primary
zocecilum contains a brown body as well asa polypide. The
smallest colonies in which I have found this condition mea-
sured *33 X ‘16 mm. and *35 x ‘21 mm. respectively. This
implies that the functional polypide is in these cases not the
original tenant of the zocecium, although in some colonies of
the same or even a larger size I have found no brown body.
The degeneration of a polypide, in fact, results in the forma-
tion of a “brown body,” which in the younger stages of the
colony nearly always remains as a distinct structure in the
zocecium. In the older stages, as we shall have occasion to
see, the brown bodies commonly disappear from the zocecia.
When present they are easy to recognise, as they appear as
bright yellow bodies even in sections which have been stained
with Grenacher’s hematoxylin. A part of the cecum of the
stomach has the same colour in sections.
I do not think that the primary polypide of z! develops a
testis, although that organ is commonly present as an append-
age of the polypide of z' in cases where the corresponding
zocecium possesses a brown body,
104 SIDNEY F. HARMER.
Stage A.—Formation of the Definitive Egg.
Figs. 17 and 18 have already been described in my pre-
liminary note (8). Fig. 17 represents a colony consisting of
three fully formed zocecia only. The section is parallel to the
surface of attachment, in what may be called a horizontal
plane. The examination of the complete series of sections
showed that both z! and z? have a brown body, a polypide, and
a testis. The level of the section figured is not such as to
show the characteristic orientation of z? (see p. 102); z> has a
brown body, a polypide, and a conspicuous cell (seen more
highly magnified in fig. 18), which obviously resembles an
egg. A few nuclei to the right of the egg probably represent
the commencement of the investments of the embryo. Sper-
matozoa were found floating freely in the body-cavity of z! and
z’, but none were found in the fertile zocecium in this parti-
cular case.
I have not observed the process of fertilisation: with any
degree of certainty, but I believe that it must occur at about
this stage. The fertile zooecium may or may not have a testis,
while testes are commonly present in other zocecia, particularly
during the early stages of embryonic development. In some
cases, and particularly in old colonies in which new broods of
embryos are developing, the testes are enormous, being some-
times as much as *30 mm. long. It can hardly be doubted
from these facts that fertilisation does occur at some period.
Considerable masses of ripe spermatozoa are found in the body-
cavity of many of the zocecia; and it is probable that they pass
thence, in some way, to the fertile zocecium. The young sper-
matozoa are always in groups of four, as in Crisia (6).
The correct identification of the large cell shown in fig. 18
is clearly of the first importance, and I have thought it
desirable to figure two other cases of the same kind (figs. 15
and 16).
Fig. 15 shows an egg attached toa polypide-bud. z!, z?, and
z° contained polypides, but none of the other zocecia were old
enough to possess more than buds. The egg seems to belong
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 105
to the primary polypide-bud of a zocecium which is younger
than z* and z°, since no brown body could be detected in it.
A precisely similar egg occurred in another polypide bud of the
same age in the same colony.
Fig. 16 shows part of the polypide of z? of a rather long and
narrow colony, measuring ‘48 mm. in length. The first three
zocecia are fully formed; and each has a brown body, a poly-
pide, and a testis. z! has either one or two eggs, and z? has
the two eggs shown in the figure. I have observed in all six
cases in which eggs were developed in z!, and in one of these
the egg was found in the recently formed brown body. It
appears to me that z! may occasionally give rise to true eggs ;
but I have observed no single case in which an embryo is
formed in that zoecium. It is probable that the eggs shown
in fig. 16 would (one or both) have developed, and that the
zocecium (z°) which contains them would have become fertile.
In another case the zocecia of a comparatively old colony
had formerly contained brown bodies and polypides. All the
latter had, however, degenerated (probably simultaneously),
forming a series of recent brown bodies, in which the remains
of testes can be detected. Each zocecium thus contains an old
brown body, and a younger, half-formed brown body; and, as
is commonly noticed in such cases, new polypide-buds have
been developed to take the place of the old polypides. One of
these, already recognisable as a young polypide, was found to
possess an egg similar to that shown in fig. 15. The colony
having undergone a complete process of degeneration, it is not
surprising to find that this egg is developed in a zocecium
which is younger than either z? or z°.
It is obvious from the foregoing statements that the first
appearance of the eggs differs in different cases. Hither one
or two eggs may be present, and they may occur with or with-
out a testis. In one case I have found them in all the first
three zocecia. More commonly I have detected an egg in only
one zoecium iu acolony. The diameter of the egg is about
14 (average of four measurements). ‘This is not very differ-
ent from the measurement (17°6 ) which I have recorded (6)
106 SIDNEY F. HARMER.
for the egg of Crisia, and the correspondence is even closer
if we take the largest Lichenopora egg (16) of which I have
a record.
In a single case I have found an egg-like cell, 9°6 u in dia-
meter, at the growing margin of acolony. This cell corresponds
closely in appearance and position with the eggs which I have
described in the growing-points of Crisia. I do not, however,
feel by any means sure that this condition really represents a
normal phase of the development. Some of my other prepara-
tions suggest that the egg is more probably differentiated in
situ from the outer layer of a young polypide bud. Moreover
the brown body with which the developing embryo is associ-
ated implies the loss of a polypide in the fertile zocecium. One
might therefore expect, a priori, that the egg would make its
appearance in a bud formed to replace a pre-existing polypide,
and not in a part of the growing edge where no polypides have
yet become mature.
I have found cells which I regard as eggs in thirty-eight
colonies. In. most cases I have been unable to detect more
than a single egg in a colony, although in some cases two or
even three eggs may occur in a single zoecium. The egg-
bearing zocecilum is commonly 2? or z®, but in several colonies
the egg occurred in a younger zocecium. In one or two of the
latter cases either z? or z> was already preoccupied with an
embryo of its own. ‘The observations I have noted down with
regard to the occurrence of these cells point strongly in the
direction of their being eggs. They do not occur promiscuously
in all the zocecia. Should a colony have been successful in
developing an embryo no further trace is ordinarily seen of eggs.
In young colonies which have no embryo these cells are of
common occurrence, and in very young colonies they are found
in just those zoccia which might be expected to produce them.
At the same time the results are not perfectly consistent, and I
can best interpret them by assuming that there is considerable
variation with regard to the first origin of the eggs. This
appears to me to be particularly the case with young colonies
in which the development has not been quite normal, whether
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 107
in consequence of the degeneration of the polypides or from
other causes.
Stage B.—Formation of the Follicle, &c.
Although the embryo and its investments can be recognised
with certainty at this stage, it is by no means easy to make out
all the details.
Fig. 19 shows a part of the polypide, with the embryo, of a
colony which possessed functional polypides in 2}, z’, and 2’,
and had two other zoccia which had reached the stage at which
the polypide-bud becomes obvious. The fertile zocecium is
either z? or z>; it possesses a brown body and a small testis,
neither of which is shown in the section figured. The body-
cavity contains some ripe spermatozoa, some of which are seen
in the immediate neighbourhood of the embryo. The latter
consists of several nuclei embedded in a mass of protoplasm
which lies in a definite cavity. This is bounded by a layer of
cells which may be termed the follicle.
Fig. 22 is from z* in an older colony which had four old and
one half-grown polypide, and two buds. No brown bodies
could be distinguished in this case—a very unusual phenome-
non in colonies of this age, although quite common in older
colonies. The four mature polypides are, however, conspicu-
ously old, and their alimentary canals have solid contents which
are probably the missing brown bodies. This question will be
considered later. The investments of the embryo are rather
more definite than in fig. 19, and it is more difficult to make
out with certainty the limit between the follicle and the em-
bryo. At least five of the central group of nuclei clearly belong
to the latter.
Fig. 21 shows a transverse section of the follicle and embryo
at a stage between the last two figures. The colony measured
‘40 by ‘21 mm., and possessed three polypides only, of which
either z? or z> was fertile. The follicle is somewhat thicker
than in the cases previously considered, and possesses more
than one layer of nuclei. A brown body is present, as usual,
in the fertile zocecium.
108 SIDNEY F. HARMER.
Fig. 20 is from a colony which possessed three polypides only,
each of them being accompanied by a brown body. The fertile
zocecium is almost certainly z?, but if not itis z’. The structures
connected with the embryo show a distinct advance in develop-
ment. The follicle is more fully formed, and there is also pre-
sent an elongated group of nuclei which is easily recognised as
the “suspensor” of the next stage.
My notes show a perfectly consistent result with regard to
the age of the colony at which this stage is passed through.
This is brought out by the following table:
Embryos in Stage B.
No. of colonies
Colonies possessing three polypides only 7 5 “ 16
+f r four or five polypides_ . : : 3
Colony larger, the fertile zocecium being younger than z? or 23 1
Sections not good and result uncertain : ; : 5
Total number of colonies examined . : : 25
It is clear that Stage B is most commonly passed through in
colonies which contain three fully developed polypides, with
perhaps one or two young buds in younger zoecia. The mea-
surements of the colonies are remarkably uniform at this
stage, the total length of the colony (in section) ranging from
‘40 to ‘48 mm.
It may be noted that although most of the above colonies
were in the three-polypide stage, it does not follow that any
colony containing that number of polypides will also contain
an embryo in Stage B. We have already seen that a consider-
able number of these colonies contain eggs. In some cases,
moreover, it appears that the polypides may be well formed in
z'—z3 before any of these zocecia undergo a histolysis of their
polypide, or at least before the polypides of z? and 2° experi-
ence that fate. Now since the occurrence of a brown body is
all but universal in all fertile zocecia, of whatever age, and
probably occurs in all fertile zocecia which are developing
normally, we can hardly expect to find an embryo in any z? or
z in which the first polypide of the zocecium has not had time
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 109
to degenerate and to be replaced by a new one. The brown
body which is found in connection with the developing embryo
is alluded to in this paper as the fertile brown body.
The exact origin of the cells forming the embryonic invest-
ments could not be made out, but I feel confident that they
are not formed as a modification of an ordinary polypide-bud.
They appear to me to be differentiated in situ from the cells
of the funicular tissue which surround the egg, aided probably
by the cells which form the capsule of the fertile brown body.
On two occasions I have, however, found young colonies in
which the appearances could be interpreted by supposing that
a polypide-bud was concerned in the production of the struc-
tures connected with the embryo. I have come to the conclu-
sion that such a view would be erroneous ; and I regard the
colonies in question as really belonging to Stage D, although
they have reached that stage rather precocicusly.
Stage C.—‘‘Suspensor Stage,’ with Brown Body and
Polypide.
I have ventured to call this stage by a name taken from the
development of the embryo in a flowering plant, in conse-
quence of the resemblance of a cord of cells supporting the
embryo to the structure known to botanists as the suspensor.
By this term I refer to the inner cord of cords (figs. 23, 24)
which is contained in an investment continuous with the cells
which immediately surround the embryo. ‘These cells, which
enclose the embryo, may be termed the “follicle ” (fig. 24).
The whole set of structures which are developed in imme-
diate relation with the embryo form the “‘embryophore,” a
structure which has been seen at a later stage of development
in fig. 11 (cf. p. 94).
The general appearance of the embryo, with its embryonic
membranes, at this stage, may be gathered from fig. 10, repre-
senting part of the fertile polypide in a colony which is about
‘67 mm. long. The colony possesses in all about nine brown
bodies, one of which is accompanied by a polypide-bud, and
110 SIDNEY F. HARMER.
the rest are in zocecia which have a functional polypide: that
is, about nine zocecia are fully formed, and of these the third
(probably) is fertile. The fertile polypide, like some of the
others, has a testis, from which the embryophore hangs down
freely into the body-cavity. The fertile brown body occurs in
the immediate neighbourhood of the embryo. This stage is
clearly not very different from fig. 20, which has already been
described.
Fig. 23 is from a somewhat younger colony (measuring ‘53
by ‘32 mm.). Five zoccia are here fully formed, with brown
bodies and polypides, and of these z? is fertile. The outline
of the fertile brown body is projected from other sections on to
the figure. In this case the fertile zocecium has no testis,
although that organ is present in some of the other zoccia.
The suspensor is, as is frequently the case, curved in such
a way as to keep the follicle in close contact with the brown
body; but this curved condition may be merely the position
assumed during the retraction of the polypide. The suspensor
obviously consists of more than a single row of cells. Its
investing cells are intimately connected with a layer which
forms a kind of capsule round the fertile brown body. The
follicle consists of a protoplasmic mass containing two more or
less distinct layers of nuclei, covered externally by a few more
flattened nuclei, which probably belong to an external flat
epithelial layer.
Fig. 24 (from a colony ‘48 by ‘45 mm.) shows the embryo
and its accessories with unusual clearness. The fertile zocecium
is 2°. The section suggests that the suspensor is really a tube,
and that the two rows of nuclei are on opposite walls of its
lumen. This is borne out by other sections, and particularly
by transverse sections. The lumen can sometimes be made
out; but it probably does not always exist as anything more
than the finest passage.
Fig. 25 represents an unusual condition, in which two
embryos are developing in the same follicle. The colony was
about ‘56 mm. long, and z? was fertile. I have found two per-
fectly similar cases belonging to the next stage (Stage D), and
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 111
it is quite possible that the phenomenon is not very uncommon.
After the commencement of embryonic fission, at any rate, it
would be quite impossible to say whether one embryo or two
were originally present. It will be remembered that two eggs
are sometimes borne by a single polypide (fig. 16). It may
fairly be inferred that under certain circumstances both eggs
may develop.
Some idea of the general appearance of the embryo at
Stage C may be obtained by the inspection of figs. 23, 24, but
it is often by no means easy to make out the structure of the
embryo, or indeed to be sure how many nuclei belong to it, and
how many may have been derived from intrusive follicle-cells.
I have previously pointed out (6, p. 215) that the follicle-
cells in Crisia do push their way in among the blastomeres ;
and I have obtained a good many preparations of Lichenopora
in which this process appears to be taking place. I can say
no more with regard to the structure of the embryo than that
it is a mass of protoplasm containing nuclei; and I am unable
to say whether, at this stage or even later, these nuclei are
arranged with any relation to embryonic layers.
Whatever doubt may exist with regard to the exact details
in Stages A and B, there cannot be the slightest doubt that the
“suspensor stage ” is a perfectly normal phase of the develop-
ment. I have examined some thirty-eight colonies in which
the suspensor was present in the condition described above.
With hardly any exception, I have found this stage in colonies
which measure ‘40—'68 mm. in length (in sections) and ‘30—
*55 mm.in breadth. The fertile zoeecium may or may not have
a testis, and I cannot lay down any general rule on the subject.
In two or three observed cases more than one zocecium was
fertile. The only other variation which need be recorded,
although it might equally well have been considered under the
next stage, is the occasional occurrence of a young polypide bud
in a zocecium which has reached the normal suspensor stage,
and has recently lost its polypide by degeneration.
In some cases this bud appears simultaneously with the
development, in other zocecia, of buds which are clearly formed
112 SIDNEY F. HARMER.
in consequence of the degeneration of the polypides. This is
a case of a phenomenon repeatedly noticed, in which an
epidemic of degeneration attacks all the polypides of a colony
simultaneously. In a large number of cases this probably has
no permanently injurious effect on the colony, as new polypide
buds are developed in all the zowcia, even while many of the
parts of the degenerating polypides are still distinctly recog-
nisable. Iam convinced that the simultaneous degeneration
of the polypides is often responsible for the degeneration of the
embryo; and in some of the above-mentioned cases in which a
young bud occurs in the fertile zocecium, the embryo would
probably have ultimately become involved in the brown body
This is borne out by examination of a colony, ‘51 mm. long by
‘37 mm. broad, in which all the polypides had moderately
recently degenerated. Polypide-buds are being developed to
replace them, and an advanced bud is present in the fertile
zoecium. ‘This bud possesses an egg, although eggs could not
be found in any of the other zocecia. The inference is that the
egg was destined to replace the (degenerating) embryo present
in the same zocecium.
Should the buds become functional polypides with sufficient
rapidity, I see no reason why the embryo should not survive.
It is possible, on the contrary, that the occurrence of a young
polypide-bud together with a degenerating polypide in the
fertile zowcium (Stage C) always implies that the embryo has
already commenced to degenerate. I am induced by the study
of one series of sections to think that this is not necessarily the
case. The fertile zocecium contains a polypide which is clearly
young. The brown body is large; near the suspensor it is
distinctly old, while at the opposite end it is as distinctly
newly formed. ‘The embryo and its accessories are normal in
character. This case suggests a fusion of the original fertile
brown body with a new one formed by the degeneration of the
fertile polypide. A new polypide has, however, been de-
veloped in the same zoccium, and the health of the embryo
does not seem to have been in any way affected by these
changes.
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA, 118
It is impossible to avoid being struck with the very remark-
able similarity between the stage which has just been described
in Lichenopora and a phase in the development of Pluma-
tella, as described by Kraepelin (11; see especially pl. i,
figs. 65—68). The figures given by that observer might almost
do duty for Lichenopora, except that the ‘‘suspensor”’ is
rather longer and the “ follicle” less complicated than in that
genus. The “suspensor’’ of the fresh-water form is, more-
over, attached to the body-wall instead of being borne by a
polypide or by a brown body.
Kraepelin and Braem (4) have investigated the origin of
this structure in Plumatella, and they both arrive at the
conclusion that the structure which supports and contains the
egg is a polypide-bud. Braem, indeed, dissents from the
statement of Metschnikoff (13), by whom the structure was
first described, to the effect that it is an ordinary bud; and
points out certain differences between it and a bud which is
destined to grow into a polypide. But the embryophore in
Phylactolemata would appear to be a two-layered bud, and
the part which corresponds to my “suspensor” is simply the
inner or ectodermic layer of that bud.
Braem further calls attention to the occurrence, in the
morula stage, of certain nuclei which are smaller than those of
the segmentation spheres. These may possibly have some
relation with the smaller nuclei which I have myself noticed
in Lichenopora during Stage C.
The earliest changes undergone by the fertilised ovum of
Plumatella, according to Kraepelin’s account, appear to be
extremely similar to those which occur in Cyclostomes. In
Plumatella, as in Lichenopora, a time arrives when the
embryo is a sharply marked spheroidal mass supported by the
end of the “‘ suspensor.”
From this stage the development of the Phylactolemata is
very different from that of the Cyclostomata. In the former
group the embryo enlarges, becomes two-layered and acquires a
central cavity which is completely closed onall sides. Oneor more
polypide-buds are invaginated from the two-layered body-wall
VOL. 389, PART 1.—NEW SER. H
114 SIDNEY F. HARMER.
into the central body-cavity, and the embryo finally escapes as
a larva provided with a larger or smaller number of polypides.
In Lichenopora, on the contrary, the morula enlarges while
retaining an embryonic, undifferentiated character in its cells.
It then undergoes repeated fission, and the definitive secondary
embryos produced in this way then acquire a two-layered
character which is strikingly suggestive of the Phylactolema-
tous embryo.
In spite of the extraordinary resemblance between the “sus-
pensor stage” of Lichenopora and that of Plumatella, I
have not been able to convince myself that the result is arrived
at, in the former, in the way described by Kraepelin for the
latter. The structure of the embryophore in Lichenopora
would indeed be most easily explained by supposing the
‘* suspensor ” to represent the ectodermic part of a bud, but I
have not been able to obtain any evidence that this is the case.
There are, however, instances among marine Polyzoa where a
new polypide-bud is formed in the immediate neighbourhood of
the brown body (cf. 7, p. 140), instead of in close connection
with the body-wall. It is possible that the Lichenopora
suspensor may belong to this little-understood class of polypide
buds, but I must emphasise the fact that I have no internal
evidence tending to prove that it is a bud.
Were it not for the very marked similarity of the early stages
of the two groups there would be little reason for anticipating
an agreement in detail between Cyclostomata and Phylacto-
lemata, particularly as the details of Lichenopora are so
different from those of Crisia, another Cyclostome.
Kraepelin (10) has, moreover, urged a series of reasons for
believing that the Ctenostomata, among the marine Polyzoa,
are the nearest allies of the Phylactolemata. As I have made
no special study of the latter group, I have arrived at no con-
clusions of my own with regard to their affinities, but the
details of their early developmental phenomena appear to me
to be sufficiently similar to those of Cyclostomes to make it
worth while to call special attention to the fact.
One other suggestion may be made with regard to the
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 115
Phylactolemata. In spite of the admirable researches of
Braem (4) and of Kraepelin (11), the mode of the first origin
of the statoblasts cannot be said to be beyond doubt. Although
in the absence of any observations of my own on this point I
do not wish to commit myself to this view, it appears to me not
impossible that some connexion exists between the formation
of the statoblasts and-the process of embryonic fission. The
natural view of older writers that the statoblasts are winter
eggs may hereafter prove to have had more truth in it than
has recently been supposed. Verworn (21), indeed, has at-
tempted to resuscitate the older view, although his results
have not commended themselves to later inquirers. Braem’s
description and figures in particular lend some colour to the
suggestion that the formation of the statoblasts may be a modi-
fied form of embryonic fission. For fear of being misunder-
stood I must, however, repeat my statement that I do not con-
sider that I have sufficient grounds either for accepting or for
rejecting this suggestion.
We may now return to the consideration of Lichenopora,
and we pass to—
Stage D.—Suspensor Stage, with Definitive Fertile Brown
Body.
The normal occurrence of this stage is supported in the
clearest possible manner by my sections.
Fig. 28 represents a typical case. The colony measures
‘43 by ‘37 mm., and it contains six brown bodies, one of which
is fertile, while each of the others is accompanied by a poly-
pide. No other zoccium is sufficiently old to have a polypide,
and the colony is one in which this stage of development has been
reached comparatively early. zis fertile; it contains a brown
body, from which the embryophore hangs down into the zoe-
cium. The brown body is not directly connected with the
orifice of the zoccium, although it is now situated near the
upper end of the body-cavity.
Fig. 27 (Pl. 9) is probably a somewhat later stage, in
which the brown body has come close up to the layer of body-
116 SIDNEY F. HARMER.
wall which is invaginated at the orifice of the zoecium. The
colony is larger than the preceding specimen, measuring 69 mm.
by ‘51 mm., and possessing at least twelve brown bodies. The
fertile zocecium is here neither z? nor z’, but a younger indivi-
dual of the colony. The figure shows that the suspensor and
the follicle are still easily recognised. The brown body has,
however, acquired a much more definite capsule of cells than it
had in the previous stage. This capsule is not so fully formed
in the somewhat earlier condition shown in fig. 28; but in
fig. 27 it has become extremely definite. Its protoplasm and
its numerous nuclei take up hematoxylin with an avidity that
shows that the tissue is actively growing, and it is indeed this
tissue which appears to be, somewhat later, closely concerned
with the development of the trumpet-shaped aperture of the
ovicell. The embryo is distinctly larger than it was in Stage C.
Fig. 26 represents the embryo of another colony (in Stage D),
which was slightly younger than that from which fig. 28 was
taken, as it had only five brown bodies and four polypides ; z?
was fertile. The embryo in this case possesses a peripheral
layer of nuclei, which surround a central group of three nuclei.
I am unable to say whether there is any morphological differ-
ence between the central group and the others.
It might be supposed, a priori, that Stage D originates by
the degeneration of the fertile polypide in Stage C. I have
distinct evidence that this is the case. The polypide is seen to
begin to degenerate, and the brown body already present
begins to become confluent with it. The old brown body and
the degenerating polypide become surrounded by a common
capsule of cells. The fertile brown body of Stage D hence
differs from that of Stage C, inasmuch as it contains the
remains of what was the fertile polypide in the earlier stage,
in addition to the original brown body. Should development
proceed normally, no further polypide is developed in the
fertile zocecium, and the brown body moves to its upper
end, and eventually comes into close relation with what was
formerly its orifice.
The size of the colony during Stage D is in most cases from
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 117
*50 to ‘70 mm. in length, and from ‘35 to ‘55 mm. in breadth,
although these limits may be passed in either direction. The
average of eight cases in which I was able to estimate the
length of the colony (in sections) was ‘53 mm., and the average
breadth (twenty-two cases) was ‘44mm. The entire colony
(fig. 4) might well have been in this stage, although still
older colonies are commonly in Stage D, as may be shown by
counting in the sections the number of their fully formed
zocecia.
The distinction between Stages C and D is obvious enough
as regards the embryo and its connected structures. But the
change from one stage to the other implies no more than the
degeneration of a polypide and its alteration into the condition
of a brown body. We should not, therefore, expect to be able
to distinguish by external features alone between colonies in
the two stages. This is all the more difficult, since the period
in the growth of a colony which corresponds with a particular
stage in the embryonic development varies within wide limits.
Stage E.—Disappearance of the Suspensor and Enlarge-
ment of the Embryo.
This stage immediately precedes the beginning of embryonic
fission. Fig. 29 represents an embryo found in a colony mea-
suring ‘56 mm. in breadth. The fertile zocecium was almost
certainly z?. The embryo has unmistakably increased in size
compared with the embryos in the previous stage. While that
of fig. 27 measures 28 in greatest length, that of fig. 29
measures 70 y, which is exactly two and a half times as much.
Other changes have also taken place. The suspensor has
ceased to be a definite structure, and the embryo has moved
towards the brown body, so as to take the place originally
occupied by the suspensor. The follicle is still distinct in this
series of sections, but the section figured only cuts its edge.
The first indication of the commencement of this stage is given
by the longitudinal elongation of the embryo, which then
begins to extend beyond the original limits of the follicle into
the stalk connecting the latter with the brown bedy. We
118 SIDNEY F. HARMER.
have seen that the suspensor is potentially a tube, and it
appears to me that the cells of the suspensor lose the regularity
of their arrangement, and that the embryo probably passes
into the midst of the altered cells of the suspensor, perhaps
by widening out the original lumen. There are, however,
some reasons for thinking that the suspensor may be pushed to
one side during the elongation of the embryo, and I cannot
pronounce definitely on this point. In any case the suspensor
ceases to be recognisable shortly after the close of Stage D,
and the embryo invariably elongates so as to approach the
fertile brown body with its distal or upper end. Towards the
end of Stage E the embryo may have completely left its original
follicle, and may lie entirely in the place of the original
suspensor.
I have found but few colonies (not more than seven or eight)
in this stage ; but those few formed a series completely bridg-
ing over the interval between Stages D and F. The length of
four of these embryos was 30, 53, 70 (fig. 29), 72 u respectively.
The average transverse diameter of the colonies measured in
Stage E (seven cases) was ‘562 mm., the extremes being °48
(‘77 long) and ‘62 mm.
The embryophore shown in fig. 29 occupies a position in
the zocecium exactly like that of the corresponding structure
in fig. 27; that is to say, the cells surrounding the brown
body are in contact with an invagination of the body-wall at
the orifice of the fertile zoecium. The invagination is, how-
ever, considerably deeper in the zocecium from which fig. 29
is taken than in that shown in fig. 27. A thin layer of cells,
not present in the earlier stage, now stretches across the mouth
of the invagination from one edge of the rim of the zocecium
to the opposite edge.
I regard this layer as the remains of the calcareous film
which occludes the mouth of the fertile zocecium (cf. pp. 86,
87). The smallest entire colony in which occlusion was just
commencing measures ‘77 mm. in transverse diameter, whereas
the largest colony in Stage E (measured from sections) was
‘62 mm. broad. This difference is quite unimportant, partly
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 119
because of the variation in the size of colonies at the same stage
of embryonic development, and partly because, in comparing
the measurements of entire colonies with those of sections,
something must be deducted from the former series of measure-
ments. The calcareous lamina which borders the entire colony
is usually not very apparent in sections, or may become some-
what crumpled. In correlating the entire colonies with the
sections it will be well to notice that sections of colonies in
Stage E give no evidence of an aperture belonging to the
ovicell.
Special attention may here be directed to the fact that a
brown body may often be seen in the occluded zocecium of an
entire colony. That brown body is of course the fertile brown
body; and the fact that it may be visible in a dry colony
when other zocecia are not seen to possess a similar structure
is of course due to the constant occurrence of the fertile brown
body near the original orifice. If brown bodies are present in
the other zoccia, they are usually placed at the end of the
cecum of the stomach, and are consequently deeply placed in
the colony, in a position where they cannot be seen.
The fertile zocecium to which fig. 29 refers has yet one
further point of interest. Atashort distance below the embryo
its cavity is completely divided by a septum which passes trans-
versely to the long axis of the zoecium. This septum, which
is commonly noticed in, and is indeed probably a normal
feature of, later stages as well, is probably calcareous during
life. It cuts off the proximal, empty end of the zocecium from
the distal end which contains the embryo.
Stage F.—Commencement of Embryonic Fission.
This stage is by far the most important of all, as demon-
strating the fundamental identity of the developmental pro-
cesses in Crisia and Lichenopora. Fortunately there is no
stage in the latter in which my evidence is clearer than in this
one. I have obtained some twenty-seven series, any one of
which by itself would have been almost sufficient to prove the
occurrence of embryonic fission.
120 SIDNEY F. HARMER.
The youngest figured (fig. 31) shows an embryo at the very
beginning of this stage. The figure is taken from a colony
which measures ‘83 mm. in transverse diameter. The diameter
of the embryo is 45, but it must be remembered, in comparing
it with fig. 29, that it is cut transversely to its long axis. The
embryo is now provided with a clear outer layer of nuclei. The
preparation for fission is indicated by the tendency of the centrally
placed nuclei to arrange themselves in groups or in regular
series, and by the appearance of the small slits or vacuoles
marked S in the figure.
Fig. 35 is a longitudinal section through an older embryo,
in which fission has definitely begun. The brown body (not
cut in this section) is unaltered, and the embryophore still
hangs down freely into the fertile zocecium. The embryo is
no longer a rounded mass, but it consists of an irregular series -
of pieces, which show clear evidence of being engaged in fission.
The embryonic protoplasm and nuclei take up hematoxylin or
borax carmine with great readiness; and there can be no doubt
from their appearance that they are engaged in active growth.
The pieces into which the embryo divides at this stage have no
very regular arrangement; but, as will be seen from figs. 33—35,
the embryonic fragments are at first arranged irregularly round
a more or less definite central space. The appearance of
figs. 31 and 83 suggests that a kind of vacuolation of the
central part of the primary embryo takes place, and that the
embryonic cells thereby become dissociated into a series of
peripherally situated groups. In tracing an embryo, such as
that shown in fig. 34, through a series of sections, the parts of
the primary embryo, where not completely separated from one
another, are seen to be connected in the most irregular manner.
Sometimes they are united with one another laterally ; some-
times an embryo is united with its vis-a-vis by a diagonal
connection passing across the centre of the whole mass; some-
times a piece of embryonic tissue is prolonged into several
finger-like processes, not unlike those of the primary embryo
of Crisia (6, pl. xxiii, fig. 11). There is, in fact, no regularity
or definite method in the breaking up of the primary embryo
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 121
of Lichenopora; and every primary embryo examined at
this stage differs in these respects from every other primary
embryo.
The main fact is, however, the same. The embryo at this
stage has ceased to be a coherent structure,and has resolved itself
into a number of complicated lobes of embryonic tissue, some
of which have actually become separate from their neighbeurs,
and some are ready to become separate at a moment’s notice.
Even now a differentiation of two kinds of embryonic cells
is not necessarily apparent. In fig. 34 there is some indication
of the occurrence of inner cells, which are probably destined
to give rise to the inner layer of the secondary embryo. Other
preparations do not show any such clear differentiation.
The general features of this stage are very characteristic.
The fertile brown body is invariably present, and it is situated
near the upper end of the fertile zoecium. The embryophore
is shaped something like the lower (closed) end of a test-tube,
and it still hangs down freely into the body-cavity of the
zocecium. ‘The embryo or its parts usually lie in spaces which
probably appear in the surrounding cells of the embryophore,
by a process of vacuolation of the protoplasm. The aperture
of the ovicell is beginning to develop.
The present is a convenient place to raise the question of
the function of the suspensor. It might have been supposed
that it was a tube carrying spermatozoa to the egg, if it had
not been for the fact that the suspensor is formed after the
development of the embryo has commenced. The time at which
one would expect fertilisation to take place is the stage before
the embryonic investments are completed, when the egg is
hardly separated from the fluid of the body-cavity. At this
time spermatozoa are commonly found in the neighbourhood of
the egg or of the embryo in its early stages (cf. fig. 19).
The function of the suspensor appears to me to be probably
connected with the nutrition of theembryo. During Stages
C and D the suspensor remains quiescent; but it probably
contributes to the formation of the mass of protoplasm in which
the secondary embryos are supported from Stage F onwards.
122 SIDNEY F. HARMER.
It may be noted that the embryophore may now contain
ege-like cells (fig. 34, a), which are similar to the giant-cells
that I have described in Crisia. I have never found these
cells showing the slightest evidence of being really eggs;
whereas the dividing primary embryo is quite unmistakable
at this stage. I therefore attribute a purely subsidiary part to
these cells, and I do not believe that they have any direct
share in the development of embryos. I have not found giant-
cells during the earlier stages of the development.
The aperture of the ovicell commences to develop at this
stage, as is distinctly shown by the sections. We have seen
that the fertile brown body of earlier stages is surrounded by
a deeply staining mass of cells which may be in immediate
contact with an invagination of the body-wall (fig. 27). In
this stage the condition closely resembles a stage which I have
described in Crisia (6, pl. xxiii, fig. 12), except that the
latter has no brown body. ‘The tubular aperture of the ovicell
of both Crisia and Lichenopora is closely connected at its
base with the deeply staining mass of cells or its derivatives.
In the latter genus these cells grow towards the distal surface
of the zoccium which contains them, just as in Crisia they
grow towards the distal surface of the ovicell. A cavity next
appears in the mass of cells, immediately above (distal to) the
brown body. This cavity, which is at first merely a series of
vacuoles, is usually perfectly definite near the brown body, and
may extend as a space, which appears crescentic in a longitu-
dinal section, halfway round the brown body. Tracing this
space upwards, it becomes less definite, and it extends into the
base of the tubular aperture of the ovicell. A transverse sec-
tion of the ‘‘ aperture ” at this stage usually shows an external
thin, body-wall which surrounds the part of the body-cavity
which extends into the tube. The centre of the transverse
section is occupied by a solid mass of cells, whose diameter is
about half of that of the entire tube. Lower down the solid
mass becomes excavated by a lumen which becomes continuous
with the cavity which occurs near the brown body. Below
the latter the cavity passes into an irregular space, traversed
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 128
by strands of nucleated protoplasm, which thus divides the
space into smaller cavities in which the derivatives of the
primary embryo lie.
I believe that the processes which take place at this stage
are practically like those which occur in Crisia. The whole
mass of the embryophore becomes highly vacuolated, and is
thereby transformed into a series of more or less definite spaces
in which the secondary embryos will lie. These spaces become
more definite near the aperture of the ovicell. In the next
stage I have often observed mature embryos in the tubular
aperture, obviously on their way out to the exterior.
Very few of the colonies which I have observed in this stage
have a smaller transverse diameter (in sections) than ‘60 mm.
The extreme measurements I have noted are ‘45 mm. and
1:04 mm. The average of the measurements of the transverse
diameter of twenty-four colonies in this stage is ‘747 mm.
There is great probability that Stage F corresponds with a
condition of the entire colony in which the fertile zocecium is
occluded, and the very beginning of the formation of the roof
of the ovicell is taking place.
Stage G.—Ovicell well developed.
The Production of the First Brood of Secondary
Embryos is at its Height.
Fig. 11, which shows a characteristic horizontal section of
a colony in this stage, has already been described (p. 94).
The fertile brown body still remains compact and conspicuous ;
and it occurs in the immediate neighbourhood of the trumpet-
shaped aperture of the ovicell. During this stage it is, how-
ever, no longer found in the fertile zocecium, which in the
section drawn (fig. 11, z?) is seen to contain some secondary
embryos at its upper end. ‘The cavity of this zoccium has
become continuous with that of the ovicell in the manner pre-
viously described (p. 87), and the brown body has passed out
into the cavity of the ovicell proper. The brown body still
forms a kind of centre from which the lobes of the embryo-
124, SIDNEY F. HARMER.
phore start ; and its position at the base of the aperture of the
ovicell materially aids in the discovery of that structure in the
series of sections.
The secondary embryos do not lie freely in the cavity of the
ovicell, but are still contained in a protoplasmic reticulum,
which is the modified embryophore. The latter has largely
increased in size with the commencement of embryonic fission,
and has become branched, its lobes extending round the zoecia,
which pass through the cavity of the ovicell (cf. fig. 6). The
general appearance of a young embryophore in this stage in
an entire colony is shown in fig. 12.
The most striking difference between the development at
this stage and that of Crisia consists in the absence of the
primary embryo of the latter, and it often requires close exami-
nation to prove that fission is still proceeding in Stage G.
Such, however, is undoubtedly the case, although there is no
coherent primary embryo left at this period.
Fig. 37 illustrates the manner in which the number of second-
ary embryos is increased during Stage G. An embryo similar
to the smallest ones seen in fig. 11 is dividing transversely.
The external and internal layers, which are so characteristic a
feature of the Cyclostome embryo, are now well marked, and
the inner layers of the two halves of the dividing embryo are
completely separated from one another. The longest diameter
of the embryo figured is 56 pn.
In some colonies it is easy to demonstrate the occurrence of
this process, and there are many cases where its recent occur-
rence may be inferred from the fact of two small embryos lying
close together in a position which suggests that they have
recently separated from one another. It is obvious that if the
embryo shown in fig. 37 had been cut transversely to its longest
diameter it would have been by no means easy to be sure that
fission was taking place. Another difficulty in proving the
occurrence of this process is due to the fact that the normal
development of a secondary embryo which is not going to
divide further may, in section, appear very similar to the divid-
ing embryos. The young definitive secondary embryo, which
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 125
is destined to form a single larva, is a small ovoid mass of cells
consisting of a well-defined external layer and an equally
clearly defined inner layer. One wall of the embryo becomes
invaginated to form the larval “sucker ”’ (cf. 6, pl. xxiv, figs.
22,23). This process is seen to be taking place in some of
the embryos shown in fig. 11.
It is not always easy to distinguish between cases in which
the sucker is being invaginated and cases of embryonic fission.
One difference may, however, be pointed out between the two
processes. The invagination of the sucker never results in
the division of the inner layer of cells into two separate por-
tions, nor does this layer even appear divided in a section of
the embryo. Fission, on the contrary, results in the complete
separation of the inner layer into two halves, one of which
belongs to each of the products of division. In some cases the
ovicell contains larger and more irregular masses of embryonic
tissue. These probably give rise to more than two secondary
embryos.
I have satisfied myself of the normal occurrence of the
process of fission, as described above, in a number of colonies ;
and I believe that the process continues, in healthy colonies,
for a considerable time. I have not made preparations of
colonies of all the larger sizes; and I am not able to say what
is the upper limit of size reached during this stage. The
largest colony among my sections, in which the embryonic
structures are in Stage G, measures 1°60 mm. in diameter,
and the smallest measures ‘80 mm. I have, however, found
mature larve in somewhat abnormal colonies which measure
only ‘56 mm. in diameter. The average diameter of fifteen
colonies measured is ‘93 mm.
Although the fertile brown body passes into the cavity of
the ovicell during this stage, it is still possible to obtain satis-
factory evidence of the identity of the fertile zocecium in well-
orientated horizontal sections. If an ordinary zoccium loses
its polypide without developing a new bud to take its place,
the previous existence of the polypide is shown by the occur-
rence of a brown body in the zocecium. There is no way by
126 SIDNEY F. HARMER.
which the brown body can be got rid of except through the
agency of a new polypide. In the fertile zocecium, on the
contrary, the brown body passes upwards to the neighbourhood
of the orifice (beginning of Stage D), and the basal end of
the zocecium is left completely empty. In examining the base
of colonies in the later stages of embryonic development, one
of the oldest zocecia is usually found to be empty; and it may
safely be inferred, in most cases, that this was the fertile
zocecium. z! commonly possesses a functional polypide even
during Stage G; and either z? or z3, as the case may be, is
usually in the same condition. The other one of these two is,
however, usually empty. In some cases the lower half of the
fertile zocecium is seen to be cut off from the upper by a
transverse septum, as described in Stage E (see p. 108). It is
probable that this is the normal arrangement, and that the
function of the septum is to restrict the embryos to that part
of the fertile zocecium which is on a level with the rest of the
ovicell.
The active production of secondary embryos during this
stage seems to have a well-marked effect on two sets of struc-
tures in the rest of the colony; namely, on the testes and on
the brown bodies.
Colonies which are in the earlier stages (A—C) have testes
in some of their zocecia in the great majority of cases. There
can be no doubt that the colony of L. verrucaria is ordinarily
hermaphrodite ; and, as we have seen, a testis may or may not
be present, with the embryo, in the fertile zocecium. Prouho
(18) has recorded some extremely interesting observations on
the succession of the polypides in Alcyonidium duplex;
in which a polypide which produces spermatozoa degenerates
into the condition of a brown body, and is succeeded, as a
normal part of the life-history, by a polypide which produces
an ovary. It is quite possible that phenomena of an analogous
nature may occur in Lichenopora, although the details are
obviously different. If continuous observations of the zocecia
of a young living colony could be made from day to day, it
might be possible to show that the apparent irregularity in the
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 127
production of testes did really follow some definite law. The
only general statements that I can make are to the effect that
testes may be present during any of Stages A—G; but that
while they are normally present in the younger stages they
are much less common in the older ones. Thus of nineteen
colonies in Stage F, in which I have noted this particular
point, seven had well-developed testes, while twelve had none
at all, or only slight remains. In some cases in which no
functional testes occur in the colony, the former presence of
these bodies can be proved by the presence of their degenerat-
ing remains in the recently formed brown bodies. Of ten
colonies in Stage G, in which I have noted the presence or
absence of testes, only one had well-developed male organs;
while in nine none were present, or at most the last remains
of testes which had previously been active.
This seems to indicate that the energies of the colony are
given as much as possible to the production of embryos during
Stage G, and the development of spermatozoa is meanwhile
suppressed.
It has already been pointed out that the healthy development
of the embryos is dependent on the functional activity of the
polypides. Were these to degenerate to any great extent
during Stage G, it is probable that the development of the
embryos would be interfered with. This is borne out by the
examination of various colonies in which degeneration is
actually taking place, and even more strikingly by the character
of certain colonies in which the embryonic development has
clearly been modified by conditions of insufficient nutrition.
These are colonies in which all or most of the polypides have
degenerated. The ovicell should be in Stage G, and it has
actually succeeded in producing perhaps three to six mature
embryos. But instead of finding the ovicell crammed with
embryos of all ages intermediate between these and the youngest
secondary embryos, the ovicell contains nothing in addition to
the mature embryos except a very few small fragments of
embryonic tissue. It may be concluded that these are not
colonies in which the production of secondary embryos is
128 SIDNEY F. HARMER.
approaching an end, from the fact that the ovicell is small and
the embryophore is compact and unbranched.
Some colonies in Stage G which have no polypides, or only
one or two polypides, appear quite healthy. It is probable
that in these cases the production of embryos would have soon
been injuriously affected, as a single polypide is probably
insufficient for the nutrition of a whole ovicell full of embryos.
This probability is increased when it is noticed, as is commonly
the case, that many of the remaining polypides have quite
recently degenerated ; showing that the colony has had a full
complement of polypides up to a very recent period.
The interdependence of the development of the embryos and
the nutritive conditions of the whole colony is, however, no-
where more clearly shown than in the very remarkable fact
that a large proportion of the older colonies have no brown
bodies at all. A brown body is of course present in a zocecium
which has lost its polypide altogether, without developing a
bud to replace it. But most of the zoccia in a healthy colony
—during Stage G, for instance—contain a polypide without
any brown body. In the earlier stages, where the mass of
embryonic tissue present in the colony is very small, polypides
can degenerate, and new buds can develop to take their place
without materially affecting the health of the colony. And as
a matter of fact we do find that most young colonies contain
polypides, degenerating polypides, fully formed brown bodies,
and young polypide-buds, which are developing to take the
place of the old polypides.
Were any extensive degeneration of the polypides to take
place in any colony containing numerous embryos, it is
probable that the colony could no longer bear the strain of
producing its embryos. In a large proportion of the colonies
in Stage G, there is no trace of the complete degeneration of
polypides, there are no brown bodies, and there are no polypide
buds, either in new zoccia or in old ones. Even the increase
of the size of the colony, by the formation of new zocecia, thus
appears to be retarded during the height of the formation of
secondary embryos.
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 129
I have formerly brought forward certain evidence (7) tending
to show that the formation of brown bodies is, to a large extent,
a kind of excretory process, and it would be surprising if
Lichenopora were altogether exempt from the necessity of
forming brown bodies, whatever their function may be. The °
explanation of the apparent anomaly of the absence of these
structures appears to me to be that parts of the alimentary
canal degenerate during the functional activity of the poly-
pide. A newly formed polypide has the inner border of the
epithelial cells of its alimentary canal clear and sharply defined.
Parts of the alimentary canal of old polypides are in the same
condition, but certain parts of the cecum of the stomach are
in a very different state. The cecum is coloured during life
with the pigment which is so characteristic of parts of the
alimentary canal in Polyzoa. In sections these parts contain
a very small quantity of protoplasm, and consist principally of
yellowish granules which are very similar to the granules which
largely compose the brown bodies—that is to say, there is no
very great difference between the latter and those parts of the
alimentary canal which are pigmented during life. I believe
that this implies that excretory substaces accumulate in the
epithelium of the alimentary canal as the polypide grows older.
In the case of the younger colonies the entire polypide usually
atrophies in course of time ; but in the case of older colonies,
in which it is of the utmost importance that every zocecium
should be contributing its share towards the nutrition of the
ovicell, the excretory granules are discharged into the stomach
continuously during the life of the polypide.
The facts are that the alimentary canal of these poly pides
usually contains solid substances which resemble fragments of
brown bodies—that certain parts of the epithelium have no
definite inner limit, but pass off into a cloud of granules which
extend into the alimentary canal, and that the polypides are
usually distinctly larger than those of young colonies. It is
not always easy to get polypides (in sections) which are well
orientated for the purpose of measurement ; but the measure-
ments I have been able to obtain support the impression one
VOL, 39, PART 1.—NEW SER. I
180 SIDNEY F. HARMER.
receives from the inspection of colonies of different ages, that
polypides in old zocecia, which have no brown bodies, are dis-
tinctly larger than those commonly met with in younger colo-
nies in which brown bodies are present in the same zoccia.
‘ The significance of the larger size of these older polypides is,
as I take it, that they have gone on growing after the time
when degeneration would have taken place in a younger colony.
The difference in the size of the polypides, at different ages of
the colony, becomes very apparent in comparing the size of
the brown bodies formed from their degeneration. The
diameter of the fertile brown body in young colonies is about
‘05 mm.; that of a fertile brown body belonging to a later
generation (fig. 36), and formed in colonies in which large
polypides are present, may be as much as ‘09 or ‘10 mm.
It is not always easy to decide, in a given case, whether the
appearance of the alimentary canal implies the passage of a
regular brown body into the canal or the partial degeneration
of part of the epithelium. I have no doubt that the old brown
bodies may be removed by passing into the alimentary canal of
the newly formed polypide, as in Flustra and some other
Polyzoa. Although one or even two brown bodies may occur
in a single zoecium of Lichenopora verrucaria, I do not
think I have ever found more than two; and in most cases
where a polypide has recently degenerated, one does not find
an older brown body as well, even though the age of the colony
makes it certain that there has been an older brown body. In
some of these cases it is probable that the old brown body
fuses with the degenerating polypide; but in others it may be
that the old brown body has passed into the alimentary canal
before the degeneration of the polypide again takes place.
Some of my sections indicate the occurrence of this process,
and the brown body seems to become attached to the wall of
the stomach, and finally to pass into it. Nuclear structures
and other parts of the brown body may in these cases occur
freely in the lumen of the alimentary canal. Even without
actual evidence of this kind it would almost be necessary to
assume that this takes place; for in no other way do I see any
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 131
explanation of the total absence of brown bodies in old zocecia.
In L. hispida the alimentary canal probably does not absorb
the brown body, since a considerable number of old brown
bodies occur in the proximal parts of the zoccia (cf. p. 97),
The old polypides are not merely different in size from the
young ones, but their several parts may be very differently
arranged, during the retracted condition at least. The polypide
of young colonies is like that of many other Polyzoa,—e. g.
that of Alecyonidium as figured by Prouho (18, pl. xxiii,
fig. 3). The short pharynx lies in the same line as the ten-
tacles, and from its lower end a long cesophagus passes upwards
(parallel with the tentacles) to open into the stomach (ef. fig.19).
The intestine and rectum continue the direction of the ceso-
phagus, while the caecum hangs down into the body-cavity
parallel with the cesophagus. The apex of the cecum is formed
of more protoplasmic cells, but its sides contain the yellow
granules which have already been mentioned. In the poly-
pides of old colonies the whole of the cecum is much swollen.
Its apex has not altered its place, as is shown by the position
of the group of protoplasmic cells. The upper end of the
stomach or the region of the intestine is, however, commonly
bent, so that the intestine and rectum may be directed down-
wards (away from the orifice of the zocecium). The walls of
the cecum are at the same time greatly swollen and very
granular. The lumen may be in places quite obliterated; and
in any case, parts of the wall have no sharp inner boundary.
The epithelium here shades off quite gradually into the granular
contents of the stomach. This might be ascribed to defective
preservation, were it not for the fact that other parts of the
same sections are well preserved. But since, in many cases,
no polypide-buds are being formed, and since none of the
zocecia are losing their polypides, I believe that this is an
arrangement by which none of the zocecia are rendered in-
efficient for nutritive purposes during the time at which there
is the greatest strain on the energies of the colony.
The phenomenon which has just been described appears to
be characteristic of old colonies (Stage G and later stages), but
132 SIDNEY F. HARMER.
it is by no means restricted to these. Brown bodies are absent
in a considerable number of colonies in Stage F, and rarely in
earlier stages, sometimes occurring even in quite young
colonies.
Whatever the fate of other brown bodies, the fertile brown
body always occurs in a normal colony. I have not examined
many colonies larger than 15 mm. in diameter, and I cannot
say how long this structure remains compact. In one colony
I found a fertile brown body which had almost divided into
two, and I have also noticed a brown body of the second
generation of embryos divided into a considerable number of
fragments which are contained in the embryophore. In all
the colonies I have examined in Stage G the fertile brown body
is still compact.
The Occurrence of Second Broods of Embryos.
My observations on this part of the development are very
incomplete, but I have noticed one or two facts of interest.
There is no doubt that new broods do make their appearance
in old colonies, and this may possibly indicate a second year’s
growth (cf. p. 89). That we are not here dealing with the
first brood is shown by the large size of the colony, and by the
fact that it possesses a large ovicell, which, however, contains
no embryos.
I have not traced the first origin of the new brood of embryos.
Fig. 36 is from a colony 2°2 mm. in diameter. No other pri-
mary embryo was found. Most of the polypides had enormous
testes, and a few had egg-cells in a corresponding position
near the apex of the cecum. The embryo is carried by an
embryophore supported by a fertile brown body, and corre-
sponds in general with Stage D or E of the first brood. Ihave
not been able to obtain a stage intermediate between the
occurrence of eggs at the lower end of a polypide and the
section figured. It does not appear to me probable that a
normal suspensor is formed. The cells which surround the
embryo are exceedingly loose and scanty, and the part of the
embryophore between the brown body and the embryo is short
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 183
and has but few nuclei. It seems to be growing down from
the brown body to enclose the embryo. The very large size
of the brown body is at once apparent, and judging by the
condition of the other polypides of the colony it may be inferred
that it has been formed by the degeneration of a large polypide
(cf. p. 130). The investment of the brown body is well deve-
loped, highly protoplasmic, and full of nuclei.
Another old colony in which I have found the same pro-
cesses taking place measured 3°4 mm. Its ovicell was well
developed, and the polypides and testes were large. The ordi-
nary zocecia were not accompanied by brown bodies, and no
eggs were found. In this case I discovered no less than eleven
primary embryvus, each of which was contained in an embryo-
phore supported by a large fertile brown body. ‘The evidence
given by these cases is quite concordant. ‘The brown body is
as characteristic a feature of the second generation as it was
of the first, and its presence indicates that the primary embryo
is developed in relation with a polypide. Moreover the space
in which the embryophore of fig. 36 lies is clearly a zocecium.
One of the eleven embryos of the larger colony has already
given rise to a good many secondary embryos, and its brown
body is broken up into fragments. The other ten have a com-
pact brown body, and are in a condition which exactly corre-
sponds with Stage F of the first brood. In all these primary
embryos, fission has commenced and is going on vigorously. In
one case the embryophore is beginning to grow out of the
space in which it was produced into the general cavity of the
ovicell.
The later development of the primary embryos of the second
brood is thus identical with that of the first brood, but it is
probable that the details of the early development are some-
what different. One of the noteworthy features of the case
is the large number of zocecia which may become fertile in a
single colony.
I have unfortunately no other observations on this interesting
part of the life-history of L. verrucaria. From the great
clearness of the embryo in all the cases observed, it is quite
134 SIDNEY F. HARMER.
possible that the earlier processes of the development would be
more easily made out in the second brood than in the first
brood.
General Conclusions.
The only genus with which Lichenopora can be compared,
as regards the embryonic development, is Crisia. In no
other case have the early processes been described at all.
In Crisia certain zocecia take on the character of ovicells.
A polypide-bud makes its appearance in the young ovicell
while the latter is at the growing-point of the branch, and is
not externally marked out in any way as an ovicell. The bud
becomes connected with an egg, and develops a tentacle-sheath
which shows it to be a modified polypide. The greater part of
the bud, however, forms what may now, on the analogy of
Lichenopora, be termed an embryophore ; and in this struc-
ture the embryo develops and undergoes its fission. There is
no fertile brown body; and the ovicell is recognisable exter-
nally as a structure which looks like a zocecium that has been
inflated.
While the ovicell of Crisia is thus clearly a modified
zocecium, which develops a polypide-bud in the ordinary way
at the growing-point, the morphology of the ovicell of Lichen-
Opora is more obscure. There is no doubt that the embryo is
formed in a zocecium; but it is formed in a fully developed
zocecium which probably is normally tenanted by two succes-
sive polypides. Each of these becomes a brown body in turn,
the second brown body fusing with the first to form the charac-
teristic fertile brown body. The exact origin of the embryo-
phore is not as clear as it was in Crisia; but there is nothing
to show that it is morphologically a bud. The later history is,
however, not dissimilar. The embryophore becomes vacuolated,
and in the vacuoles so formed the secondary embryos are later
found. Certain differences in the details of the embryonic
fission have already been pointed out. The fertile zocecium
becomes occluded, and its cavity then becomes continuous with
a series of alveoli or interzocecial spaces, which occur between
ON THE DEVELOPMENT OF LICHENOPORA VERRUOARIA. 135
the upper ends of the zowcia. The ovicell is increased in size
by the addition of fresh alveoli.
What, then, is the morphology of the ovicell? Are we to
regard it as formed by the fusion of a number of zoccia, repre-
sented by the constituent alveoli? Or, is the ovicell a highly
modified zocecium, or two zocecia in cases where two zocecia are
originally fertile? Or, lastly, is no direct comparison possible
between Crisia and Lichenopora in respect of their
ovicells? The last question may probably be dismissed with
the reply that the fundamental process of embryonic fission,
and the character of the embryophore, are so similar in the
two cases that the ovicells must be regarded in general as
homologous structures. An objection to considering the alveoli
to be suppressed zocecia is to be found in the fact that the
zocecia reach the basal lamina of the colony, and the alveoli
do not. This has previously been correctly pointed out by
Smitt (20). The alveoli do not, moreover, develop a polypide
bud at any period.
We must not lose sight of the fact that Crisia and Licheno-
pora are two very widely separated genera. It would not,
indeed, be easy to choose any other pair of recent Cyclostomes
which would be less nearly related toone another. More light
will probably be thrown on the morphology of the ovicell of
Lichenopora by the examination of genera which are less
different from Crisia.
I have obviously been fortunate in being able to study the
ovicells of L. verrucaria. Waters has more than once
(24, 26) alluded to the importance of examining the ovicells
of this genus, and has stated that in some species “large
numbers of specimens may be examined without any ovicell
being found.” Whether this is due to the absence of the
ovicells in these cases, or merely to the difficulty of recognising
them, I have no means of deciding. But it must be expressly
noted that Waters states (24, p. 261), in describing L. nove-
zelandiz, Busk, that ‘‘ where there is no ovicell the zocecial
tubes run into the centre, the central depression forming an
inverted cone without cancelli;”’ and a similar statement is
136 SIDNEY F. HARMER.
made on the next page for L. grignonensis, Busk. Waters
unfortunately does not figure any of these specimens with a
depressed centre and without cancelli; but these features
certainly point to the conclusion that no ovicell is present.
Waters’ statements are in strong contrast with my own results
on L, verrucaria, in which the ovicell is potentially present,
as shown by the alterations in the fertile zoccium in most
colonies which consist of no more than four or five, or even
three zocecia ; and it is actually present and recognisable from
the outside in all colonies which have reached a slightly later
Stage.
It would no doubt be going too far to assume that what is
true for L. verrucaria necessarily holds for other species of
Lichenopora. I am not at present able to make an inde-
pendent examination of other species ; and although the ovicells
of some of them have been described by Waters (24, 26) and
others, the descriptions refer almost entirely to the external
form, and I can find practically no points of comparison which
assist in forming a conclusion as to the morphology of the
ovicell. Waters in one place (25, p. 277) alludes to a feature
which may have some relation with some of the phenomena
described above in L. verrucaria. Speaking of Hornera
(“Idmonea”’) fissurata, Busk, Waters states that he could
not discover any opening to the ovicell, but that one of the
lateral zocecia was much larger than the others, and it appeared
that this change was “connected with the functions of the
ovicell.” In each of the ovicells subsequently examined by
Mr. Kirkpatrick one or two zowcia in proximity to the ovicell
were enlarged and altered in direction. It would be interest-
ing to know whether the enlarged zocecium in this case had
any resemblance to the fertile zocecium in L. verrucaria.
The ovicell seems to be practically identical in all species of
Lichenopora in which it has been described. It always
forms an inflated area between the upstanding zocecia, and it
is usually known to occupy the centre of the colony. In some
species it is extremely difficult to recognise the ovicell at all.
I cannot help suspecting that the ovicell normally develops at
hy
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 137
an early stage in most, if not all, of the species of Licheno-
pora, and that the presence of cancelli between the rows of
zocecia is an indication that the ovicell is developing, even
though no other evidence is obtained of its presence.
There are, however, other genera of Cyclostomata in which
the ovicell is of a simpler character than in Lichenopora.
Thus in Discotubigera lineata, MacG., as figured by
Waters (26, pl. xv, fig. 5), the ovicell is a lobed structure
which extends over a comparatively small area of the discoidal
colony near its margin. It opens by a single aperture to the
exterior. A similar form of ovicell occurs in the genus Diasto-
pora. It appears to me that these ovicells resemble those of
Liche nopora reduced to simpler terms, and they are, perhaps,
to some extent intermediate between the two genera which I
have specially studied. Taking into account the origin of the
ovicell in Lichenopora from an ordinary zocecium, the fact
that the ovicell of Crisia is demonstrably a modified zocecium,
and the existence of intermediates between these two extremes,
I venture to suggest that that of Lichenopora is also to be
regarded as formed by the modification of a zocecium. The
fertile zocecium is prevented by the position which it occupies
near the centre of the full-grown colony from expanding at its
basal end. Expansion of some kind is necessary, in order to
provide room for the swarm of secondary embryos ; andif there
is anything in this suggestion, it may be supposed that the
significance of the process is that the fertile zocecium has
become swollen at its upper end. Instead of growing out as
a free appendage of the colony, the swollen part has applied
itself to the upper surfaces of the zocecia, and has filled up all
the spaces intervening between the zocecia. The result is the
same as if the ovicell had become amceboid, its pseudopodia
extending between the zocecia, embracing them, and anasto-
mosing on the further side of each. In the cases where two
or more zocecia become fertile, the ovicell may be regarded as
being composed of as many original zoovcia.
The actual formation of the ovicell, by the formation of the
alveoli and their subsequent fusion with one another, is of
=
188 SIDNEY F. HARMER.
course somewhat different from this hypothetical process; but
I claim to have established the facts (1) that the processes
which precede the formation of the ovicell take place in a
fertile zocecium, and (2) that the development of the cal-
careous roof of the ovicell starts from the region of the orifice
of the fertile zocecium.
It remains to be seen whether the ovicells of such genera
as Hornera, Diastopora, Tubulipora, and Idmonea fit
in with this idea ; and I hope to have an opportunity of making
a more extended study of this question.
Next to the demonstration of the occurrence of embryonic
fission, a process which had been previously proved for Crisia,
the most striking fact which I have succeeded in establishing
is the restriction of the production of an embryo to one or two
of the oldest zocecia in the normal development.
Even the second brood of embryos is really dependent on
this function of the fertile zocecium, or zoccia, of the first
generation ; for they are nourished in the ovicell which was
developed in connection with the primary fertile zoccia. In
the great majority of colonial animals a large number or all
of the individuals of the colony become fertile. The phe-
nomena which occur in Lichenopora are perhaps due to the
fact that the colony in this genus is to be regarded as an
individual of a higher order. The discoidal form and the close
association of the zocecia have produced the result that the
colony behaves in this respect as if it were a single individual.
The production of fertile eggs is thus limited to one or two
individuals, and in fact to one or two of the zoccia which are
first formed as buds from the primary individual of the colony.
In the case of Crisia, where the association of the different
zocecia is much less intimate, a considerable number of the
individuals of a colony may become fertile. Here, however,
there is something analogous in the behaviour of the inter-
nodes, each of which in the great majority of species normally
produces no more than a single ovicell.
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 139
REFERENCES.
. Attman, G. J.—‘‘A Monograph of the Fresh-water Polyzoa,” 4to, Ray
Society, 1856.
. Barrots, J.—‘‘ Recherches sur |’Embryologie des Bryozoaires,” 4to,
Lille, 1877.
. Barros, J—‘ Mém. sur la Métamorphose de quelques Bryozoaires,”
‘Ann. Sci. Nat. Zool.,’ ser. 7, vol. i, 1886, No. 1.
. Bram, F.—* Unt. iib. d. Bryozven des stissen Wassers,” Leuckart and
Chun’s ‘ Bibl. Zool.,’ Heft 6, 1890.
. Harmer, 8. F.—“On the British Species of Crisia,” ‘Quart. Journ.
Micr. Sci.,’ vol. xxxii, 1891, p. 127.
. Harmer, 8S. F.—“On the Occurrence of Embryonic Fission in Cyclo-
stomatous Polyzoa,” ibid., vol. xxxiv, 1893, p. 199.
. Harmer, 8. F.—‘‘ On the Nature of the Excretory Processes in Marine
Polyzoa,” ibid., vol. xxxiii, 1892, p. 123.
. Harmer, 8. F.—“ Preliminary Note on Embryonic Fission in Licheno-
pora,” ‘Proc. Roy. Soc.,’ vol. lvii, 1895, p. 188. ‘On the Develop-
ment of Lichenopora verrucaria, Fabr.,” ibid., vol. lix, p. 73.
. Hincxs, T.—‘ British Marine Polyzoa,’ 8vo, London, 1880.
. Krazpetin, K.-—“ Die Deutschen Siisswasser-Bryozoen,” part J, ‘ Abh.
Ver. Hamburg,’ vol. x, 1887.
. Krarrewin, K.—Ibid., part 2. ‘ Abh. Ver. Hamburg,’ vol, xii, 1893.
. Levinsen, G. M. R.—“ Mosdyr,” from ‘ Zoologia Danica,” 4to, Copen-
hagen, 1894.
. Merscunixorr, E.— Beit. zur Entwick. ein. niederen Thiere,” § Bull.
Ac. St. Pétersbourg,’ vol. xv, 1871, p. 502.
. Ostroumorr, A.—‘ Remarques relatives aux recherches de Mr. Vigelius
sur des Bryozoaires,” ‘ Zool, Anz.,’ vill, 1885, p. 290.
. Ostroumorr, A.— Contribution a1l’étude Zool. et Morphol. des Bryozo-
aires du Golfe de Sébastopol,’ ‘Arch. Slaves de Biol.,’ vol. ii, 1886,
p. 8.
. Ostroumorr, A.— Zur Entwicklungsgeschichte d. Cyclostomen See-
bryozoen,” ‘ Mitt. Zool. Stat. Neapel,’ vol. vii, 1887, p. 177.
. Percens, E.— Unt. an Seebryozoen,” ‘Zool. Anz.,’ xii, 1889, p. 504.
. Prouno, H.—* Cont. 4 )’Histoire des Bryozoaires,” ‘Arch. Zool. Exp.
et Gén.,” ser. 2, vol. x, 1892, p. 557.
. Riptey, 8. O.— Polyzoa, Coelenterata, and Sponges of Franz-Joseph
Land,” ‘ Ann. Mag. Nat. Hist.,’ ser. 5, vol. vii, 1881, p. 442.
140 SIDNEY F. HARMER.
20. Surrt, F. A.—“ Krit. forteckning ofver Skandinaviens Hafs-Bryozoer,”
‘Ofv. K. Vet.-Ak. Forhandl.,’ 1866, p. 395.
21. Verworn, M.— Beit. z. Kenntnis d. Siisswasserbryozoen,” ‘Zeit. f.
wiss. Zool.,’ xlvi, 1888, p. 99.
22. VicELius, W. J.—*‘ Cont. ala Morphologie des Bryozoaires Ectoproctes,”
‘Tijdschr. d. Ned. Dierk. Ver.’ (2), i, Afl. 3 and 4, 1886.
23. VicrLtius, W. J.—‘ Zur Morphologie der marinen Bryozoen,” ‘ Zool.
Anz.,’ x, 1887, p. 237.
24, Waters, A. W.— Bryozoa from New South Wales, North Australia,
&c. Part 3,” ‘Ann. Mag. Nat. Hist.,’ ser. 5, vol. xx, 1887, p. 253.
25. Waters, A. W.—‘*On some Ovicells of Cyclostomatous Bryozoa,”
‘Journ. Linn. Soc. Zool.,’ vol. xx, 1890, p. 275.
26. Waters, A. W.—“On the Ovicells of some Lichenopore,” ibid., p. 280.
DESCRIPTION OF PLATES 7—10,
Illustrating Mr. Sidney F. Harmer’s paper “ On the Develop-
ment of Lichenopora verrucaria, Fabr.”
All the figures refer to Lichenopora verrucaria, Fabr. Slight varia-
tions in the position of the camera lucida are responsible for the fact that they
are not absolutely drawn to scale. All the figures on Plate 7, and figs. 12,
13, on Plate 8, were drawn with Zeiss A. Figs. 11, 14, and 17 were drawn
with Zeiss C, and the others with Zeiss F.
PLATE 7.
Fic. 1.—Two views of a very young left-handed colony (diameter ‘18 mm.).
1, 2, and 8 are the first three zowcia.
Fic. 2.—Somewhat older left-handed colony (diameter ‘30 mm.) seen from
above.
Fic. 3.—Left-handed colony with four zoccia (1, 2, 3, 4). (Diameter
‘48 mm.; extreme length, ‘64 mm.)
Fic. 4.—Right-handed colony with seven complete zoecia. The proximal
lip of the colony is beginning to grow towards the disc (broken), and the
commencement of two new zoccia can be seen in the space between the first
three zoccia (1, 2, and 3). (Diameter °59 mm.)
Fic. 5.—Colony in which the roof of the ovicell is developing. The colony
is probably left-handed, with occluded z*. The aperture of the ovicell is not
ON THE DEVELOPMENT OF LICHENOPORA VERRUCARIA. 141
yet completely formed. The proximal edge of the basal lamina has nearly
reached the disc. (Diameter, 1:18 mm.)
Fic. 6.—Thick radial section of part of an old colony, showing the relations
of the ovicell to the zocecia. (Diameter of the right-hand zoccium, outside
the ovicell, = ‘192 mm.)
Fie. 7.—A fully developed discoidal colony (diameter 1°92 mm.). The ovi-
cell is obviously composed of a number of alveoli. The colony is probably
left-handed, with occluded z?, some indication of which can probably be seen
just beneath the single aperture of the ovicell. The zocecium to the left of
the latter and above it is probably z?.
Fic. 8.—Two zoecia from a very old colony, with secondary thickening.
The porous roof of the ovicell is seen at the bottom of a set of honeycomb-
like spaces, the so-called “ cancelli;”? some of these are covering the bases of
the zocecia, and in the left-hand zocecium these are entirely or partially roofed.
Two of the blister-like swellings thus formed have part of their roof still
uncalcified. (Diameter of zoccia ‘14 mm.)
Fig. 9.—From another old colony. The alveoli are still distinctly visible,
and the cancelli are commencing in the interalveolar grooves, or at the base
of the zocecia. The left-hand zocecium is closed by a porous calcareous cap.
(Diameter of zocecia ‘13 mm.)
PLATE 8.
(N.B.—The measurements given below [except for figs. 12, 13] were made
from the sections.)
Fic. 10.—Part of a longitudinal section through the fertile zocecium (? z3)
of a colony, ‘67 mm. in length, with about nine fully-formed zoccia. The
elongated structure to the right of the fertile brown body is the embryophore.
Fic. 11.—Section (diameter ‘80 mm.) of a colony with fully formed ovicell.
The section is parallel to the flat surface of the plano-convex colony, and cuts
the convex side near its upper surface (cf. fig. 6). The only zocecia seen in
the section are thus the more centrally placed ones. z? was presumably the
fertile zocecium, as it is empty basally, and contains secondary embryos, at its
upper end. The fertile brown body has passed into the ovicell. z? also con-
tains secondary embryos, but has a brown body basally, and was probably
occluded without being fertile. The secondary embryos are contained in the
“ embryophore,”’ whose lobes extend, in the cavity of the ovicell, between the
zocecia. The zocecium A contained a brown body lower down.
Fig. 12.—Entire colony (diameter 1°41 mm.), stained and mounted in
Canada balsam. The shaded part shows how much of the ovicell is com-
pletely roofed. The alveoli which are still unroofed are not specially indicated.
The ovicell has an aperture, at the base of which is the fertile brown body.
142 SIDNEY F. HARMER.
The embryophore is younger than in fig. 11. Parts of the zocwcia which are
seen either through other zocecia or through the roof of the ovicell are indi-
cated by dotted lines. The colony is left-handed. z* is occluded (and was
obviously fertile), and would hardly have been visible at all in a dry
preparation.
Fic. 13.—Occlusion of the fertile zocecium ina left-handed colony (diameter
‘88 mm.). The roofing of the ovicell has only just commenced, as a calcareous
film starting from the outer border of z?, and sweeping inwards in such a way
as to leave z? open to the cavity of the ovicell. Although the figure is from a
dry mount, the remains of the fertile brown body (not shown in the figure)
are visible through the opening into z?.
Fic. 14.—Part of the growing edge of an adult colony, more highly magni-
fied, showing the way in which new zoccia are formed, and representing
various stages in the development of the alveoli. See also pp. 81, 84.
(Figs. 15—35, and Fig. 37, refer to the development of the first brood
of secondary embryos.
PLATE 9.
Fie. 15.—A polypide-bud with an egg. One other bud in the same colony
had an egg in a similar position.
Fic. 16.—From a colony, ‘48 mm. long, consisting of three zoecia. One
or two egg-cells were discovered in z}, as well as in the zocecium (z*) drawn.
Fie. 17.—Horizontal section of a colony, ‘43 mm. long and ‘25 mm. broad,
consisting of three complete zoccia only.
Fic. 18.—Part of the left-hand zocecium (? z°) of the last figure, more
highly magnified. The egg measures 12°8 » (= ‘0128 mm.).
Fie. 19.—From a colony consisting of three mature zocecia. Development
of the egg has commenced, and the follicle is becoming definite. The heads
of mature spermatozoa are seen in the body-cavity.
Fie. 20.—From a colony consisting of three mature zocecia. The follicle
is more developed, and the suspensor has become a definite structure.
(Diameter of the embryo, 11:2 p.)
Fic. 21.—From a colony, ‘40 mm. long and ‘21 mm. broad, consisting of
three mature zocecia. ‘The follicle is cut transversely, and the embryo is in
about the same stage as that of Fig. 19.
Fic. 22.—From a colony consisting of four mature and one half-grown
zocecium. ‘The suspensor is not yet clearly formed.
Fic. 23.—From a colony, ‘53 mm. long and ‘32 mm. broad, consisting of
five mature zocecia. The figure is combined from three sections of the
ocecium, the fertile brown body being projected on to the rest of the drawing.
ON THE DEVELOPMENT OF LICHENOPORA VERRUOARIA. 143
The suspensor and the cells which surround it are closely connected with a
mass of cells, a, which form part of the investment of the brown body.
Fic. 24.—Embryophore from a colony ‘48 mm. long and *45 mm. broad.
(Length of suspensor, 41°6 » ; transverse diameter of embryo, 20°8 yu.)
Fic. 25.—Showing the abnormal development of two embryos within the
same embryophore. From a colony ‘56 mm. long, consisting of five mature
zocecia.
Fic, 26.—Embryo (Stage D) from a colony °56 mm. long and ‘37 mm.
broad, consisting of five zocecia. The outer layer of nuclei is unusually
distinct.
Fic. 27.—From a colony ‘69 mm. long and *51 mm. broad, consisting of
about twelve mature zoccia. The fertile polypide has degenerated, and the
brown body has approached the orifice of the zoccium. The embryo is
28 » long, and the length of the entire embryophore with the brown body is
93 p.
PLATE 10.
Fic. 28.—From a younger colony, ‘43 mm. long and ‘37 mm. broad, con-
sisting of six mature zocecia. The brown body has not yet reached the orifice
of the zocecium, and development is consequently not quite so advanced as in
Fig. 27.
Fic. 29.—From a considerably older colony, *56 mm. broad. The upper
end of the embryophore had met the orifice, as in Fig. 27, and evidence was
obtained that the zocecium was occluded by a membrane stretching across the
orifice. Length of the embryo 70 yu.
Fie. 30.—(Stage H.) Obliquely longitudinal section through the embryo-
phore of a colony ‘62 mm. broad. The upper end of the figure was turned
towards the brown body. Diameter of the embryo 30°4 nu.
Fic. 31.—Embryo from a colony *83 mm. broad. The occurrence of the
split-like spaces s, and the arrangement of the inner mass of nuclei indicate
that embryonic fission is about to commence. Longest diameter of embryo
45°6 p.
Fies. 32 and 33 are consecutive transverse sections of the larger portion
of an embryo at the beginning of its fission. From a colony ‘93 mm. broad,
containing two fertile zowcia. The longest diameter of the embryo is ‘68 p.
A smaller piece of the embryonic tissue, not shown in the figures, is already
completely separated from the portion drawn.
Fic. 34,—Transverse section of the older embryophore from a colony
(86 mm. broad) containing two fertile zocwcia. Embryonic fission is actively
proceeding. The diameter of the giant-cell a is 17°6 p, and that of its
nucleus is 9°6 p.
144 SIDNEY F. HARMER.
Fic. 35.—Longitudinal section of the embryophore, with actively dividing
primary embryo. The brown body is not seen in this section, but would have
come at the upper end of the figure. From a colony ‘53 mm. long.
Fic. 36.—Longitudinal section of a young embryophore belonging to a
second or later generation. The brown body is conspicuously larger than that
of the corresponding stage (Fig. 28) of the first generation. The embryophore
lies in a space which is clearly a zoecium. (Diameter of the colony 2°20 mm.
Length of the embryo plus the embryophore 167°2 ». Longest diameter of
the brown body 83°6 w; of the embryo, 86 zp.)
Fie. 37.—Showing the fission of a young secondary embryo, belong-
ing to the first brood. From a colony ‘93 mm. long and ‘88 mm. broad,
similar to that-shown in Fig. 11. The longest measurement of the dividing
embryo is 56 p.
SEP <5 1896
ON NAUTILUS AND SOME OTHER ORGANISMS. 145
Letters from New Guinea on Nautilus and some
other Organisms.
By
Arthur Willey, D.Sc.
Ratu (New Brirary), German New GUINEA,
via SINGAPORE ;
April 5th, 1895.
Dear Proresson Lankester,—Although I have not seen
any mature ova of the pearly Nautilus since my last letter to
you, yet I have made some further observations on the adult,
which may be of interest.
1. Ectoparasites.—In my previous note! I accidentally
omitted to mention the occurrence of numerous Copepod
parasites in the mantle chamber of the Nautilus. They are
present in nearly every individual that I have examined, and
are found attached to the branchix, the internal surface of the
funnel, and in other regions of the pallial chamber.
The parasites are a species of the genus Caligus,’? and
possess the characteristic semilunar suckers on the first pair
of antenne. ;
When Nautili are placed in jars the Caligids emerge in
large numbers from the mantle chamber, and swim about
actively in the water, usually in close proximity to the sides
of the vessels, from which they can be removed only with
some difficulty, owing to the great adhering power of the above-
mentioned suckers.
2. Movements.—I send you a photograph of a Nautilus
in its ordinary swimming attitude,—such a figure, I believe,
1 “Natural Science,’ May, 1894.
2 C. nautili, pending a more detailed description.
VoL. 39, PART 2.—NEW SER. K.
146 ARTHUR WILLEY.
not having been previously published.1. From the photograph
it will be seen that in swimming the animal elevates itself to
such an extent that the eyes are raised above the free margin
of the mouth of the shell. As to its movements on the sur-
face of the water, there is very little to add to the excellent
account given by the late Professor Moseley in his ‘ Notes by
a Naturalist on H.M.S. Challenger.’
I can emphatically confirm Moseley when he says that “ it
is probably a mistake to suppose that it ever comes to the
surface voluntarily to swim about.’ Nevertheless a Nautilus
is not necessarily ruined by being brought up from the bottom
in a fish basket. If liberated within a reasonable time it is
capable of returning to its natural habitat. I have proved this
experimentally.
When freshly captured Nautili are placed in aquaria they
rise to the surface and sink to the bottom with the greatest
facility. The rising to the surface, according to my observa-
tions, is effected solely by the muscular activity of the animal,
and is in no way dependent on any physical modification of
the gas in the air-chambers. ‘The presence of the latter
renders the shell extremely light and buoyant, so that it is,
under normal circumstances, completely under the control of
the muscles of the animal. I say ‘under normal circum-
stances” advisedly, because there is one thing which the
Nautilus cannot do, namely, turn upside down. It is neces-
sary to insist on this, because Moseley gives a translation of
the account given by Rumphius at the beginning of last
century, which, I regret to say, so far as my experience has
gone, is very wide of the mark. Always remembering that
Rumphius was the first to demonstrate the cephalopodous
character of the pearly Nautilus, the suspicion is forced upon
me that he derived his account of its habits from the narrative
of an ill-informed and imaginative Malay fisherman. Natives
are so often remarkably well informed about the habits of
animals, that the above supposition can by no means be taken
as conveying a reproach to the old master.
1 This photograph is reproduced as Fig. 24 at the end of this article.
ON NAUTILUS AND SOME OTHER ORGANISMS. 147
“On the bottom,” says Rumphius, according to Moseley’s
translation, “the animal creeps with the other side! upper-
most, with the head and tentacles on the bottom, and makes
tolerably fast progress.” The only comment I can make on
this statement is that it is inconceivable. I wish'I had the
work of Rumphius to refer to, in order to find out how he
succeeded in seeing the Nautilus at all on the bottom. He
goes on to say, “The animals remain mostly at the bottom,
creeping sometimes into hoop-nets set for fish and lobster-pots ;
but after a storm, when the weather becomes calm, they are
to be seen floating in troops on the surface of the water... .
The floating, however, does not last long, for, drawing in all
their tentacles, the animals turn their boats over and go down
again to the bottom.”
The Nautilus can no more turn its boat over than a suc-
cessful balloon ascent can end by the cage coming down
uppermost. Anyone can convince himself of this by placing
an empty Nautilus shell in water. A remarkably small weight
is sufficient to sink such an empty shell; and when the living
animal retracts itself and ceases all muscular action, thereby
converting itself, as it were, into a dead weight, it is heavy
enough to sink several shells in addition to its own.
There is a slight error in Moseley’s account of the move-
ments of the Nautilus, which may as well be corrected. He
says, ‘On either side of the base of the membranous opercu-
lum-like headfold . . . the fold of the mantle closing the
gill-cavity was to be seen rising and falling, with a regular
pulsating motion, as the animal in breathing took in the
water, to be expelled by the siphon.” It is not a fold of the
mantle which is thus seen to pulsate, but the posterior free
membrane-like expansion of the funnel on either side.
Besides observing the movements of the Nautilus inthe narrow
limits of jars and buckets, I have also placed them in the sea in
shallow water, and will briefly describe one such experiment.
On March 16th a Kanaka brought me six Nautili. All of
them sank to the bottom of the buckets except one, which
1 That is, the ventral side.
148 ARTHUR WILLEY.
swam about on the top all the time. About an hour after-
wards I took the latter out in a canoe to observe its move-
ments in one to three fathoms of water. On placing it in the
sea it remained at first on the surface, swimming actively
backwards, i.e. with the shell directed forwards. It fre-
quently swam to the bottom and back again to the surface—
often also swimming in the middle stratum of water. I have
never found any necessity for framing an elaborate theory as
to the rising and sinking of the shell.
3. The Position of the Spadix or Hectocotylus.—
I was surprised to find that the spadix in the male developed
variably on the right and left sides of the cephalic system.
Out of thirty-seven males which were examined ad hoe,
twenty-three had the spadix on the left side, and fourteen had
it on the right side.
4. Pallial Arteries.—I will say a few words about the
pallial arteries, in order to refer to the arterial blood-supply of
the siphuncular pedicle. Anatomical evidence seems to point
to the conclusion that the latter structure has more of a
vestigial than of an actively functional physiological import-
ance. mbryology will show whether or not it has any original
relation to the primitive shell-gland.
The so-called lesser aorta of Owen, arising from the anterior
slightly incurved margin of the heart, divides immediately, as is
known, into two branches. These are respectively the anterior
and posterior pallial arteries.
The anterior pallial artery bends inwards and somewhat
downwards to the middle line, and then runs forwards below
the skin on the surface of the renal sacs. At the point where
it turns forwards a small intestinal branch is given off. Arrived
at the anterior limit of the renal sacs, the anterior pallial
artery passes into the substance of the mantle, and runs
towards the free margin of the mantle, shortly before reaching
which it bifurcates into two main marginal arteries, from which
numerous radial arteries are given off (see Fig. 1).
Exactly at the point where the anterior pallial artery passes
into the substance of the mantle a pair of branches, right and
ON NAUTILUS AND SOME OTHER ORGANISMS. 149
left, arise from it almost at right angles to it.
These may be
called the branchio-osphradial branches of the anterior pallial
A ee 7706
Nee %Q
__M.a
(ge SS, cay A erage es, WIS ees apa.
i WE ng
= \ |
/ __.boa
\ i — he
4 SS ae
KO. ‘| KO.
Zen. Ry ee es ; i + Bees yey
Se
h bee
Fic. 1—N. pompilius,
9. View of ventral surface of pallial and
visceral regions, to show the course of the pallial arteries. A Funnel.
Free edge of mantle.
7. a. Radial paljial arteries.
artery. a.p.a. Anterior pallial artery.
a. g. Region of nidamental gland.
b. 0. a. Branchio-osphradial artery. s. m. Shell-muscle. 7. 0. Region of
renal organs. /. Heart. p.p.a. Posterior pallial artery. g. Gonad.
Intestinal branch of anterior pallial artery.
Mm. e.
m. a. Marginal pallial
1. Qe
150 ARTHUR WILLEY.
artery, since among their minor ramifications they send up
branches to the tips of the branchiz supplying the integument
of the latter, and also a small branch into each of the osphradia
(i.e. into the osphradia of Lankester and Bourne, and into
those which I described in my last letter). In the female they
also supply the nidamental gland.
The posterior pallial artery runs backwards over the ventral
surface of the heart, leaves the pericardium through the orifice
described by Owen, and then passes onwards below the skin to
the left of the genital gland, between the latter and the gizzard.
Upon reaching approximately the middle point of the posterior
rounded surface of the body it, too, passes into the integument,
and immediately divides into two main branches, right and
left, which supply the dorsal and posterior regions of the
mantle, including the siphuncular pedicle. A variable number
of small secondary or tertiary branches go up into the siphun-
cular pedicle, but there is one branch which is essentially the
siphuncular artery (see Figs. 2 and 3). But even the definite
siphuncular artery is not constant in its origin, but arises now
from the right and now from the left of the two main branches
of the posterior pallial artery. The siphuncular artery is
therefore a minor ramification of the posterior pallial artery.
Owen, followed by Vrolik, described the latter as passing
in toto into the siphuncle. Keferstein, whose figure is more
accurate in this respect, says that it passes ‘‘ nach hinten zur
Korperhaut und besonders zum Sipho.”
Injection indicates that the arteries which supply the
siphuncle are homodynamous with the other ramifications of
the posterior pallial artery.
5. Pallial Veins.—It is not now necessary to go into de-
tails with regard to the pallial veins, but I will point out how
they may be seen to great advantage.
When a Nautilus becomes moribund it usually rises to the
surface, owing to an abundant production of gas in the interior
of the body. If it is allowed to die, and is then removed from
the shell, the veins are found to be injected with gas of some
sort, and the finest ramifications of the veins, in the mantle
ON NAUTILUS AND SOME OTHER ORGANISMS. 151
at least, are displayed with a clearness which could be hardly
attained by artificial injection.
Fic. 2.—N. pompilius, ¢. View of dorsal surface of pallio-visceral
region, to show the principal ramifications of the posterior pallial artery.
m.e. Free mantle edge. ¢. Region of crop. s.p. Siphuncular pedicle.
1. Region of liver. s.a. Siphuncular artery arising from the right of the
two main branches of posterior pallial artery. p.p.a. Posterior pallial
artery. giz. Region of gizzard. g. Region of gonad. 7@ Region of
intestine.
N.B.—In front of siphuncular artery are seen two small branches which
bend backwards and enter the siphuncular pedicle on its dorsal aspect.
The pedicle is here represented turned forwards to expose its ventral
surface.
The mantle is simply riddled by these veins in a manner
which defies one’s powers of draughtsmanship. The veins are
collected into two main trunks, which le on either side of the
anterior pallial artery, and proceed backwards to open into the
afferent branchial vessels. At the sides of the mantle there
are also a number of lateral pallial veins, which open into a
large sinus situated over the shell muscle.
152 ARTHUR WILLEY.
6. Miscellaneous.—In New Hanover and New Ireland
the Nautilus shell is called “ Togol,”’ and is used as a decoration
on the outriggers of the canoes and on the houses, and also as
a drinking vessel and for baling out canoes. Occasionally the
Fic. 8.—N. pompilius, ¢. Similar view as in Fig. B, to show the
siphuncular artery arising from the left of the two main branches of the
posterior pallial artery. Letters as in Fig. B.
shell of Nautilus umbilicatus is obtained, and the new-
comer then acquires the useful information that that is the
male of the Nautilus! This seems to be the idea of the
Kanakas as well as of the outlying white men.
In N ew Hanover, where primitive institutions still largely
flourish undisturbed (including cannibalism), a curious custom
prevails, of which I became aware while engaged in collecting
finger-prints. The natives employ the cheiromantic creases
for the purpose of creating factitious social groups or families,
within the limits of which marriage is tabooed. They trace a
fancied resemblance between the creases of the hand and the
ON NAUTILUS AND SOME OTHER ORGANISMS. 153
form of the wings of a given bird, and name accordingly. All
those whose hand-creases are referable to the same bird con-
stitute one family or “ta-uk,’’ apart from any question of
‘blood relationship. Thus a man of the “kanai ta-uk,” or
family of the sea-gull, cannot marry a woman of that “ ta-uk.”
It is also tabu to eat bird-flesh, but they all eat pork—exactly
the reverse of what occurs in New Britain, at least as far as
the men are concerned.
Probably most of the fighting i1 New Hanover is due to the
complicated relations of the sexes. “ Plenty fight belong
Mary.”
Finally, I may conclude these brief notes by adding that if
a Kanaka kills another Kanaka of lis own ta-uk he does not
eat him, according to my information.—Yours very truly,
Anrruur WILLEY.
Ratum, New Britatn,
via Singapore ;
June 4th, 1895.
Dear Proressor Lankester,—The following are a few zoo-
logical observations, relating chiefly to the fauna of Blanche
Bay, which I have incidentally made during the last few
months.
1. Polyclades.—The Polyclade fauna is rich both in species
and individuals. Most of those that I have hitherto had any
opportunity of observing have been taken close to the shore
of the island of Rakaiya (or Raluan), in Blanche Bay, from
the lower surfaces or in the crevices of volcanic stones.
I have, however, on several occasions also obtained them
from the reef opposite Ralum. Although a fact of no absolute
value proceeding from so limited an experience, it is neverthe-
less worth mentioning that the Polyclades fromthe reef off Ralum
have all belonged to the subdivision Cotylea, and included a
very fine Thysanozoon, attaining a length of 66 mm.; whereas
154 ARTHUR WILLEY.
those from the stones off Rakaiya, which I have collected in
much larger numbers, have all been representatives of the
subdivision Acotylea.
Several of the latter have laid their egg-plates in my jars.
There are, among others, two well-marked species of Stylochus,
both characterised by their uniformly pitted or granular ap-
pearance, but in one case of an opaque dark ashy colour, and
in the other of a light brown sandy complexion. As I cannot
match these with any descriptions in the literature at my com-
mand, it will do no great harm to refer to the former as S.
cinereus and to the latter as S. arenosus.
Stylochus cinereus has a length of 24 to 27 mm., and a
width of 10 mm. ; tentacles brownish yellow, covered with eyes,
and 3°5 mm. from anterior margin of body; cerebral eyes ex-
tending over proximal portion of outgoing nerves; anterior
marginal eyes extending backwards to level of tentacles;
margin of body cloudy, but light and unpigmented.
S. arenosus has a length of 41 to 45 mm., and a width of
16 mm.; tentacles covered with eyes except the tip, which is
orange-coloured; tentacles 555 mm. from anterior margin ;
cerebral and marginal eyesas in 8. cinereus; margin of body
nearly colourless and translucent; about twelve pairs of in-
testinal diverticula.
It is difficult to obtain uninjured examples of S. cinereus,
as it tears with the greatest readiness while being detached
from the stones on which it lives. Lacerated examples pre-
senting a ragged appearance thrive well in aquaria. The
tendency to laceration seems to be a characteristic of the
species as much as any anatomical feature. ‘There is a species
of Cryptocelis (marginal eyes all round body) common here,
which practises deliberate autotomy, this being one of its
most characteristic peculiarities.
S. cinereus laid several irregular patches of eggs on April
27th. ‘The eggs were disposed in no particular order in a
somewhat granular gelatinous matrix, each ovum being sur-
rounded by its own membrane, and measuring 12 to 18, in
diameter. Before the commencement of segmentation the
ON NAUTILUS AND SOME OTHER ORGANISMS. 155
ova of this species pass through a pronounced ameeboid stage
(Figs. 4 and 5), subsequently becoming roundagain. It is also
a favorable species for observing amceboid movements of the
polar bodies. The movements are sluggish but unmistakable,
and sometimes the polar bodies become widely separated from
one another (Fig. 6).
Fig. 4. Wie, fh.
Fic. 6.
Fies. 4 and 5.—Ova of Stylochus cinereus before commencement of
segmentation, to show the ameeboid phase. Zeiss, 3 c, cam. luc.
Fic. 6.—Ovum of S. cinereus preparing for the first division, to show
the ameeboid character of the polar bodies. Zeiss, 3 c, cam. luc.
Stylochus arenosus laid one irregularly contoured plate
of eggs on May 6th. The ova were arranged in distinct rows,
each ovum surrounded by its own proper membrane, and
measuring 9 to 10 n.
The egg-discs of two species of Planocera, occurring like
the above species in shallow water off Rakaiya, offer a strong
contrast to the egg-plates of Stylochus. I am under the
necessity of naming these, at least provisionally, on the same
principle as the above-described species of Stylochus.
Planocera discus (mihi) attains a length of 45°5 mm.,
with a width of 28mm. The dorsal surface is coloured with
delicate yellowish-brown (umber) reticulated markings, on
which are scattered black nodal spots; interspaces whitish ;
156 ARTHUR WILLEY.
margin of body of the delicate brown ground colour inter-
rupted by narrow white streaks; mid-dorsal region pitted
with numerous minute black spots; tentacles brown with
rose-yellow tip, 13 mm. from anterior margin ; large tentacular
eyes at base of tentacles, more numerous than the smaller
cerebral eyes; latter numbering about twenty-two to twenty-five
on each side; female genital aperture about same distance
from posterior end as the tentacles are from the anterior end ;
pharyngeal bursz, seen from below, dense white; intestinal
diverticula six or seven pairs.
The margin of the body is always sinuous or skirt-like, and
when at rest the animal is capable of assuming a nearly circular
form. Both this and the following species can be handled with
impunity, neither of them evincing the slightest disposition to
laceration.
Planocera discoides (mihi) is, for a Polyclade, an object
of great beauty. I obtained two specimens of it from the
bottom of a volcanic stone, on the top of which corals and
sponges were growing off the south-west shore of Rakaiya,
in some two or three feet of water at low tide.
It reaches the length of 75 mm., with a width of 36 mm.,
always with a sinuous margin. Like the preceding species, it
can assume an almost circular form. The body is remarkably
transparent ; intestinal rami, seen from above, light brown,
moniliform, anastomosing ; interspaces beset with numerous
minute rubiginous spots; larger dark brown nodal spots
scattered about dorsal surface ; margin of body light, pellucid ;
tentacles pellucid, 22 mm. from anterior margin; female
aperture same distance from posterior margin; large dense
white shell-gland between the two genital orifices; eyes about
twice as numerous as in P. discus; cerebrum in a clear pel-
lucid area, which is deeply indented or excavated ventrally ;
seven or eight pairs of intestinal diverticula, which present a
dull greyish-white colour from below.
P. discus laid a circular dise of egg-capsules on April 30th.
The disc had the appearance of consisting of a series of con-
centric circles, but closer examination showed that the rows
ON NAUTILUS AND SOME OTHER ORGANISMS. Sd
were arranged spirally, the spirals being here and there in-
terrupted (Fig. 7). The egg-disc measured 15 mm. in diameter.
Fic. 7.—Enlarged view of the egg-dise of Planocera discus. Only a
few of the egg-capsules have been inserted in the drawing.
Each row consists of numerous capsules packed closely to-
gether in what amounts to a common gelatinous tube. The
spirals can be unwound. In the outermost spiral were a few
irregularly dispersed capsules. The normal capsules contained
each from eight to eleven ova, and the latter had no other
membrane round them individually.
On May 4th the same individual, which had been kept
isolated all the time, laid another much smaller disc of egg-
capsules. I left these undisturbed, and in the course of time
many of the ova developed into ciliated embryos with a pair of
large primary eyes. The embryos swim about actively in the
capsules, three or four to seven or eight in each capsule, sur-
rounded by the fragments of those eggs which had not de-
veloped. Many of the embryos had developed abnormally,
owing probably to the fact that the water in which they were
kept had not been changed frequently enough. Before seg-
mentation commences, some of the ova in each capsule present
158 ARTHUR WILLEY.
an appearance of approaching fragmentation, and it is probable
that this is a regular occurrence. Here also the polar bodies
execute amceboid movements.
In the case of P. discoides I found two egg-discs on
May 27th on the same stone on which the adults were living.
To the unaided eye they were not to be distinguished from
those of P. discus, but microscopic examination showed
that the gelatinous matrix! in which the egg-capsules were
{>
Fic. 8.—Four egg-capsules from an egg-disc of Planocera discoides
each capsule contains four ova. Zeiss, 3 a, cam. luc.
arranged in concentric rows was more continuous, and not
divisible into concentrically disposed tubes; and, above all, that
in each egg-capsule there were only four ova (Fig. 8).
Of the four ova in each capsule, as a rule, only two develop
into ciliated embryos; frequently, however, three, and rarely
only one. I have in no instance found four ciliated embryos
in a capsule. Those ova which do not develop undergo frag-
mentation. I think the particles of the fragmented ova must
be dissolved in the fluid contained in the capsules, and not
absorbed by the remaining ova in the solid form. When two
or three embryos are revolving in a capsule, there is usually no
trace whatever of the previous existence of other ova in the
same capsule ; and when there is only one embryo in a capsule
it is no larger than other embryos.
It is chiefly before the embryos begin to revolve that the
1 A large number of amceboid bodies may be observed in the matrix in
which the egg-capsules are embedded in this species.
ON NAUTILUS AND SOME OTHER ORGANISMS. ~~ 159
actual evidence of fragmentation is to be obtained (Figs. 9—
11). Only rarely are the rounded particles derived from the
Fie. 9. Iie, IO):
CART
i
,
4
i
sab
2
Lee
AT
rear
Fics. 9, 10, and 11.—Planocera discoides. Kgg-capsules. In Fig. 9
one ovum has undergone fragmentation; in Fig. 10 two ova have undergone
fragmentation; in Fig. 11 the remains of the fragmented particles are seen
lying against an embryo which had already acquired cilia, but had not com-
menced to rotate. Zeiss, 3 c, cam. lue.
fragmentation of the original ova, to be seen floating about
in those capsules in which the embryos are revolving.
2. Cirripathes anguina.—From the reef off Ralum I
obtained a fine specimen of this unbranched Antipatharian,
upwards of 9 feet in length. It reached up close to the surface
160 ARTHUR WILLEY.
at low water. The natives here call it “a pada ur-a-ta,” or
simply “a ur-a-ta,’ which means “ the bones of the sea.”
3. Nudibranchiata.—The Nudibranch molluscs, particu-
larly the Doridide, are represented here by a great wealth of
species of all sizes up to 60 mm. by 45 mm., many of them
being brilliantly coloured. Their spiral egg-bands may be
obtained without difficulty. Several species, including two
Molids, have laid their eggs in my dishes.
On the reef off Ralum, and in Blanche Bay too, there are
to be found large numbers of very long spirally coiled ribbons,
so thickly encrusted with sand that they appear to consist of
nothing but sand. The natives call these ‘‘a pipia,” which
means “‘ earth” or ‘“‘ ground.’ Some are tough and elastic,
others brittle. ‘They are the egg-bands either of species of
Doridide, or of other Opisthobranchiate Mollusca, but I have
not yet been able to identify them.
4. Onchidium.—One of the commonest Mollusca in the
shallow water off the south-west shore of Rakaiya is a species
of Onchidium. It is often found in the very heart of large
friable stones, approximately at low-water mark, but some-
times further out. It occurs in other parts of Blanche Bay as
well. What service the dorsal eyes can be to it is not easy to
imagine. Semper supposed it was to enable the Onchidium to
perceive and escape from what he says is its chief enemy,
Periophthalmus. But although I have seen a small species
of Periophthalmus at Rakaiya, on the muddy shore of the so-
called lagoon, it is difficult to accept Semper’s view.
The papille which carry the eyes are obviously homologous
with the retractile branched respiratory papille in the posterior
dorsal region, and every transition can be observed between
them.
When kept in confinement in jars, Onchidium asserts its
pulmonate qualities by often creeping out of the jars for long
distances.
5. Larve of Polygordius and Squilla.—It is worth
recording the presence of larve of Polygordius in the “ Auf-
trieb”” in Blanche Bay. In the narrow strait which separates
ON NAUTILUS AND SOME OTHER ORGANISMS. 161
Rakaiya from the mainland I have taken one larva of Poly-
gordius in an advanced stage of metamorphosis, when the long
body has grown out behind the trochophore. On the same
occasion the “ Auftrieb ” contained also larvee of Squilla.
6. Phosphorescence.—The astonishing phosphorescence
which is to be observed when fishing with the tow-net at night
in Blanche Bay is in large measure due to Copepods. and
Ostracods.
7. Ascidians.—The Ascidiau fauna here is richer than I
expected to find it, since, as Herdman has pointed out, the
tropics are not the metropolis of the Ascidians.
The Didemnide, which form one of the most difficult groups
of compound Ascidians from a systematic point of view,
are strongly represented by red, white, yellow, and green
varieties.
Botryllus also occurs, though notcommon. I only know at
present of two species—one very thin and white, attached to
the lower surface of corals; the other thicker and purple,
found growing on a T'ridacna shell.
In regard to the compound Ascidians, I have been struck by
the apparent absence of any member of the Polyclinide.
The simple Ascidians are represented by numerous species,
one of which especially has peculiarities of such a nature that
I will describe it at some length, reserving an account of the
other Ascidians for a future occasion.
8. Styeloides eviscerans, n. sp.—In 1885, Sluiter
described, in the ‘ Natuurkundig Tijdschrift voor Ned-Indie,’
Bd. xlv, a simple Ascidian, under the name of Styeloides
abranchiata, n. gen. et sp., in which the branchial sac and
intestine were absent. Such was the condition of the animal
in other respects that Sluiter was led to suppose that this must
be the normal state of things, and founded the new genus
accordingly.
In his extremely useful ‘‘ Revised Classification of the
Tunicata” (‘Journ. Linn. Soc. Zovol.,’ vol. xxiii), Herdman,
commerting on this species of Sluiter’s, says, ‘This is such
an exceptional and remarkable case that I cannot help sus-
VOL. 39, PART 2.—NEW SER. L
162 ARTHUR WILLEY.
pecting that the single specimen examined by Sluiter was
merely an individual abnormality.”
I venture to hope that I have found the solution to this
enigma in the species about to be described.
ae
eS
LG
Ne
Fie. 12.—Group of seven individuals of Styeloides eviscerans, n. sp.
represented as lying attached to the surface of a fragment of stone. In the
large individual to the right the digestive tract is indicated in process of
extrusion through the atrial aperture. a. Anus with frilled margin. 7. In-
h, Foreign organisms attached
testine. 47.s. Branchial sac. e. Endostyle.
to test.
ON NAUTILUS AND SOME OTHER ORGANISMS. 163
The species is not common, but I have obtained three or
four examples of it from the lower surface of stones off the
south-west shore of Rakaiya, in one half to one fathom of
water.
The accompanying sketch (Fig. 12) represents a fine and
typical group of individuals of this species. The members of
the group are so intimately connected together by the mutual
fusion of their tests that one would at once suppose that they
had arisen from a parent stock by budding. Such, however,
is not the case, since by making incisions it is found that it is
only a fusion of test-substance, and not a true organic union.
Moreover sometimes isolated individuals are to be found, as in
Fig. 18.
~
eas
- Fre. 13.—Styeloides eviscerans. Outline of isolated specimen. a.
Atrial siphon. ¢. Processes of test.
The general colour of the animal or group of animals is a
characteristic dull reddish brown, the colour being more pro-
164 ARTHUR WILLEY.
nounced in the neighbourhood of the apertures, while the lips
of the latter are a pure dark red, interrupted by four light streaks
which indicate the quadripartite character both of the buccal
and atrial orifices.
As seen in the figures, the individuals are not always attached
to the rock by the same side, but sometimes by the ventral
side, sometimes by the right, and sometimes again by the left
side (Fig. 12).
The total length of the group represented in Fig. 12 was
91°5 mm., and the greatest breadth of the group 31 mm.
The animal of which an outline is given in Fig. 13 measured
51°5 mm. in length, and the atrial opening was 20°5 mm.
removed from the buccal aperture. By its external appearance
alone it is an extremely well-marked species. The surface of
the coriaceous test is in some places wrinkled and in other
places smooth.
The most remarkable peculiarity of the new species, how-
ever, is the faculty which it possesses of evisceration.
After I had had them for a short time in a vessel where
everything was fresh and in good condition, I suddenly dis-
covered a number of digestive tracts lying at the bottom. On
then inspecting the Ascidians, of which there were several
species present at the time, I found that they were all living
and in a bealthy condition.
Eventually I actually observed the process of evisceration
taking place (cf. large individual to the right in Fig. 12). It
takes a rather long time before the process is completed. Itis
effected by violent periodic contractions of the atrial siphon.
After it is over the animal presents a perfectly normal and
healthy appearance, opens and closes its siphons, and is suscep-
tible to irritation and to the influence of cocaine. So constant
is this ejection of the digestive tract that if it is desired to pre-
serve specimens intact, they must be placed in alcohol imme-
diately after capture.
The dissection of an individual which has ejected its bran-
chial sac and intestine discovers no laceration whatever; and
undoubtedly, in ignorance of the habit of evisceration, one
ON NAUTILUS AND SOME OTHER ORGANISMS. 165
would be tempted to suppose, as Sluiter did, that the absence
of an alimentary canal was the normal condition.
In adopting Sluiter’s generic name the diagnosis must of
course be amended.
Before evisceration takes place the branchial sac is found to
have the usual vascular connections with the mantle, but the
endostylar area seems to have a very loose attachment to the
mantle, and can be readily detached. When the branchial sac
is ejected the dorsal tubercle (Figs. 14 and 15) and peripharyn-
geal groove are left behind, and there is a corresponding
triangular excision in the wall of the ejected branchial sac.
Fie. 14. Iie, INK,
Fic. 14.—Styeloides eviscerans. Dorsal tubercle. yp. . Peripharyn-
geal band. d. ¢. Dorsal tubercle. g. Ganglion.
Fie. 15.—Dorsal tubercle of another individual of 8. eviscerans.
The dorsal lamina is a simple undulating or crumpled mem-
brane, and there are four folds of the wall of the branchial sac
on each side.
The genital saccules have the characteristic subeylindrical
form, and occur on both sides attached to the mantle. Curious
bodies called endocarps, whose nature I do not understand, also
occur on the inner surface of the mantle as in Sluiter’s
species.
166 ARTHUR WILLEY.
The latter is probably a distinct species from the one I have
described, although there are many features common to them
both, particularly the external form and mode of attachment.
I am a little puzzled to understand what Sluiter says about
the endostyle, and am inclined to think there must be some
mistake about it, as there is no trace of a typical endostylar
epithelium in the section figured by him. I have even
observed a line or ridge in the ventral surface of the mantle
corresponding very closely to his Taf. viii, fig. 2, but this possi-
bly represents the former line of contact between the endo-
style and the mantle.
There are naturally a great many more questions to be
answered in connection with this remarkable Ascidian, but I
have probably said enough to show that its property of evis-
ceration is its most distinguishing peculiarity, and thus to
afford an explanation for an otherwise inexplicable anomaly.
Yours very truly,
Artour WILLeEY.
Raum, GerMAN New GUINEA,
via SINGAPORE ;
September 24th, 1895.
Dear Professor LANKESTER,
1. Significance of the Siphunele in Nautilus
pompilius.
Being desirous of obtaining, if possible, experimental evi-
dence as to the physiological significance of the siphuncle
in the pearly Nautilus, I have made several successful attempts
to cut the siphuncle without otherwise injuring the animal.
The evidence supplied by the experiment cannot be regarded
as conclusive, on account of the altered conditions of depth and
temperature to which the Nautilus is exposed by being brought
up to the surface, but it may be well to consider what the
results indicate.
ON NAUTILUS AND SOME OTHER ORGANISMS. G7
At first I sawed through the shell into one of the chambers,
and then cut the siphuncle. This method has the disad-
vantages of injuriously affecting the efficiency of the chambers,
and of causing a more or less considerable loss of blood to the
animal. The latter will, however, live in confinement about
as long as untouched individuals.
A young Nautilus operated upon in this way on June 26th
was placed in the sea in shallow water, for its movements to
be watched. It sank slowly to the bottom, and then for a long
time made active revolving motions about the vertical axis,
but scarcely made any progressive movements.
On another occasion (July 10th), after several trials, I found
that the best way of performing the operation is to saw through
the shell in the neighbourhood of the posterior portion of the
body of the animal, over the cardiac region, and not to tamper
with the chambers. If the shell be held mouth downwards,
this point lies approximately in the same vertical and trans-
verse plane with the points where the free margin of the
mouth of the shell merges into the umbilicus. When a large
enough hole has been made in the shell to admit the scissors,
the shell being still held upside down, the ventral visceral
portion of the body usually detaches itself from the shell, or
can be readily caused to do so, and, sinking inwards, exposes
the root of the siphuncle, which can then be severed. On
returning the shell to its normal position the body immediately
resumes its normal intimate contact with the wall of the cavity
in which it lives, and the pressure so exerted prevents any
extensive loss of blood. Under these conditions the operation
does not, as a rule, appear to affect the vitality of the animal
in any degree.
A Nautilus! which was treated in this way on July 10th,
on being placed in the sea swam about very vigorously for
some time in the middle stratum of water, but most of the
time at a little distance from the bottom. On September 13th
I operated on four more individuals taken in Talli Bay, on
1 It should perhaps be mentioned that in this particular individual I acci-
dentally cut into the last chamber, and plugged the opening with wax.
168 ARTHUR WILLEY.
the north coast of the Gazelle peninsula. One of them showed
a tendency to sink to the bottom, which it always performed
very gradually. In this one I had accidentally punctured the
mantle over the heart. The others remained floating and
swimming about on the surface during the whole time of
observation. They did not go far in one steady direction, but
tended to go in circles, as in fact did another one whose
siphuncle was uncut. If one of the individuals floating at the
surface was forced down to the bottom with a hand-net, it
would slowly rise to the surface again. This also often happens
with a Nautilus that has not been operated on.
The results indicated by the above experiments, which, it
may be added, are worth repeating, may be summarised as
follows:
The cutting of the siphuncle (a) does not temporarily affect
the vitality of the animal; (3) does not prevent it from making
movements of translation ;! (y) does not prevent it from floating
at the surface; (6) does not prevent it from sinking to the
bottom.
It still remains to be ascertained whether a Nautilus whose
siphuncle has been cut, having sunk to the bottom of the sea
in shallow water, will undertake a journey to the surface.
My experiment of July 10th would seem to indicate that this
" might be expected to occur.
The above experiments do not appear to oppose the view
which I expressed in a former communication—that the si-
phuncle of Nautilus pompilius is, in some measure, of the
nature of a vestigial structure.
It might indeed be legitimate to suppose, on the principle of
the correlation of organs, that in the Nautiloidea the course
of evolution has led to a reduction of the siphuncle pari
passu with an increase in the efficiency of the chambers as
hydrostatic organs.
1 Tn speaking above of progressive movements I mean, of conrse, in the
usual backward direction.
ON NAUTILUS AND SOME OTHER ORGANISMS. 169
2. Some Features in the Arterial System of N.
pompilius, as determined by Injection.
(1) Cireulus Pallialis.—After successful injections a sin-
gular feature in the circulatory system is to be observed. The
marginal pallial artery, which I described and figured in a
former note, is found to pass on each side, dorso-laterally, into
a branch of the dorsal aorta, so that a complete arterial circuit
is produced.
I have even partially injected the marginal pallial arteries
from the dorsal aorta itself, but the injection fluid did not
proceed very far in this centripetal direction, owing no
doubt to the resistance it met with from the action of the
heart.
For this remarkable arterial circuit, produced by the con-
fluence of the marginal pallial arteries, which arise ultimately
as branches from the so-called “lesser aorta,” with a pair of
branches! from the great aorta, I propose the above name of
“circulus pallialis.”” It is illustrated in the accompanying
sketches (Figs. 16 and 17).
In Fig. 16 the posterior convex extremity of the body is
supposed to be somewhat tilted up, in order to show the whole
outline of the septum-producing area of the mantle. This out-
line is very distinct in fresh specimens, and the region of the
mantle enclosed by it is distinguished from the surrounding
portions of the mantle by its greater thickness and opacity.
. As already stated, this is the portion of the mantle which
manufactures the septa, and it has an abundant arterial supply
through the ramifications of the two main branches of the
posterior pallial artery. These ramifications may be grouped
together as the pallio-septal arteries; and it is surprising
to see how rigidly they are confined within the septal
contour.
It will be noticed that the latter makes on each side a
1 Yor reasons which will presently appear, these may be called the pallio-
nuchal arteries.
]
70 ARTHUR WILLEY.
posep a? ll mn
Fie. 16.—N. pompilius, ¢. Dorso-posterior aspect of visceral region, to
illustrate the circulus pallialis and the septal contour. x. m. Nuchal mem-
brane. x.a. Nuchal artery. m.e. Free mantle-edge. p.z.a. Pallio-
nuchal artery. m.p.a. Marginal pallial artery. co/. Columellar or shell
muscle. p.c.a@. Posterior columellar artery. p. p. a. Posterior proventricu-
lar artery. s.c. Septal contour. s. Siphuncle. p. s. a. Pallio-septal arteries.
s.a. Siphuncular artery. 7. Liver. 7. Intestine. 7. Testis. post. p. a.
Posterior pallial artery. g. Gizzard.
N.B.—The dorsal aorta and its branches are indicated by dotted lines.
They show dimly through the skin when injected.
ON NAUTILUS AND SOME OTHER ORGANISMS. tat
a
-- ~
~
4 \
/ \
‘ \
\
pT TTT
———————— \
Fic. 17.—N. pompilius, ¢. View of nuchal region, to further illustrate
the circulus pallialis. The dorsal free mantle-edge is reflected and a median
incision made. 4%. Hood. coz. Conecavity at base of hood, in which the
nuchal membrane (z. m.) lies. f. Dorso-posterior portion of funnel. x. a.
Nuchal artery. co/. Columellar muscle. m. p. a. Marginal pallial artery.
p.m. a. Pallio-nuchal arteries. d.a. Dorsal aorta. m.e. Free mantle-edge.
c. e. Cut edges of mantle and body-wall.
N.B.—The dorsal aorta shapes its course in this region in accordance with
the state of repletion of the erop.
172 ARTHUR WILLEY.
symmetrical figure with the outline of the great shell-muscles!
where the latter abut on the shell.
In Fig. 17 the union of the marginal pallial artery with the
left pallio-nuchal artery is represented from the inner surface
of the mantle. Here it is seen that the two arteries unite in
the dorso-lateral angle where the mantle and the funnel-flap
fuse with the body-wall; and furthermore, that from the same
point an artery is given off which passes forwards and gives
off branches to the nuchal membrane. The latter structure
was accurately described by Owen as a “semilunar ridge”
lying in the concavity at the base of the hood, and applied to
the involute convexity of the shell.
Owen thought it might serve to prevent the shell from
encroaching too much upon the hood ‘‘in the act of creeping.”
We now know that the animal does not creep on its hood with
reversed shell.
The nuchal membrane would seem to be responsible for the
dense black colour of the involute portion of the shell, and
possibly also exerts a lubricating influence. Keferstein calls
it the “ Nackenlappen.”
(2) Genital Arteries (Fig. 18).—Nolessthan three arteries
arise directly from the heart which, to my knowledge, have
hitherto escaped attention. ‘They are (a) the artery of the
genital duct or gonaducal artery ; (0) the artery of the genital
gland or genital artery; (c) the artery of the pear-shaped
gland.
The accompanying sketch (Fig. 18) obviates the necessity of
a detailed description. The genital artery is submedian, and
the main trunk lies on the dorsal side of the genital gland.
Both the gonaducal artery and the artery of the pear-shaped
gland give off a branch which passes into the perigonadial
membrane, and this apparently trifling fact, combined with the
‘subsymmetrical relations of the gonaduct and the pear-shaped
gland, may indicate that the latter is the metamorphosed
‘genital duct of the left side, and not, as I believe has been
1 For purposes of nomenclature it will be found convenient to speak of the
great shell-muscles as the columellar muscles.
ON NAUTILUS AND SOME OTHER ORGANISMS. 173
LON
WAN 7 i
es ie Z\ \ ASKEW SS aces
ee to
_—
-—
a>
Fig. 18.—N. pompilius, ¢. Genital arteries from below. r. Rectum.
r. a. Rectal artery. (N.B.—The rectal arteries are very variable.) aut. p. a.
Anterior pallial artery. 7. a. Intestinal artery. (N.B.—This artery usually
passes to the right of the rectum, as shown in this figure; but in one
instance I have observed it to pass down to the left of the rectum.) 47. v.
Branchial veins. post. p. a. Posterior pallial artery. p.s.g. Pear-shaped
gland with its artery. a’. Branches of the preceding artery and of the gona-
ducal artery, which supply the superjacent perigonoidal membrane. gez. a.
Genital artery and its branches. ¢. Testis. ¢.0. Aperture of testis. p. v.0.
Orifice of communication between the pericardial and visceral portions of
body-cavity, through which the posterior pallial artery passes. goz. a. Gona-
ducal artery. v.s. Vesicula seminalis. (N.B.—This structure, tlie testis and
pear-shaped gland are closely united to the heart by a membrane.) J. v.
Needhamian vesicle or spermatophore sac. #. Dotted line to indicate where
the pallial duplicature merges into the body-wall ventrally.
174 ARTHUR WILLEY.
suggested, the morphological equivalent of an entire left genital
apparatus.
In the female the ramifications of the genital artery pass
up on to the surface of the individual ova, and form a kind of
capillary system, the finer branches following, but not always
confined to, the reticular markings formed by the ridges of the
follicular membrane which project into the yolk (Figs. 4
and 5).
The meshes formed on the surface of the ova by the inter-
section of the follicular ridges or plications are much wider in
submature ova than in the less mature, and the ridges would
presumably be found to flatten out in completely ripe eggs,
although it has not been my good fortune hitherto to have
found any such. At the animal pole of the egg the ridges are
absent, and those which lie at the margin of this area form in-
complete meshes as described by K 6lliker in the ovarian ova
of other Cephalopods (Fig. 19).
Fie. 19.—Fresh ovarian ovum of N. pompilius, to show the reticular
markings produced by the plications of the follicle. j. a. Clear polar area, in
the centre of which lies the germinal tract.
The clear polar area of the ovum has usually a sub-
triangular shape, and from each of the corners of the
triangle what may be called a line of weakness occurs in the
follicular wall, bound on either side by incomplete meshes
(Fig. 19).
The arteries which traverse the surface of the ova give off
minute branches which pass inwards, as it were, into the
depths of the follicular ridges; and these deep-lying vessels
anastomose with one another, while the superficial branches
appear, as a rule, not to form anastomoses. It may be added
ON NAUTILUS AND SOME OTHER ORGANISMS. 175
that the impression of anastomoses is much more readily
conveyed by examination with a hand-lens than it is by the
use of the compound microscope.
Fic. 20.—Ovary removed from body, and seen from dorsal aspect. On the
right of the figure the follicular meshes of a submature and a half-mature
ovum are partially inserted to show difference in size. o. a. Neck of ovarian
sac, which bears the aperture. o. Ova. gez.a. Genital artery. Only the
more superficial branches are indicated.
The germinal tract appears in the centre of the clear polar
area as a faint whitish spot, and is turned towards the ventral
aspect of the ovary (Fig. 20). The older ovarian ova are ren-
dered shapeless by mutual pressure, with, however, a roughly
oval outline. In this condition an ovum may measure 15:5
mm. in length, with a breadth of 11°55 mm. When the pres-
sure is released by slitting open the ovary the ova round up,
and those which are submature have an average diameter of
some 10 mm. The yolk is viscous and glutinous, and pos-
sesses a translucent brownish tinge. ‘The nearly ripe ova
rupture with the utmost facility.
From a consideration of the size and relative states of
maturity of the ova, it might be expected that they are laid
singly. Every month, from December to September inclusive,
I have been able to obtain over-ripe males (with spermato-
phores in the dorsal buccal recess and in the Needhamian
vesicle) and submature females. Once in July I obtained a
male with a discharged spermatophore capsule in the buccal
recess ; in fact, I have found this more than once.
From these facts, and from the fact which I have previously
176 ARTHUR WILEY.
mentioned of the relative scarcity of the females in compari-
son with the males, I draw three provisional conclusions:
firstly, that during the process of reproduction (fecundation
Fic. 21.—Female genital organs of N. pompilius, seen from below, to
show the direction in which the polar areas of the ova lie. yp. a. Polar areas of
ovarian ova. o. a. Aperture of ovarian sac. g. Uterine portion of oviduct.
v. Vaginal portion of oviduct.
and spawning) the females live in retirement; secondly, that
the females practise what the Germans call ‘ Brutpflege;”
and lastly (what I regard as almost certain), that repreducues
takes place all the year round.
(3) Cephalic Arteries (Figs. 22 and 23).—For a descrip-
tion of these arteries it will suffice to refer to the explanation
of the figures. I will, however, call attention to the varia-
bility of the right and left anterior proventricular arteries ;
the latter was absent from the individual represented in fig. 7.
The two main trunks into which the dorsal aorta divides be-
hind the brain may be called the right and left innominate
arteries.
It is a singular fact that the great median buccal artery
always springs from the right innominate artery. The con-
stancy of this origin would seem to indicate that it is poten-
tially a paired structure.
ON NAUTILUS AND SOME OTHER ORGANISMS, iar
Fic. 22.—N. pompilius, g. Dissection, from above, of the cephalic
region to show the cephalic arteries. An incision has been made through the
nuchal membrane, the hood, and the buccal membrane. The brain-capsule has
been opened, and the median portion of the mantle behind the nuchal region
has been removed. 0.c. Buccal cone. ent. inf. Dorso-lateral inner row
of tentacles (the superior labial processes of Owen). 7. a. Labial arteries
supplying the buccal membrane and fringe. s. m. a. Superior mandibular
artery. s.7. Superior retractor muscles of the jaws (Owen). 4%. a. Buccal
artery. 7@. m.d. Paired inferior mandibular arteries. 2. m. Nuchal mem-
brane. a.c. a. Anterior columellar artery. a. p. a. Anterior proventricular
artery. f. Dorso-posterior portion of funnel. 2. e. Free mantle-edges
c. e. Cut edge of mantle. p.z.a. Pallio-nuchal artery. d. a. Dorsal aorta.
cer. Brain with cerebral arteries. 2%. Cut edge of hood.
N.B.—Apart from the cerebral arteries, all the anteriorly directed branches
of the innominate arteries pass below the cerebral capsule.
VOL. 39, PART 2,—NEW SER. M
178 ARTHUR WILLEY.
Attention may also be drawn to the two arterioles which
arise from the base of the right proventricular artery and
supply the wall of the dorsal aorta.
Fic. 23.—N. pompilius, g. Sketch to show the arteries which arise
from the innominate arteries after the brain has been cut through. 4% a.
Right tentacular artery, which gives off branches to all the tentacles of its
side. The innermost or mesiad branch (o/f a.) supplies the bipartite
laminated organ of van der Hoeven, which possibly represents a pair of
inferior labial processes, and may have an olfactory function. fa. Infun-
dibular artery, which passes through the cartilage into the funnel. The
latter also receives minor branches from the columellar arteries. oc. a. Arteries
to the eye. ped. a. Pedal artery, a convenient name for the common trunk
from which the infundibular and tentacular arteries arise. opt. Optie gan-
glion. cer. Brain cut across and drawn aside. a.c. a. Anterior columellar
artery. a. p. a. Right anterior proventricular artery. Compare its origin in
Fig. 7, where it arises from the aorta. d. a. Dorsal aorta. /. a. p. a. Left
anterior proventricular artery, usually but not invariably present; it some-
times arises from the left innominate artery, as indicated by the dotted lines ;
and sometimes, as in this example, from the dorsal aorta. zz. Innominate
arteries. d.a. Buccal artery. 7. m. a. Inferior mandibular arteries.
ON NAUTILUS AND SOME OTHER ORGANISMS, 179
8. Further Remarks,
(1) Capillaries.—As far as I have been able to observe,
there seems to be a true capillary system in the free portion
of the mantle, that is to say, in the pallial duplicature. I
have previously described the pallial arteries and veins. There
may be another system of capillaries in the funnel, which has
Fie. 24.—Photograph of a living specimen of Nautilus pompilius,
taken by A. Willey at Ralum in 1895.
an astonishingly rich vascular supply, and at whose base two
large veins, the infundibular veins, may be observed to
pass into the vena cava. But I have observed no veins in con-
nection with the genital organs, and I gathered from micro-
180 ARTHUR WILLEY.
scopic examination that the follicular arteries on the surface
of the ova possessed free openings.
(2) Blood.—I will here only mention that the blood is a
syrupy fluid with a pronounced blue tint, which becomes very
dark on exposure to the air. The corpuscles comprise amceboid
aud fusiform cells, the latter being somewhat Gregarina-like
in appearance.
Yours very truly,
Artuur WILLEY.
THE BRAIN OF A FQTAL OkNITHORHYNCHUS. 181
The Brain of a Fetal Ornithorhynchus.'
Part I.—The Fore-brain.
By
G&. Elliot Smith, Wi.D., Cia. i.,
Demonstrator of Anatomy, University of Sydney, N.S.W.
With Plate 11.
Tue foetus, with whose brain this paper is concerned, was
received already preserved in alcohol from the Australian
Museum of Sydney by Professor J. 'T. Wilson, who has
described in full the external appearances and measurements.”
It is sufficient here to mention that its extreme length, mea-
sured along the dorsal contour, is 80 mm. |
The head was split sagittally some distance from the middle
line, and the larger piece, after being embedded in celloidin,
was cut in a complete series of coronal sections, each 50 1
thick. These were stained with haematoxylin and decolourised
with alcohol acidulated with picric acid, which afforded a very
useful counter-stain. By means of the picric stain the nerve-
fibres were clearly demonstrated, so that it was easy to make
out their general distribution sufficiently accurately to compare
1 This paper forms part of a thesis dealing with the ‘ Anatomy and Histology
of the Cerebrum of the Non-placental Mammal,” which was awarded the
University Medal when presented to the examiners for the M.D. degree at
the University of Sydney, March, 1895. ‘The account of the gross anatomy
is now being published elsewhere. ‘The account of the histology of the adult
brain will be published shortly.
2 « Description (with figures) of a Young Specimen of Ornithorhynchus
anatinus,” ‘Proceedings of Linnean Society N.S.W.,’ vol. ix, part 4, 2nd
series) p. 682.
182 G. ELLIOT SMITH.
with Weigert- and Golgi-stained specimens of the adult brain.
For the opportunity of examining this valuable series T am
deeply indebted to the generosity of Professor Wilson, whose
kindly interest and valuable advice I gratefully acknowledge.
In recording the disposition of parts in a single stage of a
developing brain, the account, in the absence of other stages
to show the origin and destiny of the various regions, must be
almost purely descriptive. This paper, therefore, will indicate
the apparent differences in the specimen under discussion from
the ordinarily accepted account, rather than attempt to ex-
press an opinion upon any of the morphological questions
raised. Many interesting and important facts are elucidated,
however, by a comparison with the adult brain, a description of
whose histology forms the bulk of the thesis of which this
paper constitutes a part.
A comparison with some valuable serial sections of foetal
Perameles and Macropus in various stages of development
has thrown a considerable amount of light upon many other-
wise obscure points. These specimens were placed at my
disposal by Professor Wilson and Mr. J. P. Hill. I have also
examined some imperfect series of brains of foetal Phalan-
gista. Several foetal Echidne, which were lately received,
are unfortunately not ready for examination, but will be
described subsequently. .
The literature of this subject is very scanty and incomplete.
Before the Linnean Society of New South Wales, Dr. C. J.
Martin and Mr. J. P. Hill recently described the condition of
the neural plate in the early embryo of Platypus,! and Selenka
has described early stages of the Opossum brain.? C. L.
Herrick has published brief notes on, and figured sections of,
the foetal brain of Didelphys.? Osborn,‘ in his well-known
paper upon the corpus callosum, described the hippocampal
‘«On a Platypus Embryo from the Intra-uterine Keg,” * Proceedings
Linnean Society of New South Wales,’ vol. ix, part 4, 2nd series, p. 736.
2 * Das Opossum,’ Wiesbaden, 1887.
> * Journal of Comparative Neurology,’ 1894.
4 *Morphologisches Jahrbuch,’ Bd. xii,
THE BRAIN OF A FQITAL ORNITHORHYNCHUS. 183
region of the cerebrum of pouch specimens of Macropus of
various ages. These brief notes are all that I know referring
to the development of the brain of the non-placental mammal.
In connection with Osborn’s work, one cannot refrain from
expressing regret that he was so biassed by his previous work
upon the Amphibian and Sauropsidan cerebrum (Part I of
his paper) as to misrepresent the condition which obtains in
the young kangaroo. I am well acquainted with the decep-
tive appearance which is presented by the brain of the foetal
Macropus, especially when coloured with ordinary nuclear
stains—an appearance which lent itself so unfortunately to
Osborn’s purpose ; but the important position occupied by the
marsupial in his series should have demanded a more critical
examination than his paper indicates. For, as the connecting
link between the whole sub-mammalia—to which his argument
mainly applied—and the mammalia—to which alone can the
distinguishing names be applied a priori—M acropus formed
the link upon which the cohesion of the whole chain of his
argument depended. By his erroneous interpretation of the
condition in Macropus his argument lost all its cogency ; for
in showing, as he clearly did, the homology of that commis-
sural band, which in the non-placental mammal is undoubtedly
fornix commissure, with the dorsal commissure of Sauropsida
and Amphibia, he prepared the way for the statement that the
corpus callosum—as distinct from the fornix commissure—
exists only in Placentalia.
In describing the foetal brain there is, unfortunately, no
satisfactory account of the adult brain to which one can refer.
The accounts of Meckel and Zuckerkandl are quite valueless,
from the fact that their descriptions were based upon very
badly preserved material. Owen’s description and Garner’s
notes are reliable as far as they go, but they are very brief and
general, and deal only with the macroscopic anatomy, as also
does Turner’s paper. The only attempt at a systematic
account of the histology—that of Hill—is so clearly biassed
by its writer’s previous work, that it gives a very erroneous
and altogether misleading idea of the structures it is supposed
184 G. ELLIOT SMITH.
to describe. In this aceount, therefore, such brief references
will be made to the histology of the adult brain as are neces-
sary to a proper understanding of the foetal structures.
Forming as it does the link between the Mammalia and the
lower Vertebrates, Ornithorhynchus occupies a position of
supreme importance and interest to the morphologist. This
importance is enhanced in the case of the brain, because this
organ presents a number of very significant transitional fea-
tures, several of which shed an important light upon the
morphology.of the cerebrum in the whole Vertebrate series.
In addition there are a number of purely “ individual”? fea-
tures—notably the enormous development of the trigeminal
nerve and sensory tract—which give added interest to the
study. Although in the histological differentiation of its parts,
and in their general arrangement, the cerebrum of Platypus
clearly conforms to the mammalian type, yet numerous fea-
tures—such as the arrangement of the hippocampus, the dis-
position of the structures in the lamina terminalis, and the
“ precommissural area”’—indicate its close Sauropsidan aflini-
ties. It is of interest to note that the resemblances to the
lower brain are even more marked in the foetus under con-
sideration than in the adult.
General Account of the Fore-brain.
At the outset one is immediately confronted with the difficul-
ties of the vexed question of nomenclature. With the increasing
knowledge of brain anatomy it is perhaps only natural that
more exact divisions and distinguishing terms, more in accord-
ance with the advanced ideas, should be introduced. But
when familiar terms are used in an altogether unfamiliar sense,
or applied to entirely different parts of the brain in order to
give place to new terms no more appropriate than those they
supersede, the enormous intrinsic difficulty of this subject
becomes seriously increased, and an undesirable element of
confusion is introduced. ‘The accurate system of nomenclature
introduced by His loses much of its value by a confusing
application of Huxley’s terms ‘ prosencephalon” and “ thala-
THE BRAIN OF A FATAL ORNITHORHYNCHUS. 185
mencephalon ” in quite a different sense from that in which they
are universally used, which must prevent the general acceptance
of his suggestions. Although His’ system of nomenclature is
otherwise very convenient for descriptive purposes, the mor-
phological value of his divisions is seriously open to question,
especially the somewhat arbitrary limits to his “ Endhirn.”
The nomenclature of v. Kupffer seriously adds to the confusion.
One cannot but admire the exactness of Burt Wilder’s definitive
terms, but they are so uncouth and bizarre that the English
student has to learn a new language before he can master the
simplest description. In this paper the most unambiguous
terms will be selected from the different systems, the alterna-
tive names being indicated in brackets; but as far as possible
Huxley’s well-known nomenclature will be employed.
Hach cerebral hemisphere consists of an ovate mass, attached
to the lateral aspect of the front end of the neural tube. The
maximum length of each hemisphere, including the bulbus
olfactoril, is 5°15 mm., and at its mid-point its height and
depth are about half that measurement. The diencephalon
(with the median part of the “ Endhirn”), whose extent is
schematically represented by the shaded area in fig. 1, is of
the same depth as the posterior part of the hemisphere, but is
much shorter. Posteriorly the hemisphere overlaps the me-
sencephalon (mes.) to a slight extent.
In a median sagittal section through the fore-brain, which
is represented somewhat diagrammatically in fig. 2, the optic
nerve is seen as a prominent object (opt. n.) in the floor, so
that it forms a convenient starting-point. Immediately behind
the optic nerve the floor sinks to form the recessus infun-
dibuli (7. i.) of His (recessus postopticus of Burckhardt),
from the posterior extremity of which the tubular hypophysis
(hyp.) extends at first downwards, and then bends forwards at
its extremity. ‘The anterior wall of this extremity contains a
number of neuroblasts and a distinct layer of nerve-fibres,
which extends upwards on to the floor of the third ventricle.
The bent extremity of the tubular hypophysis is closely sur-
rounded by the large glandular (buccal) part of the hypo-
186 G. ELLIOT SMITH.
physis (pit.), which extends forwards for a considerable dis-
tance. It consists of a dense mass of highly convoluted
tubules (fig. 11, pié.), lined with cubical epithelium. Imme-
diately in front of the optic chiasma the floor again becomes
depressed (fig. 2, 7. 0.), and on either side of the middle line
forms a distinct recess (fig. 11, 7. 0.). A comparison with the
early stage of the Perameles brain shows that this recess is
the remains of the optic diverticulum, and therefore cor-
responds to His’s recessus opticus (preopticus—Burck-
hardt). According to His, the recessus opticus corresponds
to the anterior extremity of the limiting furrow (Grenz-
furche) between his alar (Fligelplatte) and basal (Grund-
platte) laminze, to which Reichert has given the name “ sulcus
Monroi,” without, however, recognising its important mor-
phological significance.
Immediately in front of the recessus opticus the median
wall takes a sudden bend into the vertical direction, and at the
same time becomes enormously thickened to form a large mass
rhomboidal in section (fig. 2, 2. inf.), which His calls the
“lamina terminalis” and Burckhardt the “lamina
infra-neuroporica.” According to His it is formed by the
meeting and fusion in the middle line of his Fliigelplatten,
but according to v. Kupffer and Burckhardt it forms part of
the floor (Bodenplatte). The researches of His seem to
indicate clearly enough that it corresponds to part of the frontal
suture (Schlussnaht), but more convincing evidence is required
before it can be granted that it is formed purely from the
Fligel-, aud not also from the Grund-, platte. This question
will be again referred to later on. In the dorsal part of the
lamina infra-neuroporica a large rounded mass of fibres will be
seen (a. c.) near the posterior surface. This is the anterior
commissure. In all Vertebrates this phylogenetically very
ancient and primitive connecting link between the two hemi-
spheres crosses the middle line in the same position in the
lamina infra-neuroporica, so that it forms an easily recognisable
basis for comparing the perplexing regions which surround it,
Immediately in front and above it there is a small scattered
THE BRAIN OF A F@TAL ORNITHORUYNCHUS. 187
bundle, consisting as yet of very few fibres, which also cross
the middle line in the same lamina. This is the first rudiment
of the fornix-commissure (fig. 3, f. c.). It will be noticed
that it lies below the foramen of Monro and in the floor of
a diverticulum of the third ventricle (figs. 2 and 3), which is
often known as the ventriculus communis (Osborn). In
the adult Ornithorhynchus brain the fornix commissure
occupies a very different position, i.e. in front and above the
foramen of Monro, and entirely dorsal to the third ventricle.
This altered position of the commissure in adult Mammalia
and many Sauropsida probably depends mainly upon the growth
of the lamina terminalis, especially that part lying between
the anterior and fornix commissures; but partly also on the
backward growth of the hemisphere, which is accompanied by
a corresponding growth of its dorsal commissure. In many
Amphibia, where the lamina terminalis does not grow to the
same extent as it does in the higher animals, the fornix
commissure maintaius into adult life a position exactly cor-
responding to that met with in the foetal Platypus. In the
arrangement of their commissures the Sauropsida are inter-
mediate between the Amphibia and Prototheria.!
Above the lamina terminalis the median wall suddenly be-
comes thin again and forms a thin plate, which appears to
spring from the anterior edge of the lamina terminalis. This
thin wall (fig, 3, 7. ”.), which is distinguished by von Kupffer
as the lobus olfactorius impar, bounds a little diverti-
culum (fig. 8*), which Burckhardt calls the recessus neuro-
poricus. The same region is distinguished by His as the
angulus terminalis, and is regarded as the dorsal limit of
the frontal suture line, and the last point to lose its connec-
tion with the ectoderm. Immediately dorsal to it there is a
1 Further and more detailed examinations of the cerebrum of the adult
Ornithorhynchus and a number of reptiles since this paper was written
have demonstrated a much closer resemblance between adult and foetus than
this paper indicates, as well as the marked similarity of both to the higher
Reptilian condition. A more detailed account of the region of the commis-
sures will be published in a short time.
188 G. ELLIOT SMITH.
small transverse layer (/. sup.), which Burckhardt calls the
lamina supra-neuroporica, and describes as ‘ ein kurzer
ependymatoscr Abschnitt, . . . der erste Abschnitt der Scheitel-
platte.” All authorities now seem to agree in regarding this
as the anterior extremity of the roof (Deckplatte). This
must be admitted if the angulus terminalis is really the dorsal
extremity of the “ vordere Schlussnaht.” The lamina supra-
neuroporica takes a sudden bend backwards (fig. 3) to form a
horizontal band, which gives origin in many lowly Vertebrates
to the plexus inferiores, and in higher animals to the
plexus laterales as well, or exclusively. In the specimen
under consideration, however, although the plexus laterales do
not actually spring from this lamina, they are formed from the
caudal prolongations of its lateral parts (vide fig. 15) on
either side of the paraphysis (par.). Immediately dorsal to
this lamina the anterior wall of the third ventricle is bulged
out to form a large sac (fig. 2, par.), which constitutes the
paraphysis of Selenka, of which a fuller description will be
given below,
The whole extent of the roof (actual) of the third ventricle
as far back as the superior commissure (fig. 2, s.c.) is deeply
invaginated to form a complex choroidal fold (fig. 2, ch. 3; also
figs. 9—15), which later on forms the “ diaplexus” of Wilder.
It is evident, therefore, that no Zirbelpolster of Edinger (par-
encephalon, prepinealis Zwischenhirndach) exists in this brain,
since the velum or origin of the median plexus extends right
back to the superior commissure. The latter is a very small
band of transverse fibres connecting the two ganglia habenule.
Immediately behind the superior commissure is the epiphysis
cerebri, which consists of a small solid clump of cells (fig. 2,
ep.) with deeply stained nuclei, lying behind a small recessus
pinealis and upon the posterior commissure (p.c¢.). The latter
is a broad band of transverse fibres, and is new generally re-
cognised as the limit of the diencephalon.
The constricted anterior pole of each hemisphere is distin-
guished as an olfactory bulb by the application of the olfac-
tory ganglion to its periphery (figs. 18 and 4, olf. g., J.g.).
THE BRAIN OF A F@TAL ORNITHORHYNCHUS. 189
On tracing the cerebrum caudally in a series of sections, the
basal aspect of the olfactory bulb will be found to be con-
tinuous with two regions which already present distinct histo-
logical features, although there is little superficial indication
of any such division. These two regions are the tuberculum
olfactorium (figs. 6 and 7, ¢.0.)—a relatively iarge basal
area next to the middle line—and the pyriform lobe, whose
surface is covered by the already well-developed ‘ external
olfactory radiation” (e.0.7r.) of Edinger. The tuberculum
olfactorium only extends back as far as the lamina terminalis,
but the pyriform lobe extends the whole length of the hemi-
sphere (figs. 6—13, pyr. and e.o.r.). In the adult the pyri-
form lobe is limited laterally by a well-defined fissure (fissura
rhinalis or ectorhinalis of Turner)—figs. 8 and 14, f.r.,—
and is separated from the tuberculum olfactorium by a deep
fissure (fig. 8, f.er.)—fissura endorhinalis of Turner. In
the foetus there is no sign whatever of the fissura ectorhinalis,
although a shallow endorhinal groove (fig. 10, f. er.) is already
to be made out. ‘This appears to correspond to what His has
wrongly called the rhinal fissure. If this is the case, he has
excluded from his ‘ lobus olfactorius” the pyriform lobe,
which does not become superficially distinguished from the
pallium (Turner) until late in ontogeny. In Notoryctes,
even in the adult brain, no rhinal fissure appears to be present,
and in Perameles nasuta it is very shallow and imperfect.
The rhinal fissure appears late, not only in ontogeny, but also
in phylogeny. It appears, therefore—and in this I am sup-
ported by His’s figures—that the so-called rhinal fissure of the
early human foetus cannot be the ectorhinal—as His believed
—but the endorhinal fissure.
The tuberculum olfactorium is not confined to the basal
aspect of the cerebrum, but crosses the ventro-mesial angle to
become continuous with the ‘‘precommissural area of the
mesial cortex ” (figs. 6 and 7, p.a.). In the foetal brain there
is no line of demarcation between these two regions, since they
present exactly similar histological features, but in the adult
(fig. 8 *) there is a clearly defined boundary line. The “ pre-
190 G. ELLIOT SMITH,
commissural area” extends from the mesial aspect of -the
olfactory bulb in front to the lamina terminalis behind (vide
figs. 2 and 4, p. a.). The tuberculum olfactorium (locus per-
foratus anticus) and the precommissural area (gyrus subcallosus
of Zuckerkandl) correspond to the “ posterior olfactory lobule”
of His. The whole of the mesial wall of the cerebrum
dorsalto the precommissural area forms part of the
rudimentary hippocampus (figs. 2, 6, and 7, hip.). It will
be seen, therefore, that the hippocampus extends as far forwards
as the olfactory bulb (the characteristic histological structure
may be recognised at a distance of 150 u behind the olfactory
ganglion). The part of the region in immediate relation to
the precommissural area is the rudimentary fascia dentata,
the dorsal part being the “ Anlage” of the cornu Ammonis.
As it is traced backwards the hippocampus is found to pass to
the dorsal side of the foramen of Monro (fig. 15) and choroid
fissure, the ependymal choroid fold being attached to the
ventral margin of the ‘‘ Anlage” of the fascia dentata. The
hippocampal rudiment is co-extensive with the fissura choroidea,
and forms its dorsal boundary (figs. 9 to 13). In the greater
part of its extent this hippocampal “ Anlage ” is bulged into
the lateral ventricle, giving rise to a shallow trough (the
“ Bogenfurche” of Schmidt, fissura arcuata). To this
trough the name “hippocampal furrow” (“ Ammons-
furche ” of Mihalkovics) may very appropriately be applied.
It has an exact morphological significance, being strictly con-
fined to the region which is to become hippocampus. It is
clear, therefore, from the relations of the anterior extremity
of the hippocampus, that the true Bogenfurche cannot form
the ‘ fissura prima” of His.
The position occupied by the hippocampus in this foetus—
entirely dorsal to the foramen of Monro and fissura choroidea
—is typical of its true and essential relations in the whole
Vertebrate series. The recognition of the homologous region
in the sub-mammalian brain constitutes the key to the whole
question of the nature of the commissures, because all the
commissural fibres arising in this area must belong to the
THE BRAIN OF A F@TAL ORNITHORHYNCHUS. 191
hippocampal or fornix commissure. In the adult Platypus
there is no difficulty whatever in recognising the hippocampus,
since its histological features, whether studied by aniline blue-
black staining or by the methods of Weigert or Golgi, exactly
resemble those of the higher mammal (figs. 8 and 14). Upon
this fact largely depends the extreme morphological import-
ance of the cerebrum of Ornithorhynchus. For while the
topographical relationship of the Prototherian hippocampus
resembles that of the Sauropsida, it at the same time presents
quite as high a histological differentiation as is found in the
Eutheria—a fact which renders its identification quite certain
and easy. By comparison it is easy to identify the correspond-
ing region in the foetal brain, where the simplicity of its
histological structure (fig. 15) and the absence of the com-
plicated foldings and inrolling of the adult structure vividly
recall the Sauropsidan hippocampus, and at a glance convinces
one of the homology. In all adult Mammalia the pallium
takes some share in the formation of the mesial cortex (figs. 8
and 14, p.), but in the foetus the hippocampus extends right
up to the supero-mesial angle, so as to exclude the pallium from
any share in the formation of the mesial wall. Similarly in
the reptile the hippocampus forms the greater part of the
mesial wall, and extends forwards above the commissures tc-
wards the anterior pole of the cerebrum, as Edinger and
Meyer have described. Not only does it form the mesial wall,
but it extends on to the dorsum of the cerebrum. Herrick,
by attempting to draw too close a comparison with the
Eutherian brain (where this region is greatly disturbed by the
development of a corpus callosum), would limit the hippo-
campus to the posterior extremity of the hemisphere. This
is a gratuitous assumption, utterly unsupported by any evidence.
The whole of the dorsal and lateral aspects of the surface
of the hemisphere, from the hippocampus above to the pyriform
below, constitutes the pallium (Turner). Both in its onto-
genetic and phylogenetic history it continues to increase in
extent after the more ancestral hippocampal and pyriform
regions have reached their full development, both in size and
192 G. ELLIOT SMITH.
differentiation of structure. Thus on the one hand it en-
croaches upon the mesial aspect of the cerebrum, and has a
marked influence in the production of the complicated in-
rolling of the hippocampus; on the other hand, it tends to
overlap the pyriform lobe, producing the rhinal (ectorhinal)
fissure (fig. 14, fr.).
The corpus striatum (fig. 10, ¢. s.) is a large mass, which is
closely related ventrally to the tuberculum olfactorium and
pyriform lobes. Below the foramen of Monro it is closely
connected (fig. 10) with the lamina terminalis, and behind the
foramen it is connected with the “ Anlage” of the optic
thalamus, which lies internal and entirely dorsal toit. The
internal capsule is a very insignificant tract, consisting of a
few scattered fibres (fig. 11, ¢.2.), which are only to be found
in one region. Even in the adult (vide fig. 8) it forms a very
scattered and relatively small system, which is indicative of
the intermediate position of Platypus between the Sauropsida,
which have no proper internal capsule (Meyer), and the mar-
supials, where it is a well-formed fibre tract.
Bebind the foramen of Monro the pyriform lobe (figs. 13
and 14, pyr.) is directly continuous with the corpus striatum
(c. s.), the connecting bridge being the rudiment of the nucleus
amygdalz (nx. a.), which in the foetus gives rise to an enormous
tenia semicircularis (¢.s.), or stria terminalis,
The foramen of Monro forms a rounded aperture situated
upon the lateral wall of the anterior extremity (Endhirn of
His) of the median cavity of the fore-brain, about midway
between the roof and the floor (fig. 2, f.m.). To be more
exact, it opens into the lateral ventricle from a small recess
of the third ventricle’ above the lamina infra-neuroporica,
which is bounded anteriorly by the angulus terminalis (fig.
3, 7. n.) and lamina supra-neuroporica (/. swp.), and dorsally by
the horizontal extension backwards of that lamina. Into the
dorsal aspect of this recess (fig. 15, a.) and in front of the
foramen of Monro, the paraphysis opens (par.). The foramen
of Monro is bounded dorsally in the whole of its extent by a
99 ¢¢ 53 ¢¢
1 © Ventriculus communis, ventricle of telencephalon, aula.”
THE BRAIN OF A FQITAL ORNITHORHYNCHUS. 193
large mass of neuroblasts (fig. LO, ¢h.), the rudimentary optic
thalamus. As this mass is traced caudally (figs. 11, 12, and
13) it gradually increases in size, forming a large bulging of
the Fltiigelplatte in the posterior part of the diencephalon.
Where it forms the dorsal boundary of the foramen of Monro
(fig. 11) this thalamic rudiment consists of three distinct
parts—a mesial ependymal layer, an intermediate neuroblastic
mass, and a lateral nerve-fibre layer. ‘The connections of the
latter cannot be accurately traced, but it appears to be the
tenia thalami.
As this thalamic rudiment is traced forwards above the
foramen of Monro the neuroblastic mass gradually diminishes
in size, until at the extreme anterior limit of the foramen of
Monro (fig. 15) it disappears altogether, only the ependymal
and nerve-fibre layers being left. The nerve-fibre layer also
soon disappears, leaving the ependymal layer only.
The point at which the thalamic rudiment ceases corre-
sponds exactly to the anterior limit of the foramen of Monro,
so that the latter is separated from the roof (Deckplatte) by
the whole width of the rudimentary optic thalamus.
The foramen of Monro is further bounded above by the
epithelial fold which forms the rudiment of the choroid plexus
of the lateral ventricle, and which is attached to the ventral
border of the fascia dentata (fig. 15, f. d.).
Below, the foramen is bounded by the lamina infra-neuro-
porica (figs. 2 and 10), which here becomes continuous with
the corpus striatum (vide fig. 10). Posteriorly the foramen is
bounded by the junction of the thalamic rudiment with the
corpus striatum (fig. 11). Anteriorly the boundary is formed
(vide the right side of fig. 15) by the fusion of the lamina
terminalis with the ventral border of the fascia dentata. If,
therefore, we leave out of consideration the choroid plexus of
the lateral ventricle, it is evident that the foramen of Monro
has no immediate relation to the Deckplatte in this brain.
Before discussing the relation of the hemisphere to the Deck-
platte the paraphysis must be described.
VOL. 89, PART 2.—NEW SERIES. N
194 G. ELLIOT SMITH.
The Paraphysis.
In describing the structures met with in a median sagittal
section it was mentioned that the dorsal part of the (actual)
anterior wall of the median cavity of the fore-brain was bulged
out to form a large sac. The corresponding structure is well
seen in the early embryo of Perameles (fig. 16, par.). Ven-
trally it opens into the cavity of the “telencephalon” of His
by a narrow neck. Posteriorly its lumen becomes suddenly
restricted by two large bulgings from the lateral walls into
the cavity. These bulgings are the anterior extremities of the
optic thalami (fig. 17). On comparing the appearance of the
Perameles brain with that of Ornithorhynchus (fig. 15)
the general resemblance is at once apparent, the only essential
difference being the presence of a large complicated median
fold of the roof of the paraphysis in Platypus. ‘This is the
anterior extremity of the ‘‘ Anlage” of the plexus choroideus
medius—(ch. 3), the ‘‘diaplexus”’ of Wilder. In the early
foetal brain of Perameles the median choroidal fold of the
third ventricle is not yet developed, although that of the
lateral ventricle (fig. 16) is well formed, so that the velum
and roof of the paraphysis form a simple transverse band. In
the Platypus embryo, however (fig. 2), a well-developed cho-
roidal fold extends from the superior commissure to the
lamina from which the lateral plexus arises, completely inva-
ginating the paraphysis (figs. 7, 9, and 15) in the middle line.
In Platypus the transition from optic thalamus to paraphysis
is a very gradual one, so that in examining a series of coronal
sections the lateral walis of the diverticulum would seem to be
merely the forward continuation of the ependymal layer of the
Fliigelplatten (fig. 15). This structure has attracted a con-
siderable amount of attention since it was first noticed, and
numerous names have been applied to it, and hypotheses
1 «Vir die vordere Epiphyse, dem Adergeflechtknoten Goettes
den Burckhardt in friiheren Arbeiten zuerst Plexus choroideus su-
perior, spater Conarium genannt hatte, braucht er neuerdings mit Kupffer
die von Selenka herrithrende Bezeichnung Parapliysis.’’—FRonizp,
THE BRAIN OF A’ F@TAL ORNITHORHYNCHUS. 195
advanced to explain it. As yet, however, its significance is
veiled in mystery. In this brain it is enclosed in a sheath of
dura mater formed by the splitting of the falx cerebri (fig. 7,
fe.), and it is closely surrounded by a number of very large
vessels. From the close relation of the organ to these large
vessels, it has been suggested that it may have some nutritive
function before the “ diaplexus” is developed (Gage).
The Relation of the Roof (Deckplatte) to the
Third Ventricle and Foramen of Monro.
In describing the foramen of Monro, it has already been
seen that the Deckplatte (“‘Scheitelplatte” of Burckhardt) is
separated from its dorsal aspect by the whole width of the
thalamic rudiment. The foramen opens into the small dorsal
recess (fig. 15, d.) of the third ventricle which is bounded an-
teriorly by the angulus terminalis and the lamina supra-neuro-
porica. The latter structure is considered to form part of the
Deckplatte, and the continuation backwards of its horizontal
part in this specimen (fig. 3) gives rise to the choroid plexus
of the lateral ventricle (fig. 15). In fact, it is generally con-
sidered that the lateral plexus is a derivative of the Deckplatte
(Minot). IRf this be so, it must be admitted that the foramen
of Monro and the hemisphere are in immediate relation to the
Deckplatte. Ifthis be so, the different parts of the morpho-
logical roof will be seen to be widely separated in the region
of the foramen of Monro (fig. 10). Above there is the un-
doubted roof of the ‘‘aula;” whereas laterally, between the
thalamic rudiment, and the hippocampal rudiment, is the
lateral plexus whose epithelial covering is supposed to bea
derivative of the Deckplatte (Minot). If, however, the hemi-
sphere, as is generally supposed, extends up to the Deckplatte,
how does the optic thalamus come to occupy the position
which is roughly represented in fig. 10, ¢h.? Kven admitting
that the choroid plexus of the hemisphere is a derivative of the
roof which has extended backwards, invading, as it were, the
mesial hemisphere wall, ouly two alternatives appear to remain
196 G. ELLIOT SMITH.
to explain the position of the thalamus in this specimen.
Hither the hemisphere brain develops from the lateral aspect
of the brain-tube on the ventral side of the anterior extremity
of the thalamus, or the latter in its growth invades the Deck-
platte. All the appearances, both in this brain and in those
of all the Perameles and Macropus specimens, are against
the latter hypothesis. It would appear that the hemisphere
develops on the ventral side of the anterior extremity of the
thalamus.) The lateral walls of the paraphysis in this speci-
men look deceptively like anterior extensions of the Fliigel-
platten.
Taken as a whole, however, the general disposition of parts
in the brain under consideration can be harmonised with the
account of His. On tracing the pars subthalamica (Forel) or
hypothalamus (Waldeyer) forwards in the series of sections, it
will be found to become continuous with a large mass lying on
the mesial side of the corpus striatum and on the ventral side
of the optic thalamus. This mass (fig. 11, *), which forms
the lateral wall of the ventral part of the third ventricle, on
being traced forwards will be found (fig. 10) to become con-
tinuous with the lamina infra-neuroporica. What is this mass?
According to the hypotheses of Burckhardt and v. Kupffer it
must be considered simply as the anterior extremity of the
Grundplatte ; whereas, according to His, it must be the anterior
end of the Fitigelplatte which has become bent towards the
base of the brain. In fig. 11 the fibres of the external capsule
(c. e.) will be seen extending into the mass to cross the middle
line (fig. 10) in the lamina infra-neuroporica, so that the mass
in question is divided into two parts—a ventral, which in the
adult is easily recognised as the tuberculum olfactorium ; and
a dorsal, which in the adult is corpus striatum. In the adult
the anterior extremity of the optic thalamus fuses with the
lamina infra-neuroporica (‘septum pellucidum”) above the
anterior commissure, so that the combined mass appears to
rest upon the anterior commissure. It is almost impossible
1 Without earlier developmental stages no opinion can be expressed con-
cerning these questions (25th April, 1896).
THE BRAIN OF A FQ@ITAL ORNITHORHYNCHUS. 197
(fig. 8) to draw any line of demarcation between corpus
striatum and tuberculum olfactorium. This being the case,
this region (fig. 11, *) may doubtfully be homologised with
the corpus striatum, as His has done. It will be noted from
the figures that the transition from undoubted hypothalamus
to this doubtful region takes place at the optic chiasma—the
termination of His’s ‘‘ Grenzfurche.” If, therefore, it is granted
that this doubtful area (*) is corpus striatum, it follows that
the hemisphere brain extends down as far as the optic recess,
as His has described in the case of the human feetus.
Having now given a general account of the disposition of
parts, there are certain regions which need further consideration.
The Olfactory Bulb.
This part of the cerebrum, whose structure is roughly re-
presented in transverse section in fig. 18, consists of four
distinct layers, as Vignal! has pointed out. Immediately sur-
rounding the ventricle there is a thick mass of small cells with
deeply stained nuclei. This is the ependymal layer (e).
Surrounding it there is a stratum of about the same thickness,
but much lighter in appearance—the layer of nerve-cells
(6). Examined with the high power, this is found to consist
of amass of well-formed neuroblasts (fig. 19). Posteriorly this
layer is found to be continuous with the essentially nervous part
of the cerebral cortex. It is probable that in the adult the
neuroblasts which compose this layer mostly develop fibres of
the pars olfactoria of the anterior commissure. The most
superficial layer is imperfect, consisting of two distinct gan-
glionic masses, each composed of neuroblasts like the second
layer. The larger mass (olf. g.) covers the whole inner surface
of the bulb, and extends on to the ventral and slightly on to
the dorsal aspects of the bulb. In this region the true olfactory
nerves terminate. Hence I have distinguished it as the
“olfactory” ganglion. On the outer aspect of the bulb there
is a second smaller mass (J. g.), presenting’ exactly similar
histological features. In this mass the fibres derived from
1 Quoted in Minot’s ‘Human Embryology.’
198 G. ELLIOT SMITH.
Jacobson’s organ terminate. Hence I have distinguished it as
the “ganglion of Jacobson’s organ.” The further
significance of this arrangement I have discussed in the
‘ Anatomischer Anzeiger.’
Separating both ganglionic masses from the second or nerve-
cell layer there is a distinct layer of nerve-fibres (y), which
form the root bundles of the “ olfactory [internal and external |
radiations” of Edinger. It corresponds to the molecular layer
or “ Randschleier” of the cortex cerebri.
The surface regions of the hemisphere are already differen-
tiated into two histologically distinct districts, (1) a basal
region (comprising the tuberculum olfactorium and the pre-
commissural area), which presents an amorphous and un-
differentiated structure; and (ii) a dorso-lateral region—the
pallium,—which already exhibits characteristically specialised
features. Interposed between these regions on either side
there are two districts—the hippocampus and pyriform lobe—
which present intermediate or transitional features. These
divisions have a further and much wider significance. The
undifferentiated basal regions—including not only the pre-
commissural area and tuberculum olfactorium, but also the
prosencephalic part of the olfactory bulb, the corpus striatum
and lamina terminalis (“septum lucidum ’”’)—are the ancient
parts of the brain phylogenetically, and develop early in
ontogeny. The transition regions appear next in phylogeny—
the hippocampus and then the pyriform lobe. The highly
differentiated pallium appears last, and continues to increase
in size and in histological specialisation long after the other
regions have reached the acme of their evolution, or are even
in their decline. Thus in the Prototheria the hippocampus
and pyriform have already reached the height of their develop-
ment, whereas the pallium progressively increases in size and
complexity in the Hutherian series up to Homo, in whom the
lowest stratum—the olfactory—the tuberculum olfactorium
and precommissural area are disappearing.
THE BRAIN OF A FQTAL ORNITHORHYNCHUS. 199
The Pallium.
Five distinct layers are recognisable in the pallium (fig. 20).
These are arranged in the following order—proceeding from
the surface, and comparing with the layers of the olfactory
bulb.
Pallium (fig. 20). Olfactory bulb (fig. 18).
suence 1. The olfactory ganglion
(an added layer).
erlbesmolecule player 7) x carunsen -sntiosede: 2. The nerve-fibre layer (y).
2, The superficial nerve-cell or “mantle”
layer (e)
3. The middle nerve-fibre layer (@)
4. The deep nerve-cell layer (c)
Depth evependymal layer {(@) cues ade acacieeent' ee ependymal layer (e).
pan
...8. The layer of nerve-cells (6).
The superficial or molecular layer of the pallium forms
a clear peripheral band containing only scattered cells and
nerve-fibres, It is the rudiment of the molecular layer of the
adult cortex, and also appears to correspond to the ‘ Rand-
schleir.” In the pyriform lobe and hippocampus (fig. 10) it
becomes greatly thickened. In the basal regions it does not
exist as a distinct layer.
The superficial nerve-cell layer (e)—the mantle layer
—consists of a very regular stratum of densely packed neuro-
blasts, from which the most important constituents of the
cortex develop.
The middle nerve-fibre layer (d) contains a large
number of nerve-fibres and scattered neuroblasts. It appears
to be the ‘ Anlage” of the corona radiata, although in the
region of the corpus striatum it is separated from the corona
proper by a large mass of cells.
The deep nerve-cell layer (c)—the Rolando-cell layer of
Ludwig Lowe—is more or less intimately joined to the epen-
dymal layer (a), but a slight interval (b) where the cells are
not so closely packed indicates the line of separation. The
destiny of this layer is uncertain,
200 G. ELLIOT SMITH.
“The Precommissural Area of the Median Cortex.”
The general relations of this region have already been
described. In a scheme (fig. 5) which indicates the hypo-
thetical arrangement of parts in the anterior wall of the
primitive brain-tube it will be seen that the olfactory bulbs
(olf. g), situated at each lateral angle, are separated by a
broad plate of nervous tissue. With the development of
the hemispheres this plate becomes bent forwards on either
side of the middle line. The lateral part, which now forms a
sagittally placed wall (p.a.), is the “ precommissural area,”
and the median band connecting the two “ aree ” is the lamina
terminalis (/.¢.), which later develops into the septum lucidum.
His calls the “ precommissural area” part of the “ olfactory
lobe.” In fact, as the homologue of Zuckerkandl’s “ gyrus
subcallosus,” it must form part of His’s “ posterior olfactory
lobule.” On morphological and histological grounds, however,
the precommissural area cannot be separated from the rudi-
ment of the “septum lucidum” (compare figs. 7 and 9), both
having equal claims to be considered as parts of the “ olfactory
lobe ” or rhinencephalon.
The Hippocampus.
As the pallium (fig. 15, p.) is traced round the supero-mesial
angle of the hemisphere the cellular elements will be found to
undergo a striking rearrangement. The molecular or super-
ficial layer (fig. 15) will be found to become very much broader
than it is in the pallium, and, by comparison with later stages
of Perameles, is readily recognised as the homologue not
only of the molecular layer, but also of the stratum lacunosum
and stratum radiatum of the adult hippocampus. The cells
of the superficial nerve-cell layer lose their densely packed
arrangement, and the cells of the second, third, and fourth
layers become directly continuous with a column of somewhat
scattered cells, which produce the pyramidal and polymorphous
cells of the hippocampus. As this cell column approaches the
margin of the hemisphere the cells become more densely packed
THE BRAIN OF. A F@TAL ORNITHORHYNCHUS. 201
again to form the rudimentary fascia dentata (fig. 15, f. d.).
Since, in the hippocampus, the cells of the fourth layer of the
pallium separate from the ependymal layer, the latter appears
as a much more sharply defined layer (fig. 15). Between the
ependymal layer and the irregular column of neuroblasts there
is a relatively wide nerve-fibre layer—the alveus. This is the
homologue of and is continuous with the fifth layer of the
pallium.
This simplicity of hippocampal formation presents a close
resemblance to the corresponding region in the reptile brain.
The anterior extremity of the hippocampus in all non- placental
mammals maintains this primitive simplicity of structure into
adult life (fig. 8). The much-disputed questions of homology
will be discussed later on in dealing with the histology of the
adult organ.
The fornix system is as yet imperfectly formed, and presents
a reptilian-like simplicity of arrangement. Although the de-
scending fibres of the fornix (figs. 7,9, 15, d.f.) are already
well formed, there are only a few scattered commissural fibres
in the lamina infra-neuroporica.
The Pyritorm Lobe:
As the pallium is traced into the pyriform lobe (figs. 6, 7,
9—12) the molecular layer suddenly increases in thickness.
This layer contains the fibres of the external olfactory radia-
tion (e. 0. 7.).. The superficial nerve-cell layer suddenly
becomes more scattered, and loses the regular appearance
_ found in the pallium proper. This layer corresponds to the
layer of “double pyramids” (Kolliker) found in the adult
pyriform lobe (fig. 8). The other layers of the pallium are
represented in the pyriform lobe merely by diffusely scattered
nerve-cells. Behind the lamina terminalis the pyriform lobe
is continuous mesially with a large mass of diffusely scattered
nerve-cells—the nucleus amy gdale.
The Corpus Striatum and Optic Thalamus.
The corpus striatum (c.s.) forms a large mass projecting
into the lateral aspect of the lateral ventricle, and lying in the
202 G. ELLIOT SMITH.
concavity of the external capsule (c.e.).Anteriorly it is
directly continued into the tuberculum olfactorium (figs. 7
and 8), which Ganser calls the “ cortex of the head of the
corpus striatum.” Posteriorly it is closely connected with
the pyriform lobe (figs. 12
14) through the intermediation
of the nucleus amygdale. Except in the one region imme-
diately behind the foramen of Monro (fig. 11), there is no sign .
of any division into caudate and lenticular nuclei by an in-
ternal capsule.
Histologically the rudiment of the corpus striatum presents
numerous well-formed neuroblasts, which give rise to nerve-
fibres, whose connections will be described immediately.
The optic thalamus presents in all sections a large mass of
neuroblasts, which varies in size and appearance in different
sections (figs. 1O—13). The only other structure yet differ-
entiated in the thalamic region is the ganglion habenule
(fig. 13, G23).
The great thalamic nucleus, which appears to cor-
respond to the “ centre mediane de Luys ” (Lowe and Minot),
is relatively very largely developed in this foetus, and, as may
be seen from fig. 14, this precocious development is an indica-
tion of the huge proportions of the adult optic thalamus,
whose enormous growth in the lateral direction reduces
the corpus striatum in this region to a very thin lamella
(fiz. 14, ¢. 8.).
Connecting the great thalamic nucleus and the corpus
striatum there is a huge bundle of fibres (fig. 12) which stands
out extremely clearly, since there is practically no internal
capsule to obscure it. By means of this fibre tract, which
apparently corresponds to Edinger’s ‘‘radiatio strio-thala-
mica,”’! a very intimate connection is established between the
corpus striatum and cortex cerebri (pallium) on the one hand,
and the ‘great thalamic nucleus”? on the other. A large
proportion of these fibres appear to arise from neuroblasts
in the corpus striatum. Others appear to spring from the
thalamus.
1 Only in part.
THE BRAIN OF.A F@TAL ORNITHORHYNCHUS. 203
The following anatomical facts concerning the brain of
Platypus appear to be associated with one another :—an enor-
mous trigeminal nerve; a huge development of substantia gela-
tinosa Rolandi in pons and medulla, and fillet in the mesence-
phalon; a large and precocious development of the optic
thalamus (where in the adult the great bulk of the fillet ter-
minates), and an exceptionally abundant supply of fibres con-
necting the latter with the corpus striatum and cortex cerebri.
The important bearing of this upon the vexed question of the
termination of the fillet (lemniscus) will be discussed later on
when dealing with the adult. It may bestated here, however,
that an examination of the Ornithorhynchus brain clearly
shows that the great majority of the fillet fibres terminate by
terminal arborisations around the large multipolar cells of the
optic thalamus, which, in turn, are connected with the cortex
cerebri and corpus striatum.
The huge stria terminalis, which arises from the neuroblasts
in the nucleus amygdalz, extends vertically upwards behind
the “‘radiatio strio-thalamica,” and then takes a sudden bend
into the horizontal plane (fig. 12, ¢.s.). The picric stain does
not permit one to accurately trace its anterior connections.
The account of the optic nerve and its central connections
will be deferred until the rest of the brain is under considera-
tion. Before then I hope to examine a series of foetal Echidne,
and to be able to supplement these fragmentary and imperfect
notes.
Literature.
The last few years have yielded an unusually rich and abun-
dant contribution to the knowledge of the difficult subject of
cerebral morphology; and in spite of the different interpreta-
tions of the vast mass of data collected, there is, beneath all
the conflicting statements, a remarkable degree of uniformity
in the essential facts of cerebral development and phylogeny.
No purpose would be served in a descriptive paper such as
this in quoting all this bulky literature. In a paper entitled
“ Die Homologien des Zwischenhirndaches und ihre Bedeutung
204 G. ELLIOT SMITH,
fir die Morphologie des Hirns bei niederen Vertebraten,”
which appeared in the ‘ Anatomischer Anzeiger,’ Bd. ix, Nos.
5 and 6, p. 152, Rud. Burckhardt began a discussion, which
was vigorously carried on in the same journal by Studnicka
and Rabl-Riickhard. Before the Anat. Gesellschaft W. His
discussed the same subject. His paper was subsequently
published in the ‘Arch. f, Anat. u. Entw. ; 1898, p. 157, “ Uber
des frontale Ende des Gehirnrohres.” In the same number
His published a paper on nomenclature, ‘ Vorschlage zur
Eintheilung des Gehirns.”
C. von Kupffer’s contribution to the argument is found in
his ‘Studien zur vergl. Entwickelungsgeschichte des Kopfes
des Kranioten,’ and also in the report of the Anat. Gesellschaft.
An admirable summary of Burckhardt’s views is found in
his paper entitled “‘ Der Bauplan des Wirbelthiergehirns” in
‘Morpholog. Arbeiten, hrsg. v. G. Schwalbe, Bd. iv, 1894,
p. 131.
Critical digests of the literature will be found in Froriep’s
paper, “‘ Entwick. des Kopfes” (‘ Ergebnisse der Anatomie und
Entwickelungsgeschichte,’ hrsg. v. F. Merkel und R. Bonnet,
Bd. iii, 1893); also the papers of Sorenson and Herrick in
recent numbers of the ‘ Journal of Comparative Neurology.’
To Dr. L. Edinger I am indebted for a copy of his paper
entitled “Die Faserung aus dem Stammganglion Corpus
Striatum ” (‘ Verhdl. Anat. Gesellschaft,’ May, 1894), contain-
ing an account of the “ radiatio strio-thalamica.”
EXPLANATION OF PLATE 11.
Illustrating Dr. G. Elliot Smith’s paper on ‘ The Brain of a
Foetal Ornithorhynchus.”
Fic. 1.—Diagrammatic representation of the brain of foetal Ornitho-
rhynchus, reconstructed from serial sections X 6. The Zwischenhirn (d/.),
which is shaded, is represented as though the cerebrum were trausparent.
The oval cerebral hemisphere is seen to slightly overlap the mesencephalon
(mes.). 6.0. Bulbus olfactorii. J. Jacobson’s organ. 2.J. Nerves from
THE BRAIN OF A F@TAL ORNITHORHYNCHUS. 205
Jacobson’s organ. n.olf. Olfactory nerves. opt.x. Optic nerve. 2.3,
Oculo-motor nerve. G.g. Gasserian ganglion.
Fie. 2.—Scheme of median sagittal section of the fore-brain, the hemi-
sphere indicated by dotted lines. 6.0. Bulbus olfactorit. p.a. The “ pre-
commissural area of the median cortex.” fdp. The rudiment of the hippo-
campus (including the fascia dentata). p.c. Posterior commissure. e.p.
Epiphysis cerebri (pineal). s.c. Superior commissure. vel, Velum. ch.3.
Diaplexus”’ of Wilder. par. “ Paraphysis” of Selenka. 2 J/, Foramen of
Monro. a.c. Commissura anterior, J/. ézf. Lamina infra-neuroporica., 7. 0.
Recessus opticus of His (preeopticus—Burckhardt). opé.z. Optic nerve. 7.2.
Recessus infundibuli of His (r. postopticus—Burckhardt). Ayp. Hypophysis
cerebri. pit. Pituitary gland.
Fic. 3.—Part of Fig. 2 enlarged.—/. sup. Lamina supra-neuroporica of
Burekhardt. 1.2. Recessus neuroporicus (angulus terminalis, lobus olfac-
toriusimpar). fc. Fornix commissure.
Fig. 4.—Scheme of horizontal section through a mammalian brain, to show
the relation of the ‘‘ precommissural area” (p.a.) to the lamina terminalis
(2.¢.) and olfactory bulb (4.0.). Olf.y. Olfactory ganglion. c.s. Corpus
striatum. ¢4. Diencephalon (His).
Fie. 5.—Scheme of hypothetical arrangement of the terminal plate of the
brain-tube, to show that the precommissural area and lamina terminalis are
parts of one structure. Compare with Fig. 4.
Fies. 6, 7, 9, 10, 11, 12, and 13 are a series of coronal sections of the
fetal Ornithorhynechus brain, each x 12. The hippocampal rudiment
(Aip.) may be seen extending through all the sections, lying in front above
the precommissural area (p.a@.), which in turn is continuous with the tuber-
culum olfactorium (¢.0.) and pyriform lobe (pyr.), on whose surface is the
“external olfactory radiation” (Hdinger), e.0.7. yp. Pallium. c.e. External
capsule.
lig. 6.—Counting the sections from before backwards (each section being
50 p thick), this is No. 24. It passes a short distance behind the olfactory
bulb.
Tig. 7.—(No. 35.) Just in front of the lamina terminalis. The falx cerebri
(fx.) splits above to enclose the paraphysis (par.) and a number of large
vessels. d./. Descending fornix fibres (Reichbundel).
Fic. 8.—Section through the corresponding region in the adult brain.
Weigert stain x 3. yp. Pallium. 4.f. Fissura hippocampi. f. d. Fascia
dentata. d.f. Descending fornix fibres. p.a@. Precommissural area, ¢. 0.
Tuberculum olfactorium. * Line of demarcation between p.a. aud Z. e.
Jer. Fissura endorhinalis (Turner). pyr. Pyriform lobe. e. 0.7. External
olfactory radiation. f.7. Fissura ectorhinalis. ¢.2. Internal capsule in
corpus striatum (c.s.). ¢.7. Corona radiata. alv. Alveus.
206 G. ELLIOT SMITH.
Fie. 9.—(No. 38). Passing just in front of the foramen of Monro on the
right side, and through it on the left. References as above.
Fie. 10.—(No. 47.) Through foramen of Monro (f 4). a.ec. Anterior
commissure. ¢h. Optic thalamus. Bf The “hippocampal furrow”
(Bogenfurche).
Fie. 11.—(No. 51.) Just behindthe foramenof Monro. /.f. Longitudinal
furrow. 7.0. Recessus opticus. pdf. Pituitary gland. G.g. Gasserian gan-
glion. c.¢. Internal capsule.
Fie. 12.—(No. 80.) y.4. Ganglion habenule. z.¢4. Nucleus thalami
(centre mediane de Luys). 7.s¢. ‘ Radiatio strio-thalamica” of Edinger.
¢.s. Stria terminalis (running horizontally). G@.g. Gasserian ganglion. Viz,
Fibres of fifth nerve proceeding from G.g. to the pons.
Fic. 13.—(No. 89.) z..a. Nucleus amygdala, from which the large stria
terminalis (¢.s.) arises and passes vertically upwards to cross on the dorsal
aspect of the radiatio strio-thalamica (7. s¢.). pos. The “ tuberculum quinti”
of the pons Varolii.
Fic. 14.—Section through the adult Ornithorhynechus brain corre-
sponding to Fig, 12 in the foetus. x 2. References as above. m.c. Middle
(soft) commissure. p. swb¢. Pars subthalamica (hypothalamus of Waldeyer).
Vie. 15.—Part of Fig. 9 enlarged. B/ Hippocampal furrow (Bogenfurche).
ch. 3. Median choroidal invagination (diaplexus). Ap. Hippocampus. /. d.
Fascia dentata. p. Pallium. c.s. Corpus striatum. d./. Descending fornix
fibres. 2. xf. Lamina infra-neuroporica.
Fic. 16.—Coronal section through the corresponding region of an early
foetus of Perameles nasuta. c#./. Choroid plexus of the lateral ventricle.
Fig. 17.—Section a short distance behind Fig. 16. ¢4. Optic thalamus.
J. M. Situation of foramen of Monro. Bf Situation of Bogeufurche.
Fic. 18.—Z.s. Olfactory bulb of foetal Platypus. x 25. olf.g. Olfactory
ganglion. J.g. Olfactory ganglion into which the nerves from Jacobson’s
organ are inserted. j. Nerve-fibre layer. d. Layer of nerve-cells. ¢. Hpen-
dymal layer surrounding the ventricle.
Vie. 19.—Neuroblasts (Zeiss D) from d, (Fig. 18).
Vic. 20.—Supra-ventricular cortex cerebri (pallium) of feetal Platypus.
x 96. a. Kpendymal layer. e. Deep nerve-cell layer (Rolando-cell layer of
Lowe). d. Middle fibre layer. ¢. Superficial nerve-cell layer (mantle layer).
J. Superficial fibre layer (molecular layer).
Fies. 6, 7, 9, 10, 11, 12, 18, 16, 17, and 18 were drawn by means of a
camera lucida. Fig. 15 was drawn from a microphotograph, for which I am
indebted to Professor Wilson.
ON ARHYNCHUS HEMIGNATHI. 207
On Arhynchus hemignathi, a new Genus of
Acanthocephala,
By
Arthur E. Shipley,
Fellow and Tutor of Christ’s College, Cambridge, and University Lecturer in
the Advanced Morphology of the Invertebrata,
With Plate 12.
ANATOMY.
In the summer of 1894 I received from Mr. Perkins, of Jesus
College, Oxford, seven small parasites which he had noticed ad-
hering lightly to the skin around the anus of a species of bird,
Hemignathus procerus, which he collected in the island of
Kauai, one of the Sandwich Island group. Each of these para-
sites was divided into three regions,—a head, a collar, and a
trunk ; and, in fact, they have an almost ludicrous resemblance
to a young Balanoglossus with one or two gill-slits (figs. 1, 11,
and 111). On investigating their anatomy it at once became
evident that the animals belonged to the group Acantho-
cephala, and, further, that they differed from the other
members of the group in the absence of what is perhaps their
most characteristic organ,—from which, indeed, they take their
name—the hooked proboscis or introvert. ‘The absence of so
characteristic a structure, and the fact that the parasites were
found outside the body, i.e. not as an endoparasite, but as an
ectoparasite, lightly attached to the skin, made me think that
perhaps the hooked introvert had been left behind in the in-
testine of the host, and that the body of the parasite had passed
out of the alimentary canal of the bird. However, careful
inspection failed to reveal any trace of a scar or mark where
the introvert might have been broken off; and although in
208 ARTHUR E. SHIPLEY.
the absence of hooks and introvert sheath, &c., the anterior
part of the body which I have called the head is as unlike the
typical introvert as possible, still in its relation to the lemnisci
and to the ligament it occupies the position of that organ,
and until we can get further information I think the best
plan is to regard this part of the body as equivalent to the
eversible part of more normal forms.
The second of the three regions into which the body is
externally divided is shorter than the head and smaller in
diameter; it may be termed the collar. The third or poste-
rior region, which may be called the trunk, is the longest and
the most slender of the three; behind it tapers to a point
where the orifice of the genital duct is situated, and this
end of the animal is always a little turned up (figs. 1, 0, 11,
and vit). The exterior of the collar and trunk are smooth
or lightly wrinkled, but the head is covered with a number of
small depressions or pits which give it a very characteristic
appearance, and which are well seen in sections. The head is
attached to the collar by a narrow neck, which is surrounded
and concealed by the edge of the collar. It is obvious in
sections (figs. v and x11). All the specimens were somewhat
shrivelled and apparently distorted. The largest measured
3°5 mm. in length, the smallest 2°> mm.; had they been fully
distended they would probably have been 1 to 1:5 mm. longer.
The body-cavity of the head is continuous with that of the
neck, and the latter opens freely into the cavity of the trunk
(fig. 13). The first-named space is by far the largest. The
lumen of the collar region is reduced by the great thickness of
the walls of this part of the body, and both here and in the
trunk much of the internal space is occupied by the lemnisci
and the reproductive organs.
The skin is one of the most characteristic features of the
Acanthocephala, and as far as I know is only paralleled by
that of the Nematodes, but it possesses certain features not
found in the last named group. The whole body is covered by
a thin cuticle which does not vary much in thickness in the
different regions of the body, and which is invaginated a short
ON ARHYNCHUS HEMIGNATHI. 209
distance into the genital pore. Beneath this is the true epi-
dermis, or subcuticle as itis called; this has in my specimens
the usual structure met with in the group so well described by
Hamann, and consists of a matrix of a fibrillar nature, the
fibrils being as a rule arranged radially, in which are embedded
a certain number of ameeboid nuclei (figs. vi and x). This tissue
is much thicker in the region of the collar than elsewhere, and
it is thicker in the trunk than in the head. It is pierced in all
directions by a series of tubes or lacunze which have no definite
lining, but which seem to be mere splits in the fibrillar matrix.
The lacune—except in the head—have a general circular direc-
tion which is very well marked in the trunk region where each
runs into a lateral longitudinal split (fig. x). They contain a
small amount of coagulum, the remnant of the fluid which
circulates in them ; during life this fluid, in other species, holds
in suspension fat and coloured oil globules. If these are
present in my species they must have been dissolved out in
the processes which precede embedding. The circular lacunze
of the trunk not only communicate with one another by means
of the two longitudinal lateral lacunee (fig. xiv), but they open
into one another by numerous small branches which have an
oblique or longitudinal direction. In the head the lacunz
have a general longitudinal course; they are not, however,
straight, but twist in and out between the pits on the surface ;
they anastomose freely (fig. iv). Thus in a transverse section
of the head the lacunze appear as round holes more or less uni-
formly arranged in the skin, and the same effect is produced
by a longitudinal section of the trunk.
In the collar region the subcuticular tissue is much thick-
ened, and the lacunar system forms a single more or less
definite ring which gives off numerous branching anastomosing
twigs (fig. v).
Although the above account attempts to give the general
course of the lacunz in the skin, it should be mentioned
that there is considerable irregularity in the arrangement,
and one is almost inclined to believe that the canals do not
remain permanent, but that they sometimes close up and
VoL. 39, PART 2.—NEW SER. O
210 ARTHUR E. SHIPLEY.
new ones appear. As they have no lining of any kind, such a
closing would leave no trace.
As Schneider,! Hamann,” and Kaiser® have shown in the
species investigated by them, the lacunar system of the intro-
vert is completely shut off from that of the neck—if it be
present—and of the trunk by a fold inwards of the cuticle
which cuts the subcuticular tissue in two. I have not been
able to find any such cuticular ring in the species in question,
but the state of preservation of my specimens does not allow
me to say definitely that it does not exist.
The lemnisci are two elongated sac-like prolongations of
the subcuticular tissue which are attached anteriorly to the
skin at the junction of the head and collar. They extend back-
wards to the extreme posterior end of the body, and are slightly
bent so that a longitudinal section may cut them in two or
three places (fig. x111). Histologically they are composed of
the same substance as the subcuticle in direct continuity with
which they arise, and they are traversed by a similar system of
canals. Physiologically they seem, as Hamann suggests, to
act as reservoirs for the fluid of the canal system of the intro-
vert; when the fluid they contain is forced into the spaces of
the introvert the latter is everted. It is withdrawn again into
the body by special muscles. In most species the canal system
of the lemnisci opens into that of the introvert in front of the
cuticular ring, and is thus completely independent of that of
the trunk. If we assume that the head of my species corre-
sponds with the introvert of other forms which have lost its
introvert sheath, the lemnisci open into the same region of
the skin as they do in other Acanthocephala.
The nuclei of tae subcuticle and of the lemnisci are very
remarkable; they cc:.espond in structure with those described
by Hamann in Neorhynchus claveceps, in which species
according to this observer both the skin and the lemnisci retain
in the adult their embryonic condition. As in Neorhynchus
1 «Arch. Anat.,’ 1868, p. 584.
2 «Die Nemathelminthen,’ Hefte 1 and 2, Jena, 1891 and 1895.
3 * Bibl. Zool.,’ Heft 7, 1892, p. 1.
ON ARHYNCHUS HEMIGNATHI. i |
the number of nuclei is very small, some twelve to twenty seem
to suffice for the whole of the subcuticle, and perhaps two to
four for each lemniscus. The structure of the nucleus shows a
most striking resemblance to an ameeba with rather short pseudo-
podia (figs. x, x11, and xtv)._ No single nucleolus can be de-
tected, but numerous chromatin particles are present, and
in some a distinct vacuole can be observed. These nuclei are
scattered about in a most irregular fashion ; not one may be
seen in a number of consecutive sections, and then perhaps
three or four may appear, and from their large size persist
through several sections. The nuclei lie, as a rule, embedded
in the substance of the subcuticle ; more rarely they are found
in the lacunz. Although there is no proof, one is tempted to
believe that the nuclei wander through the subcuticle and
lemnisci in an amceboid manner, and that the small number of
nuclei which are found in these tissues is compensated for
partly by the large size of each, but more especially by their
mobility. Similar amoeboid nuclei undoubtedly move about,
fuse with one another, and undergo fission in the subcuticle of
the larval forms of Neorhynchus claveceps.
Within the subcuticle and completing the skin on the inner
side, is a layer of circular muscles, and still more internally a
layer of longitudinal muscles (figs. vi and xv). The muscles of
these layers are but a single fibre thick, and they are not very
uniformly present. The circular layer is most complete in
the region of the trunk, and I have figured a section to show
this (fig. x11). The longitudinal layer is even less definite,
but scattered fibres can be detected here and there (figs. xiv
and xv). Hach fibre appears to be spindle-shaped, and in the
circular muscles has the striated portion only on its outer face,
forming a thin band; the inner half of the fibre consists of
vacuolated strands of protoplasm in which is a nucleus. The
longitudinal layer of muscles alone is continued over the
lemnisci (figs. 1x and xiv). These muscles are not covered
on their inner side by any layer of epithelial ceils, neither
does any such layer cover the ligament, but both tissues lie
freely exposed to the fluid of the body-cavity.
212 ARTHUR E. SHIPLEY.
In the more typical Acanthocephala the anterior end of the
body terminates in a hollow eversible portion provided with
rows of hooks whose number and shape have a certain sys-
tematic value. This introvert can be withdrawn, not into the
general body cavity, but into the cavity of the introvert sheath,
which is shut off from the general body cavity by a double
(chinorhynchide) or a single (Neorhynchidz) wall. The ex-
trusion of the introvert is believed to be effected by fluid being
forced into its lacunze by the lemnisci. It is retracted by
special muscles attached to the inside of its tip; besides these,
other retractor muscles run from the outside of the introvert
sheath, and these serve to retract the whole sheath and its
contents into the trunk. The chief nerve ganglion lies as a
rule on the posterior end of the introvert sheath, usually in
the middle line, but in the Gigantorhynchidz it is placed to
one side. From the posterior end of the introvert sheath, and
having its origin between its two walls when they are present,
the ligament runs backward, traversing the body cavity, and
ending in the funnel-shaped internal opening of the oviduct in
the female and in the vas deferens in the male.
Owing to the absence of an introvert and its sheath, the
relations of the ligament in the present species is somewhat
altered. It takes its origin from the anterior end of the head,
and at first seems to consist of a few strands of muscular
fibres which arise from the muscles of the skin (fig. x1). All
my specimens but one proved to be mature females, whose
ovaries had broken up into the egg masses which are character-
istic of the Acanthocephala. These egg masses consist of
packets of a dozen or more cells of which the peripheral layer
develop into ova at the cost of the central cells which serve
them as a food supply (figs. v1, x1, and x11). These packets
coexisted in my specimens with ova in various stages of
development, some without any egg shell, whilst others were
provided with a thick deeply-staining membrane. The whole
lumen of the head was crowded with these ova. In the region
of the collar the ova were confined by a thin-walled membrane,
and in the trunk there were two such masses of ova, which,
ON ARHYNCHUS HEMIGNATHI. 213
however, seemed less mature than those lying in the head.
Lying amongst the various organs in the body-cavity were
a number of very finely granular masses, which I take to be
the masses of spermatozoa (figs. vi and x). Of the complex
system by means of which the ova leave the body, little could
be made out beyond the fact that a well-marked funnel is
present opening into the posterior end of the body-cavity
of the trunk (fig. 1x). I failed, however, to find a second
opening near the narrow end of the funnel such as occurs
in other forms, but this may have been due to the poor state
of preservation. 'The funnel leads into a duct which opens
on the posterior end of the trunk.
The testes are two in number, and lie one behind the other
in the ligament, though owing to its looping both may appear
in the same transverse section. The spermatozoa do not
escape into the body of the male as the ova do into that of the
female, but pass down a duct in the ligament which opens
at the end of the body. Traces of accessory glands were seen,
but the details were not clear.
The brain lies on or in the ligament just behind its point
of attachment to the skin of the head (figs. x1 and x11).
Owing to the disruption of the ovaries in my female specimens
the ligament could not be traced very far, but in the only
male it reached from one end of the body to the other. ‘The
brain consists of a few large ganglion cells with a clear homo-
geneous cytoplasm and deeply-stained nuclei; the divisions
between the cells were very sharp and straight (fig. xr). In
the females this mass of cells lay in the ligament; in the male,
on the other hand, it occupied the centre of the fibrous and
muscular strands which compose that body (fig. xv). In the
former I could trace no nerves leaving the brain, but in the
male two nerves surrounded by muscles pass backward ; these
obviously correspond with the retinacula of other forms,
Classification.
Until recently the group Acanthocephala included but one
genus, Echinorhynchus, which comprised several hundred
214 ARTHUR E. SHIPLEY.
species. Recently, however, Hamann! has pointed out that
these species present certain differences which enable him to
divide the group into three families, each with a corresponding
genus. ‘To these I venture to add a fourth, to include the
remarkable form above described. This family may, I think,
be called the Arhynchide, and the new genus Arhynchus,
which name refers to the absence of the eversible introvert ;
and, inasmuch as it is convenient in naming a parasite to have
some indication of its host, I think the specific name may
be hemignathi.
If these terms be adopted, the classification of the Acan-
thocephala will be as follows, the characteristics of each of
the first three families being taken from Hamann’s papers.
ACANTHOCEPHALA.
I. Family EKcuinoruyncuipa. ‘The body is elongated and
smooth. The introvert sheath has double walls, and the
introvert is invaginated into it. The nerve ganglion is in the
introvert sheath, mostly embedded in it and central in posi-
tion. ‘The hook papille are only covered with chitin at their
apex, and the hooks have a process below.
Genus Echinorhynchus, with the characters of the
family.
The vast majority of Acanthocephala belong to this family ;
a few may be mentioned. EH. proteus, found in many fishes
aud varying in size with its host ; its larval forms inhabit the
Amphipod Gammarus pulex, and are also found in the body-
cavity of numerous fresh-water fishes. E.clavula occurs in
many fishes and in the intestine of a species of Bufo. E.
angustatus is found also in fishes, with its larval form in the
Isopod Asellus aquaticus. EK. moniliformis is said to
attain maturity in the human intestine; its usual host is a
mouse, and its larval host the larva of a beetle, Blaps mucro-
nata. KE. porrigens invests the intestine of the rorqual
and H. strumosus that of a seal. There are many others.
1 Loe. cit., and ‘ Zool. Anz.,’ Bd. xv, 1892, p. 195.
ON ARHYNCHUS HEMIGNATHI. 215
If. Family Gicanroruyncnipm. Large forms, whose body
is ringed and flattened during life like that of a Tenia.
The hooks are like those of a Tzenia, the hook-papilla being
entirely covered with chitin. There are two root-like pro-
cesses to each hook. The introvert is muscular, has no lumen
and the introvert cannot be retracted into it, but the whole
retracts into the body-cavity. The ganglion is excentrically
placed to the side, behind the middle of the so-called sheath.
The body-cavity is enclosed in a structureless membrane, and
is traversed by membranes stretched transversely. The
lemnisci are long, coiled, with a central lacuna.
Genus Gigantorhynchus, with the characters of the
family.
Hamann includes three species in this family—G. echino-
discus, G. tenioides, and G. spira; and points out that
EK. gigas agrees with them in all points but that of the external
annulation. The first of the above-named species occurs in the
intestine of anteaters, and has been found in Myrmecophaga
jubata and Cycloturus didactyla. G. tznioides
has been found in a species of Cariama, Dicholophus
cristatus; and G. spira lives in the king vulture, Sarco-
rhampus papa. E. gigas in the adult stage occurs in the
small intestine of swine, and its larval host is believed to be
the grubs of Melolontha vulgaris and Cetonia aurata in
Europe and of Lachnosterna arcuata in the United States.!
It is recorded once from the human intestine.
III. Family Neoruyncuipa#. Sexual maturity is reached
in the larval state. The introvert sheath has a single wall.
A few giant nuclei only are found in the subcuticle and in
the lemnisci. ‘The circular muscles are very simply developed,
and the longitudinal muscles only present in places.
Genus Neorhynchus, with the characters of the family.
This genus includes but two species, N. claveceps and
N. agilis. They both present interesting cases of pzdo-
genesis, the large embryonic nuclei of the young larva do
not break up into numerous nuclei as they do in the com-
1 C, W. Styles, ‘Zool. Anz.,’ Bd. xv, 1892, p. 52.
216 ARTHUR E. SHIPLEY.
moner species. N. agilis is found in Mugil auratus and M.
cephalus; N. claveceps in the Carp, Cyprinus carpio,
its larva form according to Villot! in the fat bodies of the
Neuropterous insect Sialis niger; it has also been found in
the alimentary canal of the leech Nephelis octocula, and
specimens of the water-snail Limnzea have been articially in-
fected with it.
IV. Family Aruyncuip#. Short forms, with the body
divided into three well-marked regions,—head, collar, and trunk.
The head is pitted, the collar smooth, and the trunk wrinkled,
not annulated—in spirit specimens. There is no eversible in-
trovert, and no introvert sheath and no hooks. The sub-cuticle
and the lemnisci have a few giant nuclei, and the lemnisci are
long and coiled.
Genus Arhynchus, with the characters of the family.
This family in the length and curvature of its lemnisci re-
sembles the Gigantorhynchide, and in the persistence of
the embryonic condition of the nuclei in the sub-cuticle and
the lemnisci, the Neorhynchide ; but in the shape of the body,
its division into three well-marked regions, the absence of
eversible introvert, introvert sheath, and hooks, it stands alone,
though to some extent nearer to the Neorhynchide, in which
the introvert is relatively small, the introvert sheath simple,
and the number of hooks reduced, than to either of the other
families.
The single species Arhynchus hemignathi was found
attached to the skin around the anus of a Sandwich Island
bird, Hemignathus proceros. This bird is a member of a
family Drepanidide, which is entirely confined to the Sand-
wich Island group. Professor Newton tells me that it is
probable that the “ food of Hemigunathus consists entirely
of insects which it finds in or under the bark of trees;”’ hence
it is probable that the second host of this parasite, if such
exists, must be looked for amongst the Insecta.
Tue ZootocicaL LABORATORY, CAMBRIDGE ;
March, 1896.
1 «Zool. Anz.,’ Bd. viii, 1885, p. 19.
ON ARHYNCHUS HEMIGNATHI. 27
DESCRIPTION OF PLATE 12,
Illustrating Mr. A. E. Shipley’s paper on “ Arhynchus
hemignathi, a new genus of Acanthocephala.”
In some cases the names of the various structures are written on the figures,
in others the following abbreviations have been adopted. czve. mus. The layer
of circular muscles in the skin. gez.d. Genital duct. gez.p. The external
opening of the duct. dace. The lacune in the skin. Jda/. lac. The large lateral
lacune of the trunk. dem. The lemnisci. Zig. The ligament. dong. mus. The
longitudinal muscles of the skin. mus. The muscles from which the ligament
arises. uc. The amoeboid nuclei of the skin and the lemnisci. sperm. Coagu-
lated masses of spermatozoa in the body-cavity of the female.
Fies. 1, 0, and 111.—Three views of three different specimens of Arhyn-
chushemignathi. Hach x 20. The division of the body into three regions
is well marked. The details are shown in Fig. 1. Figs. 1 and 1 are rough
sketches.
Fig. 1v.—A transverse section through the head of a female, crowded with
ova and egg-masses; the ligament is shown in section. x 40,
Fig. v.—A transverse section through the same, just below the edge of the
collar. In the centre is the neck, which fuses with the collar a few sections
further back. The big circular canal of the collar is shown at dae. x 40.
Fic. vi.—A transverse section through the trunk of the same. The upper-
most lemniscus is cut in two places. The ovary is double, and shows egg-
masses as well as eggs; some coagulated masses of spermatozoa are lying in
the body-cavity. x 40.
Fic. vit.—A surface view of the external opening of the genital duct.
x 40.
Fic. vill.—Some developing ova, highly magnified.
Vie. 1x.—A transverse section through the trunk near the genital pore,
taken from the same series as figs. Iv, V, aud vi. It shows part of the funnel-
shaped internal opening of the genital duct, gex. d. x 40.
Fic. x.—A transverse section from another specimen taken behind the
opening of the genital duct. This shows the arrangement of the lacune and
their communications with the lateral lacune, /aé. lac.
Fic. x1.—A longitudinal section through the central part of the skin of the
head, showing the origin of the ligament and the ganglion cells of the brain,
lying in a mass of ova and egg-masses.
218 ARTHUR E. SHIPLEY.
Fite. x1.—A small portion of the skin in section, showing the single layer
of circular muscle-fibres. x 40.
tre. xui.—A median longitudinal section through a female. The whole
body-cavity full of ova and egg-masses. The ligament is seen in the head, and
the genital duct near its opening in the trunk. The left lemniscus, cut twice,
is alone seen. X 80.
Fic. xtv.—A transverse section through the trunk of a male, showing one
of the testes. This section shows also the longitudinal muscles on the
lemnisci and the large lateral lacune, Jat. lac. x 40.
Fic. xv.—A transverse section through the head of a male, showing the
brain in the ligament, and the longitudinal muscle-fibres very well developed.
x 40.
ZOOLOGICAL OBSERVATIONS IN THE SOUTH PACIFIC. 219
Zoological Observations in the South Pacific.
by
Arthur Willey, D.Sc.Lond.,
Balfour Student of the University of Cambridge.
With Plate 13.
In the course of a brief stay in Sydney I had the opportunity
of working up for publication some material obtained by me.
My warmest thanks are due to Professor W. A. Haswell, who
has generously placed space and appurtenances at my disposal
in his laboratory at the University of Sydney.
ip
On a New Ampuioxus FROM THE LoutstapE ARCHIPELAGO
(Asymmetron caudatum, n. sp.).
(With Plate 13, figs. 1 to 4.)
Last January, as a result of my dredging operations in the
magnificent lagoon of the Deboyne groups of islands, of which
Panaieti is the largest, I obtained, on two occasions, specimens
of the family Branchiostomide (a single specimen only on each
occasion), which proved to be a new species, nearly related to
Asymmetron lucayanum, Andrews.
‘They were dredged up from a depth of one fathom and seven
fathoms respectively, living in clean coral sand. In the same
habitat a species of Ophelia is very abundant, and to the
casual gaze it bears a striking resemblance to the Amphioxus,
with which it is associated. This Ophelia burrows in the sand
with its pointed snout, ike Amphioxus, but more deliberately,
220 ARTHUR WILLEY.
and swims with exactly the same motions of the body as
Amphioxus, which is not the case with most other Annelids,
At first I thought I was obtaining a fine series of Amphioxus,
but the great majority turned out to be Ophelids.
The species now to be described, for which I propose the
name Asymmetron caudatum, n. sp., adopting Kirkaldy’s
system! of subdivision of the Branchiostomide, is perhaps
interesting on account of its geographical distribution relative
to other species of Amphioxus, rather than from the possession
of novel features.
That it should be entirely distinct from its relatives in
Torres Straits, from whom it is removed by a distance of less
than six hundred miles, and, on the other hand, closely allied
to a species residing in the Bahamas, upwards of eight thousand
miles away, on the other side of the American continent, is
certainly a noteworthy fact in distribution.
The specific differences between species of Amphioxus are
frequently of apparently little moment, but may be of im-
portance when taken in conjunction with geographical distri-
bution. On this principle the Japanese Amphioxus recently
described by Andrews?” should at least be regarded as a marked
variety of rather than identical with Amphioxus Belcheri.
As mentioned above, I only succeeded in obtaining two
individuals of Asymmetron caudatum, because the unpro-
pitious season of the year, when one is liable to be overtaken
by westerly gales, rendered it imadvisable to delay for any
considerable length of time in the lagoon.
The larger specimen was an immature female, and measured
29°5 mm.; the other was a mature or submature male,
measuring 20 mm. in length. Other differences, particularly in
respect of the caudal fin (cf. figs. 1 and 3), occurred between
the two individuals ; and these differences, combined with the
marked contrast in size, would seem to point to a sexual
1 Kirkarpy, J. W., * A Revision of the Genera and Species of the Branchio-
stomide,” ‘ Quart. Journ. Mic. Sci.,’ vol. xxxvii, 1895, pp. 8303—323.
2 AnpRews, HE. A., “An Amphioxus from Japan,” ‘ Zool. Anz.,’ 1895.
This should be named A. Belcheri, var. japonicus.
ZOOLOGICAL OBSERVATIONS IN THE SOUTH PACIFIC. 221
dimorphism, which, if found to be constant among a large
number of individuals, would be a fact of some importance.
My two specimens alone prove that an appreciable sexual
dimorphism may occur, whether as a variation or as a fixed
property.
The following are ali the numerical facts that I have ascer-
tained regarding the two specimens.
1. Female. Length, 29°5 mm.; number of myotomes, 60;
formula, 40—9—11; length of tail or urostyloid process be-
hind myotomes, 3 mm.; distance of anus from root of tail,
2°5 mm.; distance of atriopore from root of tail, 6°5 mm.
2. Male. Length, 20 mm.; number of myotomes, 64;
formula, 44—9—11; number of gonads, 30, disposed uni-
laterally on right side.
The characteristic generic features of Asymmetron, exhibited
by the new species, are as follows.!
The right metapleur is continuous with the ventral fin; the
latter has no fin-chambers and no fin-rays; the right and left
metapleura pass equally into the rostral fin in front ; an inter-
tentacular membrane is present betweenthe ventral buccal cirri ;
the buccal cirri are plain; and lastly, there is a pair of post-
atrioporal, or, better, post-atrial caeca, which are, however, very
short, ending abruptly between the posterior lips of the
atriopore.
The dorsal fin-spaces become very irregular posteriorly, and
cease at about the 54th myotome.
In addition to the length of the body and the number of
myotomes, an important specific feature of Asymmetron
caudatum is the rostral fin, which is marked off from the
dorsal fin by a pronounced constriction of somewhat varying
depth (cf. figs. 2 and 4). The rostral fin is indeed a striking
feature in the fresh object, but is less marked after preservation.
The caudal fin also varies considerably in the two individuals,
its urostyloid portion being much more constricted off from
the succeeding portion in the region of the myotomes in the
male specimen than in the female (figs. 1 and 3). It is, how-
' T have verified all the points mentioned by sections and preparations.
222 ARTHUR WILLEY.
ever, important to note that in any case there is a distinct
caudal fin extending to the posterior extremity, and Kirkaldy’s
opinion that there is no caudal fin in Asymmetron is somewhat
misleading.
From the above description it is evident that A. caudatum
comes very near to A. lucayanum, but differs definitely and
distinctly from it in size, number of myotomes,! and in the
possession of a prominent rounded rostral fin, marked off above
and below by a constriction.
i
On tHE NEPIONIC SHELL OF THE Recent NavtiLus.
(With figs. 5 and 6 on Plate 13.)
While conversing recently with Mr. Charles Hedley,
Conchologist to the Australian Museum, Sydney, I had
occasion to refer to the very large size of the ovarian ova
of Nautilus pompilius, apparently larger than the ova of
any other known Cephalopod; while the fluidity of the yolk
would lead to the presumption that the ova, when deposited,
would be provided with a firm covering, whether horny,
coriaceous, or calcareous.
Mr. Hedley then drew my attention to two young shells of
N. pompilius in one of the museum cases, in each of which,
at the same region in both, there was a very distinct discon-
tinuity in the formation of the shell by the animal, and added
that this line of demarcation probably represented the limit
reached by the shell at the time of hatching from the egg. As
1 The average myotome formula for A. lucayanum as given by Andrews
is 44—9—13 = 66, and the average length 13 mm. The maximum length,
according to Andrews, is 16mm. out of a large number of specimens.
Kirkaldy gives 19 mm. as the maximum length of specimens sent to Prof.
Lankester by Prof. Agassiz.
ZOOLOGICAL OBSERVATIONS IN THE SOUTH PACIFIC, 223
the size of this young shell behind the line of division between
it and the later formed shell-substance seemed to correspond
roughly with the dimensions of the ovarian ovum, as stated by
me in a former communication,! I at once became interested
in Hedley’s suggestion.
The term “nepionic ” was invented by Hyatt in 1890, and
embodied by Jackson in his important memoir on the ‘ Phy-
logeny of the Pelecypoda,’? and has been adopted by Pilsbry
in the ‘Manual of Conchology.’ As I understand it, it is
intended to be applied to the larval stages, sensu stricto, of
all animals. To quote Hyatt, “ This term should be confined
to the designation of stages of growth which are immediately
continuous with later stages, and parallel or referable in their
origin to the adults of allied existing or fossil forms which are
not so remote as those from which the embryonic stages are
derived.”
The nepionic period is succeeded by the adolescent or neo-
logic stages, which “‘ are of as great importance for tracing the
genealogy of small groups as are the [nepionic] characters in
larger groups.”
In the case of those terrestrial Gastropods which lay large
hard-shelled eggs full of yolk, the development and growth of
the shell take place within the egg until the close of the
nepionic period, by which time the shell has in many in-
stances attained a very respectable size. Thus the nepionic
shell of the above-mentioned molluscs is that portion of the
true shell (as opposed to the embryonic shell) which develops
within the egg.
1 “Letters on Nautilus, &c.,” see ante, p. 175.
* Jackson, Kopert Tracy, “Phylogeny of the Pelecypoda, the Avi-
culidse, and their Allies,” ‘ Memoirs Boston Soc. Nat. Hist.,’ vol. iv, No. 8,
1890.
Proressor Hyarr introduced the term “nepionic ” in a foot-note to the
preceding memoir as a substitute for the term “silphologic” suggested by
him ina previous paper (Hyarr, ALpuEvs, ‘ Values in Classification of
the Stages of Growth and Decline, with Propositions for a New Nomen-
clature,” ‘Proc. Bost. Soc. Nat. Hist.,’ vol. xxiii, Part 3, 1887).
1 am indebted to Mr, Hedley for these references.
224 ARTHUR WILLEY.
The nepionic shell frequently stands in marked contrast to
the post-nepionic or (with the large terrestrial Gastropods)
post-natal shell, both by a definite line indicating a tem-
porary discontinuity of growth, and by its different colour
and minute sculpturing.! In fact, it is obvious that a
definite break in the continuity of shell-formation is more
likely to occur in those cases in which, after the completion of
the nepionic shell, the young mollusc has to go through the
operation of hatching, after which it recommences to form
post-natal shell-substanee. This supposition is, I believe, in
accordance with the facts.
With regard to Nautilus, about whose reproduction we are
so much in the dark, it is a distinct point gained to know, at
least with respect to its shell, at what stage the animal
hatches.
The existence of the line of discontinuous growth, at a
particular point, in young shells of N. pompilius, might not
be regarded as conclusive in itself, but I have since found a
precisely similar and constant discontinuity of growth in a
number of shells of N. umbilicatus and N. macrom-
phalus.
In the two last-named species the nepionic shell differs
from the post-nepionic shell by the presence in the former of
a minute plication, and of a peculiar glossy sheen. In N.
pompilius the nepionic shell also has a peculiar sheen, and
presents a contrast to the post-nepionic shell in the colour
of its markings, which are brick-red, those of the latter being
crimson.
Fig. 5 (Pl. 13) represents a drawing of the umbilical region
of N. umbilicatus, Lam. On looking down into the um-
bilicus at the first whorls of the shell, it will be seen that
after the young shell had completed rather more than one
entire whorl, there occurred a sharp and abrupt interruption
in the growth of the whorls, and (in the actual specimen,
although not shown in the figure) the character of the shell
1 Compare Hepiry, Crartzs, “On the Structure and Affinities of Panda
atomata,” Gray, ‘ Records Austral. Mus.,’ vol. ii, 1892, pls. v and vi.
ZOOLOGICAL OBSERVATIONS IN THE SOUTH PACIFIC. 225
added after this line,! is very distinct from the previously
formed shell, as already mentioned. Sometimes there is a
visible depression in the shell at the nepionic line.
In the shells of all species of Nautilus there are sometimes
to be seen one or several strongly marked lines indicating
occasional, but not regularly periodical, interruptions in the
growth of the shell,—periods of rest, in fact. These, however,
are neither constant in position nor in their occurrence in
different individuals. On the other hand, the nepionic line
occurs in a constant position in all individuals of a species,
although in N. pompilius it cannot be seen beyond a certain
stage, owing to the overgrowth of the whorls.
Altogether I have seen the nepionic line in eighteen shells
of the three above-named species of Nautilus.
Fig. 6 (Pl. 18) is an outline sketch of a young shell of
N. pompilius? with still perforated umbilicus, to show the
nepionic line marking off the prenatal from the post-natal
shell. ‘The following are the measurements of this shell:
(a) Length of complete shell (from the anterior free margin
to a point directly opposed to it on the posterior convex surface
of the shell), 32 mm.; greatest width of complete shell, 19 mm.
(b) Length of nepionic portion of shell, 27 mm. (this varies
slightly in different individuals from 25 to 27 mm.); greatest
width of nepionic portion of shell, 16 mm.
It may be observed that the above dimensions of the nepionic
shell of N. pompilius are a good deal in excess of those of
the ovarian ovum which I previously described ; but not more
so relatively, I think, than is the size of other newly hatched
Cephalopods to the original size of the freshly deposited ova or
of the mature ovarian ova.
Knowing the size and character of the nepionic shell of
recent Nautilus, and the size and character of its submature
ovarian ovum, I think there is ample justification for the con-
clusion that the nepionic line marks the period at which the
young individual hatches from the egg.
1 T shall refer to this in future as the “ nepionic line.”
* The sketch was taken from a specimen in the Australian Museum by
kind permission of Mr. Charles Hedley.
VOL. 39, PART 2.—NEW SER. P
226 ARTHUR WILLEY.
There is strong presumptive evidence that the animal of
Nautilus at the time of hatching already possesses the main
features of the adult, with the possible addition of a yolk-sac,
This follows both from the size of the nepionic shell, which
comprises a number of chambers, and from the consideration
of a very small specimen of Nautilus, which I obtained in
New Britain and mentioned in a former publication.!, I have
unfortunately mislaid the shell of this specimen, but it had a
perforated umbilicus, through which daylight could be seen,
and was at least as small as the shell from which Pl. 13, fig. 6,
was drawn.
Jackson (loc. cit.) refers to Hyatt’s figures of a young
Nautilus Koninckii, in which is indicated “a smooth ne-
pionic period, succeeded by a fluted neologic stage.” Ihave .
been unfortunately unable to refer to Hyatt’s well-known
memoir on the “ Embryology of Fossil Cephalopoda ”’ (‘ Bulletin
Mus. Compar. Zool. Cambridge, Mass.,’ vol. iii, 1872).
In the first part of his “ Beitrage zur Entwickelungsge-
schichte der fossilen Cephalopoden” (‘ Paleontographica,’
vol. xxvi, 1879-80) Branco figures a number of minute shells
of Ammonites, which sometimes, when they measure even less
than 1 mm. in diameter, exhibit a deep circular constriction ;
and this not only in cases in which the adult shell is grooved,
but also in many instances in which the latter is not grooved.
If this groove were to be identified or compared with the
nepionic line of recent Nautilus, which is itself sometimes
depressed, as mentioned above, it would indicate that the ova
of the Ammonites referred to were much smaller and poorer in
yolk than those of tne recent Nautilus. This would be an
interesting conclusion if it could be substantiated.
1 * Natural Science,’ June, 1895.
ZOOLOGICAL OBSERVATIONS IN THE SOUTH PACIFIC. 227
IDOE
On some VARIATIONS IN THE SHELL oF NAUTILUS, WITH
Description or A New Variety (N. POMPILIUS, VAR.
MoRreETONI, NOV. VAR.).
(With figs. 7, 8, and 9 on Plate 13.)
Of the obvious differences that exist between the shell of
N. umbilicatus and that of N. pompilius the following
may, for my present purpose, be particularly remembered.
1. In N. umbilicatus (see Pl. 13, fig. 5)—
a. The umbilicus is perforated, the shell being so loosely
coiled that a small aperture, which can be looked through, is
left in the centre of the whorls.
(3. The umbilicus is large and open, so that all the whorls
can be distinctly seen. This holds good also for N. macrom-
phalus, but here the umbilicus is not perforated.
y. The free margin of the shell does not merge into the
umbilicus, but tends abruptly inwards and downwards, so as to
form a prominent shoulder, nearly at right angles to the
lateral free margin. It finally fuses laterally with the anterior
convexity of the shell, which is pigmented black.
2. In N. pompilius— :
a. The umbilicus is imperforate (excepting very young
individuals).
. [3. The umbilicus is practically obliterated through a depo-
sition of callus by the dorso-lateral angles of the mantle.
y. The free margin of the shell merges directly into the
umbilical region.
The different characters tabulated above are usually very
distinct, but they may be considerably weakened by variation.
One variety, for instance, of N. pompilius will present all
the characteristics of the species with the single exception
that the free margin of the shell does not pass directly into
the umbilicus, but forms a shoulder.
In another variety the free margin of the shell will pass
gradually into the umbilicus, but the latter will be perforated.
228 ARTHUR WILLEY.
Finally, in a third variety the deposition of callus will have
been entirely left out and the whorls loosely coiled, so that we
shall have the above-mentioned shoulder, a perforated um-
bilicus, and visible whorls.
The latter variety is not identical with the so-called N.
stenomphalus, but is, in a sense, intermediate between the
latter and N. umbilicatus. I obtained a single example
from New Guinea waters, kindly given to me by the Hon. M.
H. Moreton, of Government House, Samarai, British New
Guinea. I have pleasure in naming it N. pompilius, var.
Moretoni.
In acollection of shells of N. pompilius it may be ob-
served that the extent of deposition of callus over the um-
bilicus is very unequal; and that while there is usually a thick
prominent plug of white callus, sometimes the deposition of
the latter has been so deficient as to leave a deep umbilical
depression on cither side.
A peculiar feature about the variations now being dealt with
is that they may occur on one side of the shell only, the other
side being more or less normal.
Pl. 13, fig. 7, represents the umbilical region of a variety of
N. pompilius, whose umbilicus was only partially perforated.
On the right side of the shell (the latter being considered with
the convex side downwards, as it is in the fresh condition)
there was a deep hole, about a centimetre in depth; on the
other side there was a deep umbilical depression, which, how-
ever, was closed over with callus, with the exception of a
minute pin-hole aperture—perhaps half a centimetre in depth.
This latter could only be made out by close inspection, and
at first sight the Jeft side of this shell appeared to be practi-
cally normal.
It is clear that if this variation were carried a little farther
in the same direction, we should have a form possessing the
essential features of N. stenomphalus, and there is no
reason at present to regard the latter as anything more than a
variety of N. pompilius with a persistent perforated um-
bilicus. In a specimen of N. stenomphalus in the Austra-
ZOOLOGICAL OBSERVATIONS IN THE SOUTH PACIFIC, 229
lian Museum only a portion of the coil was dimly visible, and
the free margin of the shell passed directly into the anterior
border of the umbilicus without a prominent shoulder.!
In Pl. 13, fig. 8, an interesting variation is shown in which
the umbilicus was imperforate, yet nevertheless the free margin
of the shell did not pass simply into the umbilicus, but bent
round in such a manner as to form a prominent shoulder,
behind which there was a deep umbilical depression. I have
two examples of this variation from New Guinea, and in both
cases the above-described shoulder was confined to the left side
of the shell, the right side being normal.
It will be found convenient to name these varieties. That
first described above, with partially perforated umbilicus, may
be called N. pompilius, var. perforatus, nov. var.
The second variety, with the shoulder, may be called N.
pompilius, v. marginalis, nov. var.
Now we come to the third variety to be described, namely,
N. pompilius, var. Moreton, nov. var.
In this variety, which is a very well-marked one, there is a
prominent shoulder as shown in Pl. 13, fig. 9; the umbilicus is
somewhat widely perforated, and a considerable portion of the
whorls can be distinctly seen. ‘This shell was symmetrical in
every respect. At the first sight of it one might well hesitate
whether to regard it as a loosely wound pompilius, or as a
closely wound umbilicatus, but the existence of intermediate
varieties shows, I think with certainty, that it belongs to the
pompilius group.
All the variations described above belong to the category of
substantive variations, according to Bateson’s nomencla-
* It may be added here that there also seems no reason for regarding
N. scrobiculatus as other than a variety of N. umbilicatus. On this
view the species of recent Nautilus would be reduced to three,—N. pom-
pilius, N. umbilicatus, and N. macromphalus.
2 All the varieties described in this paper were obtained in New Guinea,
and with one exception from the natives. I have certain evidence of the
living Nautilus occurring in New Guinea waters, but the natives know
nothing of it, and it is not in a workable condition for the zoologist
230 ARTHUR WILLEY.
ture,! and it is important to notice that they can be arranged
in three groups as follows :
1. Single variations, in which the variation affects only
a single character, such as the umbilical shoulder or the um-
bilical aperture.
2%. Collective variations, in which a whole group of cha-
racters are affected, as in the variety Moretoni.
3. Incomplete variations, as in the two examples of my
variety marginalis, where the umbilical shoulder was con-
fined to one side of the shell.
Finally, it may be pointed out that the species and varieties
of the recent Nautilus seem to fall naturally into the following
scheme, which, however, does not aim at suggesting any
particular direction of evolution.
N, macromphalus. N. umbilicatus.
N. umbilicatus, v. secrobiculatus.
N. pompilius, v. Moretoni.
N. pompilius, v. stenomphalus.
N. pompilius, v. perforatus.
N. pompilius, v. marginalis.
Ve
fr
N. pompilius.
University oF SYDNEY ;
May 18th, 1896.
1 Bareson, W., ‘Materials for the Study of Variation,’ Macmillan, 1894.
ZOOLOGICAL OBSERVATIONS IN THE SOUTH PACIFIC. 231
EXPLANATION OF PLATE 13,
Illustrating Dr. Willey’s “ Zoological Observations.”
Figs. 1—4 refer to Asymmetron caudatum, n. sp. See p. 219.
Fic. 1.—Asymmetron caudatum, n. sp. Posterior extremity of
female from right side. d. Dorsal. v. Ventral. From living object.
Fic. 2.—A. caudafum. Anterior extremity of female from right side.
From living object.
Fic. 3.—A. caudatum. Posterior extremity of male from right side.
From living object. :
Fie. 4. A. caudatum. Anterior extremity of male from left side.
From living object.
Fig. 5.—N. umbilicatus. Drawing of umbilical region to show the
nepionic line, which is situated a little beyond the first complete whorl of the
shell. From right side.
Fic. 6.—N. pompilius. Outline sketch of young shell with perforated
umbilicus, to show nepionic line. From right side.
Fic. 7.—N. pompilius, v. perforatus, nov. var. Umbilicus partially
perforated. The dotted line is merely to indicate the limit of the black
pigment on the anterior convexity of the shell. Right side.
Fic. 8.—N. pompilius, v. marginalis, nov. var. Umbilical region of
left side, to show the umbilical shoulder and depression.
Vie. 9.—N. pompilius, v. Moretoni, nov. var. Umbilicalregion, showing
shoulder, perforation, and whorls. From right side.
N.B.—The drawings are placed in such a way that the convex surface of
the shell would be directed upwards and the mouth of shell downwards,
because it is thought that they would be more intelligible in this position ;
but the right and left sides are named with the shell considered in the
natural position, with the convex surface directed downwards and the month
upwards.
} ; pays atk me ee a4 . 4 iva ue:
z : eu
5 valde oaihs OE, ee eet) Woe ety
7 ‘
a D Din a we) << "
= it Cm <> * i ’ be
: me : ; “ew ire Rey =
ae? CYA os j
7 a : ° C Fe tour’
whe a a
y : ; 'S O2 as eS 3 ali ate
; see (he (he =
te
; “ a 6 * ¢ -
: } a; a ae genes |
, : PD j . ‘ a :
; 3 -
. « 7 (NIAMS 7h aay
Ps = ©
> 2
: ; es ty et SETS)) 7
a a be , pe ‘SAD Gare J
: ‘ih
s ; eta Jc
: "oe res
. .
i 0 tLe Pr i
th fi a |
ee ee et,
ne ae ae Hig Bex
hal £ ; 7 7 _ Lhe 5 hae “eo s Ht
fai ald | i Paci &. aie mit ies he val hy By:
* - i a
7 .
= . q - ce
> Sista
CHLAMYDOMYXA MONTANA. 233
Chlamydomyxa montana, n.sp., one of the
Protozoa Gymnomyxa.
By
E. Ray Lankester, M.A., LL.D., F.R.S.,
Linacre Professor in the University of Oxford.
—_—-
With Plates 14 and 15.
I vo not know of any naturalist who has seen the beautiful
Chlamydomyxa labyrinthuloides of Archer in a state
of “ expansion ”’ since he described it in this Journal twenty-
one years ago. Mr. Thisleton Dyer, now Director of Kew
Gardens, saw it in its active streaming condition when Archer
exhibited it at one of the meetings of the Dublin Microscopical
Club five-and-twenty years ago. And the other members of
that admirable band of naturalists saw it and testify to the
correctness of Archer’s description and figures.
Some seven years after the publication of his paper Archer
sent both to me and (through Prof. Percival Wright) to Mr.
Patrick Geddes samples of Sphagnum with the cysts of
Chlamydomyxa attached to the fronds of the moss. I kept
the specimens which I received in conditions which I hoped
would favour the rupture of the cysts aud the out-crawling
of the Chlamydomyxa-network, but entirely failed to obtain
such a result.
Mr. Patrick Geddes made a careful study of the cysts received
by him, and wrote a very interesting article on them, accom-
panied by a coloured plate, in this Journal, vol. xxii, 1882,
p. 30. Mr. Geddes especially dwelt upon the remarkable
laminated formation of the cellulose cyst-wall, and oun the divi-
234 HK. RAY LANKESTER.
sion and multiplication of the encysted mass. He pointed out
the importance of Chlamydomyxa for the theory of the vege-
table cell-wall and its laminated structure, and he was led to
compare the encysted condition of Chlamydomyxa with the
encysted condition of such forms as Heematococcus and Glao-
capsa. He also made important observations on the red-
coloured oily substance formed in the cysts.
I frequently searched for Chlamydomyxa in the years fol- _
lowing the publication of Archer’s paper whenever I found
myself in a moorland country with Sphagnum bogs. In the
neighbourhood of Lervik, on the Stavanger Fjord in Norway,
I searched in vain, as also on Dartmoor. I was, however,
rewarded for my coutinued efforts by at last finding
Chlamydomyxa in abundance on August 22nd, 1886, on the
surface of Sphagnum gathered in small ditches cut in the
bog which occupies clearings in the pine-wood at Pontresina
in the Enghadine; and on two subsequent visits to Switzer-
land at the same time of year, once at Zermatt (1890) and
later again in the Enghadine (1892), I found and studied
Chlamydomyxa.
On the first occasion when I found Chlamydomyxa, viz. in
August, 1886, at Pontresina, I was fortunate in being able to
observe and draw specimens which expanded their remarkable
networks under the microscope, and I showed such active
specimens in the Hotel Saratz to my friend Ernst Haeckel,
who happened to arrive there for a short holiday, and to Sir
Edward Fry—a judge not only of legal but of biological
matters. My drawings then made are reproduced in Plates
14 and 15, figs. 1—4. I was able to recognise the cysts
which accompanied the naked Chlamydomyxa as belonging to
the latter, and, indeed, should have recognised them from my
previous familiarity with the encysted specimens sent to me by
Archer. But I have never seen expanded Chlamydomyxa
since that first occasion. At Zermatt and at the Maloja in
subsequent years I found only the cysts, and no persuasion
which I could offer was sufficient to induce the Chlamydomyxa
to leave its cyst. The cysts, however, were always very abun-
CHLAMYDOMYXA MONTANA. 235
dant, and varied in shape and size. It is not at all improbable
that the month of August is either too late or too soon for the
best chance of observing Chlamydomyxa in the motile condi-
tion, and I should recommend other observers in the same
locality to try the latter part of June and commencement of
July, as affording the best chance of observing Chlamydomyxa
in the unencysted condition.
The Chlamydomyxa which I studied in Switzerland, whilst
agreeing in essential features with Archer’s genus, appears to
me to be certainly a distinct species, for which I propose the
name Chlamydomyxa montana, ‘The grounds for this
distinction will be apparent from the following description.
Mode of Occurrence.—The specimens of Sphagnum which
yielded the Chlamydomyxa had a peculiar ruddy-brown appear-
ance and a glistening surface, differing from ordinary healthy
growths of the moss. They were old and in a state of incipient
decay. I scraped the surface of the leaflets of the moss and
teased them on a glass slide, so as to remove and spread as
much as possible for observation the matters adhering to the
surface.
General Appearances.—I then found, when the preparation
was examined with the microscope, numerous olive-brown disc-
like or ovoid bodies from the 535 to the =, of an inch in
diameter, which very scon showed a movement of the colour-
less border by which each was surrounded. The colourless
border seemed to open out and spread itself in the form of a
network of threads.
The Threads.—The gradual separation and spreading of
these threads is a very curious phenomenon, and certainly
gives the impression that the threads are pre-formed, closely
packed together, and that they gradually separate and straighten
out. Iam not prepared to assert that such is the case, but I
think it very likely. A fully expanded specimen of fair average
size 1s represented in Plate 14, figs. 1 and 2. It is represented
in two consecutive phases of expansion, separated by about five
minutes, as seen under a high power—No. 10 immersion of
Hartuack. If one taps the cover-glass the whole branching
236 E. RAY LANKESTER.
aud outstretched system of threads suddenly contracts, and
forms again a clear colourless “ border” to the pigmented
disc. But immediately the outpushing of the threads recom-
mences, and seems to proceed by a straightening and disen-
tanglement of the individual threads, which grow in length
as one watches them, apparently being extruded from the
central mass. The commencement of the process of expansion
of the threads is shown in a drawing of another specimen
(Plate 15, fig. 4).
In all cases the process of expansion is accompanied by the
formation of “ vacuoles”? in the colourless border-substance
(see figs. 1—4), which change their shape and position as
the spreading goes on. But they have not the character of
the ‘contractile’? vacuoles of Heliozoa, although Archer
speaks of such “ contractile ” vacuoles (perhaps by inadvertent
use of the term “ contractile”) as occurring in his C. labyrin-
thuloides. Archer also observed frequently solid food par-
ticles such as diatoms to be entanged in the substance of the
colourless material. I never observed any food particles thus
engulfed in C. montana.
The threads are of extreme tenuity, of equal diameter
throughout, and appear to me not to undergo any measurable
change in dimension. I did not see them shorten and thicken,
but they appeared to become flexed and gathered together
when a stimulus was applied to them. Further, I never saw
any thread either fuse with a neighbouring thread or divide
into two. It appeared to me (but the observation is difficult)
that when two threads come together they may be very closely
apposed, but nevertheless retain their distinctness; and con-
versely that, where a thread seems to divide into two longi-
tudinally, the case is really one of the separation of two pre-
existing threads.
My general conclusion is that the threads do not really
form either a dendritic branching figure or a network, but are
merely apposed so as to form when less expanded, or, to speak
more accurately, when less straightened, an apparent meshwork,
aud when more straightened and separated from one another
CHLAMYDOMYXA MONTANA, 237
an apparent tree-like structure, the appearance in both cases
being illusive. A gentle to-and-fro swinging movement of an
outstretched thread is sometimes seen. Such movement may
be due to currents in the surrounding liquid.
As to the further interpretation of the nature of the threads
I will say more below.
The Oat-shaped Corpuscles.—Travelling upon the threads and
closely packed against one another in those parts of the hyaline
border-substance which are, so to speak, not unravelled, are
very numerous oat-shaped or fusiform corpuscles. They are
represented in figs. 1—4, and one is drawn on a greatly
enlarged scale in fig. 5. These corpuscles are described by
Archer as leading characteristics of his C. labyrinthuloides,
and are compared by him with the nucleated spindle-shaped
bodies which travel upon the threads of the Labyrinthula of
Cienkowski. In Archer’s species of Chlamydomyxa, however,
these corpuscles were homogeneous, showing neither an envelope
of protoplasm nor a central nucieus. The same is true of the
fusiform corpuscles of C. montana. They are structureless.
It is important to note that they are very much smaller in
C.montanathaninC.labyrinthuloides. Archer states that
in the latter they are ;,4,5 inch long and half as broad, whereas
0
I find that in C. montana they are not more than the ;,1
12000
ofan inch or two microns in length. Nevertheless they are
perfectly distinct and uniform both in size and shape. I found
that they were stained more strongly than the threads by the
addition of a solution of iodine ; and since neither by this nor by
any other method could I detect anything indicating a central
nucleus or nuclei in Chlamydomyxa (Archer failed equally), I
think that these oat-shaped corpusclesin both species of Chlamy-
domyxa should be regarded as nuclei, the particles of a frag-
mented scattered nucleus. In encysted specimens of Chlamy-
domyxa montana I entirely failed (either by direct exami-
nation or by use of reagents) to detect any trace of these oat-
shaped corpuscles.
Movement of the Oat-shaped Corpuscles—The movement of
the oat-shaped corpuscles is the most interesting and character-
238 E. RAY LANKESTER.
istic feature presented by Chlamydomyxa. It must be dis-
tinguished altogether from the straightening and expanding
movement of the mass of filaments; at the same time it is not
manifested until the filaments have become—some at least of
them—straightened and free. Then as such a filament separates
itself, and as it were slowly pushes itself forth in a straight
line, first one, then another, and finally many of the oat-shaped
corpuscles are seen to advance along it. They move slowly in
one direction as a rule, stopping sometimes after a consider-
able advance, and then resuming movement. They do not all
travel at the same rate on one filament. I saw on several occa-
sions one corpuscle overtake another and glide over
the back (so to speak) of its more slowly moving companion,
and advance in front of it. Archer also witnessed this pheno-
menon in his larger species. Further, the corpuscles do not all
travel in one direction on one and the same filament. Some
are advancing towards the free extremity of the filament, whilst
others are travelling away from it. The corpuscles travelling
in opposite directions meet and pass one another, or sometimes
on meeting come to a standstill, then after a time the two
separate from one another, reversing their previous direction
of movement. These movements are not novel in themselves,
but similar to the movements of the granules in the psendo-
podium of an Actinospherium or in the threads of the vacuo-
lated cell protoplasm of the Tradescantia hair. What is
peculiar in the case of Chlamydomyxa is this, viz. that the
moving corpuscles are oat-shaped bodies of definite and uniform
size, and that there is no visible coating of streaming proto-
plasm embedding both them and the filament upon which they
move (as there is in the case of the pseudopodium of Actino-
spheerium).
I am inclined myself, from a careful consideration of my
own observations on Chlamydomyxa and Actinospherium,
and of Cienkowski’s observations upon Labyrinthula, to adopt
the view that the filaments of Chlamydomyxa are inert pro-
ducts of the metamorphosis of its protoplasm, which have a
certain amount of durability, but can be rapidly absorbed by
CHLAMYDOMYXA MONTANA. 239
the protoplasm which also gives rise to them, as is the case
with the axial fibres of the pseudopodia of Heliozoa.'! The
oat-shaped corpuscles also are to be regarded as inert nuclear
bodies, inert so far as motility is concerned.
The movement of the oat-shaped corpuscles along the threads
is, I believe, produced by an exceedingly delicate coating of
hyaline protoplasm. Of the existence of such hyaline proto-
plasm, which is neither filament nor corpuscle, we have
evidence, firstly in the agglutination of the filaments when not
extended, and secondly in the movement of contraction and
expansion of the mass of filaments. I do not think that it is
possible anywhere actually to see as an isolated substance this
delicate ‘‘ varnish ” of hyaline protoplasm, but it seems to me
reasonable to infer its existence.
Thus I should regard the filaments of Chlamydomyxa as the
delicate skeletal supports of a still more delicate streaming
protoplasm, in which, as in coarser pseudopodial expansions,
parallel currents of different rates or opposed in direction
may arise.
The Central Pigmented Corpuscles.—The centre of the disc or
irregular body, which is the form taken by Chlamydomyxa
montana when in the free motile state, consists, as shown
in figs 1—4, of a number of rounded corpuscles closely
pressed against one another, and of a yellow-brown colour
with a tendency to a greenish tint. There is no doubt that a
colourless protoplasmic substance invests each of these cor-
puscles, and whilst holding the mass together is continuous
with the colourless marginal substance or border of the disc.
These corpuscles are of fairly uniform size (1, of an inch in
diameter), and are of viscid consistency. I consider them as
identical in character with the green vesicles described by
Bourne as forming the bulk of the structure of his Pelomyxa
viridis (see this Journal, vol. xxxii ,1891), and I entirely agree
with him in separating them altogether from relationship to
‘ I am inclined to think that such an elastic filament, one-sided in posi-
tion, must be present also in all cilia and other forms of vibratile protoplasm.
24.0 E, RAY LANKESTER,
ordinary chlorophyll corpuscles such as those of the leaves of
plants.
I consider that these coloured vesicles, which differ from the
conception of “corpuscles” or ‘ granules” only in their
relatively greater fluidity, are identical with the colourless
‘“‘ Glinzkorper” (Greef) or refringent corpuscles of Pelomyxa
palustris, as I have stated in a note to Gould’s paper on the
structure of that Protozoon, published in vol. xxxvi of this
Journal, 1894; and they all, no doubt, have an important
chemical function in relation to the surrounding protoplasm.
The colouring matters which develop in such vesicles appear to
have a special character, and without necessarily agreeing
exactly with what is called ‘“ chlorophyll,” are related to it
and to such substances as diatomin, which accompany chloro-
phyll in other more highly developed organisms. The
colouring matter of Professor Bourne’s Pelomyxa viridis
is clearly not simple chlorophyll, or if such chlorophyll is
there it is mixed with other pigmented bodies. ‘The pre-
dominant yellow-brown colour of the vesicles (granules or
corpuscles) of the disc of Chlamydomyxa montana is
suggestive of diatomin. Very possibly it masks chlorophyll.
This is probable, because Archer represents his Chlamy-
domyxa labyrinthuloides as containing vesicles like those
in question, but coloured bright green; and further, because
after encystation my C. montana undergoes the most re-
markable changes in colour, developing a brilliant grass-green
tint (in some examples), whilst accompanying this chlorophyll-
like colouring matter there are always found in the cysts of
C. montana irregular or central droplets of a brilliant crimson
oily fluid, soluble in ether (see Pl. 15, figs. 8 and 9).
Encysted Condition.—The cysts of Chlamydomyxa mon-
tana closely resemble those of C. labyrinthuloides, as
described by Archer and Geddes. ‘The substance of which the
eyst-walls are composed yields the blue colour characteristic of
cellulose when treated with H,SO, and iodine. The cysts are
further remarkable in two features, and unlike the capsules
usually termed ‘‘ cysts” among Protozoa. ‘They have (a) a
CHLAMYDOMYXA MONTANA. 241
most definite laminated structure, consisting of a series of
“ shells ” or complete coatings of cellulose substance, deposited
one within the other, on the surface of the protoplasmic
organism. The protoplasmic organism apparently shrinks
after the formation of a first cyst, and having grown smaller
forms a second deposit, and so on, in some cases to a ninth or
tenth. In the second place, (b) the living organism enclosed
in the cyst appears to be by no means quiescent, but to be
undergoing important chemical changes, as shown by its
remarkable change of colour (to a bright green) and its de-
velopment of “droplets” of red-coloured oily material.
The activity of the encysted Chlamydomyxa is further
and very markedly demonstrated by its movement,
change of form, and division into separate masses.
The fact that the shells or coatings of cellulose are so freely
and abundantly deposited by the living matter enables one to
follow these changes in the encysted Chlamydomyxa with
certainty. Thus in Pl. 15, fig. 6, we have evidence of division
of an original mass of living Chlamydomyxa into three; in
fig. 7, part of the cyst at first formed and marked a has been
deserted, whilst the original outer cysts are much larger than
those now occupied by the living material, which has divided
into two masses. In fig. 9, three cysts occupied by green-
coloured Chlamydomyxa which have given rise also to the red-
coloured oil drops, are represented. These three cysts pro-
bably originated from the division of one parent mass. The
division of the encysted Chlamydomyxa has been described in
considerable detail by Professor Patrick Geddes in his paper
already referred to.
In reference to differences between C. montana and C.
labyrinthuloides, I may point out that I have not seen the
body or mass of C. montana when it is in the dendriform
streaming condition, to be partly lodged in the ruptured cyst,
as Mr. Archer saw and figured that of C. labyrinthuloides.
Nor have I seen green pigment (presumably chlorophyll) and
red oil drops in that phase, as Mr. Archer describes for C.
labyrinthuloides. The dendriform streaming specimens of
VOL. 39, PART 2.—NEW SER. Q
242 E. RAY LANKESTER,
C. montana which I observed (in considerable numbers for
several days) were always entirely free from any cyst-wall, and,
moreover, their central granules or vesicles were of an uniform
yellow-brown colour. This may be due to specific difference,
but it would be interesting to know what appearance C. mon-
tana presents early in the season,—for instance, in June.
The very regular circular cyst drawn in fig. 10 is a fairly
common form. As shown in the figure, the cyst contents are
yellow in colour, almost bright yellow, with radiating struc-
ture resembling yolk-columns. As many as eight concentric
lamin were observed in the cyst drawn in fig. 10.
Affinities of Chlamydomyxa.—I do not think that either the
observations made by me on C. montana, or the progress of
our knowledge of Protozoa since Archer described C. labyrin-
thuloides twenty-one years ago, enables us definitely to assign
to Chlamydomyxa its position in relation to other Protozoa
Gymnomyxa.
I cannot agree with Professor Geddes that it should be
regarded as related to the Algz, since affinities have to be
determined by a consideration of all the circumstances, and I
cannot see how Chlamydomyxa would fit in with known Alge.
I fully agree with Archer that the nearest ally of Chlamy-
domyxa is the Labyrinthula of Cienkowski, and that any con-
sideration of affinities must be based on this alliance. Archer
was inclined to regard the “ threads”’ of Chlamydomyxa as
protoplasmic, whilst Cienkowski regarded those of Labyrin-
thula as of a horny nature. I do not think that there is any
essential difference between the threads of Labyrinthula and of
Chlamydomyxa. I regard them both as a“ formed material,”
differing from streaming protoplasm, and comparable to the
axial pseudopodial filaments of Heliozoa, The fusiform travel-
ling nuclei of Chlamydomyxa differ very greatly from the oat-
shaped nucleated corpuscles of Labyrinthula both in size and
structure. But it seems to me a reasonable view that the con-
dition of Chlamydomyxa is derived from that of Labyrinthula,
aud that the fusiform nuclei of the former represent the oat-
shaped corpuscles of the latter in a reduced condition. The
CHLAMYDOMYXA MONTANA. 243
fact that in C. montana as compared with C. labyrinthu-
loides the fusiform nuclei are still further reduced, being but
one third their size, tends to indicate a progressive reduction
of these bodies. The protoplasm surrounding the nuclei of
the oat-shaped corpuscles of Labyrinthula is represented in
Chlamydomyxa by an exceedingly delicate and practically
invisible layer, which is also extended over the threads, and is
the seat of the movement which is rendered visible by the
translation of the nuclei.
The Protozoa which come nearest to Chlamydomyxa and
Labyrinthula are certain of the Mycetozoa; but each of the
two genera in question differs in its own way from the typical
Mycetozoa, especially as to reproduction, and nothing would be
gained by sinking them taxonomicaily in that assemblage.
The most remarkable feature in which Chlamydomyxa differs
from Labyrinthula and from all other Protozoa is its “ encysted
phase.’’ The enclosure of the general protoplasm in a cyst-
wall may be compared with the fruit-formation and other
cyst-like productions of the Mycetozoa. But the physiological
character of the cysts of Chlamydomyxa, the activities of the
encysted organism, and the great relative duration and im-
portance of the encysted phase are peculiar to Chlamydomyxa,
and may be explained by the fact that this organism is an
inhabitant of fresh water, and subjected to the vicissitudes of
temperature and evaporation of the inhabited water which we
know are frequently associated with special protective struc-
tures and aberrant phases of growth and activity.
A point in which both Chlamydomyxa and Labyrinthula
agree with the Mycetozoa is their epiphytic habit.
In his paper on Chlamydomyxa labyrinthuloides,
Archer draws attention to a terricolous plasmodium (found in
a flower-pot) of an unknown Mycetozoon described by Cien-
kowski. It is not possible to decide from his (Cienkowski’s)
description whether he had before him Archer’s Chlamy-
domyxa, but it is most probable that he had not, since
he does not describe the characteristic structures of that
organism. Similarly the Biomyxa vagans described and
244 E. RAY LANKESTER.
figured by Leidy, in his large volume on Rhizopoda in the
United States Geological Survey of the Territories, 1879,
presents some points of agreement with Chlamydomyxa, and
was discovered by Leidy upon Sphagnum. Yet inasmuch as
neither the delicate filaments nor the fusiform nuclei, nor the
central coloured vesicles (granules), nor cellulose laminated
cysts are ascribed by Leidy to his Biomyxa, we must suppose
that it indicates a distinct organism, not even closely related
to Chlamydomyxa.
OxrorD; July 8th, 1896.
EXPLANATION OF PLATES 14 & 15,
Illustrating Professor Ray Lankester’s memoir on “ Chlamy-
domyxa montana, n. sp.”
All the figures except Fig. 6 are drawn to the same scale.
Fic. 1.—Chlamydomyxa montana, n. sp., in an unencysted expanded
condition.
Fic. 2.—The same specimen-after an interval of five minutes. Drawn at
Pontresina from a living specimen as seen with objective No. 10 immersion
of Hartnack, August 22nd, 1886.
Fic. 3.—Another and larger specimen observed and drawn at the same
time.
Fic. 4.—A specimen which is beginning to expand its filaments. After
five minutes the threads would have a much wider extension.
Fic. 5.—A single fusiform corpuscle (nucleus) greatly magnified, showing
its shape and its relation to the filament. The actual length of the corpuscle
iS =5455 Inch.
Fig. 6.—An encysted specimen of C. montana which has divided into
three. Pontresina, August, 1886.
Fic. 7.—An encysted specimen of C. montana which has divided into
two. Pontresina.
Fic. 8.—An encysted specimen of C. montana which has developed
chlorophyll and a crimson oil drop. The cyst-wall is simple. Pontresina.
Fic. 9.—Three similar encysted specimens. Pontresina. These and the
last were brought alive to London and drawn there.
Fic. 10.—A spherical cyst with eight lamine to the cyst-wall and a golden
yellow pigment. This last specimen observed and drawn at the Maloja in~
1892.
JAN 4 1897
OONSTITUTION AND DEVELOPMENT OF TERMITES. 240
The Constitution and Development af the Society
of Termites: Observations on their Habits;
with Appendices on the Parasitic Protozoa
of Termitide, and on the Embiide.
By
Professor B. Grassi im collaboration with Dr. A. Sandias.
With Plates 16—20.
Tue memoir, of which Ihave made the following translation, was originally
published in the ‘ Atti dell’ Accademia Gioenia di Scienze Naturali in Catania,’
ser. 4, vols. vi and vii (1893-4), but has become more generally known
through the appearance of a separate edition in the former year. But though
its value was immediately recognised, the work has not yet become familiar
to an extent commensurate with the importance of its contributions to natural
science.
The object of the treatise, itself an expansion and completion of certain
preliminary papers on Termitidee by Professor Grassi, is set forth in that
author’s introduction, and requires no further explanation; but it is perhaps
permissible to point out that, over and above the results obtained in the
pursuit of that object, the memoir is a signal and instructive example of a
class of work but too seldom resorted to—the union of morphological re-
search with an inquiry of the most prolonged and persevering character into
the habits and bionomics of the living form. In these latter respects it forms
a worthy parallel to the work of such great pioneers in the investigation of
the social systems of insects as Smeathman, Huber, and Réaumur. And it
is for this reason that no study of an abstract which merely summarises the
results obtained can prove as convincing or suggestive as that of the original.
An excellent general summary of the various writings on Termitide, which
includes references to such facts as have been observed since the publication
of the present memoir, is to be found in vol. v of the ‘Cambridge Natural
History,’ by Dr. Sharp. In view of its existence it has been thought un-
necessary to make more than a very few additions to the original text; these
will be found indicated in the usual way by square brackets.
Wa ter F. H. Branprorp, M.A.Cantab.,
Lecturer on Entomology at the Royal Indian
Engineering College, Cooper’s Hill.
VOL. 389, PART 3 —NEW SER. R
246 B. GRASSI AND A. SANDIAS.
To Fritz MUtier,
on his Jubilee.
Catania, 1892.
INTRODUCTION.
Ir has been recognised by Darwin, in his immortal work on
the ‘ Origin of Species,’ that the existence of insect commu-
nities, composed of different castes (kings, queens, workers,
soldiers, and the like), furnishes his opponents with a trenchant
weapon. Now these communities actually consist not merely
of fertile forms (kings and queens), but of sterile forms
(workers, soldiers, &c.), distinguished by important modifica-
tions of structure and marvellous instincts, none of which are
to be found in their parents. The sterile progeny would there-
fore appear to be uninfluenced by one of the prime factors in
the struggle for existence, that of heredity, owing to the im-
possibility of transmitting to the offspring such modifications
of structure or instinct as may be gradually acquired; and
Darwin himself admits that it required great confidence in his
theory not to renounce it in the face of this objection.
I have had to deal shortly with this question on another
occasion with regard to bees, for which I have observed that
the cause of their apparent deviation from the normal rule
may possibly be found in the existence of workers capable of
oviposition, which apply themselves to all the tasks of the
colony, and possess the characteristics of true workers, with
the single difference that they lay parthenogenetic male ova.
The males which hatch from these eggs would then transmit
to the female ova—laid, as we know, by the queen—the cha-
racteristic properties of the worker.!
1 Darwin, however, explains the phenomenon by a very ingenious com-
parison. ‘ According to M. Verlot,’ he writes, “some varieties of the
double annual stock, from having been long and carefully selected to the right
degree, always produce a large proportion of seedlings bearing double and
quite sterile flowers ; but they likewise yield some single and fertile plants.
These latter, by which alone the variety can be propagated, may be compared
CONSTITUTION AND DEVELOPMENT OF TERMITES. 247
As it is probable that the same explanation is equally ap-
plicable to other Hymenoptera, I put aside that order of
insects to devote myself to another, of remote relationship—
namely, the Corrodentia, of which the Termitide constitute
the typical form. But I found myself here in a much less
cultivated field than that of the Hymenoptera, and it took
me several years before I was able to see my way.
In the meantime a school has arisen under Weismann,
which denies the heredity of acquired characteristics ; this has
somewhat modified, and perhaps rendered more interesting
the general questions which have inspired my particular re-
searches, as I will explain.
It has been known for many years that the same bee larva
may develop either into a queen or a worker in accordance
with the nutriment it receives. Nutrition, therefore, possesses
a most remarkable influence on the bee’s generative organs,
and on many other characters correlated with their greater or
less development (e.g. the faculty of producing wax, or of
collecting honey, which is possessed by the worker only, and
not by the queen). This would indicate that the environment
has a powerful and direct influence on the genitalia, and would
therefore tend indirectly to show that the much-disputed in-
heritance of acquired characteristics is a possibility.
with the fertile male and female ants, and the double sterile plants with the
neuters of the same community. As with the varieties of the stock, so with
social insects, selection has been applied to the family, and not to the indi-
vidual, for the sake of gaining a serviceable end.” [‘Origin of Species,’
ed. 6, p. 230. The whole chapter should be consulted.] But it may be
objected that the difference between the queen and worker forms is far more
profound than that between the simple and double stocks.
Biichner [* Aus dem Geisteslebens der Thiere,’ translated under the title
‘Mind in Animals,’ London, 1881], on the other hand, thinks that the
explanation of the phenomenon should be found partly in atavism, and partly—
as far, that is, as regards the marvellous instincts—in the instruction which the
young receive from the colony. That atavism certainly plays a part in instinct
is demonstrated ad evidentiam by a fact discovered by myself. It is known
that certain silkworms become pup and moths without spinning a cocoon.
Now my experiments show that the offspring of such moths may spin perfectly
constructed cocoons.
248 B. GRASSI AND A. SANDIAS.
But this argument in its favour would be greatly weakened
if the circumstance indicated by me above—namely, the pos-
sibility of hereditary transmission by means of oviparous
workers—should have played the part which I suppose it to
have done.
I determined, therefore, to investigate the origin of the
workers and soldiers in Termitide,—in forms, that is, whose
phylogenetic source is certainly absolutely distinct from that
of the social Hymenoptera.) Such an inquiry would at
least throw light on my hypothesis with regard to the bee,
and in any case should lead to results of some utility for the
problem I have several times referred to, the heredity of
acquired characteristics, and for the subordinate and related
question of the direct influence of environment on the gene-
rative organs.
The recent researches of Van Beneden, Boveri, O. Hertwig,
and others on the ovum and spermatozoon had afforded yet
another problem for solution, which would be modified accord-
ing as to whether there should or should not exist special eggs
or spermatozoa for the workers and soldiers of Termitidee.
In short, the theory of evolution, the disputed heredity of
acquired characteristics, and lastly, the theory which postu-
lates the existence in every somatic cell of elements derived
from both parents, these alike have all furnished me with
motives for regarding the elucidation of the origin of caste
forms in the social Termitide as a problem of the highest
interest. This, then, is the main object of the present memoir,
and it has been arrived at only by dint of prolonged observation
and preliminary experimentation, the results of which are
fully related here as a necessary corollary.
The development of my argument has, so to speak, com-
1 It must be recollected that, though we are acquainted with forms which
perhaps can be ascribed to the Blattide from the Silurian strata, the Termi-
tide, according to Scudder (in Zittel’s ‘Handbuch der Palaontologie,’ ii, 772),
are absent in all the paleozoic, and appear only for the first time in mesozoic
strata. Nevertheless the phylogenetic independence of the Termitide and
Hymenoptera is indisputable.
CONSTITUTION AND DEVELOPMENT OF TERMITES. 249
pelled me to discuss neoteinia! and the simplification and
degeneration of forms, and therefore to express my present
views on the Thysanura, which for a long time formed the
special object of my studies.
In these preliminary remarks I have prominently set forth
the fundamental ideas which have directed my work, and now
pass at once to the consideration of the Termites.
As yet two species only of Termitide, Calotermes flavi-
collis, Fabr., and Termes lucifugus, Rossi, have been
found in Italy. The latter forms the subject of Lespés’
admirable monograph,? but Calotermes flavicollis is known
to us only by a few accounts of merely systematic importance,
and not even complete.®
However, Fritz Miller has made important investigations*
1 (The term neoteinia has been introduced by Camerano (‘ Bull. Soc.
Ent. Ital.,’ 1885, pp. S9—94) to denote the persistence during adult life of
part or all of the characteristics normally peculiar to the immature, growing,
or larval stages (e.g. the persistence of gills in the axolotl). It therefore
covers much the same ground as is denoted by the term pedogenesis, but
appears, so far as can be gathered from Camerano’s paper, to include a some-
what wider class of facts than those comprised under the latter term, which
would fall under his definition of total as opposed to partial neoteinia.
Neoteinia, or the persistence of larval characteristics, does not necessarily
imply that anticipation in time of sexual maturity which is usually connoted
with the use of the term pedogenesis,—which, moreover, is strictly applied
to agamic reproductions.—W. F. H. B.]
2 « Recherches sur l’organisation et les meurs du Termite lucifuge,”
‘Ann. Sci. Nat.’ (4), v (1856), pp. 227—282, pls. v—vii.
3 (Hagen, “ Monographie der Termiten,”’ ‘ Linn. Entom.,’ x, 1—144, 270—
395; xii, 4—342. The reference to C. flavicollis, op. cit., xii, pp. 54—
61, pl. i, fig. 12; pl. ii, fig. 15.]
4 “Beitrage zur Kenntniss der Termiten,” ‘Jen. Zeitschr.,’ vii (1873).
I. “Die Geschlechtstheile der Soldaten von Calotermes,”’ pp. 333—340,
pls. xix, xx. II. “Die Wohnungen unserer Termiten,” pp. 341—358.
III. “‘ Die Nymphen mit kurzen’ Fliigelscheiden ? (Hagen), ‘ Nymphes de la
deuxitme forme’ (Lespés). Hin Sultan in seinem Harem,” pp. 451—463.
Id., ix, 1875. IV. “Die Larven von Calotermes rugosus, Hag.,”
pp. 242 —264, pls. x—xiii.
250 B. GRASSI AND A. SANDIAS.
on another species of Calotermes, and these, as will be seen
later, are in entire agreement with our observations on the
European species.
Nevertheless, as I have already mentioned, the study of
Termitide in general is as yet very incomplete ; and although
books of travel certainly abound in notices of these interesting
insects, they mostly deal with detached and imperfect obser-
vations.
I have repeated Lespés’ investigations on Termes luci-
fugus with particular reference to the so-called ‘“ nymph of
the second form” which that distinguished naturalist de-
scribed.
He attributed to it an entirely different significance from
that which it really possesses, owing mainly to the circum-
stance that his period of study was confined to about nine
months—from November, 1855, to August, 1856. But how
difficult the elucidation of the true function of this nymph of
the second form has been the reader will gather from the
pages of this memoir; and I may state at once that it has
necessitated the removal of many hundred cubic metres of
earth and the cutting up of hundreds of trees, a task that has
had to be carried out little by little, and has consequently
demanded no small amount of patience.
I am indebted to a small but fortunate discovery for the
most interesting observations on Calotermes flavicollis,
which I have to relate in the present work.
If from three to twenty Calotermes of different ages are
placed in a glass tube three to eight centimetres long (Pl. 19,
fig. 11), closed with a cork and kept warm—for example, in the
waistcoat pocket, unless in summer-time, they continue to live
and constitute a family, or better, an independent colony ; they
rear a fresh king and queen,’ if orphaned, and they rear soldiers,
1 [The statements made on the sexual forms of Termitide will be more
clearly understood if anticipated by a summary and definition of the terms
employed in the course of the work. Sexual or royal forms are of two
kinds, true and neoteinic. The true royal forms are imagos, or
perfect insects, which acquire a complete development of the wings and
CONSTITUTION AND DEVELOPMENT OF TERMITES. 251
&c., if in want of them. In short, after a certain time the tube
will contain a complete little nest, if it did not oxiginally
do so. However, the insects do not hesitate to bore through
the cork, and, unless they are watched, one is surprised to find
some day that they have all made their escape. But fre-
quently they remain in the tube, even though they have made
holes putting them into communication with the outer air
(PL 19, fig. 11).
Some of these little nests can be kept alive for several
months, but many die off after a few weeks, though not until
they have afforded sufficient opportunity for making numerous
observations through the glass walls by aid of a lens. It is
desirable to use tubes of various lengths and calibres, because
certain points are better seen in a wide than in a narrow
tube.
In addition to the insects, the tubes are partly filled with
fragments of wood, which should be neither too dry nor too
moist. In the former case the insects gradually shrivel, con-
tract, dry up and die; and in the latter case there is a deposit
of water vapour on the inner walls, and they are evidently
killed by over-dampness. Death ensues more or less rapidly
according to the amount of water deposited, and is sometimes
almost as sudden as if the insects were suffocated or chloro-
formed. Slow death due to over-dampness may be accom-
panied by distinct cedema or reddish discoloration of the
darkening of the chitin concomitantly with the maturation of the gonads after
the last ecdysis. They leave the nest by “swarming.” The neoteinic
royal forms undergo a premature maturation of the gonads whilst in a late
arval or nymphal instar, or that of an immature and pallid imago. This
maturation is accompanied by an arrest of development of other parts of the
body; the chitin does not darken normally, and the wings do not grow
further. They do not leave the nest. The latter class is further subdivided
into complementary and substitute royal forms. Complementary
forms are not found in Calotermes, but are normal components of the Termes
nest, in which they are the only reproductive individuals. Substitute
forms are developed to supply the loss of the true royal forms in Calo-
termes, or the complementary royal forms in Termes. ‘The three varieties
of royal examples all comprise kings and queens,—that is, individuals of
each sex. ]
252 B. GRASSI AND A. SANDIAS.
body ; the latter is accompanied by the presence of a bacterium
which I have not investigated.
If the wood is too damp the colony may generally be saved
by the simple precaution of leaving the cork out of the tube ;
after a few days the moisture diminishes, and the cork must
then be re-inserted, or the opposite extreme of undue dryness
is quickly reached. With time and patience an exact estimate
can be formed of the amount of moisture necessary, and it
can then be easily regulated. The cork must fit closely, and
want of oxygen need not be feared. At one time I was
accustomed to uncork the tubes three or four times a day,
but I subsequently found this to be quite unnecessary; if the
Calotermites require change of air, they are able to provide it
for themselves by boring through the cork.
When these tiny nests are examined the tube may conve-
niently be laid flat and left quiet, but an occasional shake is
sometimes useful] to rouse the inmates to activity.
Unfortunately Termes lucifugus does not flourish in
these tubes, but drags on a feeble existence for a week or ten
days at most; I have made many but quite unsuccessful
attempts to establish a moderately suitable environment for
them therein. The species is all the more difficult to study
because, in comparison with Calotermes, the members of a
colony are smaller and more rapid in movement.
Many as have been the details made out through these glass
tubes in my laboratory, there are others which will be more
successfully undertaken by anyone who has the opportunity of
applying my method to the large exotic species.
Considering that the termites of tropical countries are
among the most injurious of insects, I must point out a
practical result of my investigations, and one of the highest
importance. Contrary to the belief of residents and the
accounts of travellers, such as Major Casati,! a nest of termites
cannot be destroyed merely by killing the king and queen.
If there remain alive some eight, ten, or twenty examples,
which include any undifferentiated larve, or larve of perfect
1 («Ten Years in Equatorial Africa,’ i, p. 165.]
CONSTITUTION AND DEVELOPMENT OF TERMITES. 250
insects or nymphs, these eight, ten, or twenty will form a new
colony, which slowly but surely will become as flourishing as
the original stock.
It follows, therefore, that the extirpation of these insects is
excessively difficult in practice; and their entry into any
situation where they can prove destructive must be prevented.
This is the strategy to adopt against them, for when once they
have gained admission there is no way of getting rid of
them.
The present work has no claim to be a complete monograph
of the Termitidz, and merely furnishes materials for whomso-
ever will undertake so large a task. It is confined, in short,
to the discussion of certain fundamental points.
Before concluding I must express my thanks to Dr. Calan-
druccio, who has given me no little help in this research.
Many observations, particularly those on Calotermes kept
in tubes, have been made under my supervision by my pupil
Dr. Sandias, and have afterwards been carefully verified by
myself.
Tur CoLony or CALOTERMES FLAVICOLLIS.
1. Situation and Nest.
As with all Termitide, the life of Calotermes is intimately
connected with the vegetable kingdom, for it lives exclusively
on woody matter. At Catania I have found it only in the
stems or roots of living but partially decayed, and therefore,
as a rule, old plants, and never in those of which the stem or
tap-root measures less than one or two centimetres in thick-
ness. But when once established in a stem or main root it
can extend into decayed portions of the branches or side-roots,
even if of lesser diameter. Though always absent in entirely
sound plants, it can be found in partly decayed examples of
many kinds,—of many, but not of all, for I have never met
with it in lemon, orange, or Agave americana, &c. It is
also relatively scarce in cactus (Opuntia), even though large
portions of the plant are dead; but if present, it may infest
254 B. GRASSI AND A. SANDIAS.
the rotten phylloclades, which in fact are sometimes the first
portions to be occupied.
It is difficult to find a tree with any decayed portion which
has not been attacked by Calotermes in Catania and the
adjoining provinces, at least in the low country, for I have
not searched at a higher elevation than Nicolosi. At Castro-
giovanni the species appears to be entirely absent.’
As mentioned above, it inhabits still living plants. Should
the plant die the Calotermites survive until it has become
completely dry, when they perish—a fact that anyone can
verify by examination of the vine-stocks annually turned out
from old vineyards. If the dead trunk-does not dry up, as
is the case in marshy situations, the insects continue to
flourish.
Lucas and some other writers state that they have found
Calotermes flavicollisin buildings. I have never observed
this at Catania, but Dr. Sandias has found flourishing colonies
at Trapani in the woodwork of verandahs, doors and stair-
ways, &c., ten years old; probably the wood harboured the
insects before being worked up for domestic purposes. It
should be observed that the climate of Trapani is very damp,
so that wood probably dries less there than at Catania.?
In order that a plant may harbour Calotermes, it must
(necessarily), as I have stated, exhibit some amount of decay,
because such decayed portions alone are occupied. To proceed
to details, Calotermes never invades the healthy parts, but
encroaches at most on their boundaries. If a partly decayed
vine-stock is infested, it is usually easy to make out that the
healthy tissues are respected; yet should they contain an
internal channel of decay barely larger than the body of a
Calotermite, the insects can enter and excavate a gallery, which
will then present but a very thin lining of decayed matter, and
1 [In a foot-note at the end of the original work the authors state that, since
it was printed, Signor Giuseppe Corona has discovered the species at Castro-
giovanni. |
2 The Marchese Doria informs me that he has confirmed Dr. Sandias’s
observation at Genoa, although the climate there is certainly not moister than
that of Catania.
CONSTITUTION AND DEVELOPMENT OF TERMITES. 255
will therefore appear to run in the sound structures. But, so
far as I have observed, this lining of decayed tissue is never
absent.
It is undeniable that the process of decay is hastened by
these burrows, particularly because they allow the infiltration
of rain-water ; but a detailed study of these phenomena would
lead too far from the subject.
Calotermes greatly prefers to live in the deepest parts of
the dead wood,—that is, the parts nearest the living tissues,
and consequently the most recently decayed. Portions which
have been long dead are usually too dry for it, and are chosen
as a tenement by ants.
Fertile pairs, unaccompanied by eggs or with very few
offspring—that is to say, nests in process of formation or
recently formed,—are to be found by special search at the
places where a plant has been pruned a year or two previously
and has subsequently rotted, provided that the pruning has
not followed on antecedent decay. ‘This can be easily observed
in the fig or vine. Fresh nests, therefore, are usually estab-
lished where the dead portion is still so limited in extent as
not to afford space for a numerous colony, and these situations
are rarely preoccupied by other insects, such as ants, beetles,
or other Termitide, &c. As the colony gradually increases
and requires more room, the decayed area spreads, but to all
appearance independently of the insects. That this is the
case is rendered probable by an examination of progressive -
decay in a district where Calotermes is absent, such as
Lombardy, where the winter is too severe for them.
Turning once more to the formation of the fresh termitaria,
I do not wish to deny the possibility of its taking place in
trunks in which extensive decay already exists; but such an
occurrence is unusual, either because the wood is preoccupied
by termites or other insects, or because the royal pairs must
traverse parts which are already dry and swarming with enemies
(ants), in order to arrive at the required spot where the decay
is most recent.
The corks of the tubes are attacked probably because they
256 B. GRASSI AND A. SANDIAS.
retain a certain amount of moisture; for if the insects are
shut up with nothing but particles of cork they all die in a
few days.
To sum up, two conditions are essential for the life and
well-being of Calotermes—a suitable temperature and a
suitable amount of moisture. But whereas increase of tem-
perature is favorable, at least up to a certain point, the degree
of humidity can vary only within very restricted limits.
Owing to the severity of the winter, Calotermes is absent
from Lombardy, Piedmont, and Venetia, but it is found in the
province of Genoa; its northward distribution is not accurately
known. At Catania it is much more sluggish in winter than
in summer, the ova do not develop, and the larve and nymphs
do not moult. The influence of moisture has been repeatedly
referred to; it is correlated with the nature of the cuticular
structure (chitinous layer), which except in a few regions, such
as the mandibles, is very thin, even in the adults and royal
pairs, in both of which it is brown in colour. The younger
the individual, or the more recent its ecdysis, the thinner is
the chitin. In general the rule holds good for Calotermes
that white and semi-transparent examples have a thinner
cuticle than those which are more opaque and inclined to
yellow (very old soldiers or substitute-queens). Young or
freshly moulted specimens, usually distinguishable by their
greater transparency and whitish or rarely faintly yellow colour,
always require more moisture than those which are older or
have not cast their skin for some time, and possess a more or
less evident yellow tint. The imagos, before or after the loss
of their wings, can endure a greater degree of dryness; but
even these generally die in a short time if removed from the
wood and exposed to._the open air.
These conditions of temperature and moisture, naturally
with some variation in degree, especially of the latter factor,
must hold good for all other species of Termitide, and are
probably correlated with their limitation to warm countries,
their much wider European distribution in epochs when the
mean temperature was presumably higher, and with their
CONSTITUTION AND DEVELOPMENT OF TERMITES. 257
choice of evening or night, the interval after a shower or a wet
day as a swarming-time (Casati).!
The nest of the Sicilian Calotermes resembles those of its
congeners, which show, as a rule, very little architectural
skill. Strictly speaking, they do not build any nest, but con-
tent themselves with burrows excavated in wood and never
quitted. These burrows vary very much in dimensions, and
the same gallery may be narrow in one part and very wide in
others. In one of the enlargements, towards the heart of the
nest, the royal couple, with a numerous surrounding, is usually
to be found. There is never a true royal chamber, such as
has been described for many other species. The galleries are
very variable in direction, but the widest and longest are
generally subparallel with the long axis of the stem. A large
number of transverse passages is commonly found, and these
are often too narrow to allow room for more than a single
individual at a time. According to Fritz Miller,? the burrows
of Calotermes possess an inner lining of excrement. This
appears at first sight really to be the case when they are exca-
vated in rotten and somewhat softened wood. But if they
are examined in cork, or in wood which, though dead, is still
hard, the absence of such a lining is easily established, and
this fact once determined, one can satisfy oneself that burrows
made in damp wood are also unlined.
In common with other termites, Calotermes avoids the
outer layers of the cortex (as may be readily observed in vine-
stocks), and thus protects the nest from the direct action of
the atmosphere or infiltration of water.
As previously stated, the colony provides for an increase in
numbers by penetrating deeper into the wood. This penetra-
tion is also determined by the fact that the superficial wood
becomes too dry. The abandoned portion becomes then com-
monly occupied by ants, the pitiless enemies of Calotermes.
The termitarium is separated from the ant-burrows, or from
openings caused by wind-cracks or axe-wounds by means of
1 (‘Ten Years in Equatorial Africa,’ i, p. 166.]
2 (‘Jen. Zeitschr.,’ vii, p. 343.]
258 B. GRASSI AND A. SANDIAS.
barricades of masticated wood, or more commonly of excre-
ment, which is cemented together by saliva, with or without an
admixture of disgorged wood. This disgorgement is normal,
and not exceptional, as Fritz Miller believes.
The following facts and other details of interest may be
easily observed through the glass tubes.
If some fragments of wood and a few Calotermites are
put into a tube and left uncorked, the insects will be found
on the following day to have established themselves either in
the whole of the space filled with wood, or in its lower por-
tion, at the bottom of the tube. In either case the occupied
area is delimited by means of disgorged matter deposited in
the interstices between the particles of wood. As far as can
be seen the boundary wall is complete, but it lies at different
levels and not entirely in the same plane; its margin can
easily be made out at the point of contact with the glass, and
will be found on revolving the tube to form a complete circle,
but with some irregularities of level, obviously due to a
selection of the smallest interstices for cementing up without
reference to their higher or lower elevation.
[The process of construction can be watched in the tubes ;
the insect regurgitates a pasty mass, which is spread out, if
necessary, by the antenne, so as to form a rounded pellet on
the glass of about a millimetre in diameter, and of the colour
of rotten wood. |
The boundary wall is undoubtedly designed for protection
against direct contact with the atmosphere ; this is proved by
its not being constructed if the tube_is closed with a cork
instead of remaining open.
Such a wall may occasionally be built even in a corked
tube, should part of the nest contain any offensive substance.
Thus if flour be added to a small tube-nest, after a few days
the upper part containing the flour, which will have become
mouldy, is found to be cut off in the manner described’from
the lower portion, in which all the insects have collected.
Sometimes the Calotermites,in a tube may be seen to cement
1 [‘ Jen.®Zeitschr.,’ vii, p. 3438.]
CONSTITUTION AND DEVELOPMENT OF TERMITES. 259
together pieces of wood which are unduly loose; but at other
times they do not do so, although much incommoded by their
mobility. In certain cases they line the lumen of the tube
with disgorged matter, so as to form a continuous layer, save
for a few irregular patches, but they generally spare them-
selves this task, or merely begin and immediately abandon it.
It should be added that if the tube is too damp its inner
surface may be covered with moisture, to which the insects
adhere, particularly when young, and so die. This is guarded
against, at least partially, by the lining of disgorged substance.
When the supply of wood in the tube is very small, the
Calotermites usually have recourse to the cork, in which they
bore one or more galleries a little wider than their body ; they
gnaw away and hollow out the cork, carrying the débris into
the tube, which is soon filled up, as the cork-dust is allowed
to remain as a loose mass.
If a corked tube is surrounded with sawdust, some of it can
generally be found in its interior, having been carried in
through galleries in the cork but little wider than the insects’
bodies.
[The corks of tubes kept in the waistcoat pocket are often
channelled with burrows leading to the exterior, through which
the insects escape and are lost, in the belief that they are
enlarging their nest. Sometimes books near which the waist-
coat was laid at night were attacked, and the paper was found
to be gnawed in places the next morning. |
Burrows made in corks may often be seen to possess a sort
of rounded lid at their external opening; this is cut out of the
superficial layer of cork, and fits tolerably closely, though it
usually falls in a little. It may be hinged to the cork by a
more or less wide attachment.
[Calotermites may be frequently seen to carry excrement to
the bottom of the tube, and to accumulate it there, mixed
up with wood and cork-dust ; and sometimes they keep the eggs
in the same place. |
If the stopper is too narrow the space between it and the
tube is gradually filled up with excrement.
260 B. GRASSI AND A. SANDIAS.
Calotermes flavicollis differs materially from Termes
lucifugus in the fact that it confines itself to excavating
burrows, and is thus merely a borer; whereas the latter
species not only bores, but builds tunnels, in order to connect
pieces of wood at a considerable distance apart.
2. Number of Individuals in the Colony.
In these observations a normal development of the colonies
isassumed. For should the queen, for example, die before the
community has reared a successor, several months pass during
which oviposition is suspended and multiplication is conse-
quently brought to a standstill.
A colony of Calotermes rarely consists of more than a
thousand members, and is relatively numerous when it contains
five hundred. This is correlated with the fact that the queen
of our species is very far from attaining the colossal dimen-
sions which are well known to occur in the queen-termites of
tropical countries.
After fifteen months of common life the king and queen
may be surrounded with fifteen or twenty young, after another
year with about fifty, and in the two or three following years
the population increases till it reaches a maximum at which it
becomes nearly stationary. This is attained when the king
and queen have reached the largest possible size.
Eight or ten winged adults may depart from a two-year-
old nest, and the number leaving in successive years increases
concomitantly with the increase in the population.
At the time of maximum oviposition a queen of three to
four years old lays as a rule four, five, or six eggs a day.
3. The Different Castes (Plate 16).
By way of preliminary, it should be stated that for the sake
of orderly arrangement a knowledge is assumed in this part
of the work of certain conclusions, the demonstration of which
is postponed to the succeeding chapter.
A colony of Calotermites contains—
1. Undifferentiated larve (fig. 1), capable of becoming
CONSTITUTION AND DEVELOPMENT OF TERMITES. 261
either soldier larvee, or larve of perfect insects,—that is, larve
possessing the first indications or rudiments of wings.
2. Larve of perfect insects (fig. 3), and nymphs (fig. 4)
derived from them.
3. Larve of soldiers (fig. 2), and soldiers (figs. 5 and
16); the latter derived from the former, and those derived
from undifferentiated larve, from the larve of perfect insects,
or from young nymphs.
4. Perfect insects derived from nymphs (fig. 6).
5. A royal couple, properly so called, derived from perfect
insects (figs. 7—11).
6. Larve of substitute royal forms (fig. 14), and
substitute forms (figs. 12, 13, and 25) themselves derived
from the former; those in their turn having originated from
undifferentiated larve with fourteen or fifteen antennal joints,
from larvee of perfect insects, or from nymphs.
To proceed to details: a colony contains newly hatched
larvee about 1 mm. in length,! with ten antennal joints of
which nine are pilose, and one not. The glabrous third joint
is relatively very long, and presents indistinct traces of a
tripartite division. 'These traces soon become more marked,
and the separation of the distal joints appears constantly to be
clearly defined before that between the middle and the proximal
joints, and possibly may actually precede it.
As a result we have larve which are a little over 2 mm. in
length, and exhibit twelve distinct antennal joints, of which
the 3rd, 4th, and 5th are short and bare. In succession the
5th joint, as far as I have seen, becomes pilose and relatively
longer; then the 4th and finally the 3rd exhibit the same
phenomena.
We thus obtain examples with twelve antennal joints, all
pilose; and in the meantime the body increases in size, so that
such specimens measure about 4 mm. in length.
The newly born larve are all perfectly alike,—that is, they
1 In Termitide the length of the individual as a rule has only a relative
value, as it depends largely on the degree of moisture in the atmosphere, the
food supply, &c.
VOL. 39, PART 3.—NEW SER. S
262 B. GRASSI AND A. SANDIAS.
are undifferentiated; but when they have attained an
average length of 2 mm. they are divisible into two groups,
one with a large head, little narrowed anteriorly (Pl. 16,
fig. 2), the other with a small head, more evidently narrowed
towards the apex (PI. 16, fig. 1).
The former have become soldier larve, and will ulti-
mately become soldiers. The latter are still undifferentiated ;
they reach alength of 4mm. and acquire twelve pilose antennal
joints (vide supra); the character of the head either remains
unchanged, and continues to be undifferentiated, or be-
comes modified by an increase in size and in the width of the
anterior portion, so that they become soldier larve.
Small soldiers (Pl. 16, fig. 16) are to be found which are
derived from these very larvee ; they are less than 5 mm. long,
and possess twelve pilose antennal joints, with an ill-marked
suture between the 4th and 5th.
It may be well to observe at once that soldiers of me-
dium size with thirteen or fourteen pilose antennal joints,
and large soldiers (Pl. 16, fig. 5) with fifteen to seventeen
pilose antennal joints also exist. Further, the antenne of
many soldiers are evidently mutilated.
Moreover I have often found soldiers with antenne of twelve,
thirteen (Pl. 16, fig. 19), or fourteen all pilose joints, of which
the third is in process of division into two, both pilose ; some-
times the division is completed in one antenna and hardly
indicated in the other. In short, all intermediate stages exist
between small, medium-sized, and large soldiers, and the fore-
going distinctions have therefore a relative value alone.
For a considerable time I supposed that the soldier of
medium size originated from the small soldier, and grew itself
into a large soldier; but as I have not succeeded in proving
my hypothesis, in spite of long-continued research, I can no
longer consider it to be well founded.
Turning to the undifferentiated or small-headed larve, we
have already seen that they can acquire antenne of twelve
entirely pilose joints, and a length of 4mm. Subsequently a
thirteenth joint is added, and their length increases to 6 mm.,
CONSTITUTION ANI) DEVELOPMENT OF TERMITES. 2638
when also the thirteenth joint becomes pilose ; then a fourteenth
joint is developed, and the whole fourteen are always found to
be pilose in examples of 7 mm. and upwards in length.
Iixamples with fourteen joints may exhibit evident traces of
wings, which indeed may be already present in individuals
with thirteen pilose joints, and a very indistinct rudiment of a
fourteenth glabrous joint at the base of the third. But, on the
other hand, examples with fifteen joints may be found without
trace of wings.
Those which have acquired wing-rudiments may be called
larve of perfect insects.
Undifferentiated larve, with thirteen or fourteen joints, as
well as larve of perfect insects with thirteen (the fourteenth
very imperfect) or fourteen joints may undergo a direct trans-
formation into soldier larve with thirteen- or fourteen-jointed
antenne, and from these originate the soldiers of medium
size with the same number of antennal joints.!
Lastly, undifferentiated larve with fourteen joints may
either become larve of perfect insects without increase in
that number, or they may remain unchanged till it has been
increased to fifteen, all of which are pilose; and the latter
condition may also be reached without change by the larvee of
perfect insects.
Both the undifferentiated larve and those of perfect insects
with fifteen joints may give rise, without increase in that
number, to the same kinds of forms that have been indicated
for the larvee of fourteen joints,—soldier-larve, larve of royal
substitutes, and, provided that they are as yet undifferentiated,
larvee of perfect insects; and the last may acquire a sixteenth
joint without increase of the wing rudiments (PI. 16, fig. 3).
As a rule, therefore, the development of the sixteenth joint
is accompanied by an increase in the wing-rudiments, so that
the larva of the perfect insect becomes a nymph.
I have found no undifferentiated larva with more than fifteen
antennal joints; but, as I have said, we may have sixteen
1 | have been unable to decide whether the number of joints is capable of
increase in the soldier larve.
264 B. GRASSI AND A. SANDIAS.
completely pilose joints in the larve and nymphs of perfect
insects. During this stage both of the latter may undergo
the same fate as in the preceding stages, becoming soldier
larvee, larvee of substitute royal forms, or, if already larve of
perfect insects, nymphs.
Both the larvee of perfect insects or nymphs may pass with-
out change to the stage in which the antennz possess seven-
teen joints. During this stage both may become larve of
soldiers or royal substitutes. If this does not happen the
nymphs acquire larger wing rudiments, and the larve of perfect
insects become nymphs (Pl. 16, fig. 4).
Before proceeding further it is desirable to give a fuller
definition of the term nymph.
It may be applied to such examples as are 8 or 10 mm. in
length, with sixteen or seventeen antennal joints, and with
wing-rudiments easily distinguishable by the naked eye. The
expression, it must be observed, is incapable of precise defini-
tion, inasmuch as there are no characters in insects with an
incomplete metamorphosis which distinguish a larva from a
nymph, except the wings, which have already begun to develop
in larve of a certain age. It might possibly be adopted as
connoting the earliest rudiments of wings, but this presents
the difficulty that they are not easily detected. I therefore
follow the terminology of Lespés and Hagen, and conventionally
indicate as nymphs those individuals which have the beginnings
of the wings readily visible to the naked eye.
From the account just given it follows clearly that large
soldiers are derived from the large soldier larvze, just as the
moderate-sized and small soldiers originate from larve of
corresponding sizes, and that the large soldier larve have arisen
in their turn from undifferentiated larve with fifteen antennal
joints, or from larvee of perfect insects with antenne of fifteen,
sixteen, or seventeen joints, or lastly also from nymphs. The
latter, possessing sixteen or seventeen joints, may be trans-
formed into soldier larve, and subsequently become large
soldiers in which the wing-buds can be distinctly made out
with the naked eye (Pl. 16, fig. 26). At a later period these
CONSTITUTION AND DEVELOPMENT OF TERMITES. 265
rudiments are reabsorbed until hardly a trace is discoverable
with the microscope. This origin of the soldier from the
nymph is certainly infrequent.
Nymphs with seventeen antennal joints may become imagos
without increase in that number, or with an antecedent addition
of one or even two joints; and the perfect insects may thus
exhibit antennz with from seventeen to nineteen pilose joints,
a numerical difference which has no relation to sex. Nymphs
with seventeen to nineteen joints may be also transformed to
royal substitute larve, but I do not believe that those with
eighteen or nineteen joints can become soldiers, as I have never
found so large a number in any example of that caste.
The change from the nymph to the imago is accompanied
by the development of pigmented compound eyes, while the
wing-rudiments, from being vertical and closely appressed to
the sides of the body, become dorso-lateral at their origin,
nearly horizontal, and divaricate at the apex.1 These speci-
mens have at last reached the adult stage (imago), and (Pl. 16,
fig. 6) possess fully developed wings; at first white, they
gradually become black and capable of flight.
It will be seen from this account that examples with a
number of antennal joints varying from twelve to seventeen,
and therefore of very different lengths, can be transformed to
soldier larve, and consequently to soldiers; and that those in
which the number of joints varies from fourteen to nineteen
may become larvee of substitute royal forms. The latter larve
are not easily separable from the others; but in those with
from fifteen to seventeen antennal joints the pigmentation of
the compound eyes is evident. But it is absent or very scanty
in substitute larve with fourteen, and some with fifteen joints.
Ocular pigment is sometimes present in nymphs with seventeen,
and always in nymphs with eighteen or nineteen antennal joints
before their metamorphosis into substitute royal larve.
The larvee of perfect insects or young nymphs are customarily
selected for development into substitute forms (Pl. 16, fig. 14),
1 In these specimens the wings have really reached their full development,
but are enclosed in a chitinous sheath.
266 B. GRASSI AND A. SANDIAS.
and the wing-rudiments of the resulting royal examples are
therefore absent or very ill-developed. [When a nymph with
more evident wing-rudiments is selected for the throne, one or
more of these (usually that of the right fore-wing) is bitten off
CPLIG, tig. TS)
Ihave repeatedly stated that the royal forms may originate
in two ways; in one case they are derived from perfect black
examples, of which the wings are fully developed and become
detached along a special line of fissure, so as to leave a short
stump (the Schuppe or Squama of authors). These ex-
amples constitute the true black or normal kings and queens.
In the other case the royal forms are substitutional, and origi-
nate from examples which have suffered an arrested develop-
ment of the wings, and in which the compound eyes are
usually but not invariably pigmented! (Pl. 16, figs. 17, 25). It
is remarkable that the antenne are never found intact in any
king or queen, whether true or substitute, however young it
may be. And the majority of royal forms possess a different
number of joints, varying from thirteen to six or even four on
either side.
A most striking feature of the Calotermite colony is the
entire absence of workers, in which this species agrees fully
with the American form studied by Fritz Miller.
It may, therefore, be concluded that the kingdom of
Calotermes is composed of three castes: that of the soldiers ;
that of individuals which reproduce without becoming black
imagos ; and lastly, that of forms which lay eggs after the
acquisition of fully formed wings,—that is, after reaching the
stage of the perfect insect.
The phenomena related in this section may be recapitulated
in the following synthesis.
The normal development of Calotermes up to the
perfect stage may undergo deviation at different
1 Perfect insects which are still white may also become substitute forms.
They do not darken, and the wings are torn off, rarely along the special line,
but usually irregularly as in Termes lucifugus (q. v.). These observations
were made after the present work was completed.
CONSTITUTION AND DEVELOPMENT OF TERMITES. 267
ages. This deviation may lead to the formation of
substitute royal forms, or of soldiers, after passage
through the respective larval stages. The soldier
larva may originate by modification of examples with
from twelve to seventeen antennal joints; and the
larva of the royal substitute forms by modification
of examples with from fourteen to nineteen antennal
joints. Perfect insects which are still white may
also become substitute forms.
I can confirm the fact discovered by Lespés' in Termes,
aud by Fritz Miller?in Calotermes, that the caste of soldiers
is composed of examples of both sexes.
A minute description of all the forms in the different stages
of development would be of little interest, and I shall therefore
confine myself to a few points which have a special bearing on
the problems I have undertaken to solve.
Beginning with the sense-organs:
(1) All trace of eyes is wanting in newly hatched larvee.
The soldiers possess compound eyes, unpigmented and not
prominent. The time of their appearance has not been de-
termined; they are present in a rudimentary condition in
larve of 3—4 mm. in length, with twelve antennal joints—the
third glabrous, but they are not clearly distinguishable except
in sections. Ata later period they are evidently faceted, but
remain destitute of pigment. The compound eyes acquire
pigment, as previously mentioned, in nymphs about to become
imagos and in most examples selected for the dignity of sub-
stitute kings or queens. In the perfect insect the eyes become
more prominent and abundantly pigmented, and between them
is developed an ocellus devoid of any trace of pigment, which
I therefore regard as rudimentary.
(2) Sensory hairs (Tastborste), characterised by their
shortness and their connection with the nervous system, are
very abundant over the antenne and the whole of the mouth
1 «Ann. Sci. Nat.’ (4), v, p. 244.
? (Jen. Zeitschr.,’ vii, pp. 333—340. ]
268 B. GRASSI AND A. SANDIAS.
parts, which possess no other form of nerve-ending, such as
cones (Kegel) or papille (Zapfen). These hairs are espe-
cially numerous on the apex of both pairs of palpi and on the
antenne ; on the latter they are most abundant on the apical
half of the terminal joints. There appears to be no marked
difference between the sensory hairs of the mouth parts and
antenne, and experiment shows that the latter organs are
constructed so as to remain functional even when deprived of
a certain number of joints.
(3) The tibize exhibit the peculiar sense-organs discovered
by Fritz Miller,! evidently tympanal organs (PI. 19, fig. 10).
This is shown by the presence of the usual terminal rods, of
a characteristic tracheal branch, the lumen of which is not
accurately cylindrical, and which opens at either extremity
into the main tracheal trunk of the tibia, and lastly, of tym-
panic membranes.
Tactile hairs are present on various parts of the body, and
the so-called abdominal appendices (cerci) also appear to be
essentially tactile. These appendices are really identical with
the caudal cerci of Thysanura, reduced to a short basilar piece
and a long terminal joint, of which the apical extremity is
glabrous. ‘The remainder of their surface is covered with very
long, fine, and readily vibratile hairs, in addition to others,
such as are scattered over the body.
The description just given of them goes to show that they
correspond in Calotermes with those found in other insects.
All the sense-organs here described, with the aforesaid excep-
tion of the eyes, are fully developed at the time of hatching.
The visual structures are certainly more or less imperfect ex-
cept in the adults, and become functional concomitantly with
the wings. Their unimportance in other stages is evident from
the fact that pigment may be either present or absent in the
eyes of substitute royal forms, and that individuals without,
or with more or less imperfect eyes, apply themselves equally
well to the work of the colony.
[Calotermites move their antennz freely, and employ them
1 [‘ Jen. Zeitschr.,’ ix, p. 234, pl. xii, figs. 32, 34.]
CONSTITUTION AND DEVELOPMENT OF TERMITES. 269
just as a blind man does his stick, as Lespés observes. But
they are not accustomed to use them in mutual caresses, like
other social insects. If the antennz are cut off at the base
the insect becomes inert, stands in a fixed attitude, almost dis-
regards the difference between light and darkness, quivers (vide
infra) at rare intervals, and then for a shorter period and less
violently than usual. It is not successful in soliciting excre-
ment (vide infra), or does not directly attempt to do so. All
these inconveniences are only partially exhibited if the an-
tenne are detached more or less remotely from their base.]
Fritz Miiller’s observation! that the number of antennal
joints in Calotermes rugosus is increased by the successive
formation of new joints at the base of the third is well known,
and has led to the division of the antenna into two components
(base and flagellum), as in other Arthropoda.
I have attempted to determine the origin of the new joints
in Calotermes flavicollis, but have not been successful in
obtaining a clear idea of the process.
In deciding whether any given joint is the most recent, its
smallness, its freedom from hairs, and the indistinctness of
the line of demarcation between it and the parent joint must
be taken into account. But it will be readily understood that
all these criteria are apt to fail.
In the present case, the one which appears most practicable
(the presence or absence of hairs) may certainly lead us
astray. In fact, antenne of 13, 14, 15, 16, and 17 joints, with
the third and fourth joints glabrous, are to be found, as well
as completely pilose antenne of twelve to sixteen joints (fig.
22). The natural inference is that the third or fourth joint,
which was pilose in the latter examples, has become denuded
in the former (e.g. for an antenna of fourteen joints, the 3rd
and 4th pilose, to become fifteen-jointed (fig. 21) with the 3rd
and 4th hairless, one or other of the latter joints must neces-
sarily have lost its clothing) ; this may have taken place in
connection with ecdysis, as we shall see later.
1 [* Jen. Zeitschr.,’ ix, pp. 246, 247.]
270 B. GRASSI AND A. SANDIAS.
The broad fact remains that individuals exist with from
thirteen to seventeen antennal joints, with either the third or
fourth (fig. 20) or both joints glabrous, or without a single
glabrous joint.
How are the new joints formed? I regard it as certain
that the 13th arises by unequal division of the 3rd, at the
base of which it appears as a bud. The 15th and 17th appear
to arise from the 4th, and the remainder like the 13th (Pl. 16,
figs. 20—22). But closer study is necessary before arriving
at a conclusion.
It is of fundamental importance to notice that the new
joints do not take origin from a zone of embryonic (undiffer-
entiated) tissue, but from already differentiated structures
(hy podermis, nerves, &c.).
Calotermes moults periodically, and it is quite untrue
that the number of antennal joints increases coincidently with
the moult. It is more probable that the latter takes place
after the process of joint formation, and in this we may per-
haps find the explanation of the just related facts respecting
the loss of antennal hairs.
The number of ecdyses cannot be specified ; I have found
examples of all sizes in process of moulting, and can go no
further than to fix it at not less than five. The adults and
soldiers do not moult ; and the latter are derived from soldier
larve by an ecdysis, so that while the exuvie are larval, the
new instar is that of a fully-developed soldier.
As with other insects, the operation is accomplished by
means of an anterior medio-dorsal fissure, through which are
drawn out first the head and thorax, and lastly the abdomen
(Pl. 16, fig. 15). In rare cases an example has been seen to
be assisted in the operation by his comrades.
Two ecdyses merit special attention. The first is that
undergone by the nymph, furnished with wings apparently
small in size (really of full dimensions and elaborately folded
up under the old ensheathing cuticle), and with genital appen-
dices ; from this it emerges with wings of full amplitude, and
CONSTITUTION AND DEVELOPMENT OF TERMITES. 271
without appendices should it be a female. The other impor-
tant ecdysis is that in which the larva of a substitute royal
form loses its genital appendices if a female, and this leads
us to a detailed consideration of the latter organs.
The genital, often wrongly termed the anal appendices,! are
homologous with those of the ninth abdominal sternite of
Thysanura, and are attached in Termitide to what is appa-
rently the eighth, but is really the ninth sternite, the first
being fused with the metasternum. They possess hairs which
do not differ from the ordinary hairs scattered over the body.
They are present in all sexually immature examples, but
in the males only when mature. Further, the sternites differ
in the sexes ; in the female the true (not the apparent) seventh
sternite is strongly developed and semicircular, and the true
eighth and ninth are small and possess a median fissure. In
the male the true seventh is rather small, as are the true eighth
and ninth, in which the fissure is wanting. ‘There is no penis.
I have but little to say about the internal organs. The
tracheal system agrees with Fritz Miiller’s description,’ and I
have likewise observed the stigmata, tracheal trunks, anas-
tomoses, &c. Further, this species exhibits the blind tracheal
branches figured by Miiller, which I take to be suppressed
trunks or tracheal vesicles.
The alimentary canal presents the following features. In
newly born larve the teeth of the proventriculus are colour-
less and quite soft,—that is, covered as yet with a delicate
cuticle; and the Malpighian tubules are four in number.
Four others appear at the epoch when the antennez possess
eleven distinct joints, of which the third and fourth are
glabrous, and the former shows traces of division into two;
these new tubules require a certain time in which to attain
the size of the original four, and I have found them to be still
1 [The genital appendices must not be confused with the abdominal appen-
dices or cerci, referred to on p. 268.—W. F. N. B. |
2 [‘Jen. Zeitschr.,’ ix, pp. 257, 259, pl. xiii.]
Die B. GRASSI AND A. SANDIAS.
much the smaller when the number of antennal joints has
reached twelve, with the fifth joint already pilose, or with the
third and fourth alone bare. I may note here that the appear-
ance of intestinal Protozoa (vide infra) coincides -with this
last stadium, though Joenia is confined to the large-headed
larvee, whereas Monocercomonas is common both to these
and the small-headed forms.
The development of the four secondary Malpighian tubules
proceeds in such a way that each is placed midway between a
pair of the primary tubules; that is to say, the latter are
equidistant from each other, and the secondary tubules are
intercalated at equal distances between them, so as to produce a
series of alternate large (primary) and small (secondary)
tubules.
Their mode of development is shortly as follows: they
spring from the proctodeum at its junction with the mid-gut,
exactly at the same level of the primary series. I have been
unable to detect any special layer of embryonic tissue destined
to give origin to them, and they may therefore be regarded as
a direct derivation from the proctodeal epithelium.
When the antenne possess twelve joints, all pilose, the
difference between the earlier and later Malpighian tubules
has ceased to exist.
The salivary glands are highly developed in all castes and
at every stage of growth. There is a single pair, as well as a
large salivary reservoir, such as Miller has described (Pl. 19,
fig. 7). There is an unpaired external opening in connection
with the labrum.
[The supra-cesophageal ganglia are situated as in Thysanura,
with the olfactory lobes anterior, and the fungiform bodies
posterior as in Termes lucifugus (see figs. 27—383c, fung.).
The latter are relatively well developed when compared with
those of Embiide (fig. 34) or Thysanura; there are two
on either side, or four in all, not well separated from
each other. As in other insects, they are characterised by
1 [*Jen. Zeitschr.,’ ix, pp. 256, 257, pl. xii, fig. 42.]
CONSTITUTION AND DEVELOPMENT OF TERMITES. 273
the possession of small cells, of which the nucleus stains
more intensely than in the other nerve-cells.
Should a characteristic feature of these organs be sought,
by way of contrast with those of less intelligent insects, it
will be found in the abundance of these particular small
nervous cells. |
The visceral nervous system is well developed, and resem-
bles that of Blattidz (fronto-labral commissures ; frontal and
stomato-gastrie ganglia, &c.). The salivary glands are supplied
(only ?) by branches from the subcesophageal ganglion, coming
off from those to the labium or lower lip.
The ventral ganglionic chain possesses six abdominal ganglia,
and thus agrees with Lespés’ description; the sixth is very
large, and is in correspondence with the true seventh and
eighth abdominal segments.
At the level of the fourth abdominal ganglion—that is, in
the true fifth segment—there open numerous unicellular glands
of unknown function. The retro-cerebral gland, which exists
in Termes lucifugus,' is absent.
Certain features of the different castes and larval forms will
now be described.
[Newly born examples are semi-transparent and almost pure
white. If the mouth parts are detached the apex of the
mandibles and the inner lobe of the maxille can just be seen
with the microscope to be tinged with yellow, a feature which
cannot be made out by examining intact specimens even with
a good lens. After a few days these parts acquire a more or
less pronounced yellow colour, owing to the thickening of the
investing cuticle, and at this time a yellow line, caused by the
approximated tips of the mandibles and maxille, may be
distinguished at the front of the head even with the naked eye,
though better with a lens.
This yellow line appears early in forms which develop a
large head (soldier larve), and is delayed in those of which the
head remains small. At the time of its appearance the animal
1 [Vide Pl. 16, figs. 28—33, and description. ]
274 B. GRASSI AND A. SANDIAS.
is seen to adopt as food a material which is usually of the same
dirty yellowish colour as that of more developed examples, and
is apparent through the translucent abdomen, so that the
general white colour of the body is blotched with yellow. The
importance which attaches to this fact will be explained in the
subsequent chapter. |
The youngest soldier larve are therefore distinguishable by
the greater size of the head and the lesser constriction of its
anterior portion, as well as by the yellow line in this region;
moreover the thorax and the abdomen, which possesses
the aforesaid yellowish blotching, appear to be somewhat wider.
The body of small soldiers (5 mm. in length) does not become
fully yellow, with exception of the head, which is golden-
yellow, and the mandibles, which are brown. Their head is
subglobose, and at first sight these small soldiers greatly
resemble those of Termes lucifugus, but are distinguishable
by the indistinctly marked neck. The relative width of the
head and pronotum varies in soldiers of different sizes, as is
shown in the plate (Pl. 16, figs. 5, 16).
The body of medium-sized or large soldiers is golden-
yellow, much deeper anteriorly, while the mandibles are
coloured as in the small soldiers. The head is rectangular,
and longer than broad, subquadrate. At the time of differen-
tiation all the soldiers, large and small alike, are white, as I
have stated.
True kings and queens (6'5 to 7 mm. in length) of recent
development are black, except for the apical portions of the
legs and antennze, which are light yellow, and the anterior
three fourths of the pronotum, which are golden yellow.
With increase in bulk, at the first moment of the assumption
of the definitive habit, the white intersegmental linear spaces
become evident, and are most conspicuous when the royal
forms have reached the maximum dimensions of which they
are capable (Pl. 16, figs. 7,11). I may note here that a length
of 10 mm. for the king and of 14 mm. for the queen is the
greatest that I have observed.
These white lines, which are not as yet evident in royal
CONSTITUTION AND DEVELOPMENT OF TERMITES. 275
forms as much as two years old (figs. 8 and 9), correspond with
the interspaces between the first and second, and in succession to
between the sixth and seventh abdominal somites. The second
line is usually the widest, but is sometimes similar to the
others. They occur alike in both sexes, but as the king in-
creases in bulk less than the queen, they are naturally less
marked in the former than in the latter sex at the same age
and never ultimately reach the same degree of accentuation.
The substitute king and queen are distinguished not alone
by the almost constant presence of pigment in the compound
eyes, but also by their pale-yellow colour, which deepens with
age to a golden tint.
In these royal forms alike the abdomen becomes enlarged,
and I have found' examples of both sexes almost equalling
the largest true kings and queens in size (PI. 16, figs. 12, 13).
With the natural dilation of the abdomen, certain interseg-
mental spaces, namely, those which are indicated above for
the true king and queen, appear as whitish lines.
I conclude by mentioning that I have looked in vain for
any difference between the legs of the various forms. The
tarsi are normally four-jointed; and the apex of the tibia
possesses three spines, toothed on one side, and common to
all the legs (Pl. 16, fig. 24). A plantula is present in the
true imagos, but is wanting in all the other members of the
colony, including the substitute kings and queens.
These observations probably hold good for all species of the
genus Calotermes; I infer this from the many incomplete
notes which are scattered in the pages of various writers.
4. Relative Numbers of the Castes.
A nest contains a single king and queen, neither of which
can be normally wanting. They may be either true or sub-
stitute forms. At times one of the pair, either king or queen,
is a perfect and the other a substitute form; and in such
cases we have the strange phenomenon of courtship between
a black and a yellow example.
The existence of but a single royal couple is an assured
276 B. GRASSI AND A. SANDIAS.
fact, but it is not uncommon for two or more nests of Calo-
termes to co-exist without (at least for us) well-defined limits
in the same stem, which therefore appears to contain more
than a single royal pair. In connection with this condition are
certain facts which prove the existence of the terrible jealousy
which is so remarkably shown, e. g., between separate families
of the hive-bee. But for the sake of convenience I shall defer
them to a succeeding section on Habits.
When a colony is deprived either of king or queen, or of
both, it furnishes a certain number of substitute royal forms,
of which only one, if a single true form is missing, or two,
that is, a pair, if both are absent, is called to the throne.
In every nest, forms of different ages are always to be
found; and as a general rule small individuals are more, or
not less, numerous than large.
Soldiers are relatively scarce, not more than from two
to four being found in nests of from eight to fifteen inhabit-
ants. In large colonies they exist in the proportion of one to
‘very fifteen or twenty examples.
5. Seasonal Variations in the Colony.
It must be recollected, as a general rule, that the develop-
ment of Calotermes is arrested during the winter months,
that is, from the middle of November to the middle of April;
and this condition explains the absence of individuals in pro-
cess of moulting during that period.
In these months the nest contains eggs, sometimes a hun-
dred or more, which are always in the gastrulation stage.
Their development remains stationary during the greater part
of May, to reeommence towards the end of the month, so that
newly hatched young are to be found from about the 10th of
June till the end of July,a few eggs, from five to ten, hatching
every day. In July it is evident that the number of young
larvee increases, so does that of the eggs diminish until
their final disappearance. Towards the end of July it is
difficult to find a nest which still contains eggs, proving
that oviposition is suspended at this time. The same holds
CONSTITUTION AND DEVELOPMENT OF TERMITES, 277
good for the interval from November to the middle of May;
while the opposite condition exists in the second half of
-May, in June, September, and October. Consequently
oviposition remains suspended during winter and a large
part of the summer. At the end of October many eggs
are still to be found in the gastrula stage, and it follows
from the preceding statements that they remain thus during
the winter; but at the same month a certain number can be
found in an advanced stage of development. From this, and
from the fact that many new-born larve are to be found in
that month, I can confidently state that the eggs laid in
September (and perhaps in the early part of October) develop
immediately without need of hybernation.!
At all seasons, except from about April to the middle of June,
larvee, with the features of being newly hatched, are to be found
in the nests of Calotermes. The larve which are born in
the second half of October make no progress in development
until the following April.
Larve in stages of development which succeed those with
the characters of recent hatching are to be found at all times
of the year.
Nymphs are to be found in every nest during nearly all the
year, being absent at most only in August and September.
The perfect insects develop from July to October, a few
_stragglers appearing in the spring. Their swarming is quite
different from that observed in bees, and they leave the nest a
few days after they become black. [As development is not
simultaneous, they swarm in small groups of at most thirty
examples, and occasionally singly or in pairs. A colony may
therefore swarm, so to speak, ten or twelve times in all, from
July to October. ]
This fact explains the existence in many nests of a few
winged examples still at the end of the swarming season,
that is in October.
1 TI must admit that there is a certain gap in my observations in the
summer months, when I am unable to reside at Catania owing to the exces-
sive heat.
VOL, 39, PART 3.—NEW SER, T
278 B. GRASSI AND A. SANDIAS.
[As a rule a certain number of both sexes become mature at
the same time, and the males invariably take flight two or
three hours later than the females.| Further details on the
subject of swarming will be found in the later section on Habits.
Soldiers and soldier larve are to be found the whole year
round, but only small soldiers are present in nests under two
years old.
6. Duration of Development, of Life, &c.
These matters are very difficult to ascertain, but the follow-
ing facts are certain :
1. Owing to the interruption of development from November
to April, individuals in the same stadium may be of different
ages.
2. Eggs which pass the winter in the gastrula stage will
give rise in the following summer at most to soldiers with
15-jointed antenne, or to larve with similar antennze, and with
or without very short wing-rudiments. These larve become
perfect insects and swarm in a later summer.
These conclusions are the result of minute examinations of
numerous small or orphaned nests, &c., some of which I shall
proceed to record.
By searching in the situations which have been described
as the points of origin of new nests, from August to April,
some two to twenty perfect insects, with the wings reduced to
stumps, may easily be found; the majority are grouped in
pairs, male and female, each of which may be accompanied by
a few eggs. These pairs are recently formed, and, if originally
numerous, are subsequently reduced to one or two (see the
succeeding section on Habits).
A pair is established, say, in August, and remains to the
end of the autumn with only fifteen or twenty eggs; twelve
months later, at the end, that is, of the following autumn, it
will be surrounded with fifteen, twenty, or at most thirty
young of different ages, the most advanced being large soldiers
with at most fifteen antennal joints, or larve with a similar
number, all pilose, and with a very slight indication of wings,
CONSTITUTION AND DEVELOPMENT OF TERMITES, 279
I must regard certain nests which may be found at the end
of October, as having arisen from pairs which had only begun
to lay eggs in May of the year following the swarm-period.
In these the number of inmates is less than in the nests pre-
viously referred to, and no larva possesses more than fourteen
antennal joints, the third being glabrous, and the constric-
tions between the 3rd, 4th, and 5th respectively being ill-
marked. The wing-rudiments of these larve are not yet
discernible, and the soldiers of such nests do not possess more
than twelve antennal joints, and are small.
The hypothesis that such nests may have originated from
royal pairs disclosed only during the preceding summer,
instead of the previous year, and consequently only three or
four months old, is quite untenable ; as is indicated by the
fact that nests are never found to contain only newly hatched
or little-grown larve at the end of autumn, as must of neces-
sity be the case if young can be born in the year in which the
nest is founded. In short, the fact remains that at the end of
autumn certain nests are found to possess undeveloped eggs
only, whereas others already contain small soldiers, and larvee
with fourteen antennal joints: nests intermediate between
these two classes are entirely wanting, though they should be
present if the above hypothesis were correct, inasmuch as the
swarm-period lasts from the end of July till that of October.
In March, 1891, I observed an absence of nymphs in the
examination of several nests orphaned two years previously
(occasionally they were present in very small numbers).
These nests contained ova and forms in all other stages. It
therefore appears that destruction of the royal pair in spring
results in the suppression of swarming during July and
August of the second succeeding year, but in those months
only, because examples in which the wings are only just
indicated in March will have become imagines by September
and October.
Nests orphaned between February and June sometimes con-
tain no eggs, and usually no new-born larve in the following
winter.
280 B. GRASSI AND A, SANDIAS.
It is consequently evident that the development of substi-
tute royal forms proceeds very slowly.
The data furnished by orphaned and young nests, and by
many other observations which I have made, indicate that
perfect and fully winged examples are not obtained before
August from eggs laid in July of the preceding year. In
short, Calotermes passes part of two years in the larval
and nymph stages before taking flight. The soldier may com-
plete its development in the same year in which it is hatched.
It is difficult to say anything about the duration of life.
I must deny the existence of any particular season when the
soldiers die off, as Lespés has claimed for Termes. The
life of the king and queen may be estimated at four or five
years at least.
7. Situation of the Different Forms in the Nest.
I have already mentioned that Calotermes does not
possess a royal chamber. The king and queen, whether true
or substitute, usually remain in close proximity in the heart
of the nest, where the inmates are always most crowded.
They readily change their situation. The eggs are mostly
near them or a little way off, and are never heaped together.
Larve, nymphs, and winged forms, if present, are irregu-
larly commingled ; but larve newly hatched tend to cluster
together. The soldiers also are irregularly scattered, but a
few are often found in close attendance on the royal couple.
The soldiers are generally the first to make their appearance
when a nest is opened,
Substitute forms in process of development occur separately,
or in groups of two to four in different parts of the nest.
It thus follows that the component members of a Calotermite
colony have no special situations in the nest.
8. Certain Habits.
I propose to deal here only with those habits which could
not appropriately be dealt with in Section 3.
Calotermites work, feed, and rest indifferently by day or
night. When resting, they remain motionless without adopt-
CONSTITUTION AND DEVELOPMENT OF TERMITES. 281
ing any special attitude. They are fond of darkness, and
when kept in a tube occupy the portions furthest from the
light. They certainly work more actively in the dark, but are
capable of doing so even when exposed to daylight.
Oviposition continues both by day and night.
Swarming takes place in the morning, usually from 9 to
12; and it must therefore be recognised that Calotermes
no longer avoids the light when it has reached the perfect
stage.
Before swarming the winged insects collect habitually in a
spot which careful observation shows to be in the neighbour-
hood of an exit-hole, putting the nest into communication with
the outer air.
Swarming takes place through this hole. The insects issue
by ones or twos, so that the twenty or thirty examples ready
to take flight emerge in perhaps a quarter of an hour.
Once outside the nest, they run upwards for a few metres if
the locality admits of it, and then only do they take wing. In
a room they evidently fly towards the light, and if a wind is
blowing they follow its direction. Some soon become tired and
settle on the trunks or branches of neighbouring trees; the
majority become lost to sight, but many certainly end by
alighting on trees. It is here that they group themselves into
pairs, the males and females of which must frequently be
derived from separate nests, for, as I have mentioned, the sexes
swarm separately ; this acts as a safeguard by which Calo-
termes habitually avoids in-breeding.
The winged forms have not been observed to pursue cach
other in the curious way which will be spoken of under
Termes.
I may describe more precisely the manner in which the
males and females come together when settled on a tree. The
winged forms habitually search for a decayed spot, and when
found they dig it out, after losing their wings, in order to bury
themselves ; it is in this act of excavation that the meeting and
subsequent pairing take place.
The wings may be shed, merely by striking against an
282 B. GRASSI AND A. SANDIAS.
obstacle, or by becoming damp and adhering to some spot,
while the insect continues to move about.
But if not favoured by chance the Calotermite rids itself
of its wings, as the following observation shows. Four perfect
insects, which had recently left the nest, were captured by
hand after flying about a room for some time, and were put
under a piece of rotten wood. They had hardly settled down
before they began to strip off their wings by resting their tips
against some projecting corner of the wood and then moving
backwards a little, so that the wings buckled towards the base,
broke, and dropped off. When rid of them they began to
gnaw the wood, at first along, and then across the grain; each
worked by himself and at some distance from his fellows.
Subsequently several chance encounters took place between
them; they threatened to bite each other, and then ran off in
different directions. They were of the same sex. If they had
been of different sexes they would certainly sooner or later
have copulated.
In the colony of Calotermes all members work for the
common welfare. The soldiers serve for defence, but as a rule
only when some important enemy has to be combated; at
other times nymphs and possibly the older larve assume the
task.
Cremastogaster scutellaris, Ol., which is abundant
here, is one of the most formidable enemies of Calotermes,
near which it makes its own nests. This ant enters the termi-
tarium to massacre, whereas its own nest is never invaded by the
Calotermite soldiers. If some examples of Cremastogaster
are put into a tube containing a Calotermite nest, the follow-
ing phenomena can readily be followed. The soldiers place
themselves with gaping mandibles, waiting for any enemy that
may come within reach. They then snap their jaws rapidly,
shearing off antenne and legs, tearing the abdomen, or even
cutting the ants in two at the level of the abdominal petiole.
The soldier’s mandibles are seen to act like extremely sharp
shears.
CONSTITUTION AND DEVELOPMENT OF TERMITES. 285
The ants themselves attack the Calotermites indifferently,
but habitually avoid the heads of the soldiers, only daring
occasionally to attempt to lop off their mandibles. As a rule
they attack the soldiers from behind by biting the abdomen,
and to protect it the soldiers creep under pieces of wood so as
to leave the head alone free.
If a few Cremastogaster and a larger number of Calo-
termites are put into a tube, peace is usually concluded after
about an hour’s conflict, with a certain number of dead and
wounded on both sides. The ants take up a position in one
part of the nest and the Calotermites in another.
[Besides the ants, the soldiers of Termes are terrible
enemies, but being small they are easily cut in two through
the thorax by the Calotermite soldiers. The workers of
Termes are much less dangerous enemies than the soldiers of
that species. One, put into a tube-nest of Calotermes with
soldiers, was at once placed hors de combat by a nymph,
which cut off part of the buccal apparatus. Then sundry large
larvee and other nymphs hurried up, bit off its legs, and tore
its abdomen until the viscera protruded. Thesoldiers took no
part except towards the end of the struggle, when one gave it
a bite. Similar observations have been made several times,
and show, as we said before, that the soldiers purposely reserve
themselves for more important foes.
Another Termes worker was put into a Calotermes
nest containing no soldiers. The inmates took flight, pro-
bably terrified by the knowledge that they were unprotected
by soldiers; and the Termes succeeded in throwing the
nest into confusion, until after some lapse of time a nymph
plucked up courage to bite its abdomen, and thus killed it.
A substitute queen of Termes lucifugus was introduced
into another tube containing a Calotermite nest. A soldier
promptly despatched her by decapitation, and then only did
the nymphs intervene to tear the body. During the rest of
the day the soldier never moved from the spot where he had
killed the queen. |
Besides protecting the nest the soldiers fulfil other duties,
284 B. GRASSI AND A. SANDIAS.
such as that of carrying the young and eggs on their
mandibles. These organs are useless for gnawing wood, and
their possessors therefore remain idle for hours together,
while the rest of the colony is in full activity. Except for
these tasks of defence from great dangers and wood-gnawing
and consequently excavation of galleries, it is, so to speak, an
absolute rule that all labour necessary for the community can
be undertaken by any of its members.
Newly hatched larve can be seen carrying a fragment of
wood heavier than themselves. All forms except the soldiers
excavate galleries, provided that their mandibles are sufficiently
strong, which naturally is not the case just after birth or after
a moult. Both king and queen, whether true or substitute,
gnaw up wood and transport excreta, eggs, or wood-meal.
Oviposition appears to be a very laborious process; in one
case the egg was not extruded until an hour, and the succeeding
egg until half an hour after appearance at the vulva. Once I
saw a soldier assist the queen by raising and gently stroking
her abdomen, but as a rule she lays without assistance.
Hatching is effected without need of any assistance from
inmates of the colony. The chorion is tolerably thick, and
the eggs can be kept in a watch-glass without drying up, and the
process of hatching observed. Moulting also is accomplished
as arule without assistance.
If the eggs and young are exposed by opening a nest, it is
striking to see how the other inhabitants disperse without
paying them the least attention, in contrast to the behaviour
of ants; and yet one must recognise that the colony is as
deeply interested in their welfare as ants are in that of their
own offspring, and that it scatters simply because it is panic-
stricken. ‘This may be proved by shaking a small nest made
of loose pieces of wood in a glass tube or jar. For a moment
all the inmates are thrown into disorder by terror, but they
quickly recover themselves, become persuaded that it was
merely an earthquake, so to speak, and devote themselves to
the restoration of order by carrying the eggs back to their
place at the bottom of the jar, and removing the young on the
CONSTITUTION AND DEVELOPMENT OF TERMITES. 285
mandibles of the soldiers, or in the mouth-parts of their other
fellow-inmates.
It must be recognised that Calotermites are perfectly well
aware of the presence or absence of the royal pair, and they
start about providing fresh ones directly they are orphaned.
The following observation is important in this connection.
A nest was divided, say, into three smaller nests, in one of
which the royal pair was retained; they were put into separate
tubes, kept together and uncorked in the same waistcoat
pocket. After three or four days larve of substitute royal
forms were found in them, except in the one containing the
royal pair. The colony in the tree from which the original
nest was taken had not turned their attention to raising sub-
stitute forms, though they were 30 or 40 centimetres distant
from any king or queen; whereas those in the tubes, though in
close proximity to a royal pair, had at once begun to provide
substitutes. It follows that the existence of the royal couple
is certainly not perceived by means of any odour which they
emit; and this affords us another marked contrast with what
is observed in the case of bees.
The copulation of Termitidze has been extensively dis-
cussed ; some writers believe it to be accomplished in the
open air, but the more general opinion is that the presence of
the king in the nest shows that it takes place there, and is
repeated from time to time. I regarded the latter supposition
as possibly correct directly I found that the king is invariably
present in the nest, and that his bulk increases, though to a
less extent, concomitantly with that of the queen.
At last, on April 17th, 1891, about 11 a.m., I detected the
king and queen in coitu in a glass jar containing a small
Calotermes nest. The pair appeared to be about three years
old, judging by their size. They stood end to end inastraight
line with the tips of their abdomens applied to each other ;
their attitude was normal, with the dorsum uppermost, except
for a slight upward flexure of the apex of the abdomen. After
some half-minute they separated, perhaps owing to the ex-
286 B. GRASSI AND A. SANDIAS.
posure to lght. The contiguous parts of their abdomen
exhibited a white substance which, when the king and queen
became detached from one another, remained adhering under
the hinder extremity of the queen.
It is therefore certain that connection takes place in the
nest and is repeated at intervals; and all my observations
satisfy me absolutely that it cannot be accomplished in the
open air, and is practicable only after loss of the wings.
[I must regard certain relations which were observed to
subsist between two substitute forms still far from maturity,
because they were only about a fortnight old, as amatory in
nature. One stood still while the other gradually approached,
and when sufficiently near brought its antenne into contact
with those of its fellow; it then quickly retired for some dis-
tance, and returned later to repeat the pastime. This took
place at least four times. On the fifth occasion the one which
had been standing still, made movements as if to detain its
companion ; they then remained together, and very rapidly
stroked various parts of each other’s body, especially the apex
of the abdomen. Their position during the reciprocal palpation
of this part was almost that assumed during the act of coitus. ]
Several writers have mentioned the convulsive movements
characteristic of Termitide. These movements or quiverings
are easily observed in Calotermes, and may be repeated
periodically at very short intervals, almost at the frequency
of the pulse-rate.
In the act of quivering the tarsi are held motionless, while
the body is shaken forwards and backwards; there may bea
simultaneous slight lateral or vertical oscillation. Sometimes
an example may stop whilst running, in order to quiver one or
more times.
Occasionally these convulsive movements are repeated a few
times only, and then stop altogether; but at other times they
recur after a few seconds or at most a few minutes’ rest, and
may thus be continued sometimes for hours with many similar
intervals of rest.
CONSTITUTION AND DEVELOPMENT OF TERMITES. 287
In the intervals between successive convulsions the insect
remains still or progresses for a short distance only. These
movements are executed by all members of the colony ex-
cept those newly hatched.
I have satisfied myself by careful observation of the phe-
nomena exhibited in tube-nests that these convulsions serve
as a cry to summon help or give alarm, or as a lament; in
short, as a mode of intercommunication.
When the insects are suddenly annoyed or disturbed by
any cause, such as a rough shake of the tube, its change from
a vertical to a horizontal position, sudden illumination, or the
prolonged effect of too bright a light, all the members of the
colony begin to quiver, except those that are running briskly
about in search of a better situation. Moribund examples
sometimes perform these movements at intervals of a few
minutes, in some cases for a couple of hours.
[If a few Termes lucifugus are put into a tube contain-
ing the usual little nest of Calotermites, some of the latter,
evidently excited by the very rapid movements of the intruders,
run off rapidly in the opposite direction, and stop at intervals
to quiver with much more energy than usual. This phenome-
non is exhibited alike by larve, nymphs, and soldiers, but the
latter quickly turn back to face the supposed enemy. It seems
evident in such a case that those which are first aware of the
presence of the Termes quiver violently in order to alarm
the population. |
[Sometimes the convulsions of an insect in the neighbourhood
of the cork are quickly followed by the exit therefrom of
others which have been burrowing inside it. Such cases
appear to prove the utility of the action as a mode of
signalling. |
Members of the same nest clearly recognise each other.
[As a proof, a few examples are removed from the nest and
returned after five or six hours. The population is not dis-
turbed or alarmed, and does not scurry about at their re-entry.
It is a possible objection that these specimens have immediately
recognised the nest, and therefore create no disorder on their
288 B. GRASSI AND A. SANDIAS.
return; to meet it a new nest was provided, from which
certain individuals were excluded. They were introduced a
few hours after the fresh nest was put in order and quiet, but
they caused no disturbance although it was unfamiliar to
them. As a control experiment, a few strange Calotermites
were put into the same nest; the inhabitants took fright at
once, and scattered in different directions. But after a little
time all became quiet, and no fighting was actually observed.
We may therefore conclude that examples taken from different
nests readily fraternise ; and this applies to the soldiers as
well as to the larve and nymphs. ]
[I should mention with respect to the soldiers that if too
large a number is added to a nest, these supernumeraries, as
they may be called, are found to be killed or eaten one by one
during the next few days.]
[The following observation was made on one occasion. Half
a dozen strangers, including a substitute royal example, were
added to one of the usual nests in a tube. The royal specimen
showed signs of hesitation and remained in the same spot,
merely turning round and round, and straightening its legs as
much as possible, as though to raise itself above the level of
its neighbours. When an inhabitant of the nest approached
and touched it, it drew back at once; the other did the same,
so that they separated as rapidly as if they were stung. Some
hours later the new-comer became quiet, and several inmates
of the nest approached it to caress the antenne, &c. I must
mention that the nest was orphaned. ]
[We have already stated that Calotermites furnish a certain
number of substitute forms when deprived of the king and
queen. Why then is only a single royal pair to be found ?
One day one of three substitute forms, which inhabited a
small nest, was seen rapidly to pursue another with gaping
jaws, with the evident intention of killing it, in which it was
unsuccessful. Next day the nest contained only two such
forms ; very probably the one which was pursued the previous
day had been killed.]
[Two true royal forms, a king and a queen, were introduced
CONSTITUTION AND DEVELOPMENT OF TERMITES. 289
into one of the usual nests, constructed the day before and
containing only nymphs. One of the pair began at first to
advance hesitatingly, while the other stood perfectly still ; but
shortly one after the other gradually moved forwards.
The nymphs then began to scatter in different directions in
the endeavour to keep as far away as possible from the royal
pair; these gradually retreated to the bottom of the tube,
where they were left alone. But now and again a nymph
approached with open jaws, and savagely bit one of the pair
in the head or thorax, getting bitten with equal fury in
return, and consequently retreating.
Next day the royal couple was still left unattended, but was
no longer actually disturbed. Then a second similar pair was
introduced ; at once the nymphs all became greatly excited,
attacked them and promptly reduced them both to helpless-
ness by biting off their legs. The second pair was alive two
days later, and the queen of the first pair was seen to attack
the second king with open jaws; he retreated by dragging
himself along the tube, being unable to run through the loss
of all his legs. A day later he was dead, and the first queen
was observed nibbling the stumps of his legs ; on the following
day the queen of the second pair was also found dead.
Several repetitions of this experiment were made, and always
gave similar results. Whenever two or three supernumerary
royal pairs are put into a nest, a single pair is all that can be
found at the end of a few days.]
I infer from all these facts that Calotermites exhibit those
phenomena of jealousy and hostility which are so well known
in bees ; however, their manifestation is less rapid.
The observations just recorded, and many other facts
omitted for the sake of brevity, lead us to the fundamental
conclusion that the colony of Calotermes tolerates
neither supernumerary royal examples nor super-
numerary soldiers. Both one and the other are
slaughtered. On the one hand, then, the colony can
provide itself with royal forms or soldiers when they
are required (vide infra); on the other hand, it rids
290 B. GRASSI AND A. SANDIAS.
itself of them when they are over-abundant. These
facts imply the possession by Termites of a faculty
which may be termed a sense of proportion or nume-
rical sense.
By way of conclusion, I must add that though I have had
Calotermites under observation for several years, I may still
have failed to detect a great part of their marvellous instincts.
This is owing to two circumstances: 1, they are often slug-
gish ; 2, when a nest is opened the population is thrown into
such a state of astonishment that it usually does nothing but
run away and give signals of alarm. And almost all the
observations here recorded are due to the method of employ-
ing tube-nests.
Tue CoLony oF TERMES LUCIFUGUS.
1. Situation and Nest.
Several writers, particularly Lespés, have published observa-
tions on this subject ; but I shall summarise merely my own
investigations.
Here in Sicily Termes lucifugus usually inhabits plants,
rarely furniture or the wooden beams of buildings. It is most
common in plants of which the stem or tap-root measures at
least three quarters of a centimetre in diameter; but when it
has once entered a stem or large branch it will pass on into
the very smallest twigs and roots. It mines irregular galleries,
and often avails itself of old beetle-burrows (Bostrychus,
&c.); Calotermes does the same thing. Like that species,
Termes leaves the outer layers untouched, so that a trunk
may be completely mined out and yet appear sound, while the
hand can easily be thrust into it by breaking through the thin
intact superficial layer.
Owing to the extreme tenuity of this layer in the smallest
roots invaded, the galleries may appear at first sight to be
tunnelled directly in the ground ; but I have never been able
definitely to establish the existence of such a mode of con-
struction.
CONSTITUTION AND DEVELOPMENT OF TERMITES. 291
I have never found Termes in the orange, lemon, or vine;
but, unlike Calotermes, it is very common in cactus. It
continues to live and flourish in perfectly dead and dry wood,
even when employed in the construction of roofs, doors,
furniture, casks, &c.
It is characteristic that whereas Calotermes confines itself
to the original host plant, Termes successively attacks fresh
plants or timber, and may thus pass from the furniture of a
house to a tree, or the reverse, or from one piece of furniture
to another. The following facts are related to this peculiarity :
A few workers of Termes, which are probably explorers,
are sometimes found to make a spontaneous appearance in the
open air and in broad daylight.
On other occasions Termites travel by availing themselves
of natural cracks, e. g. in lava, or by hollowing out small dry
rootlets underground, or in reeds by means of the tubular
lumen. Frequently they have recourse to galleries, which are
quite distinct from the burrows made in wood, because they
are fabricated or built up, so to say, by the insects themselves.
They are therefore not merely miners, but also builders.
These galleries are of two kinds, tubular or D-shaped in section.
Galleries are usually constructed in tubular form in the absence
of any suitable base on which to build them; if such be present
the gutter-shaped gallery (semicircular in section) is resorted
to, but even in this case that part of the base which is enclosed
between the walls of the tunnel is incompletely cemented over.
Galleries of the latter kind are made by preference in the angle
of junction of two walls (Pl. 17, fig. 40).
Gutter-shaped galleries may reach the considerable length
of eight or more metres. In the choir of the principal church
of Pedara I have seen such a tunnel leading from the stalls to
a crack in the wall adjoining the wooden ceiling, in which the
crack disappeared. The insects travelled between the choir
and the ceiling by means of the tunnel and then of the crack,
which they made use of apparently without modification. In
such cases it is sometimes difficult to distinguish the track of
the insects right up to the point where the gallery leaves off.
292 B. GRASSI AND A. SANDIAS,
Similar galleries also occur in the spacious ceiling of the
Benedictine church at Catania.
In the building of the Botanical Garden at Catania the
Termites have invaded the benches of the school, the book-
cases and window-frames, &c. Here I found two gutter-shaped
tunnels over 40 cm. in length on the wall of a room; they
began from the timber invaded by the insects, and ran along
the surface of the wall, to finish at a point where no fissure
was discoverable.
The direction of the gutter-shaped galleries may vary as
required ; they are usually vertical or oblique, less often hori-
zontal, and they may branch in various ways. The tubular
galleries are mostly short, rarely exceeding 5 cm., and are
narrower than an ordinary pencil. They serve to connect two
gutter-shaped galleries or two portions of a nest; and a
gutter-shaped gallery may become tubular for part of its
course.
On one occasion I found a much flattened tubular gallery
about 15 mm. in width, 4 mm. in depth, and 5 cm. in length,
and somewhat irregular. This fragile structure was suspended
from the ceiling, and contained a certain number of Termites;
there were some apertures at its free extremity (Pl. 17, fig. 38).
At other times I have found much shorter flattened galleries
hanging from the ceiling (Pl. 17, fig. 39; Pl. 18, fig. 15).
The purpose of these structures escapes me, but, recollecting
the excrescences built on plants by certain tropical Termites,
I suspect them to be a rude attempt at a nest. A similar
explanation may perhaps be advanced for the galleries found
at the Botanical Garden of Catania, which terminated in a
free extremity.
Both forms of gallery usually have a diameter from 2 to 6 mm.
Their lumen varies at different points, and is generally large
enough to allow several insects to pass at once; the internal
surface is tolerably smooth, while the outer is irregular and
rugged. Their colour is variable, but is usually of a chalky-
grey tint. They are composed of fecal and disgorged matter,
and of triturated wood. When connected with a plaster cor-
CONSTITUTION AND DEVELOPMENT OF TERMITES. 293
nice they contain distinct scattered white specks, which are
fragments of plaster.
These Termite galleries are invariably very light, porous,
and friable. When the Termites meet with large empty spaces
while in process of enlarging their nest, they may fill them up
by building a complicated labyrinth, as many ants are known
to do (Pl. 18, figs. 14, 16); and they readily adapt such pro-
jecting pieces of wood as the space may contain to what they
are building by covering them over with the materials of
construction, or cementing them suitably together.
Excellent specimens of both classes of gallery can be obtained
by putting Termites into a glass jar half full of broken-up
cactus-phylloclades, and closed with a cork or merely witha
sheet of paper (Pl. 18, fig. 17).
If many such nests are formed, one or more can generally be
kept alive for six, eight, or more months. I have published
elsewhere an account of one of these nests, which I repeat
textually.
“For eight months I have kept a colony of Termes luci-
fugus without king or queen in a jar half full of crushed-up
phylloclades of Opuntia, and closed with a sheet of paper tied
over the mouth instead of with a bung. ‘The jar holds three
litres, and its mouth is wide and polished.
“ At the beginning of April the Termites were seen to have
settled in the bottom layer only of the rubbish, and the
remainder, some 7 cm. in thickness, was quite uninhabited.
It was not till the 20th of May that a few specimens appeared
init. Some days later a semicircular gutter-shaped tunnel was
found adhering to the walls of the empty part of the jar (the nest
material occupying barely a half). This tunnel put the rub-
bish into communication with the paper cover, which presented
a small aperture large enough to admit the body of a Termite
at the extremity of the tunnel.
“ All kinds of forms in the colony (larve, nymphs, soldiers,
workers, and winged adults) went backwards and forwards by
this gallery, which in its greater part would only allow room
for a single individual at a time, but which widened here and
VOL. 89, PART 3.—NEW SER. U
294, B. GRASSI AND A. SANDIAS.
there so that two could pass simultaneously. This primary
tunnel, as it may be called, was the only one constructed during
five days, but it was made to bifurcate by the addition of a
lateral branch. This came off at an acute angle from near the
middle of the primary gallery, and ran upwards to terminate
on the margin of the vessel, where the polished surface began.
During the next few days the uninhabited part of the rubbish
was tunnelled with numerous burrows, opening on its upper
surface by several small holes, some of which were continued
upwards by tubular chimney-like galleries of different heights
(the largest measured 5 cm.), vertical or oblique in direction,
and varying in width, usually just capable of being traversed
by a single Termite. The free ends of these chimneys were
sometimes closed, sometimes open, and in that case an inmate
of the colony would not infrequently peep out as from a watch-
tower (Pl. 18, fig. 13).
“ Swarming took place on June Ist. On the following days
the Termites did very little work, to all appearance. They
lengthened a chimney and built another gallery like the first,
but not connected with it and not reaching the lip of the
vessel ; and lastly, they made an incomplete extension of the
primary tunnel by carrying it along the junction of the lip and
the paper. The latter exhibited two fresh holes at this point.
“One day I destroyed the newly formed section of the
primary gallery by taking off the paper, which I purposely
put back so that it did not accurately cover the margin of the
vessel, but left a gap; next day the primary gallery was con-
tinued horizontally outwards for about 1 cm. from the margin
of the vessel ; this new portion did not run on the surface of
the glass, and, like the chimneys, was tubular instead of being
semicircular (Pl. 18, fig. 17). A day later it was dismantled
and destroyed by the Termites; this did not surprise me,
as I had already observed them make and unmake portions
of galleries.
**On June 20th the tops of the chimneys and openings were
closed up, the galleries were unoccupied, and the whole of the
colony had once more retired to the bottom of the vessel.”
CONSTITUTION AND DEVELOPMENT OF TERMITES. 295
This colony was unfortunately killed at the end of June by
over-dryness.
The structures built before swarming were certainly made
chiefly for the purpose of facilitating that procedure. But
similar erections, with the exception of the chimneys, can be
obtained at a time remote from the swarm-period, or after it
has elapsed, as in the case just described, where the Termites
evidently found the vessel unsuitable, and attempted to
abandon it, but were unsuccessful, and therefore all died.
I state that they endeavoured to quit the jar because on
other occasions I have observed a general migration from a
vessel which had been inhabited some time, the contents of
which were found on examination to be unsuited to their
welfare (too damp, mouldy, over-dry, &c.).
Termes does not line the galleries hollowed out of wood
with excrement. The wood-meal produced by burrowing
varies in colour with the material from which it is derived.
Burrows are very easily made in certain cactus-phylloclades
which keep their white colour after death, but become very soft ;
the inner surface of these burrows and the dust removed are
both white.
Termites customarily select the softer parts of the wood, but
when these are all destroyed they attack the hard parts, and
thus form spacious chambers, openings, &c.
Temperature and humidity are as important for Termes as
for Calotermes; and as far as is known at present the
geographical distribution of these forms coincides.' But it
must be remembered that.the former can flourish at a some-
what lower mean temperature than the latter, so at least I
imagine, for the following reasons :
Oviposition commences as early as the beginning of May.
In October and November forms in intermediate stages between
the nymph and the larva with the earliest indication of wings
1 (Termes, however, extends farther north, occurring in France at
Toulouse, Bordeaux, Rochefort, and La Rochelle, and in Italy in Tuscany
and Venetia, the present writer having lately found a winged example at
Venice.—W. F. H. B.]
296 B. GRASSI AND A. SANDIAS.
are very scarce; they begin to appear in December, and
become abundant in January and February. The nymphs
accomplish the imaginal transformation by April or May.
But the fact that examples with the characteristics of those
recently born are to be met with in spring, though the nest
may have contained no eggs since the month of September,
shows that the development of the earliest stages is arrested
during the winter, as in Calotermes. Termes, it must be
noted, requires a less degree of moisture than Calotermes,
and can therefore live with comfort in dry and seasoned timber,
and in desiccated portions of trees, &c.
During the warmer months they bury themselves deeper and
deeper in dead roots, so that their nests appear to be depleted,
and it becomes difficult to procure complementary royal forms,
eggs, new-born larve, &c., without digging to a great depth.
Termes and Calotermes often share the same tree, but
the former habitually confines itself to the dead and drier, the
latter to the moister parts. But it will readily be understood
that there is no sharp demarcation between these two regions,
and therefore none between the two colonies.
2. Number of Individuals in the Colony.
It is practically impossible to make any accurate estimate
of the limits of a nest of Termes, as will be seen later. How-
ever, a single tree, which certainly does not harbour more
than one nest, will contain at times as much as a litre of Ter-
mites. Asa rule, the offspring of one nest extends to several
trees, and the population of a single colony may therefore be
reckoned as upwards of two litres—that is to say, very many
thousand individuals.
3. The Different Castes (Plate 17).
The society of Termes lucifugus differs widely from that
of Calotermes, or of such other Termitidze as have hitherto
been adequately studied.
Its characteristic feature is the invariable absence of a true
CONSTITUTION AND DEVELOPMENT OF TERMITES. 297
royal pair,—that is, a pair derived from winged imagos, which
have lost their wings (except the stumps). This statement of
mine will appear bold, but I have examined thousands of nests
during a period of about seven years, and am in a position to
make it without the least fear of contradiction. Small colonies,
founded by a true royal pair, are to be obtained only by arti-
ficially enclosing winged Termes in glass jars partly filled
with wood. Nothing of the kind is ever found in nature. On
one single occasion I lighted on a true royal pair, though
without eggs, in January,—that is, about six months after
swarming.
For the present we may leave out of sight these artificial
nests in glass jars to consider those found under natural
conditions.
The principal differences between the Termes colony and
that of Calotermes are as follows:
1. Termes possesses the caste of workers, which is wanting
in Calotermes.
2. On the other hand, Termes has no true royal pair, but
its place is supplied by a large number of sexually mature
individuals, which I term complementary royal forms
(Pl. 17, figs. 16, 17, 21). These complementary forms have
the characters of larve just about to become nymphs—that
is, with the wing-rudiments relatively shorter than in the
nymph. Their length differs a little, however, in different
examples, and occasionally they are entirely wanting.
3. Orphaned nests—that is, nests from which the comple-
mentary forms have been abstracted—contain numerous substi-
tute royal forms. These may resemble the complementary
forms, but their wings are frequently entirely wanting (Pl. 17,
fig. 15), or else developed as in the nymph (id., fig. 23).
Sometimes they have the characters of an imago which has
become brown in a few places only, and has the wings mu-
tilated (id., fig. 24).
The ordinary nest of Termes may evidently be compared
with the orphaned nest of Calotermes, with this difference,
that the former is much richer in royal examples, which
298 B. GRASSI AND A. SANDIAS.
usually possess some trace of the wings, whereas these are
entirely absent in most of the substitute forms of Calo-
termes.
The orphaned nest of Termes has a still closer resemblance
to that of Calotermes, for the royal forms, as I have said,
are frequently destitute of wing-rudiments.
In short, in the nest of Termes, as in the orphaned nest of
Calotermes, individuals of which the wings have never been
fully developed are invariably raised to the throne.
There is a further important distinction: in the Calo-
termes nest the king is always to be found beside the queen,
whereas in numbers of Termes nests examined I have only
twice found a single king associated with troops of queens.
These kings were observed in the hot season,—that is, at the
time when substitute queens are most difficult to find, because
then the insects habitually bury themselves deep in the ground.
I cannot doubt that I might have found many others if I had
been able to continue my investigations in August and
September.
The complementary or substitute king is certainly present
in the nest about the time when the queens of either kind
reach maturity, and he disappears after pairing. The colony
must therefore rear fresh kings every year, which become
mature in August and September, fertilise the queens, and die.
By way of confirming this inference, I may say, in addition to
the facts just related, that recently orphaned nests contain as
many examples in process of becoming substitute kings as
those about to become substitute queens. Complementary
kings in process of development can be found from the middle
of March onwards, but are always very rare in non-orphaned
nests, while developing complementary queens are entirely
wanting.
Finally, I may add that all possibility of parthenogenesis is
excluded, as will be seen farther on, by the constant presence
of abundant spermatozoa in the spermathece of the substitute
queens.
CONSTITUTION AND DEVELOPMENT OF TERMITES. 299
A nest of Termes contains !—
I. Very young larve, the head of which is alike in
those of equal length (PI. 17, fig. 1).
They include forms from the smallest (scarcely 1 mm. long,
with antenne of eleven joints, the third bare, the rest pilose)
(Pl. 17, fig. 1) to those a little over 2 mm. in length (with
antenne of twelve joints, the third bare, the rest pilose, some-
times with the fourth bare in one antenna only). Larve
between these two groups are intermediate in length, and have
either eleven pilose antennal joints or twelve, the third and
fourth bare and the rest pilose. There are four Malpighian
tubules in the smaller larve (Pl. 18, fig. 11), eight (four large
and four small) in the larger (id., fig. 12). I may add at once
that the smallest larve possess no parasitic Protozoa.
II. Examples 2°25 to 3°75 mm. long, with twelve
entirely hairy,or thirteen antennal joints; some with
large (Pl. 17, fig. 2), others with small heads (id., fig. 3).
Those with large heads may be regarded as young
workers, and may become either adult workers or
soldiers.
The smallest members of this group still possess four small
[secondary] Malpighian tubules, and may be free from Pro-
tozoa (always from Trichonympha).
III. Examples with fourteen antennal joints, 3°75
to 4 mm. in length, with or without very short
wing-rudiments.
They fall into the following category :
A. Forms without a trace of wings, with the head
relatively large, the abdomen stout and rather short, and
the colour of the body less conspicuously white (Pl. 17, figs.
5, 78). These are more or less immature workers, capable of
becoming adults, or of transforming into soldiers. They are
derived from either the large- or small-headed forms of the
preceding stage (11).
1 As before, I must assume a knowledge in this section of certain experi-
ments which will be described later.
300 B. GRASSI AND A. SANDITAS.
B. Soldiers with all antennal joints pilose. De-
rived from large-headed forms of the preceding stage (11).
c. Examples without trace of wings and with a
relatively small head (Pl. 17, fig. 4). Derived from small-
headed forms of the preceding stage (11).
p. Examples with very short wing-buds (Pl. 17,
fig. 6). Also derived from small-headed forms of the preceding
stage (11).
IV. Individuals 4 to 6 mm. or more in length, with
fifteen or sixteen antennal joints. Some possess very
short or partly developed wing-buds, and belong to
groups c and p of Stage 11 (Pl. 17, fig. 9). Others have no
trace of wings, and these may be of three kinds: a”, more
or less youthful workers; 3B’, soldiers; c*, larve of
royal forms, complemental or substitutional, with-
out sign of wings,and with the head small. a? are
derived from forms a or c of Stage 111; B* from forms 4,
possibly also from forms B, c, and p of Stage 111; c?, lastly,
from form c of that stage.
V. Individuals with seventeen or eighteen anten-
nal joints, incapable of flight, and infertile, or at
least far from maturity.
These may be of five kinds: a°, soldiers (with not more than
seventeen antennal joints, all pilose) (Pl. 17, fig. 14); B%,
adult workers (circa 5 mm. in length) (id., fig. 13); c’,
“nymphs ofthe first form’! (7—8 mm. long), with long
wing-pads, the genital organs little developed (id., fig. 10) ;
pb’, “nymphs of the second form”? (4—8 mm. long,
with more or less short wing-pads, the genital organs well
developed) (id., figs. 19,20, 22); n3, larve of complementary
or substitute royal forms, without trace of wings (7—
9 mm. in length). As will be seen later, the forms p® are also
larve of royal substitutes. Form a* may originate from a? of
Stage rv, and perhaps from 8’ as well ; B® is derived from form
1 [Lespés, ‘ Ann. Sci. Nat.’ (4), v, pp. 248—251, pl. v, fig. 6.]
2 [Id., pp. 251—254, pl. v, fig. 7.]
CONSTITUTION AND DEVELOPMENT OF TERMITES. 301
A? of Stage 1v; c® and p® from the forms with wing-buds of
Stage tv; and lastly, u* always from form c? of Stage rv.
To this fifth group belong further certain individuals
which I regarded as abnormalities before I ‘was acquainted
with the facts I have recorded about the soldiers of Calo-
termes.
They are nymph-soldiers, or nymphs with the buccal ap-
paratus of soldiers. They probably lose their wings and
become simple soldiers, as do those of Calotermes.!
VI. Perfect insects—that is, with fully developed
wings and capable of flight (Pl. 17, fig. 11). The number
of antennal joints remains, as in the nymphs, at seventeen or
eighteen, but they are always entirely pilose. These specimens
are distinguished by a general piceous colour, save for the
mouth parts, tarsi, and apices of the tibie, which are testa-
ceous. They originate from the nymphs of the first form.
VII. Complementary (Pl. 17, figs. 16—18, 21) or sub-
stitute (ad., figs. 15,22, 25) royal forms, sexually ma-
ture, or nearly so, but in the guise of the larva or
nymph, or else resembling an imago, partly infuscate
and with the wings torn.
The latter spring from immature, not fully darkened perfect
insects ; the others, in the larva or nymph form, arise from
the royal larve 2n® of group V, or from nymphs of the second
form, or finally from nymphs of the first form (fig. 23).
The royal forms derived from nymphs of the first form or
from not fully darkened perfect insects are all substitute,
never complemental forms.
The wing-rudiments may be wanting, or more or less short.
The body may be of the whitish yellow of old paper, or may be
more or less extensively blotched with brown. The abdomen
is much inflated, so that these forms are generally recognisable
at a glance.
They exhibit a further characteristic in the possession of
‘ It now seems to me probable that they should be compared with the
egg-hatching workers of the honey-bee.—G. B. Grassi, October, 1896.
302 B. GRASSI AND A. SANDIAS.
longer abdominal hairs than those of the nymphs and perfect
insects; these hairs are generally transverse in direction,
whereas they point obliquely backwards in the nymphs and
imagos.
The number of antennal joints may increase in the adult
workers, larvee of complementary or substitute forms, and
perhaps in the soldiers, to the maximum figure stated above.
The general law laid down for the castes of Calotermes is
equally applicable to Termes, with the introduction of a slight
modification, due solely to the existence of the workers, which
are wanting in Calotermes.
It will stand thus:—The regular development of
Termes up to the perfect insect may undergo a devi-
ation at various periods of life, which leads to the
formationof workers, of complementary or substitute
royal forms, or of soldiers; the last passing through
the stadium of the young worker. The deviation in
question may take place at various periods.
As the anatomical structure of Termes is in general agree-
ment with that of Calotermes, I shall restrict myself to a
few very brief remarks thereon.
Turning to the sense-organs, it is noticeable that the soldiers
of this genus do not possess compound eyes, nor do the larve
which become sexually mature, until the wing-rudiments begin
to appear. Consequently the complementary and substitute
royal forms, which have no trace of wings, are equally desti-
tute of compound eyes. But these are present and pigmented
in all the other complementary and substitute forms. The
pigment is also acquired by the nymph of the first form when
just about to change to the perfect insect.
In the nymph of the second form there is a distinct structure
in the neighbourhood of each compound eye, which I interpret
as a rudimentary pigmentless ocellus. This may also be easily
seen in nymphs ready for the final change and in the perfect
insect; but I have been unable definitely to find it in sub-
stitute forms (I have not looked for it in sections).
CONSTITUTION AND DEVELOPMENT OF TERMITES. 3808
The well-known law of Fritz Miller as to the increase in the
number of antennal joints is certainly inaccurate for Termes
lucifugus.
The 11-jointed antenna acquires its 12th joint by subequal
division of the 38rd, so that it cannot be said whether the
resulting 3rd or 4th is the younger. When the number of
joints exceeds twelve it is an indisputable rule that the third
is the youngest if the number is odd, and the fourth if the
number is equal. And when the number of joints is odd no
fresh joint develops until its predecessor has become pilose ;
for I have never found an example with an odd number of
joints which were all pilose, while it is easy to find examples
with an equal number of joints which are or are not all pilose
(Pl. 17, figs. 28—37).
I shall now describe the substitute and complementary royal
forms somewhat fully.
They rarely exceed 11 mm. in length, and those without
trace of wings probably never attain a length of more than
6 or 7mm. The only two kings I have found measured about
7 mm. in length. The shape both of the head and thorax in
the examples possessing wing-rudiments is exactly similar to
that of the nymph, and therefore of the perfect insect. On the
other hand, those which have no sign of wings find their
parallel in the small-headed larve; their pronotum has a
characteristic shape, which will be better understood from the
figure than from a laborious description, and which distin-
guishes them from the workers without a shadow of un-
certainty (Pl. 17, figs. 26, 27).
The colour of the body, as before mentioned, is generally of
the pale yellow tone of old paper, sometimes with an aureous
tinge. The head is dull aureous-yellow, with brown com-
pound eyes, which are wanting, as I have said, in the com-
pletely wingless forms. Many old examples exhibit large
areas which appear brown or sepia-coloured to the naked eye,
but are seen under the microscope to be yellow sprinkled with
minute black spots. These areas are as follows:
304 B. GRASSI AND A. SANDIAS.
1. The thoracic and abdominal terga (fig. 17). The latter,
however, exhibit an immaculate longitudinal median vitta and
a similar rounded spot on either side. The vitta may or may
not be evident on the metathoracic, and is wanting on the
mesothoracic tergite; it is present on the pronotum, where it
forms a cruciform mark with a similar transverse vitta towards
the apex. As a rule the side margins of the pronotum are
also immaculate.
2. The lateral limb of the abdominal sternites; but a few
black spots may also be seen to exist over the median portion.
3. The thoracic pleurze (side-pieces).
4. The basal portion of the legs.
It is important to notice that completely yellow examples
may at times have the wing-rudiments more developed than in
those with brown markings, although the greater abdominal
development, or their known history, may show the latter to
be the older.
The wing-pads are dirty white and exhibit abundant trachez,
arranged as in the fully formed wing.
Hairs are distributed everywhere except on the interseg-
mental spaces. As I have already said, those on the abdomen
are longer and transversely directed (figs. 17 and 18), and thus
distinguish these forms from those in which the wings are
fully developed (fig. 12). The genital appendices are always
absent in the female.
The abdominal hairs and black maculation furnish
characteristic points of difference from the fully
winged forms. The latter are found to become black by
uniform darkening,—that is, without first presenting the black
spots peculiar to the complementary or substitute examples.
The number of antennal joints in this latter group of royal
forms is variable, and may differ on each side of the head ; it
may be 14, 15, 16, or 17; but the last number is exceptional.
In no case are the antenne intact, as the examination of the
last joint as well as the third and fourth shows (PI. 17, fig. 29).
I have found queens without indication of wings and with
sixteen-jointed antenne, which were obviously truncated at the
/
CONSTITUTION AND DEVELOPMENT OF TERMITES. 305
apex ; and the condition of the third and fourth joints being
pilose, led me to infer that eighteen would have been present
if they were intact. But this is never the case in these forms.
There remain for description those substitute forms which
are derived from perfect insects which have not become black,
and have the wings torn off (Pl. 17, fig. 24). They measure
about 6 mm. in length, and possess the customary long out-
standing hairs. The dark compound eyes are conspicuous,
and the ocelli can also be made out. The antenne are
curtailed as usual, and the wings are rarely torn off exactly
along the hind margin of the squama, but so as to leave an
additional portion of varying, usually small size, the laceration
following a very irregular course, as if the wing had been
gnawed off.
The body is generally of a yellow colour, and is not spotted
with black, but the margin of the pronotum, especially the
posterior, and the hind margins of the meso- and meta-notum
are of a uniform brown, even when seen through the micro-
scope.
Occasionally the head and the entire meso- and meta-notum
are brown ; frequently, also, the thoracic pleurz and the outer
face of the basal portion of the legs. In some examples the
apex of the abdomen is brownish. In some the wings are of a
uniform dirty white, but in many others the squama, the costal
margin, and perhaps part of the torn edge are brown.
The genital appendices are present in the male, but are
wanting in the female, as in other complementary and sub-
stitute queens.
The stages of growth of substitute or complemental forms
with longer or shorter wing-buds are important, and require
notice. They are to be found by selection of the examples
with seventeen or eighteen antennal joints and rudiments of
the wings. According to Lespés’ classical researches, these
examples are of two kinds, with the wing outgrowths respec-
tively strongly and feebly developed. The former (fig. 10) are
his “nymphs of the first form;” the latter (figs. 19 and 20),
306 B. GRASSI AND A. SANDIAS.
b
his ““nymphs of the second form,” are further characterised
by the bulkier and more ovoid abdomen. The eyes of the
latter group begin to become pigmented and prominent, and
their ocelli are visible. Their antennz are intact, and the
hairs resemble those of the nymphs of the first form.
These nymphs of the second form become complementary
royal forms by a moult in which they acquire the characteristic
direction of the abdominal hairs and, if of female sex, lose the
female genital appendices.
They exhibit a marked development of the genital organs
which will be subsequently described.
In the examples, previously mentioned, of which the head
begins to enlarge, much the same development of the man-
dibles and maxille takes place as in Calotermes flavi-
collis.
With respect to the general colour of the inmates of the
colony, I should add that the workers are normally dirty white
or yellowish, and the soldiers more distinctly yellow ; freshly
moulted or very small specimens, and most undifferentiated
forms, or those destined for sexual maturity, are pure white.
The legs are alike in all the forms; the anterior tibiz
possess three, the others two apical spines.
Adult and fully winged examples exhibit the well-known
sexual differences of the seventh, eighth, and ninth abdominal
sternites, viz.—1. The seventh (the apparent sixth) is strongly
developed and semicircular (with the rounded edge posterior in
the female), very short in the male. 2. The eighth is reduced
to two lateral lobes in the female, and is small and entire in
the male. 38. The ninth nearly resembles the eighth. A
similar disposition is found in the mature substitute and com-
plementary forms. As in Calotermes, the ecdyses are rather
numerous in Termes lucifugus, and do not bear the sup-
posed relation to the increase in the number of antennal joints.
CONSTITUTION AND DEVELOPMENT OF TERMITES. 307
4. Relative Numbers of the Castes.
It is impossible to indicate the relative numbers of the dif-
ferent forms with any degree of certainty, owing to the exces-
sive difficulty of fixing the limits of a colony. But, as before,
the soldiers of Termes are far less abundant than the other
forms. The workers occur in enormous numbers; the young
and larve are also very numerous, the “nymphs of the first
form ” relatively much less so.
Examples in process of development into royal forms are
common only in trees which, though densely populated,
contain no or very few royal forms. Nevertheless a certain
number of male “ nymphs of the second form” can easily be
found from March to June, especially in trees containing com-
plementary or substitute queens.
Complementary forms occur in 6 or 8 per cent. only of trees
invaded by Termes, but in these there may be upwards of a
thousand, though the number usually oscillates between fifty
and two hundred.
Substitute royal forms, varying in number from about ten to
two hundred, are confined to those nests which have been par-
tially or completely orphaned, either by destruction of the com-
plementary kings and queens, or by cutting down a tree and
removing it a kilometre or so away from its original situation.
In the nests which I have mentioned as having been ob-
tained in glass jars, a single true royal pair was found.
5. Seasonal Variations in the Colony.
The colony differs very much at different times of the year.
Eggs are present in May, June, and July, and, in all proba-
bility, in August and September as well.
The youngest larve are never present in April and May.
Nymphs of the first form are not met with in June and
July. Forms with tolerably well-marked wing-buds, and with
fourteen to sixteen antennal joints, are absent or very scarce
in October and November; they increase in number in
December, to become abundant by January or February.
308 B. GRASSI AND A. SANDIAS.
Winged imagos occur from the beginning of April to the
middle of June (a few stragglers being exceptionally found as
late as September).
The remaining forms are present, as a rule, all the year
round, except the kings, or incipient royal forms of either sex.
The latter have already been repeatedly referred to, and
must receive further notice in the following chapter.
Lastly, I must not omit to mention that nymphs of the first
form are absent in some years from certain nests, which have
probably been orphaned at a previous period.
6. Duration of Development, Life, &c.
The eggs hatch fifteen or twenty days after they are laid.
Many observations lead me to conclude that the very young
larvee found in the winter do not develop farther than the
nymph of the first form in the following summer, and there-
fore must certainly live through a second winter before acquir-
ing wings; e.g. larvee born in October, 1889, will not have the
wings fully developed until April, 1891.
Thus, too, the complete or almost complete absence in
October and November of examples with 14—16-jointed an-
tennz and distinct wing-buds compels me to believe that those
hatched in May have become nymphs already, or else that
they do not yet possess fourteen antennal joints. The latter
hypothesis is correct; for if the former were, a much larger
number of nymphs ought to be found in particular nests than
is actually the case.
A small colony obtained in a glass jar was furnished with
a number of fully winged individuals in the early part of May ;
on the 20th of December it contained, beside the other in-
habitants, five workers with 12—14-jointed antenne.
Several orphaned nests were placed in large glass jars in
January ; next October they still contained small specimens
with 13—14-jointed antennz, while no substitute forms had
developed.
These further facts agree sufficiently with the hypothesis
which I regard as correct.
CONSTITUTION AND DEVELOPMENT OF TERMITES. 309
Tn conclusion, a lapse of eighteen to twenty mouths may be
estimated to take place between the times of hatching and of
reaching the perfect state. The workers and soldiers probably
require a much shorter time, and both can apparently be
developed by the autumn from eggs laid in May of the same
year.
The duration of life of single specimens cannot easily be
estimated. The soldiers and workers certainly do not die off
about the middle of June, as Lespés pretends. The king,
whether complementary or substitute, does not live more than
a couple of months after reaching sexual maturity, whereas
the corresponding queens will live for several years.
7. Situation of Different Forms in the Nest.
The fertile or nearly fertile royal forms usually live in a very
remote part of the nest, often collected together in a deep root
or in the heart of a large trunk.
The only two mature kings I have found accompanied the
queens; the latter may be surrounded with numerous new-born
larvee and clumps of 30 or 40 to 100 eggs.
Eggs and larve can be carried long distances, as is shown
by the occasional presence of numerous young larve and some-
times of eggs as well in trees destitute of queens.
The larvee, nymphs, winged forms (if present), soldiers, and
workers all occur mingled together, both in the midst of the
queens and elsewhere. Recently invaded situations contain,
as a rule, chiefly workers and soldiers; and isolated workers,
which one is tempted to look on as explorers or pioneers, may
sometimes be found under stones, or in reeds, &c. Certain
parts of the nest often contain principally workers and soldiers ;
others larvee with wing-rudiments and nymphs.
[I would add further that when a nest is opened certain parts
will contain nothing but soldiers in the neighbourhood of the
egg-clumps. It is a mistake to regard this as a normal state
of things; the disturbance and noise of opening the nest have
put all the other inhabitants to flight, while the valiant soldiers
alone remain to protect the eggs. |
VOL. 39, PART 3,—NEW SER. a8
310 B. GRASSI AND A. SANDIAS.
VIII. Certain Habits.
The swarming will be first described.
A swarm was observed on May 14th, 1891. A host of
Termites issued in groups of two to five from a single crack in
a tree-trunk, rose on the wing for about three metres, and
then followed the direction of the wind. Many fell on the
neighbouring plants, and others were seized by ants as they
emerged. The swarm began at 10 a.m., and lasted for more
than two hours.
This incomplete observation was made by the laboratory
servant, who chanced to see a similar occurrence on May 24th,
1891. In the second case the winged forms were seen to issue
from several holes, at which a few soldiers and workers also
appeared. The imagos were all destroyed, at first by two
lizards, which remained on watch near the nest and devoured
them directly they appeared. When the lizards were driven
off the insects were seen to come out in groups of six or eight,
run a certain distance, spread their wings, and then take
flight. At first they rose a certain height, and then followed
the direction of the wind (a light scirocco). A few struck
against branches and fell to earth, but they quickly got up
again. This swarm lasted from about 9.30 to 11 a.m. When
it was over the nest was opened, and found to contain many
imagos ready for swarming about half a metre from the exit-
holes. These were all females, as were those (twenty-six)
collected during the swarm.
Sinilar swarms were observed on various other occasions.
One of special interest took place in the laboratory from the
nest in a glass jar, which I have previously described, men-
tioning its occurrence on June Ist. The details are now given.
Towards the end of April I examined the nest at the bottom
of the jar, and observed the appearance of perfect insects,
some brown, others still white, but becoming brown in the
course of afew days. After May 20th, suspecting that swarm-
ing would take place soon, I covered the vessel with a wider
and taller glass bell-jar. Both were placed on a pane of glass,
CONSTITUTION AND DEVELOPMENT OF TERMITES. 311
and the jar was not completely closed, because the paper did
not accurately fit its mouth.
On June Ist, about 9 a.m., I found a number of winged forms
with a few soldiers and workers on the part of the sheet of
glass which lay between the two jars; the poor little animals
had taken advantage of the spaces left by the ill-fitting paper
to emerge, and were then vainly trying to escape from the
bell-jar. Nothing was disturbed, and twenty-four hours later
no change had taken place, except that the winged forms were
more numerous, and some were moribund.
I then decided to liberate them, and removing the bell-jar I
put them by means of a quill pen into a receptacle containing
suitable materials for a nest—but without success, for they all
died in less than forty-eight hours. No more winged examples
could be detected through the walls of the glass jar; it
was replaced unopened on the pane of glass, and once more
covered with the bell-jar. Twenty-four hours later a large
number of larve and several soldiers were again found on the
pane. They were then returned to the nest by means of a
quill, and did not escape again.
The observation here recorded was made some years ago,
and though imperfect is not wanting in significance.
Another swarm was observed on a May morning of the
present year (1892). There happened to be a small level
place in front of the holes from which the perfect insects
issued, and on this they ran to and fro before taking wing.
They were accompanied by soldiers and workers which had
emerged into daylight, evidently to protect the swarm.
Several writers have recorded that black Termites, when
ready to fly, perform certain movements which can be best
followed by putting the insects on a sheet of paper; these are
the so-called love passages, often described, and especially by
Fritz Miller.
They are not exhibited by Calotermes, but may easily be
observed in Termes lucifugus.
The majority of examples dispose themselves after the loss
of their wings in pairs, one behind the other. More exactly,
ole B. GRASSI AND A. SANDIAS.
the one in front attempts to run away from the other, which
pursues it and palpates the extremity of its abdomen, and
sometimes the sides as well. In some cases the pair is com-
posed of a wingless individual in front and a winged one
behind ; or a male in front and a female behind; or the
opposite; or both may be of the same sex, whether male or
female.
If a few workers are put among the imagos, one of the latter
may often be seen to pursue a worker in the same manner.
And occasionally three examples, instead of merely a pair,
may be seen, one following the other.
I believe that the meaning of these supposed amorous dis-
plays is entirely different from that usually assigned to
them, and that the pursuer wishes to solicit the dejecta of
the one pursued; this will be explained in the following
chapter.
I have said that the adults lose their wings (the persistence
of the squama being understood) (Pl. 17, fig. 12), and I must
now explain more minutely how this occurs. Suppose that the
wings of a specimen are accidentally allowed to touch the moist
walls of the glass jar, they stick to it, and readily break off as
soon as the owner tries to run away.
[The insects perform various movements on their own account
in order to tear off the wings. I have seen one raise and
lower them, and at the same time put the hind leg over them
so as to hold them down to the surface on which it was stand-
ing. Another example got rid of them by violent fluttering ;
and a third, which had only one wing left, tried to tear it off
at first by forcible flapping, and then succeeded by holding it
firmly with one of the hind legs. ]
It often happens that the imagos lose their wings while still
in the nest, but they nevertheless abandon it, as do those of
Calotermes.
In fact, the perfect insect has an imperious craving to quit
the nest in which it develops.
Winged examples artificially enclosed in a corked tube
CONSTITUTION AND DEVELOPMENT OF TERMITES. 313
quickly gnaw through the cork and escape. This happens even
if the wings have fallen off. ;
It will be recollected that the two sexes swarm at different
times. This is demonstrated by—
1. The fact that all members of a swarm are found to be
of the same sex.
2. The presence in many nests, late in or at the end of the
swarm-period, of black winged forms, which are all males or
all females only (by a rare exception a single male may be
found to every twenty or thirty females), whereas nests in which
the imagos are still white contain a male to every two or three
females.
The convulsive movements spoken of under Calotermes
are exhibited alike by Termes, and are common to all mem-
bers of the colony except the newly born, and have the same
significance. Moreover the soldier is able simultaneously to
produce a special crepitus (creaking), which arises whenever
the head is held horizontally during the act of quivering by
friction between the hind margin of the occiput and the ante-
rior margin of the pronotum. But whenever the head is held
in the normal position during the act—that is, somewhat de-
flexed—no perceptible sound is produced, owing to the absence
of friction.
The soldiers, therefore, possess two distinct modes of commu-
nication ; and itis noticeable that those of Calotermes always
hold the head obliquely deflexed when quivering and produce
no sound,
I may add that this characteristic crepitus may be heard at
very short intervals by applying the ear to a trunk containing
a nest of Termites. This proves that the quivering motions
are a constant feature in normal and undisturbed nests, in
which they are therefore not employed to give indications of
alarm or distress; and I conclude that, besides these signifi-
cations, the convulsive movements must also have the value of
ordinary speech ; that they constitute, in short, a means of in-
tercommunication. The same conclusion holds good for Calo-
314 B. GRASSI AND A. SANDIAS.
termes; and I imagine that the quivering of both species
produces a sound which is perceptible to the tympanal organ
of the tibia, but is inaudible to the human ear.
Termites, moreover, may communicate by means of the
antenne. Thus, if a few are placed on a table, they usually
arrange themselves in single files, which circle round the
objects standing thereon ; and, in such a case, if two Termites
moving in opposite directions chance to meet, they recipro-
cally touch their antennz and then continue each on its own
course (vide also previous statements).
Tasks necessary for the common welfare, with a few excep-
tions, are undertaken by all the inmates of the colony ; but the
soldiers are unable to gnaw wood, owing to the great elonga-
tion of their mandibles.
Substitute or complemental royal forms have never been
seen to prepare wood-meal, or to transport it, or ova,&c. Yet
all these duties are carried out by the perfect insects with fully
developed wings, before or after they have been shed. Newly-
born larve may easily be found carrying about wood powder.
[The soldiers serve for defence, like those of Calotermes.
Two soldiers, one a Termite, the other a Calotermite, were put
together in a small glass vessel. They accidentally came into
contact and began to fight. The Termite, having the advan-
tage of great quickness in movement, whereas Calotermes is
sluggish, bit off some of its enemy’s legs, and was proceeding
to further hostilities, when the other seized an opportune
moment and cut its head off. On other occasions the soldiers
of Calotermes tore the abdomen of the soldiers of Termes
to pieces.
The soldiers’ mandibles may appropriately be likened to a
powerful pair of shears. Termite soldiers become formidable
when put into one of the customary little nests of Calo-
termes deprived of soldiers, rapidly cutting off the antennz
of numerous examples, and biting them in various places.
But if they are few in number, the Calotermites eventually
reduce them to helplessness by shearing off their mandibles,
and then pursue, tear, and kill them.
CONSTITUTION AND DEVELOPMENT OF TERMITES. 315
A fight is invariably provoked by putting soldiers of both
species into a jar; if large larve or nymphs of Calotermes
are introduced into a colony of Termes they are usually left
alone by the soldiers, which probably fear them ; whereas Calo-
termites (see p. 283) always kill any workers or nymphs of
Termes which have been added to their nest. Termes
soldiers fraternise, and do not fall out, even when taken from
different nests.
If triturated wood, soldiers, workers, and young of Termes
are put together in a jar, the soldiers are soon seen to post
themselves on the top of the rubbish, evidently on guard.
Sometimes inmates of the same nest (soldiers and workers,
or the latter inter se) come to blows, and wound each other
ferociously in the thorax or abdomen, and do not stop unless
others interfere to separate them. These internecine battles
can be provoked, e.g. by overturning the contents of a nest,
and are perhaps due to each individual imagining that his
neighbour is the cause of the disturbance.
Termites shun the light, and prefer to collect in the darkest
parts of a vessel. ]
(To be continued.)
316 B. GRASSI AND A. SANDIAS.
EXPLANATION OF PLATES 16—20,
Illustrating Professor B. Grassi’s and Dr. A. Sandias’s paper
on “The Constitution and Development of the Society of
Termites: Observations on their Habits ; with Appendices
on the Parasitic Protozoa of Termitide, and on the
Embiide.”’
The first number after the explanation of each figure indicates the ocular,
the second the objective of the microscope employed. Kor. = Koritska
microscope, withthe tube in. Hart. = Hartnack microscope, with the tube in.
T. = Termes lucifugus. C. = Calotermes flavicollis.
Instead of the expression that a given individual possesses, e.g. seventeen,
antenual joints, the abbreviation ‘‘ with seventeen joints” is employed.
PLATE 16.
Calotermes flavicollis.
Fic. 1.—Small-headed larva, with twelve joints. The third, fourth, and
fifth indistinct ; the former not pilose; the fifth with short hairs (distinguish-
able with a higher amplification only).
Fie. 2.—Large-headed larva, with thirteen joints. The third short, and
not pilose.
Fie. 3.—Small-headed larva, with sixteen joints. The third and fourth
scarcely indicated, and not pilose. Wing-rudiments distinctly present, but
very short, and visible only with higher amplification.
Fic. 4.—Nymph, with seventeen joints. The third pilose, the fourth not.
Fic. 5.—Large soldier.
Fig. 6.—Perfect insect, with fully developed wings.
Fie, 7.—True queen, in the fourth year of maturity.
Fic. 8.—Outline of a true king or queen, in the second year of maturity.
Fic. 9.—Outline of a true queen, in the third year of maturity.
Fic. 10.—Outline of a true king, in the fifth year of maturity.
Fic. 11.—Outline of the abdomen of a true queen, in the fifth year of
maturity (drawn approximately to the same scale as Fig. 10).
Fie. 12.—Young substitute queen.
Fie. 13,—Outline of the abdomen of a substitute queen, in the third year
of maturity (drawn to the same scale as Fig. 11).
CONSTITUTION AND DEVELOPMENT OF TERMITES. 317
Fic. 14.—Developing royal substitute form, with fifteen joints; the third
the shortest, and not pilose.
Fic. 15.—Exuvie of a developing substitute form.
Fie. 16.—Outline of the body of a small soldier (the antenne are omitted),
Fie. 17.—Outlines of the bind angle of (4) the mesonotum and (a) the
metanotum of a substitute queen, with very slight traces of the wing-
outgrowths. 3, 4, Kor.
Fic. 18.—Outline of the right half of the meso- and metanotum of
developing substitute form. The anterior wing-rudiment is torn off.
Fic. 19.—Antenna of a small soldier. 3, 4, Kor.
Fie. 20.—Base of a fifteen-jointed antenna. 3, 4, Kor.
Fic. 21.—The same in a different stage, but still fifteen-jointed.
Fic. 22.—The same after development of a sixteenth joint.
Fie. 23.— Ovary and oviduct of a soldier. 3, 4, Hart.
Fic, 24.—Posterior leg of a young nymph.
Fic. 25.—Old substitute queen, with wing-rudiments.
Fic. 26.—ialf of the thoracic terga, exhibiting wing-rudiments, in a young
soldier. 3, 5, Hart.
Termes lucifugus.
Fic. 27.—Brain of a worker, by transmitted light. 3, 5, Hart.
Fies. 28 to 33.—Series of horizontal sections of the brain of a young
nymph, to exhibit the fungiform bodies (interpreted as psychic centres). They
are represented by the darker and more closely dotted portions (which possess
small, deeply staining nuclei).
Fig. 28 represents the most superficial section, and is followed by the
others in numerical order.
Fig. 34 represents, by way of comparison, a similar section through the
brain of Embia, at the point where the fungiform bodies attain the maximum
dimensions. gl. Retro-cerebral gland. [This gland of unknown function
exists (only ?) in the nymph of the first form, the perfect insect, and the
soldier. It eliminates a transparent secretion, which can be spirted out for
some distance. ]
PLATE 17.
Termes lucifugus.
Fic. 1.—Larva, with eleven joints, the head undifferentiated.
Fic. 2.—Large-headed larva, with twelve pilose joints.
Fic. 3.—Small-headed larva.
Fie, 4,.—Small-headed larva, with fourteen joints, the fourth not pilose.
318 B. GRASSI AND A. SANDIAS.
Fig. 5.—Diagram of a similar but large-headed larva. The hair-lines
between 2 and 3, and between 4 and 5, apply to each of the two figures.
Fic. 6.—Larva similar to 4, but larger and with wing-rudiments.
Fie. 7.—Larva similar to 5, but equal in size to that of Fig. 6.
Fie. 8.—Anterior portion of a larva similar to that of Fig. 7, but with a
somewhat larger head ; probably a soldier larva.
Fic. 9.—Larva, with fifteen joints; the third alone not pilose, and with
evident wing-rudiments.
Fie. 10.—Nymph, with long wing-pads.
Fic. 11.—Fully winged perfect insect.
Fic. 12.—Perfect insect after shedding the wings.
Fic. 18.—Adult worker.
Fig, 14.—Soldier.
Fic. 15.—Young wingless substitute queen.
Fie. 16.—Old wingless complementary queen.
Fic. 17.—Old complementary queen, with slight rudiments of wings.
Fig. 18.—The same, with more distinct rudiments.
Fie. 19.—Male nymph of the second form (March Ist).
Fic. 20. —Outline of a male nymph of the second form (April 11th).
Fig. 21.—Young complementary queen, derived from a nymph of the second
form.
Fie. 22.—Thorax of a nymph of the second form, with the wings more
developed than those of the example in Fig. 21.
Fie. 23.—Substitute queen, derived from a nymph of the first form.
Fic. 24.—Substitute queen, derived from a perfect insect; partly infuscate,
and with the wings torn (the infuscation is not shown).
Fic. 25.—Posterior extremity of a very old substitute queen, without sign
of wings.
Figs. 1 to 25 are all drawn by the camera lucida to an equal scale, with
exception of Figs. 16, 17, and 21, which are somewhat more enlarged.
Fic. 26.—Pronotum of a worker. 3, 4, Kor.
Fic. 27.—Pronotum of a completely wingless substitute queen. 3, 4, Kor.
Fic. 28.—Antenna of a nymph of the second form. 3, 4, Kor.
Fic. 29.—Antenna of a complementary queen. 3, 4, Kor.
Fic. 30.—Base of a twelve-jointed antenna.
Fie. 31.—Base of a thirteen-jointed antenna.
Fig. 32.—Base of a fourteen-jointed antenna.
Fic. 33.—Base of a fifteen-jointed antenna.
CONSTITUTION AND DEVELOPMENT OF TERMITES. 319
Fic. 34.—Base of a sixteen-jointed antenna.
Fic. 35.—Base of another sixteen-jointed antenna.
Fic. 36.—Base of a seventeen-jointed antenna (nymph of the first form).
Fic. 37.—Base of an eighteen-jointed antenna (nymph of the first or
second form).
The figures 30 to 37 are all copied by the microscope, 3, 4, Kor., with the
tube drawn out.
Fic. 38.—Flattened tube constructed by T., and suspended from a plaster
cornice. The lower end (to the right of the figure) was open; the upper end
is broken.
Fic. 39.—A similar flattened tube—short, enlarged and flask-shaped. A
small portion of the cornice to which it was suspended is represented.
Fic. 40.—Portion of a gallery formed by T. in the angle of a wall; seen
from the inside.
PLATE 18.
Termes lucifugus.
Fie. 1.—Ovary of a perfect insect before loss of the wings. 3, 6, Kor.
Fic. 2.—Ovaries of a nymph of the second form. 38, 4, Kor.
Fic. 3.—Ovary of a very old complementary queen. The tubules are evi-
dently atropied, and the spermatheca was empty. 1,4, Kor. Por. ant. =
anterior portion.
Fic. 4.—Left ovary of a nymph of the first form. 3, 4, Kor.
Fig. 5.—Testis of a perfect insect before loss of the wings. 3, 6, Kor.
(Spermatozoa are present in the vas deferens.)
Fie. 6.—Testis of a nymph of the second form. 3, 4, Kor.
Fic. 7.—Testis of a nymph of the first form. 3, 4, Kor.
Fic. 8.—Testes, vasa deferentia, and vesicula seminalis (the two latter
not containing spermatozoa) of a perfect insect before loss of the wings.
3, 4, Kor.
In the above eight figures, and in some on the following Plate, the efferent
ducts are partly shown, in addition to the generative glands.
Fic. 9.—Developing spermatozoa, in the fresh state. The one indicated
with the letter a is fully developed. 4, 3 Kor.
Fic. 10.—Oviduct (¢u.), uterine portion (a. ov.), spermatheca (spé.), seba-
ceous (colleterial) glands (g/. seb.) of a complementary queen. 3, 4, Kor.
Fic. 11.—Portion of the intestine at the origin of the four Malpighian
tubules in a newly-born larva. 3, 4, Kor.
Fic. 12.—Similar portion, with four large and four small tubules. 3, 4, Kor.
320 B. GRASSI AND A. SANDIAS.
Fis. 13.—Superficial portion of a nest of T. constructed in a jar, with exit-
holes and chimneys for swarming.
Fic. 14.—Portion of a termitarium found in a bench in the church
Pesara.
Fie. 15.—Tubular gallery, similar to that of Fig. 39, Pl. 2, but not flask-
shaped.
Fic. 16.—Portion of a partly excavated and partly built nest, found in a
large cavity in the root of a cactus swarming with T.
Fie. 17.—Galleries constructed by T. kept in a glass jar; D-shaped, except
the one projecting over the lip, which is tubular, The shading at the bottom
indicates the rubbish which harboured the colony.
PLATE 19.
The left side of the plate refers to Calotermes flavicollis, except
Fig. 6, which refers to Termes lucifugus; the right side refers to Embia.
Lert Sipe (CALOTERMES),
Fie. 1.—Ovary and colleterial gland of a true queen in the second year of
maturity. 3, 4, Kor.
Fie. 2.—Ovaries of a very bulky queen, enlarged about three diameters
(the tubes do not actually lie in one plane as in the figure). eat. ant. =
anterior extremity.
Fie. 3.—Colleterial glands of a true queen. 3, 4, Kor.
Fic. 4.—Left testis (¢est.), vas deferens (con. def.), and vesicula seminalis
(ves. sem.) of a true king. 3, 6, Kor.
Fie. 5.—Testis of a substitute king, several years old. 3, 6, Kor.
Fie. 6.—Portion of the spermatheca, showing the ducts and orifices of the
underlying glands (fresh). (Termes lucifueus.)
Fie. 7.—Salivary glands of a small larva. ves, = reservoir. 38, 8, Kor.
Fig. 8.—Relations of the hinder part of the chylific ventricle and intes-
tine. The former contains a blackish, the latter a yellowish detritus. Their
junction is indicated by the insertion of the Malpighian tubules.
Fie. 9.—Stomodeum (zt. ant.) and part of the chylific ventricle (xt. med.).
inv. = invagination of the former into the latter.
Fic. 10.—Tympanic organ in the fresh state. 3, 75, Kor.
Fie. 11.—Tuhe-nest. Large portions of the glass are obscured by dis-
gorged matter. The cork is riddled with burrows, and is represented separately
on the right.
Fie. 12.—Spermatozoa from the spermatheca, in the fresh state. 5, 54,
Kor.
CONSTITUTION AND DEVELOPMENT OF TERMITES. 321
Rieut Sipe (Empta).
Fig. 1.—Ventral ganglionic chain; L. 1 = anterior, L. 2 = middle, L. 3
= posterior leg.
Fic. 2.—Dorsal view of the tracheal system (the stigmata are numbered).
Fig. 3.—Alimentary canal. gl. sal. = salivary glands. int. ant. = oeso-
phagus and proventriculus. zz¢. med. = chylific ventricle. zt. post. = in-
testine.
Fic, 4.—Ovarian tubes and oviducts (ovd.). seg. med. = segment médiaire
(the successive abdominal segments are numbered).
Fig. 5.—Male generative organs.
Fic. 6.—Posterior extremity of the male. proc. dext. = right, proc. sin.
= left apophysis. cer. dext. = right cercus.
Fic. 7.—Labium and labial palpus of one side.
Fic. 8.—Maxillary lobes and palpus.
Fie. 9.—Anterior tarsus (seen obliquely from the side).
Fic. 10.—Middle tarsus (idem).
Fic, 11.—Posterior tarsus (idem).
PLATE 20.
Parasitic Protozoa of Termitide.
The figures are all drawn with a Koritska microscope, ocular 5, objective +,
Leitz. The dotted lines represent the continuation of the unbroken lines on
the remote face.
Fig. 1.—Trichonympha agilis, in optical section, nearly corresponding
with the middle plane of the body.
Fic. 2.—The same, at a somewhat different level.
Fic. 3.—The same, in a superficial plane.
Fic, 4.—The same, partly schematic, to show the arrangement of the
flagella.
Fie. 5.—The same, to give an idea of the complicated arrangement of the
spicules in some examples.
Fic. 6.—Joenia annectens.
Fic. 7.—Idem.
Fic. 8.—Idem.
Fie. 9.—Endoskeleton and nucleus of a very large example of Joenia in
optical section.
Fic. 10.—Microjoenia hexamitoides.
322 B. GRASSI AND A. SANDIAS.
Fie. 11.—Dinenympha gracilis.
Fie. 12.—Idem.
Fic. 13.—Idem.
Fic. 14.—Idem.
Fie. 15.—Idem.
Fic. 16.—The same, with spirilla attached to one extremity.
Fic. 17.—The same, covered with spirilla.
Fic. 18.—Pyrsonympha flagellata (the majority of the flagella are
omitted).
Fic. 19.—The same, in superficial view.
Fie. 20.—The same, exhibiting the nucleus and rodlets.
Fic. 21.—Holomastigotes elongatum (the majority of the flagella are
omitted).
Fic. 22.—The same, in superficial view.
Fic. 23.—Plan of the lines of origin of the flagella on both faces of
Holomastigotes elongatum.
Fic. 24.—The same, in another individual.
ON OTENOPLANA. 323
On Ctenoplana.
By
Arthur Willey, D.Sc.
With Plate 21.
Tue discoverer of the remarkable genus Ctenoplana, which
presents affinities both to the Ctenophora and to the Turbel-
laria, was, as is well known, Professor Alexis Korotneff,
who obtained only a single specimen off the west coast of
Sumatra, and described it in the ‘ Zeitschrift fiir wissenschaft-
liche Zoologie *! for 1886.
Korotneff found his specimen drifting in a current of the
sea, in the company of a large number of Porpita. It was
distinguished by its deep red or crimson colour, and was
named C. Kowalevskii in honour of the discoverer of
Celoplana.
Since 1886 no second record of the occurrence of the genus
has been made.
In January of this year (1896), while cruising among the
islands which form the Hastern Archipelago of British New
Guinea, in pursuance of zoological work, I was fortunate
enough to pick up a cuttle-bone which had evidently been
afloat for a long time, and was being carried along by the
current off the group of the islands named on the chart the
Conflict Group.? On the cuttle-bone were numerous minute
1 Vol. xliii, pp. 242—251, Taf. viii.
? These islands surround a magnificent lagoon.
824 ARTHUR WILLEY.
organisms, including young green-tinted Polychetes, young
Planaria, Anthozoan larve, young Cirripedes in the Cypris
stage, and four specimens of Ctenoplana. Three of the last
were of a pronounced green colour, and the fourth was
crimson.
The green Ctenoplana, both on account of its colour and its
shape, is an entirely new kind, and I shall name it C. Korot-
neffi, The crimson variety may or may not be specifically
identical with C. Kowalevskii; but as my sketches of the
external form differ somewhat from Korotneff’s figures, I think
it will be well to give it a provisional name with the object of
engaging the attention of any zoologists who may have future
opportunities for studying the genus. I propose, therefore, to
call my crimson specimen C. rosacea.
As TI had no chance of getting ashore, I had to make the
best of the limited accommodation supplied by a small cutter,
in order to observe the appearances presented by the living
animals and their movements. Although I omitted to make
accurate measurements of the living expanded animals, their
average diameter would correspond closely in length with that
of C. Kowalevskii, which, as stated by Korotneff, mea-
sured 6 mm.
Many of the external features of Ctenoplana were correctly
described and figured by Korotneff, but his specimen appears
to have been not very active; and although, as he says, he had
it under observation in the living condition for three to four
hours, it did not once extrude its tentacles, so that he only
became aware of their existence in section. The consequence
of this was that Korotneff was completely wrong in localising
the plane of the tentacles. He placed them in the plane at
right angles to that in which they actually occur.
At the ends of one of the principal diameters of the disc-
shaped body of Ctenoplana the margin of the body is incurved.
Korotneff, without the smallest hesitation, describes these
marginal bays as anterior and posterior in position, while,
according to him, the tentacles were situated along the diameter
at right angles to the “‘ antero-posterior ”’ diameter—that is to
ON OTENOPLANA. oe
say, in what would correspond to the transverse plane. This,
however, is quite wrong.
As will be seen later on, I agree with Korotneff in his iden-
tification of the antero-posterior axis (in comparison with
bilateral animals), but it is along this axis that the solid
tentacles lie.
Furthermore, Korotneff’s specimen did not give him an
exhibition of its swimming powers, so that he could not
observe the movements of the ctenophoral plates, and the
result was that he formed the opinion that these plates had
become altered as to their function, and that they hardly
appeared to serve for the progression of the body. This
is a curious repetition of the old error with regard to the
means of locomotion of the Ctenophores, as set forth in
Chun’s monograph. As a matter of fact, when Cteno-
plana swims, the ctenophoral plates are its sole means of
locomotion.
I now pass on to a systematic account of my own observa-
tions.
1. Shape and Movements of the Body.—Like the
Planarians, the body of Ctenoplana comprises a thickened
median ridge-like area and two lateral thin skirt-like areas,
the “Seitenfelder” of Lang. In the attitude assumed when
crawling, the body is nearly round with the exception of
the above-mentioned marginal bays, from which I observed
the muscular pinnate tentacles being constaxtly protruded and
retracted while the animal was crawling (fig. 1). As also
observed by Korotneff, Ctenoplana crawls with one of the
rounded margins of the body directed forwards as a rule. The
tentacles thus appear remarkably like transversely paired
structures, and one would naturally at first describe them as
such. But it must be remembered that the terms anterior,
posterior, and transverse, as applied to bilateral animals,
are not applicable to Ctenoplana.
The tentacles, when extruded, are found to be white struc-
tures, thus making a marked contrast to the green or red
colour of the body. They are provided with small secondary
VOL, 39, PART 3.—NEW SER. Y
326 ARTHUR WILLEY.
tentacles or pinne, arranged somewhat irregularly, but in a
single series. Like the tentacles of Coloplana, described by
Kowalevsky, they are strictly comparable with the tentacles of
a Cydippid. Both the tentacles and their pinne are quite solid,
being completely filled up with a muscular core. Within the
body each tentacle is enclosed within a hollow sheath which
opens to the exterior at the end of a small papilla at the
base of the marginal bay. When retracted, therefore, the
tentacles form a median axial skeletal support for the body,
interrupted in the middle region of the body by the aboral
sense-organ.
The aboral surface of the body may be at once called the
dorsal surface, and the oral the ventral surface.
The possession of a relatively wide Planarian-hke skirt not
only permits Ctenoplana to crawl about on firm surfaces, but
enables it also to attach itself, in a highly characteristic
manner, to the surface-film of water by its ventral surface.
In this position it greatly resembles Planarians, which are also
fond of assuming the same position. When lying thus attached
to the surface of the water the round central oral opening can
be seen. The mouth can be protruded so as to form a slight
cone.
When swimming, Ctenoplana brings the two halves of the
skirt together so as to form a bell-shaped, or better, a Pilidium-
shaped structure which progresses very rapidly by means of
the ctenophoral ‘plates. In swimming, the aboral pole is
directed forwards as it is in the Ctenophora.
The ctenophoral apparatus consists of eight small oval
plates, placed four on each side of the tentacle axis.‘ Across
each plate run six or seven shallow grooves, from which the
long cilia arise (figs. 1 and 5). The cilia of each groove appear
in section to be united usually for some distance from their
base, and then to separate out into the individual cilia (fig. 5).
The ctenophoral plates alternate with the lobes of the central
gastric system (figs. 1—3). I only had a fleeting view of the
1 The line joining the bases of the tentacles may be called the “ tentacle
axis.”
ON OTENOPLANA. 327
peripheral anastomosing ramifications of the gastric system,
and have not indicated them in the sketches of the external
form.
When Ctenoplana wishes to sink from the surface to the
bottom it doubles itself up in the usual way, and so sinks
apparently without employing the combs. This was also ob-
served by Korotneff.
2. The Aboral Sense-organ.—As already described by
Korotneff, the aboral sense-organ consists essentially of an
otolithic mass, suspended by stiff processes from adjacent cells
in a cupule, and surrounded by a ring of ciliated tentacles.
Korotneff figures the latter in the form of a simple circlet.
This, however, is not the case. The circlet of sensory
tentacles surrounding the otolith consists of two
distinct and separate halves, with about nine ten-
tacles in each half. The one half is placed on one
side, and the other on the opposite of the tentacle
axis (fig. 1).
This is perhaps the most important observation that I was able
to make on the living animal, and it is a crucial one for deciding
upon the homologies of the axes of Ctenoplana with those of
bilateral animals. The division of the circlet of sensory
tentacles into two portions was remarkably distinct and un-
equivocal. The sensory or apical tentacles (as distinguished
from the muscular or terminal pinnate tentacles) are usually
carried extruded (figs. 2 and 3), but they can be completely
retracted.
3. Cilia.—I cannot confirm Korotneff in his statement as
to the general distribution of cilia over the surface of the body.
The places where I have observed cilia (apart, of course, from
the ctenophoral plates) are as follows: (i) on the sensory
tentacles, (11) on the cells lining the sheaths of the pinnate
tentacles, and (iii) over a large area of ventral surface (fig. 5).
I must deny the presence of cilia on the general dorsal
surface.
1 As Ctenoplana is semi-opaque, it is difficult to discern much of the
internal structure in surface view.
828 ARTHUR WILLEY.
4. Gastro-vascular System.—The central main portion
of the gastro-vascular system presents the lobed appearance
shown in the figures, the lobes being paired about the tentacle
axis. The middle and largest pair of lobes belong to the
stomach, and thus serve to mark out the stomachal plane
(Magenebene of Chun). The stomachal plane, therefore, as in
the Ctenophores, lies at right angles to the plane of the ten-
tacles, which corresponds to the funnel plane (Trichterebene)
of Ctenophores.
My identification of the stomachal plane in Ctenoplana is
just the reverse of Korotneff’s, who erroneously placed it in
the true tentacular plane.
From the two opposed sides of the stomach a narrow median
canal leads into the two terminal end-lobes of the central
gastric cavity (cf. the schematic fig. 11). The two end-
lobes are in open communication with the peripheral canal-
system.
I do not find such definitely circumscribed peripheral canals
as those figured by Korotneff, but they appear to me in section
merely as the spaces partitioned off by the dorso-ventral
trabeculee, which Korotneff describes as dorso-ventral muscles
(fig. 5).
The median funnel-vessel was correctly figured by Korotneff.
It arises from the stomach immediately opposite to the
mouth, and, proceeding aborally, embraces the sense-organ
without opening to the exterior. It is very clearly shown in
section.
5. Tentacle Sheaths and Musculature.—The tentacle-
axis is occupied by the sheaths of the tentacles, which are
hollow tubes lined by ciliated cells lying immediately beneath
the dorsal surface, and completely separated from one another
by the aboral sensory complex. The muscles of the tentacles
form part of the voluminous musculature, which, so far as I
can make out, effects the retraction of the aboral sense-organ
and of the ctenophoral plates, which can be completely with-
drawn into the body (ef. fig. 5). The tentacles were retracted
in my preserved specimens, and so it was impossible for me to
ON OCTENOPLANA. 329
analyse this very complicated longitudinal musculature. The
musculature! on one side of the stomachal plane is completely
separated from that on the other side; so that in sections
parallel with the stomachal plane, passing through the region
of the sense-organ, no muscles are visible. But at a short
distance on either side of the median stomachal plane the
sections in contracted specimens are almost entirely occupied
by the convoluted bundles of muscles. Again, beyond the
region of the ctenophoral plates the sections merely show the
dorsally placed muscular tentacle lying in its sheath (fig. 10).
6. The Gonads.—By the discovery of the male genital
organs of Ctenoplana I have brought a welcome additional
piece of evidence as to the adult character of the organism.”
The testes are placed at the bases of the two end-lobes of the
main portion of the gastro-vascular system. Their position is
indicated by crosses in fig. 1. They thus consist of two pairs
of organs, paired about the tentacle axis. They may be either
simple or lobed and subdivided. They may contain practically
nothing but mature spermatozoa as in fig. 7; or they may
contain both mature and developing spermatozoa as in fig. 9.
Finally, they may possess one or several ducts opening to the
exterior on the dorso-lateral surface of the body below the
level of the ctenophoral plates.
The male genital ducts are merely tubular extensions of the
tunica propria which encloses each testis.
In the centre of that portion (always the ventral portion) of
the testis in which the immature sperm-cells (spermatogonia
and spermatocytes) occur there is usually to be observed a
cavity surrounded by large clear cells exactly like those which
line the cavities of the terminal gastric lobes ; and, in fact, I
have traced this cavity into communication with the gastro-
vascular system (cf. figs. 5, 6, and 9).
1 IT do not include the dermal musculature described by Korotueff, about
which I am at present in the dark.
2 Unfortunately I can say nothing about the female reproductive organs.
It seems unlikely that Ctenoplana should be unisexual. More probably it is
a protandric hermaphrodite.
330 ARTHUR WILLEY.
These central cavities in the immature testes may therefore
be called the genital ceca, and the genital products appear
to arise as proliferations of the walls of the ceca (figs. 5
and 6).
What is very puzzling is the fact that similar proliferations
occur on the walls of the terminal gastric lobes themselves
(fig. 5). On the dorsal walls of the terminal gastric lobes the
cells of these proliferations appear to assume the properties of
chloragogenous cells, and numerous yellowish refringent con-
cretions occur in and amongst them. Somewhat similar re-
fringent particles are to be seen in the cells of the subjacent
true endoderm. Finally, in connection with this subject I can
only mention the fact that as the median walls of the neigh-
bouring terminal lobes fuse together on nearing the median
canal which connects them with the stomach, the minute
cellules which compose the greater part of the proliferations
in question are replaced by long pyramidal cells which com-
pose a compact gland, having a radiating structure due to the
peculiar arrangement of the cells. I will call thisa gastric
gland, and hope that at some future date light may be
thrown upon its nature.
What distinguishes the genital proliferations from the
above-described gastric proliferations, apart from their
different topographical relations, is the fact that the nuclei of
the cellules of the former are of different sizes (fig. 9). The
larger nuclei I interpret as belonging to spermatogonia and
the smaller to spermatocytes. Unfortunately 1 am unable
to make out any nuclear structure in my preparations, although
they are otherwise well enough preserved.
The spermatozoa form dense clusters with characteristically
deeply stained heads and unstained tails. The tails are directed
both outwardly and mesially. When a testis contains only
mature spermatozoa there is no longer any trace of the genital
cecum (fig. 7).
With regard to the ducts in one individual, I counted no
less than twelve ducts, which were distributed equally between
the four testes. Of these ducts 1 was able to see the actual
ON OTENOPLANA. 331
opening to the exterior in six. In another individual I counted
seven ducts altogether.
It should be added that the above description of the male
gonads applies exclusively to C. Korotneffii, all three speci-
mens of which possessed them.
7. Axial Relations—Comparison of Ctenoplana with
Planaria and Ctenophora.—As already known from the
work of Korotneff, Ctenoplana agrees with the Ctenophora
in the possession of a main axis(Hauptachse) which connects
the aboral pole with the oral pole, the mouth with the sense-
organ, and that this main axis forms the line of junction of
the two principal planes—namely, the tentacular plane and
the stomachal plane.
Ctenoplana presents remarkable Planarian affinities in re-
spect of its dorso-ventrally flattened body, in the possession of
a definite dorsal surface, and a definite ventral or locomoter
surface, in its habit of creeping, and especially in its habit of
attaching itself to the surface-film of water. This enumeration,
to which may be added the partial ciliation of the ectoderm,
nearly exhausts the list of its strictly Planarian affinities.
Besides the coincidence of the main axis and principal planes
of Ctenoplana with those of the Ctenophora, the chief points
of affinity are the possession of two pinnate tentacles which are
each retractile within a sheath, the possession of the eight cteno-
phoral plates, and the presence of the median funnel vessel.
The two series of sensory tentacles placed on opposite sides
of the otolith, whose epithelium is directly continuous with
the epithelium of the cupule of the otolith, are directly com-
parable with the polar plates (Polplatten) as described by
Chun in the Ctenophores. In the first place they agree with
the latter in lying in the stomachal plane, in so far that they are
paired about the tentacle axis. This is the most important
point of agreement morphologically, but they also agree in
1 According to the remarkable observations of Chun, some Ctenophores
possess this power, effecting it by spreading out the wall of the stomach,—
sometimes, as in Lampetia Panceri, nearly everting the stomach as far as
to the origin of the peripheral vessels.
332 ARTHUR WILLEY.
some details. The relation to the otolith-bearing portion of
the sense-organ is identical in both cases. The polar plates of
Ctenophores are ciliated, as are the sensory tentacles of Cte-
noplana. Moreover in the Beroide, according to Chun, the
thickened margin of the polar plates does not form a simple
ridge, but is raised up into a series of lappets. This is a very
remarkable correspondence, and after my observation of the
double, paired character of the sensory tentacles of Cteno-
plana I think there can be no doubt that the latter are homo-
logous with the polar plates of Ctenophores.
The comparison of the gonads of Ctenoplana with those of
other forms is not such a simple matter. They agree with
those of the Ctenophora in being developed about the walls of
diverticula of the gastro-vascular system, and with those of the
Potyclades in being enveloped in a tunica propria. But they
differ from both in the possession of ducts opening directly to
the exterior. In the Ctenophora the genital products fall into
the meridional vessels, and are discharged through the mouth ;
while in the Polyclades, according to Lang, the tunice pro-
prize which envelop the innumerable testes open into a system
of intra-cellular genital capillaries which eventually convey the
sexual products to the vas deferens on each side, by which
they are ultimately led to the ventrally placed external genital
pore.
We now come to the critical consideration of the axial rela-
tions of Ctenoplana. The problem to be solved is the follow-
ing :—To what do the planes of the tentacles and of the -
stomach respectively correspond in bilateral animals? Does
the tentacle plane of Ctenoplana (‘Trichterebene of Ctenophores)
correspond to the sagittal plane of bilateral animals or to the
transverse plane ?
We shall find, if it has not already appeared evident in the
foregoing pages, that Ctenoplana unequivocally proves, as I
think, that the tentacle plane or funnel-plane of it and the
Ctenophores corresponds to the sagittal plane of bilateral
animals, and not to the transverse plane.
At present there exist two interpretations of the axial rela-
ON CTENOPLANA. 3993
tions of the Ctenophores, namely, that of Chun! and that of
Lang.? These may be briefly tabulated as follows.
According to Chun—
1. Tentacle or funnel-plane of Ctenophores = Sagittal plane of Bilateralia.
. Stomachal plane of Ctenophores = Transverse plane of Bilateralia.
3. Main axis of Ctenophores*® = Longitudinalaxis of Bilateralia.
[a
According to Lang—
1. Tentacle or funnel-plane of Ctenophores = Transverse plane of Polyclades.
. Stomachal plane of Ctenophores = Sagittal plane of Polyclades.
3. Main axis of Ctenophores becomes bent (geknicht) in Polyclades.
tw
From the above it will be seen that, as regards the tentacle and
stomachal planes, Lang’s interetation is the exact reverse of
that of Chun; and yet it is singular that there is no mention
of such a fundamental discrepancy in Lang’s monograph.
Selenka (quoted by Lang) held the view that the anterior
end of a Polyclade corresponded to the aboral pole of the
Ctenophore, and the posterior end of the former to the oral
pole of the latter. Lang says he himself formerly held this
view, but afterwards gave it up as being erroneous. It would,
however, necessarily follow if the main axis of the Ctenophores
corresponded to the long axis of Polyclades as stated by Chun.
The latter view, however, is irreconcilable with Chun’s own
identification of the tentacle plane of Ctenophores with the
sagittal plane of Bilateralia, and, in fact, it may be dismissed,
once for all, as erroneous.
Chun’s other homologies, however, in respect of the tentacle*
and stomachal planes are fully confirmed by the conditions
1 Carl Chun, ‘ Die Ctenophoren des Golfes von Neapel,’ 1880.
2 Arnold Lang, ‘ Die Polycladen des Golfes von Neapel,’ 1884.
3 Carl Chun, “Die Verwandtschaftsbeziehungen zwischen Wiirmern und
Ceelenteraten,” ‘ Biol. Centralblatt,’ Bd. ii, 1882-8. In this paper Chun
intimates that the main axis of Ctenophores becomes the long axis of Poly-
clades; but I cannot find out how he reconciles some of the views here
expressed with the previous statements as to the homologies of the planes
contained in his monograph.
4 Chun denominated the plane in which the tentacles of Ctenopkora lie
the Trichterebene, because there are no tentacles in the Beroide.
334 ARTHUR WILLEY.
observed in Ctenoplana. Chun was at first in doubt as to the
criterion by which to homologise the planes of Ctenophores
with those of Bilateralia, as the axes passing through these
planes in Ctenophora were equipolar (gleichpolig).
The way by which he finally arrived at the conclusion that
the tentacle plane corresponded to the sagittal plane was so
remarkable that I will give a free translation of his descrip-
tion.
“ Naturally,” says Chun, “ we must disregard all accidental
conditions of asymmetry by which one of the axes (Kreuz-
achsen) becomes inequipolar. For example, one seldom finds
a Cestus veneris in which the two band-like halves of the
body are equal in length. . . . Should, however, one of the
axes prove to be inequipolar in such a way that constantly an
essential organ-complex failed to develop on the one half of
the axis, then we should have a transition to bilateral symmetry
which would enable us to speak of a dorsal and a veutral sur-
face. . . . How surprised was I to find a larva which pre-
sented a remarkable axial disturbance in the funnel-plane [2. e.
the tentacle plane]! I give it the provisional name of Thoe
paradoxa, as I have not succeeded in associating it with cer-
tainty with any adult Ctenophore. It possesses, in fact, only
a single tentacle apparatus and tentacle [Fangfaden]. Only
in the course of the later development is a second tentacle
apparatus differentiated at the other pole of the axis, so that
the original disturbance becomes gradually levelled out.”
In Ctenoplana the tentacle axis and the stomachal axis
are equipolar ; but if we consider about which axis the paired
structures are situated, we are simply forced to acknowledge
that the plane of the tentacles corresponds to the sagittal
plane,—in other words, that the tentacle axis of Ctenoplana
and Ctenophora corresponds to the longitudinal axis of Bi-
lateralia.
Lang’s theory of the origin of Polyclades from Ctenophores
rests in the first instance on the assumption that the pinnate
tentacles of Ctenophora and Celoplana are homologous with
the sensory tentacles of Polyclades; and his above-quoted in-
ON CTENOPLANA. 335
terpretation of the axial relations is framed in accordance with
this assumption.
In the first place the fact should be emphasised that under
no circumstances and from no point of view are the tentacles
of Ctenoplana bilaterally disposed, but they are biradially
disposed.
As mentioned above, it cannot be denied that, in the creeping
attitude, the tentacles of Ctenoplana present to the onlooker
the appearance of ordinary transversely paired structures, and
it may seem difficult to imagine an ancestor of bilateral
animals with an unpaired tentacle in front and an unpaired
tentacle behind. But the point is that we have not got to
imagine this, because in the animals with which we are dealing
there are no such relations as anterior and posterior, right and
left.
As regards the particular homology of the pinnate tentacles
(Greiftentakel) of Ctenophores with the nuchal tentacles of
Polyclades, so strongly and, it must be added, plausibly upheld
by Lang, I venture to think that my observations on Cteno-
plana, especially as to the double character of the aboral
circlet of sensory tentacles, justifies me in frankly denying its
accuracy. From their close relation to the central sensory
apparatus, and the fact that they are paired about the tentacle
axis, which I regard as equivalent to the longitudinal axis of
Polyclades, I suggest that it is much more probable, from their
relations and function, that the paired multiple sensory ten-
tacles of Ctenoplana and the polar plates of Ctenophora are
homologous with the sensory nuchal tentacles of Polyclades,
than that the latter are homologous with the pinnate tentacles
of Ctenophora, whose chief function is that of seizing objects
for food.
Moreover, from their structure and function, their extreme
retractility within definite sheaths, and their worm.like
mobility, it would appear that the pinnate tentacles of Cteno-
plana and Ctenophora belong to a category of structures
totally different from that of the nuchal tentacles of Poly-
clades. They belong, namely, to the same category of struc-
336 ARTHUR WILLEY.
tures as the proboscis of Nemertines and of certain Rhabdocele
Planariauns.
Finally, we may recapitulate the organs in Ctenoplana which
are paired about the tentacle axis (cf. fig. 1). There are four
pairs of ctenophoral plates, three pairs of gastric lobes, two
pairs of gonads with their ducts, and one pair of multiple
sensory tentacles.
From what has been said I regard it as proved that—
1. The tentacle axis of Ctenoplana = the longitudinal axis of Planarians.
2. The stomachal axis of Ctenoplana = the transverse axis of Planarians.
3. The main axis of Ctenoplana = according to Lang, the primary main
axis of Planarians, which becomes bent as the ganglion is shifted
towards the anterior end of the body.
4. The main axis of Ctenoplana and Ctenophores = the dorso-ventral axis
of Bilateralia.
8. Synopsis of Species of Ctenoplana.
(i) C. Kowalevskii, Korotneff—Colour crimson, body in
swimming attitude shaped like a truncated pyramid ; median
dorsal surface concave ; free margin of skirt frilled. Habitat,
west coast of Sumatra.
(ii) C. rosacea, n. sp.—Colour crimson; body in swim-
ming attitude of a quadrilateral form; median dorsal surface
convex; free margin of skirt plain. May be merely a variety
of preceding species. Habitat, Eastern Archipelago of New
Guinea.
(iii) C. Korotneffi, n. sp.—Colour green; body in swim-
ming attitude roof-shaped; median dorsal surface upraised
with two upright end-knobs; free margin of skirt slightly
frilled. Habitat, Hastern Archipelago of New Guinea.
In Korotneff’s fig. 13 there are only eleven sensory ten-
tacles figured for C. Kowalevskii,! while C. rosacea had
1 Korotneff describes an aperture to the exterior on each side in the
neighbourhood of the tentacles in C. Kowalevskii which leads into a
“system of canals which branch in the body parenchyma.” Korotneff him-
self says that after he had preserved his specimen in sublimate he was unable
properly to orientate it. I have already shown that he completely misplaced
the tentacles. The apertures in question are obviously the openings of the
ON CTENOPLANA. aol
about eighteen. But I should hesitate to insist upon this as a
specific difference.
9. General Conclusions.—With regard to the systematic
position of Ctenoplana, which we now know to be an adult
animal, I am strongly of opinion that it is an ancestral form,
and not, as some zoologists seem to suppose, a highly modified
ereeping Ctenophore. By ancestral form I simply mean a
primitive archaic form belonging to an ancestral type, and of
course I do not imply that it is the actual ancestor of any-
thing in the world. That the Planarians and Polyclades in
particular have close affinities with the Ctenophora there can
be no doubt, but it is very much open to question whether
the former are derived from the latter. The view that the
Polyclades are so derived seems a reversal of the natural order
of events, which point to the littoral fauna as the origin both
of the pelagic and of the abyssal fauna.
Are we to regard the immediate ancestors of the Turbellaria
as amorphous forms, like Trichoplax, or forms without any
kind of symmetry, like Planule or the Mesozoa? Or, on the
contrary, are we not rather to regard their immediate ancestors
as forms with some kind of radial symmetry ?
Having regard to the complete bilateral symmetry of the
flat-worms, and more particularly their well-developed nervous
system, with cerebral ganglion in even the lowest forms, I
cannot imagine them to be derived directly from amorphous
organisms, but rather from animals which possibly, like Cteno-
plana, possessed a biradial symmetry.
Ctenoplana approaches more nearly to a condition of bilateral
symmetry than the Ctenophores do, in that it possesses very
clearly differentiated dorsal and ventral surfaces. And this is
exactly what we should expect to find in the littoral or sub-
littoral ancestor of such purely pelagic forms as the Cteno-
phora, the pelagic habit, as is well known, often tending to
produce a more or less radial symmetry.
tentacle sheaths, and the “system of canals” are the sheaths themselves,
which do in fact send off occasional diverticula, possibly due to the contraction
of their walls during preservation (ef. figs, 5 and 10).
338 ARTHUR WILLEY.
On the other hand, a biradial form, like Ctenoplana,
possesses the potentiality of assuming a strictly littoral
life, in which the ventral surface is the permanent locomotor
surface, such an existence leading to a condition of bilateral
symmetry, according to well-understood physiological prin-
ciples.
The ctenophoral plates must have put in their appearance
for the first time in some form or other; and although it is at
present beyond the limits of our knowledge to explain how
they arose, yet it is not right to conclude that the ctenophoral
plates of Ctenoplana are degenerate or reduced structures
merely because they are smaller than the ctenophoral rows of
the Ctenophora.
It is a groundless assumption to say that Cceloplana and
Ctenoplana are modified creeping Ctenophores. Ctenoplana
is an expert crawler, it is expert at hanging on to the surface
film of water, and it is indeed an expert swimmer. Everything
it attempts it does well in the old primeval fashion, and there
is nothing degenerate about it.
If Celoplana and Ctenoplana are neither Ctenophores nor
Planarians, what are they? I think it is necessary to create
a new order of Plathelminthes for their reception; and I
propose to call the new order the Archiplanoidea, and to
regard it as equivalent to the orders Turbellaria, Trematoda,
Cestoda, and Nemertina.
Furthermore, I should look to the Archiplanoidea for the
ancestors of all the Plathelminthes (including the Nemer-
tines) on the one hand, and of the Ctenophora on the
other.
The resemblance in form and shape, however superficial,
between Ctenoplana and the Pilidium larva of Nemertines
should not pass unnoticed; and it is a remarkable fact that
the main axis of Pilidium passes through mouth and apical
sense-organ as in Ctenoplana.
In the Archiplanoidea, therefore, we have organisms pre-
senting a transition from radial to bilateral symmetry.
In the Cerianthide, as we know especially from the works
ON CTENOPLANA. 339
of Carl Vogt! and Ed. van Beneden,? we have also forms
belonging to the category of radial animals, which are un-
doubtedly physiologically radial, and nevertheless present a
pronounced bilateral symmetry.
Van Beneden’s views are clearly set forth in the following
quotation from his memoir above quoted:—‘‘ Je partage
entiérement l’opinion de Sedgwick et de Caldwell d’aprés
laquelle le disque qui porte la bouche et les tentacles, chez
les Actinozoaires, répond morphologiquement a la face neu-
rale des Annéles, des Arthropodes, et des Chordés. Je pense,
comme ces auteurs, que la bouche des Cnidaires est homo-
logue 4 la fente blastoporique des Arthrozoaires. Les diver-
ticules coelomiques qui sont, ontogéniquement parlant, la
cause de la segmentation, répondent aux loges mésenteriques
des Anthozoaires et les cloisons intersegmentaires sont ana-
tomiquement équivalentes aux sarcoseptes.”
If we accept these conclusions side by side with those
derived from the study of Ctenoplana, we are compelled to
frame the hypothesis, which I believe to be highly probable,
of the diphyletic origin of Bilateralia.
The following scheme will make this view clear, and will
save a long discussion :
TC tenophora.
_A rchiplanoidea————__
ian __——Plathelminthes.
Sheet Sn ee OO
Cer Se ks
ee —— CL lenin
I believe this view will be found to be a natural one in every
respect ; and if it be regarded by morphologists as substantiated,
it will certainly relieve the science of morphology of several
burdens. For instance, Hubrecht’s original speculations as to
a relationship between the Nemertines and the Chordates, as
well as Bateson’s comparison of the Nemertines with Balano-
1 Carl Vogt, “ Des Genres Arachnactis et Cerianthus, f. Arch. de.Bioli
t. ve 1888.
* Hd. van Beneden, ‘“‘ Recherches sur le développement des Arachnactis,”
ibid., t. xi, 1891.
340 ARTHUR WILLEY.
glossus and many other such like theories, which at the time
no doubt appeared to be logical necessities, will be quite ruled
out of the field of possibilities.
The descendants of the Archiplanoidea have no ccelom and
no preoral lobe. The descendants of the Cerianthide have a
celom and also a proral lobe (excluding the Anthozoa).
It is an interesting parallelism that the criterion for the
antero-posterior axis, both in the Cerianthidze and in the
Ctenophora (Thée paradoxa), was provided by what may be
called a directive tentacle. It may, indeed, be something
more than a mere parallelism.
10. Summary of Principal Results.—(1) Discovery
of one very distinct new species of Ctenoplana, and of another
somewhat doubtful new species.
(2) Observation of the movements and of the pinnate tenta-
cles of the living Ctenoplana.
(3) Accurate localisation of the pinnate tentacles.
(4) Discovery of the double character of the circlet of sensory
tentacles surrounding the otolith.
(5) Discovery of the male genital organs and ducts of Cteno-
plana, thus proving that Ctenoplana is an adult animal.
(6) Description of the tentacle sheaths.
(7) Account of the genital ceca of the gastro-vascular
system, about whose walls occur the genital proliferations.
(8) Chloragogenous tissue and gastric gland.
(9) The tentacle axis of Ctenoplana corresponds to the longi-
tudinal axis of Planarians, the stomachal axis of the former to
the transverse axis of the latter, and the main axis of the
former to the dorso-ventral axis of the latter.
(10) The solid pinnate tentacles of Ctenoplana are not dis-
posed bilaterally, but biradially.
(11) The ctenophoral plates, gastric lobes, gonads, gonaducts,
and aboral sensory tentacles are paired about the tentacle
axis,
(12) The aboral sensory tentacles of Ctenoplana are homo-
logous with the polar plates (Polplatten) of Ctenophora, and
with the nuchal tentacles of Polyclades.
ON CTENOPLLANA. 341
(13) The testes of Ctenoplana are enclosed within a tunica
propria.
(14) The pinnate seizing tentacles (Greiftentakel) of Cteno-
plana and Ctenophora, retractile within definite sheaths, belong
to the same category of structures to which the proboscis of
Nemertines and certain Rhabdoccele Planarians belong.
(15) Creation of a new order, the Archiplanoidea, for the
reception of Cceloplana and Ctenoplana. The order will thus
contain two families, the Celoplanide and the Cteno-
planide.
(16) Hypothesis of the diphyletic origin of Bilateralia.
In conclusion, as I am about to leave Sydney, I wish to
repeat my thanks to Professor W. A. Haswell for his kindness
and hospitality to me.
SypNEY; June 29th, 1896.
EXPLANATION OF PLATE 21,
Illustrating Mr. Arthur Willey’s paper “‘ On Ctenoplana.”
N.B.—My material was preserved in sublimate and in a sublimate-acetic
mixture, and stained with alum-cochineal.
Fic. 1.—Ctenoplana rosacea, n. sp., from dorsal aspect. The two
pinnate tentacles are extended, the ctenophoral plates alternate with the
gastric lobes, the two middle large gastric lobes mark the stomachal plane ;
in the dorsal centrum is seen the otolith, surrounded by an incomplete circlet
of ciliated sensory tentacles paired about the tentacle axis; the spots round
the margin of the skirt represent crimson pigment spots. From living speci-
men. N.B.—The crosses indicate the positions of the male genital organs.
Fie. 2.—C. rosacea. In swimming attitude. From living specimen.
Tentacles retracted. ¢.0. Opening of tentacle sheath.
Fic. 8.—C. Korotneffi, n. sp. In swimming attitude. The space indi.
cated by dotted line below the aboral sense-organ was dimly seen through
the body, and probably represents the space into which the sense-organ is
withdrawn on retraction, although it might represent the funnel vessel.
VOL. 39, PART 3,—NEW SER. Z
342 ARTHUR WILLEY.
From living specimen. Tentacles retracted. ¢.0. Opening of tentacle
sheath.
Fic. 4.—Sketch of Ctenoplana as it may be seen when attached to the
surface-film of water. In the centre is seen the mouth. The two end-lobes
of the central gastric system show up white.
Fie. 5.—C. Korotneffi. Section parallel to stomachal plane (i.e. trans-
verse to tentacle axis) to show the origin of the genital cecum, &c. The
section is somewhat oblique. cé/. Ciliated epithelium of ventral surface.
ce. Genital cecum. ch. Chloragogenous cells. c¢.p. Ctenophoral plate, re-
tracted. d.e. Dorsal spongy vacuolar non-ciliated epithelium, with mucous
granules at external surface. These cells would seem to be comparable to
the so-called “ Glanzzellen” of Ctenophora. ext. Colenteron. ep. Diges-
tive epithelium, nuclei placed near free end of cells. These are clear faintly
staining cells with indistinct cell outlines, and with a sharply defined non-ciliated
free margin. gez. Genital proliferation on the wall of the genital cecum.
g.p. Gastric proliferation. mes. Mesenchymatous tissue. ¢.0. Opening of
tentacle sheath. ¢.s. Tentacle sheath with its ciliated epithelium; the
tentacle itself is retracted further back. v.e. Non-ciliated glandular epithe-
lium of the ventral surface.
Fic. 6.—The same. Succeeding section to preceding through gonad to
show the conversion of the genital cecum into a canal. ce. Genital cecum.
s. Genital proliferation broken up into polygonal groups of sperm mother-
cells.
Fie. 7.—C. Korotneffi. Section through the region of the gonad of
another individual to show the testis full of mature spermatozoa and the
opening of the genital duct. d. Tangential section of portion of duct. d.e.
Dorsal epithelium. g.o. Genitalaperture. ¢.p. Tunica propria with flattened
nuclei in its walls.
Fig. 8.—The same. Succeeding section to preceding to show junction of
genital duct with tunica propria.
Fic. 9.—Section through a gonad of same individual as that from which
fig. 5 was taken, to show subdivision of the testis. c@. Genital cecum.
d. Genital duct. s'. Spermatogonia. s*. Spermatocytes. s*. Spermatidia.
s4. Spermatozoa. ¢.p. Tunica propria.
Fie. 10.—Section through tentacle and its sheath. c.s. Cavity of sheath.
s. Wall of sheath composed of ciliated epithelium. ¢. Tentacle with peri-
pheral nuclei and central muscle-fibres. In the centre of the tentacle runs a
core of mesenchyme-cells, mes. p. Branch or pinna of the tentacle.
Fie. 11.—Schematic figure to explain connections of the lobes of the
central portion of the gastro-vascular system. /. Position of funnel vessel.
ATTEMPT TO DEDUCE VERTEBRATE EYES FROM THE SKIN. 343
An Attempt to deduce the Vertebrate Eyes
from the Skin.
By
H. M. Bernard, M.A.Cantab., F.L.S.
With Plate 22.
In the February issue of the ‘ Annals and Magazine of
Natural History’ I published an outline sketch of a new
theory of vision. It was suggested that eyes arose as local
modifications of tissue induced by the excessive crowding of
pigmented granules at spots most frequently and brilliantly
illuminated, and that this crowding of the pigmented granules
might be made to explain both the origin of the eye as a
structure and its functions as an organ. I now propose, as
time and opportunity allow, to compare a few eyes with the
tissues out of which they have, or according to the theory they
should have. arisen, in order to ascertain how far such com-
parisons support the theory.
I take the Vertebrata first, not only because the general
facts of the structure and development of the eyes in this class
are most widely known, but because of the paramount interest
attaching to any questions relating to our own highest sense.
Now, according to the theory alluded to, eyes must have
developed out of that tissue in which, under the action of
light, pigment accumulates. That tissue is the skin; hence
the eyes in the Vertebrata must have arisen as modifications of
the skin. There are two kinds of eyes in the Vertebrata; the
more or less vestigial ‘ pineal” eye, and the Vertebrate eye
344 H. M. BERNARD.
proper. These eyes are constructed on different plans, and
must either, according to the theory, be different modifications
of the same kind of skin, or else modifications of two different
types of skin. I shall endeavour to show that the latter
assumption best accords with the facts ; that the change which
undoubtedly took place in the character of the skin as the
Vertebrata arose out of their Invetebrate ancestors is sufficient
to explain the difference between these two types of eye.
Following the lines of the theory, I make two assumptions :—
(1) The retina is but a specialised portion of the epithelial
layer of the skin, between the cells of which the pigment
granules from the subjacent chromatophoral layer stream
outwards under the action of light. (2) The retina and the
chromatophoral layer must have been in intimate and insepar-
able association through all the stages of the evolution of the
eye. I have, then, to try to show how far the facts relating
first to the structure, and secondly to the development of the
Vertebrate eyes can be harmonised with these assumptions ; how
far, indeed, it is possible to deduce the Vertebrate eyes directly
and continuously from the skin.
A glance at the diagrams (Plate 22) shows that a striking
parallel between the eyes and the skins, out of which I assume
them to have arisen, can be easily instituted, but the details
are by no means easy to work out, and still more difficult to
demonstrate. The greatest difficulty, of course, lies in the
fact that, according to their ontogenetic histories, the eyes
either wholly or partly developed from the brain. It is true
that no one has yet succeeded in elucidating these embryological
records ; nevertheless they hold the field, even though amount-
ing to little more than bald assertions that the eyes developed
from the brain and not from the skin. Before critically exa-
mining this adverse testimony I propose to marshal all the
available facts which seem to connect the eyes directly with
the skin, and to suggest possible explanations of the various
modifications and specialisations of the tissues of which our
eyes are constructed. Only after having shown that it is
possible to deduce the eyes as continuous structural modifica-
ATTEMPT TO DEDUCE VERTEBRATE EYES FROM THE SKIN. 349
‘tions of the skin shall we examine the embryological records,
and endeavour to show that the facts are, in the main, in har-
mony with the argument based upon structure.
I take the pineal eye first, in spite of the disadvantage
involved in this choice, arising out of the many debatable
details concerning this organ. Its claim even to be an organ
of sight at all has been disputed. Assuming it, for the present,
to have been an eye, we are, it seems to me, justified for many
reasons, which will be more apparent later on, in regarding it
as having preceded in course of time the Vertebrate eyes
proper. Its place in this discussion is, therefore, that which
is suggested by its history as here interpreted.
The close resemblance of the pineal organ to an eye of the
Invertebrate (Molluscan) type was first pointed out by de
Graaf.| The suggestion has not been accepted, for reasons
chiefly embryological, nevertheless I am convinced that it is
correct. If the pineal organ ever was an eye—and all the
evidence points that way,—it should, according to our theory,
have developed directly from the skin. With this hypothesis
it entirely agrees, its structural relationships being precisely
those which they should be had it arisen simply as an invagina-
tion of a skin of the Invertebrate type, i.e. of a skin consisting
of an external palisade layer supported internally by a layer of
connective tissue (Diagram1). Regarding its structure alone,
then, and this is the only point that at present concerns us,
there is no difficulty in believing that the pineal eye deve-
loped first as an optic pit from the skin of the ancestors of the
Vertebrata, before that skin had assumed the Vertebrate type,
i.e. before the palisade layer had become protected externally
by the mucous and horny layers.
Dealing with the development of the pineal eye more in de-
tail,—according to the theory, the retina arose by a secondary
multiplication of the epithelial sensory cells for the apprecia-
tion of the variations in intensity of irritation caused by the
movements of the pigmented granules towards the exterior ;
while the dioptric apparatus was due to the accumulation
1 ¢ Zool, Anz.,’ 1886.
346 H. M. BERNARD.
at the surface of slime wholly or partially produced by the
excessive discharge of the pigmented granules. This refrac-
tive mass, at first probably lying merely at the surface, would
eventually be enclosed in an invagination, a phenomenon so
common (e.g. in the formation of glands) that a discussion
of the mechanics by which it might be brought about need not
here detain us.
The formation of a lens, the definite structure of which
is now perhaps only indicated by the lengthening of the pali-
sade cells as shown in the diagram, brought the pineal eye as
an organ of sight to a high functional level. From this it has
steadily declined, until it persists as a vestige which, in many
cases, can no longer function in any way for the appreciation
of variations of light intensity.
In support of this interpretation of the morphology and
physiological origin of the pineal eye, I would call attention
to Spencer’s figures,! a study of which leaves no doubt what-
ever that the pigmented cells are streaming from the connective-
tissue capsule round the eye through the retina, just as in the
rest of the skin they stream out from the cutis through the
palisade layer of the epidermis. In the series of sections of
Hatteria, kindly lent me by my friend Mr. Martin Wood-
ward, in addition to the pigmented cells, there are others
making their way alike through the retina and the palisade
layer of the epidermis. These are cells containing enormous
vacuoles which force the nuclei to one side. On reaching the
horny layers these cells flatten out, and their vacuoles form
flat spaces which give the cuticle a sort of false lamination.
A still closer parallel between the pineal eye and the skin is
observed in Uromastix, for the loan of sections of which I
am again indebted to Mr. Woodward. Streaming through the
cutis are innumerable cells laden with concretions which are
black by transmitted, white by reflected light. They remind
one of the guanin granules found in the tissues of Arachnids.
Now, while the cells containing these white concretions do not
seem to be able to pass through the palisade layer, but stop
1 «Quart. Journ. Mier. Sci.,’ xxvii.
ATTEMPY! TO DEDUCE VERTEBRATE EYES FROM THE SKIN, 347
short within the cutis beneath it, the cells which contain the
red pigmented granules pass freely on, and are seen everywhere
forcing their way into the epidermis, and, I believe, helping to
build up the horny layer.
On turning to the pineal eye, what do we find? A close
network of cells filled with similar white matter enveloping
the eye externally. The white matter here also fails to climb
up into the retinal cells, just as in the skin it fails to climb up
into the epidermis, while on the other hand passing up between
these cells are others laden with red pigment, which makes its
way into or between the retinal cells, just as the pigment from
the cutis passes between and into the palisade cells of the
epidermis.!
In addition to the pineal eyes of Hatteria and Uromastix,
I have examined this organ in Petromyzon planeri, ammo:
coetes of which I obtained near Jena and fixed in corrosive
sublimate. As is now well known, there is no black pigment
in these eyes; they appear to me to have become a receptacle
for concretions similar to those already noted in the cutis of
Uromastix.? The condition of the eye itself appears to indi-
cate extreme degeneracy. Having sunk below the skin it
1 Comparison of the pineal eye and skin of Uromastix reveals another
fact of some importance for the theory of light sensation which I have
ventured to put forward. The pigment granules are in this case much
larger and more clearly defined in the eye than in the skin. It is a noticeable
fact that in eyes in general the pigment grains are much more sharply and
distinctly granular, and also apparently not seldom rather larger than are the
pigment granules in the rest of the body. ‘This fact is certainly in favour of
my suggestion that their passage up and down between the rods of the
retina causes a mechanical stimulation.
2 [ therefore doubt the wisdom of Gaskell’s description of this matter as
“white pigment.” It is worth noting also that the white concretions,
which have no definite shape in the cutis, are pointed and fusiform in the cells
round the eye in Uromastix. ‘The guanin granules have the same or a
very similar shape in the argentea of fishes. I have seen similar white
long-oval plates in the eyes of spiders. These facts suggest the possibility
of associating these last-named granules with the guanin “crystals” found
in such quantities in the bodies of these as of other Arachnids. (On their
origin cf. ‘Journ. Roy. Micr. Soc.,’ 1893, p. 427.)
348 H. M. BERNARD.
seems no longer in a position to have much pigment pass out-
wards through it, and, if any travels through, it probably passes
right through without being even temporarily arrested, pre-
sumably because the resistance of the retina to the passage
of the pigment which in the functional eye is, according to
the theory, the cause of the irritation and resulting sensation
has ceased. In the same way the resistance of the retina to
the entrance of the white matter observed in Uromastix has
also ceased, so that it streams outwards through the cells in
single shapeless or angular granules [which are of very differ-
ent sizes, and have no resemblance to true pigment granules
either in appearance or distribution], to accumulate at the distal
ends of the degenerate retinal cells. Beyond this point they
seem unable to travel. I know of no evidence which would
lead me to believe that the coagulum in the cavity of the eye
is due to any transformation of these granules; on the other
hand, it is quite possible that it may result from the occasional
passage and conversion into slime of true pigment.
With regard to Leydig’s doubts! as to whether the pineal
eye was ever an eye at all, his objections are based upon the
very facts which, it seems to me, establish the point beyond
dispute. He appeals to the presence of pigment in the lens.
But this is exactly what we might expect in an eye going out
of function. The pigmented cells ball together as they reach
the free spaces among the retinal cells outside the palisade
layer; then streaming through the retina they ball again on
reaching the cavity of the eye, and may accumulate as globular
masses against and within the lens.2 For some reason or
other, as the eye ceased to function, the pigmented granules
ceased to clarify on reaching the cavity of the eye, as they
appear to do, say, in the eyes of molluscs, of Petromyzon
(see below), on passing into the vitreous humour through the
1 © Abh. Senk. Nat. Gesch.,’ xvi, 1890, p. 531.
2 In some cases the pigment seems (in Spencer’s figures, |. c.) to be forcing
its way between the lens cells; this we should imagine would be the most
natural. It is not easy to understand why, in other cases, the cells should
ball together in the heart of the lens.
ATTEMPT TO DEDUCE VERTEBRATE KEYES FROM THE SKIN. 349
cells in the pars ciliaris retine of Vertebrate eyes (see below),
and as they certainly do in the rete mucosa of the Vertebrate
skin. That they formerly clarified in the cavity of the pineal
eye also we have some evidence in another of the facts which
Leydig adduces as a reason for disbelieving the original ocular |
function of these structures. The remains of the homogeneous
clear substance which I assume once filled the cavity of the
eye still here and there persists, as Leydig himself has shown.
It appears in some cases as bristle-like streaks or threads of
clear matter streaming outwards from the surface of the retina,
or even in a thick layer like a cuticle. That this substance
ever formed a definite system of rods turned towards the
cavity of the eye, such as has been suggested by Gaskell,! I
think somewhat doubtful, although it is quite possible that
here and there some such differentiation may have taken
place (cf. the “ rods ” of the Cephalopod eye).
Summing up this brief sketch of the pineal eye, we note
that structurally (its embryology will be discussed later
on) it is quite explicable as a simple invagination of a skin
of the Invertebrate type, and may well have arisen in the
manner suggested by our theory, while our ancestors still
possessed such an undifferentiated epidermis.
An eye arising in this way would necessarily be what is
known as a direct vision eye,—that is, the nerves would end in
the retina without any bending back upon themselves.
The Vertebrate Eye Proper.—In process of time the
skin lost the simple character it possessed when, according to
the foregoing, the pineal eye arose. Cells budded off from
the palisade layers, while others (e. g. pigment-bearing cells)
migrated through the palisade layer, and these together, the
1 *Quart. Journ. Micr. Sci.,’ vol. xxxi. At the same time it is obvious
that the interpretation of the facts here adopted is hardly reconcilable with
the deduction of Vertebrates from the Arthropods; a chitinous exoskeleton
does not lend itself to such simple invagination as we have here assumed.
I do not think the depression in the centre of the retina of the pineal eye of
Petromyzon planeri at all justifies Gaskell’s comparison of that eye with
the eye of the Acilius larva. A comparison with other pineal eyes shows that
that depression is hardly primitive ; it may even be a result of degeneration.
390 H. M. BERNARD.
former being the more important element, built up a layer of
cells, the outermost of which typically harden into horny
scales. We thus get the Vertebrate type of skin as shown to
right and left of Diagram II. The cutis is richly provided with
blood-vessels (shown diagrammatically as loops), the chroma-
tophores are numerous, and those which have reached the
epidermis are seen forcing their way up between the palisade
cells. Lastly, the ultimate ramifications of the integumentary
nerves no longer end, as in the primitive skin, in or among
the palisade cells, but, penetrating that layer, terminate
among the cells of the rete mucosa.
In this change in the character of the skin we can perhaps
find, on the one hand, a partial explanation of the degenera-
tion of the pineal eye, and, on the other, a clue to the chief
differences between that eye and the Vertebrate eye proper.
The pineal eye, sunk below the palisade layer and no longer
in organic connection with it, would probably suffer by any
change such as that described in the character of the outer
skin. The separation of the ocular vesicle from the palisade
layer would hamper its control over the development of that
layer, which might or might not run a course favourable to
the eye as an organ of vision. As far as we can see, the
secondary thickening of the outermost. layers by stratification
from the palisade layer was not calculated to benefit the pineal
eye. It would impede its function, and therefore cause it to
degenerate. A still more important factor making for degene-
ration is probably to be found in the rise and development of
the skull. A review of the available facts relating to these
eyes involuntarily suggests that they—there were, it seems,
originally two!'—succumbed, first one and then the other,
before the advancing edges of the bony plates which developed
to protect the ever-enlarging brain. This degeneration would
be hastened if new and perhaps more efficient eyes developed
to replace and more than compensate for the loss of the old.
The eyes which appear actually to have replaced the pineal
1 Owsianikow, ‘Mem. Akad. St. Petersburg,’ 7, xxxvi, 1888; also Locy,
‘ Anat. Anz.,’ 1894, p. 169.
ATTEMPT TO DEDUCE VERTEBRATE EYES FROM THE SKIN. 301
eye (or eyes) would again, according to our theory, have
developed out of the skin; but, inasmuch as the character of
the skin had changed, the type of eye would differ from that
of the pineal eye. The same physiological principle would,
however, come into play; the cells of the palisade layer
would be irritated by the variation in the pressure of the
pigmented matter travelling between them under the action
of light, and these cells could become sensory cells by asso-
ciation with the nerve-endings now found among the cells of
the rete mucosa. The new eye might thus from the first be, as
indicated in Diagram II, an inverse eye, i.e. the nerve-fibres
(n), passing outwards towards the surface, would have to bend
back on themselves in order to become associated with the
palisade-retinal cells.
The gradually thickening nerve-strands of this hypothetical
epidermal eye may be assumed to have ultimately passed in a
group through the palisade layer at the (? posterior) edge of
the new specialised sensory area (see diagram). This ever-
thickening and broadening nerve-strand appears to have carried
through with it one or more loops of the vascular system of
the cutis, together with strands of connective tissue.
The specialisation of a portion of the palisade layer to form
the retina of a purely epidermal eye such as that figured in
the diagram would, in course of time, prevent that portion
from yielding any mucus or horn cells towards the exterior,
more especially as the palisade cells of the retina must be
supposed to have early specialised into cuticular rods. The
deficiency caused by this more or less effective barrier to the
free passage of pigment granules through the retina towards
the surface would have to be made up somehow, not only
to form a dioptric apparatus, but also for the protection of
the retina. It was along time before I could see how the
palisade cells surrounding the retina could cover it over with
a horny layer. The solution to the problem suggested in the
diagram was arrived at by working backwards from the lens
of the definitive eye (cf. Diagrams IV, III, II). The palisade
cells round the periphery of the primitive retina appear to
ope H. M. BERNARD.
have lengthened greatly, and bent over the specialised area
somewhat in the manner indicated. These long cells would
therefore be merely a modification of the ordinary prickle-
cells. By the detachment and death of their distal portions
they would be able to yield a succession of horny cells, so as
to maintain a continuous hard covering over the retina. This
covering appears eventually to have formed a primitive surface
lens, moved by fibrils from the intrusive connective tissue.
The next stage in the development of this hypothetical
epidermal eye was probably an invagination of the whole of
the modified portion of the skin, including the lengthened
palisade or lens cells. This invagination is shown in Diagram
III. Two factors may have helped to bring it about. 1. The
retina becoming bulged inwards by the great accumulation of
slimy fluid in the cavity of the eye, the tendency of the
surrounding skin would be to constrict off the vesicle so
formed. 2. The lens being attached by fibrils to the place of
entrance of the optic nerve, and probably slightly moveable,
would tend to be drawn in with the retina. Any way, if the
eye started at all as we have suggested, we are justified in
assuming an invagination such as that described, for that alone
appears able to explain the arrangement of the tissues, among
which I would call attention to the loops of the vascular system
of the cutis, which, diagrammatically represented, assume the
positions. shown by faint dotted lines in Diagram III.!
Further, the remains of the neck of such an invagination may
still perhaps be seen in the fibrous ring which persists after
the formation of the aqueous chamber as the ligamentum
annulare (fishes) or ligamentum pectinatum (other Vertebrates),
and which, it is important to notice, connects the iris (and not
the lens) with the cornea (lig. an., Diagram IV).
If we picture to ourselves the change of shape which the
spoon-shaped retina shown in Diagram II would have to
undergo in passing to the form shown in Diagram III, the
origin of the choroidal fissure is at once apparent. As the
1 In this I am following the familiar diagrams of the blood-vessels of the
Vertebrate eye to be found in text-books.
ATTEMPT TO DEDUCE VERTEBRATE EYES FROM THE SKIN. 353
retina bulged further and further inwards, the nerve-strand
connecting it with the brain shifted down towards the base of
the cup. It is quite indifferent whether we speak of the nerve-
strand cutting down the sides of the ever-deepening retinal
cup, or of the sides of the cup growing up on each side of the
nerve. The vascular and connective-tissue elements which
accompanied the nerve (see Diagram II) would now protrude
into the cavity of the eye. Inthe completely invaginated eye
(shown in Diagram III) it is obvious that the choroidal fissure
could originally only have extended from the edge of the long
palisade or lens cells to the entrance of the optic nerve.
In Diagram IV I have indicated some of the adaptations
which would be necessary to complete the Vertebrate eye. The
lens had to be isolated, and its long cells rolled under it from
all sides. We need not assume that the earliest isolated lens
had its fibres arranged in the highly specialised manner
characteristic of the higher Vertebrates; any folding of the
long palisade cells under the more hardened central portion
would be sufficient. The freedom of the lens and folding
under of its long cells could, it seems to me, be brought about
if the lens itself, under the action of contractile fibres in-
truding with the optic nerve into the cavity of the eye, was so
far moveable that its axis deviated through small angles from
the optic axis of the eye.
With regard to this method of accounting for the isolation
of the lens, the following considerations are worth notice.
The eye itself would have been immoveable through all the
earliest stages of its evolution, but the Vertebrate eye is, as
we know, a highly moveable structure. We are justified in
concluding that its present powers of movement are but more
perfect ways of attaining ends which, in its earliest stages, it
must have striven to attain. It is not unlikely, therefore, that
the connective-tissue strands which entered the eye in its
earliest stages became attached to the edges of the primitive
surface lens, and effected some simple movements. As the
retina bulged inwards, these simple lens movements might be
supplemented by slight movements of the whole eye,—these
354 H. M. BERNARD.
double movements probably playing some part in bringing
about the invagination depicted in Diagram III, an invagina-
tion which, as above insisted upon, must have included the
lens. Whether the eyeball formed by this invagination ever
moved with any degree of freedom under the cornea may be
doubted ; the fibrous connection persisting in the ligamentum
annulare was almost certainly strong, and may well have been
rigid from the first. Certain it is that the movements of the
eye asa whole, i.e. of the ocular invagination together with
the skin above it (cornea), not only gave rise to the conjunctival
folds by which the eye is suspended in the skin, but also ren-
dered the directive adjustment of the lens by contractile fibrils
unnecessary. As is well known, fibrils for the movement of
the lens (“ m. retractor lentis ”—Beer) still persist in the eyes
of the bony fishes, not only for the adjustment of focus, but
also, according to Beer,! for slight changes in the direction of
the line of vision. It is in keeping with these facts that the
eyeball itself in the fishes is much less moveable, and in this
respect also more primitive, than it is in the higher Verte-
brates. We shall have presently to note other primitive
features connected with the eyes in the bony fishes.
On the isolation of the lens and the consequent abstraction
of a large portion of the palisade layer from the front of the
ocular globe, we may assume that this layer was regenerated,
as shown by dotted lines in Diagram IV.
The aqueous chamber developed as a space among the fibrous
tissue forming the neck of the invagination: it would thus
be morphologically a lymph space in the cutis. I have indi-
cated the splitting apart of the fibres in this place in Diagram
III. The aqueous chamber would thus from the first be quite
distinct from the vitreous chamber, which was primarily a cleft
in the epidermis (see below).
The iris could be developed by the thinning away of the
fibrous substance above the lens, and its radial contraction
away from the optic axis. This contraction may be supposed
to have given rise originally to the ciliary processes, as shown
1 « Pfliig. Arch.,’ Bd. liii, 1894.
ATTEMPT TO DEDUCE VERTEBRATE EYES FROM THE SKIN. 355
in the diagram. These are, however, now secondarily spe-
cialised.
It was an integral part of my theory, not only that the
accumulation of pigment granules in the illuminated spots
gave rise to the sensory visual areas in the epithelium by the
irritation caused by their passage between the epithelial cells,
but also that the excessive discharge of such granules supplied
or helped to supply the refractive matter for a dioptric
apparatus. This supposition entirely harmonises with this
suggested deduction of the Vertebrate eye directly from the
epidermis.
When a certain area began to form a retina, i.e. in the
earlier stages of the condition shown in Diagram II, the excess
of pigment would pass out between the sensory cells, and
would supply so much matter to the epidermis cells outside
the retina; here it would be clarified, as the pigment is
typically clarified in the epidermis, and added to the thickness
of the refractive layers. As the sensory cells multiplied, and
were further cuticularised as rods, and the relation. between
them and the irritating chromatophores became more special-
ised, the way to the exterior would be barred, and the necessary
material would have to be brought in from the circumference
of the retina.
We have probably to look to the long cells depicted in the
diagram for the bringing of the matter used in forming the
harder refractive portions of the primitive lens. But, in
addition, pigmented cells would almost certainly travel into
the cavity of the eye along the connective-tissue strands
which accompanied the nerve and blood-vessels. The rapid
clarifying of this pigmented matter in the fluid contained in
the space between the lens and the retina would give rise to a
slimy mass, the rudiments of the vitreous body. The deriva-
tives of this intrusive connective tissue would be found in the
processus falciformis and pecten, which still convey pigment
into those eyes that retain such structures. As the eye ad-
vanced in specialisation, and this important channel for con-
veying pigment into the eye to supply material for the vitreous
356 H. M. BERNARD.
humour was almost entirely lost (Diagram IV), compensation
would be found in the great increase of the inner surface of
the eye, the posterior half of which alone could function as
retina; the other, anterior half, consists of the undifferentiated
palisade epithelium, which allows the pigment to pass
through as slime into the vitreous humour. It is also pro-
bable that some material for the same purpose finds its way
along the optic nerve into the cavity of the eye.
This suggested origin of the vitreous humour, viz. that it is
largely due to pigment granules passing into the eye cavity
and dissolving into slime, is not only in accord with our
homology of the retina with the epidermal palisade layer which
gives rise to the mucus layer, but it is also supported by
certain facts. In sections of the eyes of embryo chickens,
dendriform exudations of slime, obviously forming the vitreous
humour, are seen coming from the ciliary portion of the retina.
The vitreous body is known to be easily detachable from all parts
of the interior of the eye, except the ciliary region and the
place of entrance of the opticnerve. The pigment granules are
known to pass from the pigmented epithelium into the palisade
layer in the iris, and the process can be seen beginning before
the lens is reached. The actual microscopic appearances in
fortunate sections are all in favour of their supplying slime
for the vitreous humour. In my series of sections of the larva of
Petromyzon planeri the retina is only in contact with the
pigmented epithelium in the axis of the eye, only here are any
rods and cones found ; where the retina is not in contact with
the pigmented epithelium, not only are no rods and cones
developed, but the pigment does not accumulate in the choroid
epithelium. It appears as if, there being no layer of rods and
cones to hinder its. advance, the pigment granules pass
freely through, and form masses of semi-clarified granules
within the space between the retina and epithelium. In the
ciliary region of the retina, where the two layers—the retina
(without rods) and the epithelium—come gradually into con-
tact, all the appearances are as if the pigment granules are
streaming through the retina into the vitreous humour. The
ATTEMPT TO DEDUCE VERTEBRATE EYES FROM THE SKIN. 307
eye is, however, so immature that this latter is not yet a
clear slimy fluid, but a dense mass of fine granules, quite
referable, as far as microscopic appearances go, to pigment
granules only partially dissolved and clarified.!
This, in brief, is the way in which it is possible to conceive
that the eye might have been developed directly from the
skin, as our theory demands. That it actually has developed
in this way is perhaps difficult to prove in face of the very
different history suggested by its embryology ; still, when the
evidence is summed up, it appears to be of considerable
weight.
I claim, for instance, that by this method of deriving the
eye, the arrangement and character of all the more important
tissues are fully accounted for.”
As examples I may draw attention to the following points :
1. The pigmented epithelium is not an ordinary epithelium,
but rather a close layer of chromatophores, some of which
are large ten-sided cells with two nuclei,’ which latter feature
is not infrequent in chromatophores.
2. The presence of the stellate chromatophores in the
choroid would be a vestige of the dense chromatophorai layer
that seems to have been very generally present in the cutis of
many pre-mammalian Vertebrates.
3. The fibrous connection between the front surface of the
iris and the cornea, specially pronounced as the ligamentum
annulare of fishes, might well be the remains of the neck of
the assumed invagination. It is attached to the iris and not
to the lens, which is of significance for this method of deducing
the eye from the skin.
4, The difference between the vitreous and aqueous humours
1 I propose to enlarge on this subject in an illustrated paper dealing
specially with skins and pigments.
? I have omitted all mention of the sclerotic because I believe its develop-
ment was concerned with that of the conjunctival folds, and of the specialised
muscles for the movement of the eye, which latter did not come within the
range of the inquiry.
3 Boden and Sprawson, ‘ Quart. Journ. Mier. Sci.,’ vol. xxxiii.
VOL. 39, PART 3.—NEW SER. AA
358 H. M. BERNARD.
is also accounted for. A comparison of the diagrams shows
at a glance that, while the latter is simply a lymph space (cf.
Fuchs’s diagram of the lymphatics of the eye!) arising within
the connective tissue of the cutis, the former is an epidermal
space, and is accordingly filled with slimy matter, presumably
identical with that which gradually turns the cells of the rete
mucosa into the horn cells.
5. The shape and curious arrangement of the cells forming
the lens, the gradual loss of their nuclei, and their progressive
transformation into hard refractive matter as the centre is
reached are simply explained by regarding them as the prickle
and horny cells which formed at one time a primitive surface-
lens, subsequently engulfed in the invagination. That these
fibres are closely comparable with epidermal prickle-cells can
be demonstrated under the microscope ;* they have all the
known characters of such cells, in addition to their gradual
decay and conversion into hard refractive matter. The pig-
mented granules of the iris on the one hand and the vitreous
humour on the other could obviously supply them with abun-
dance of slimy matter for this purpose.
6. I may be allowed again to allude to the arrangements
of the blood-vascular loops as shown in the diagrams; to the
ciliary process, which is accounted for as a simple mechanical
result of the formation of the iris (cf. Diagrams III and IV) ;
and to such structures as the pecten and processus falciformis,
and the intra-bulbar blood-vessels still persisting in different
eyes among the lower Vertebrates.
In addition to the weighty evidence afforded by these purely
morphological arguments, it is worth noting that the recent
discovery of Wolff,? fully confirmed by Miller,‘ as to the re-
generation of the lens in Triton from the iris, is in the main
in accord with Diagrams III and IV. The details as given
by these authors are, however, very curious. The outer of
' Reproduced in Morris’s ‘ Treatise on Anatomy,’ 1893, p. 889.
2 Cf. the first note, p. 357.
3 * Archiv fiir Entwick. Mechan.,’ i, 1895.
4 Eric Miller, ‘Arch. f. micr. Anat.,’ p. 23, 1896.
ATTEMPT TO DEDUCE VERTEBRATE EYES FROM THE SKIN. 399
the two inner pigmented layers of the iris, viz. that con-
tinuous with the pigmented epithelium, becomes the lens
epithelium on the distal face of the lens, while the inner
yields the long lens-cells. This naturally raises the question
whether the lens epithelium may not be the homologue of
the pigmented epithelium, the lens fibres alone representing
the palisade layer. It may be that this regenerative
process is strictly recapitulatory, and that this is the right
explanation of the facts. On the other hand, the lens epi-
thelium passes so gradually into the lens fibres round the
equator of the lens, that it is difficult to believe that the whole
structure is not the result of the folding of a single con-
tinuous epithelium. Again, the lens in the Vertebrate eye can
hardly be such an entirely new structure, dating, as in the case
of the lens of the pineal eye, simply from the time of the
invagination, which this new suggestion would compel us to
believe. It seems to me far more probable that it was an early
modification of the external horny layers of the epidermis,
which, as illustrated in the diagrams, became involved in an
invagination. A comparative study of lenses, including those
in degenerate eyes, might throw further light on this subject.
The embryological development of the lens is, according to
our view, so purely adaptive that it can hardly help us.
It is obvious, further, that this method of deducing the
Vertebrate eye from the skin has the advantage that this organ
need no longer be any exception to the rule which obtains in
the animal kingdom, that organs of sense develop directly and
continuously out of the skin. Lastly, we must not forget the
significance of the fact that the change of type which can be
seen to have taken place in the Vertebrate eyes is easily refer-
able to the change which’we know must have taken place in
the character of the skin from the Invertebrate to the Vertebrate
type.
Having, then, shown that it is possible physiologically and
morphologically to deduce the Vertebrate eyes from the skin,
the pineal eye from the skin in its Invertebrate condition, the
definitive eye from the skin in its Vertebrate condition, we
360 H. M. BERNARD.
have to face the embryological histories of these eyes and the
interpretations usually put upon them.
Taking the definitive eye first, it is found to arise ontogeneti-
cally by the union of two distinct structures. The ‘ primary
optic vesicle,” as it is called, develops as an invagination of
the brain towards the skin. This collapses into itself so as
to form a cup or spoon, open towards the skin,and joined to
the brain by a stalk or handle. The anterior wall, lining the
hollow of the cup or spoon, becomes the retina and retinal
palisade layer; the posterior wall, forming the outside wall of
the cup or spoon, becomes the pigmented epithelium in contact
with the palisade layer. Into this cup there dips down from
the outer skin an invagination which becomes constricted off
and forms the lens. The outer rim of the cup or spoon grows
over the lens so as to form, with other elements, the iris.
The question we have to try to answer is, are these pro-
cesses even approximately historical, or are they purely em-
bryological adaptations in order to obtain a desired end by a
series of short cuts? The answer must depend entirely upon
the weight of the evidence. It must be remembered that we
have no direct evidence whatever. Our sole guides are infer-
ences to be drawn from known facts, and these we are only
roused up to use when discussing some proposed clue to the
right understanding of the said facts. In the present instance
a theory of light sensations compels me to assume that the
Vertebrate eyes have developed from the skin, and in the fore-
going pages I have endeavoured to show that the known facts
of structure and function are explicable on this assumption.
The morphology and physiology of the eye, as the latter is
interpreted by the theory, go throughout hand in hand. When
we turn to the embryological development our connected
story is thrown into confusion. Some of the details, it is true,
are in most satisfactory agreement, as we shall presently see;
but the most essential processes, viz. those just described,
differ entirely from our scheme.
In the face of this serious difficulty, are we to withdraw or
boldly to examine the embryological processes in order to
ATTEMPT TO DEDUCE VERTEBRATE EYES FROM THE SKIN. 361
ascertain (1) whether any plausible phylogenetic explanation
can be given them? and (2) whether they cannot be more
simply explained as adaptations? With regard to the former
test, I confess at the outset that it seems to me almost waste
of time to discuss the possibility of these complicated in-
foldings and fusions of separate parts indicating the history
of such a singularly compact organ as the eye, especially if we
are to assume, what we are justified in assuming, an unbroken
continuity of function from the earliest rudiments to its most
complicated development. We are, in the first place, pre-
cluded from believing that the functional eye developed con-
tinuously from the brain towards the skin which then formed
a lens, as we should have no explanation of the bending back
of the nerve-fibrils. In order to get over this difficulty it is
suggested that the primary optic vesicle, i.e. the invagination
from the brain, may represent a primitive eye of the type
known as an “optic pit.” It is true that, according to our
interpretation of the facts, we have evidence in the pineal eye
or eyes that such structures did at one time exist on the heads
of the ancestors of the Vertebrata. Ifa pair of such eyes were
caught and eventually folded in by the medullary groove, they
might, it is thought, give rise to these “ primary optic vesicles,”
with the retine in the right position to become the retine of
the paired Vertebrate eyes. And, at first sight, there seems
to be some possibility of deducing the Vertebrate eyes from a
pair of engulfed, and therefore vestigial, and for the time
being functionless eyes, which may have started into life once
more by coming into contact on each side with some thicken-
ing, or perhaps glandular invagination, of the ectoderm. This
latter, by condensing the light upon the pigment still present
in these buried eyes, might once more set it in regular move-
ments, which would record the variations in the intensity of
the light. Some such hypothesis, it is claimed, would explain
what is called the inversion of the eye, for the light would
now shine through what was formerly the under side of the
retina.
The closer, however, this suggestion is examined, the more
362 H. M. BERNARD.
unworkable it becomes. Whence, for instance, came the optic
nerve of such an eye before it was engulfed in the medullary
groove? If from some primitive ganglionic centre, how came
it to be transferred to the definitive centre formed by the
medullary groove? How did this optic nerve manage to
coincide with the stalk of the invagination? Further, it is
not easy to deduce in this way the specialised association of
the chromatophoral layer with the retinal layer. The actual
evidence against the primary optic vesicles having been me-
chanically forced in by the lens is very strong, quite apart
from the difficulty of imagining why a lens-like body should
crush in the optic vesicle, and, as if by a happy accident, create
a highly organised eye. Why, again, was the pigment con-
fined to the (originally) external hemispheres of these buried
optic pits? Lastly, we do not know to what extent the me-
dullary groove is historical. Nervous systems usually develop
from thickenings of the ectoderm; and if a large amount of
material is required, the thickening may easily become an
invagination or a groove, and it is by no means necessary to
suppose that such invagination has any phylogenetic significa-
tion at all; it may be purely an adaptive process for the supply
of formative tissue. Indeed, in the Cyclostomata and bony
fishes, which are very low down in the Vertebrate phylum,
the central nerve-strand actually arises as a solid thickening.
But, admitting for the present that the medullary groove may
have some historical value, eyes are not very likely to let
themselves be tucked in so as to be functionless, unless there
is some obstacle to prevent them from shifting. According to
our theory, the pigment will always be drawn in the direction
of the strongest illumination, and on changes taking place in
the form of the body the eyes will shift—through this attrac-
tive action of the light upon the pigment—into the positions
in which they can best function as organs of sight.) This
1 Perhaps one of the most remarkable instances of the power of the eyes to
follow light is that recorded by Carl Chun (‘ Biologisches Centralblatt,’ xiii,
1893). Certain Crustaceans have luminous organs which throw light upon
the ground beneath them, This light has actually drawn down a portion of
ATTEMPT TO DEDUCE VERTEBRATE EYES FROM THE SKIN. 3638
renders any theory of the infolding of a pair of functional
eyes into the medullary groove, to my mind at least, highly
improbable.
Indeed, I think it will be freely admitted that it is not easy
to find any plausible interpretation of the embryological pro-
cesses as recapitulatory of the actual evolution of the eye. We
turn, therefore, to ask whether they can be better explained as
adaptations ?
At the outset it is obvious that the second assumption with
which we started, that the pigmented epithelium must have
been from the first in intimate and inseparable association with
the retina, compels us to assume that the primary optic vesicle,
in which the retinal half only secondarily comes in contact
with the pigmented half, is simply a developmental adaptation.
Let us see whether the hypothetical history of the eye above
sketched offers any interpretation of this optic vesicle. The
great importance of the eye in the economy of the organism,
leads to its development earlier and earlier in the embryo,
each shift back admitting of shorter cuts for the attainment of
the desired end. This may well have been the rule, even when
the eye was at the simple stage shown in Diagram II, in which
the retina was spoon- or ladle-shaped, with the nerve represent-
ing the handle. The most necessary requirement for the eye
at this stage was undoubtedly the rapid formation of the
specialised retina involving a great increase of nerve tissue,
and its intimate association with the chromatophoral layer.
Secondary infoldings of ectodermal tissue in order to supply
by a short cut the great increase of sensory cells required in
the formation of retinz is a well-known phenomenon in the
development of other eyes, which are more obvious modifica-
tions of the skinthan arethe Vertebrate eyes. I may mention
the larger eyes of scorpions and spiders, the specialised retinz
of which develop embryologically as invaginations in a plane
the eye on each side, so as to look at the ground illuminated by the animals
themselves, All stages occur; in some animals the eyes are only elongated
downwards, in others a piece.of each eye has become detached so as actually
to form a separate pair of eyes,
364 H. M. BERNARD.
more or less at right angles to the optic axis of the definitive
eye, the two layers of the invagination becoming applied to
form the more specialised elements of the organ. This is what
we may assume to have been the first adaptive embryological
process in the evolution of the Vertebrate eye. Its end was
not to produce the definitive eye, but the hypothetical eye at
the stage shown in Diagram II. Hence the primary optic
vesicle may have originally been a simple ectodermal invagina-
tion apart from the central nerve system.
The earlier this retinal invagination is laid down, the more
chance it would have of being involved in the medullary
groove. Connection between them could be found, first of all,
in the rudiments of the optic nerve, which would have been
laid down as an ectodermal thickening joining the optic invagi-
nation with the central medullary invagination. A further
shortening of the process might soon lead to the primary optic
invagination developing as a lateral offshoot from the medul-
lary groove.}
We are not altogether without evidence to support this in-
terpretation of the origin of the primary optic vesicle, as an
adaptation for the production of the eye when it was in a
primitive condition,—such, for instance, as we have depicted in
Diagram II. I have already called attention to what, accord-
ing to the history of the eye here described, are certain very
primitive features in the eyes of fishes. I refer to the accommo-
dation of the eye by means of the m. retractor lentis (Beer),
to the persisting power of moving the lens so as to vary the
direction of sight, to the slight powers of movement of the
eye itself as a whole, to the small development of the iris,
and to the persistence of a prominent ring of tissue—the liga-
mentum annulare—further forward on the iris than is the
homologous ligamentum pectinatum of higher Vertebrates,
which represent, according to the foregoing, the neck of the
ocular invagination (Diagram III). Now it is surely of some
1 The arrangement sometimes found, in which the optic vesicles only grow
out after the groove has closed over, can be easily regarded as a secondary
specialisation.
ATTEMPT TO DEDUCE VERTEBRATE EYES FROM THE SKIN. 3865
significance to find that, in these same animals, not only does
the nervous system arise as a solid strand of ectoderm, but the
eye and the optic nerve arise “‘ quite high up on the side of
the brain,” ? and only sink down later into the ventral position
(with respect to the medullary axis) typical of the higher
Vertebrates. It seems to me that these facts belong to one
another, and may be claimed as primitive features retained
by these lowly Vertebrates.?
It is, I am aware, usual to regard this development of the
medullary axis as a solid strand as secondary, but the argu-
ments are not conclusive. For the reasons above given, I am
inclined to believe that the solid medullary axis in these cyclo-
stomes and bony fishes is a direct inheritance from their Inver-
tebrate ancestors ; and that, conversely, the widely open medul-
lary groove is a secondary specialisation for the purpose of
supplying a still larger quantity of material as the nervous
axis of the Vertebrates became more and more pronounced.®
According, then, to our interpretation of the facts, the
1 Hoffmann, ‘ Arch. mik. Anat.,’ xxiii, 1883, p. 45.
* Evidence of this secondary connection of the optic invagination with the
brain has, further, been recently deduced from a study of the developing
brain itself. Waters, from a study of the primitive segmentation of the brain
(‘ Quart. Journ. Micr. Sci.,’ vol. xxiii, 1892, p. 457), arrives at the conclusion
that the optic nerve was once serial with the other segmental nerves. The
fact that the larger proportion of the nerve fibrils develop from the retina
towards the brain [Assheton, ‘Quart. Journ. Micr. Sci.,’ 34 (1892), p. 85;
Robinson, ‘ Journ. Anat, Phys.,’ 30, 1896, p. 319] may perhaps represent
the enormous secondary multiplication of the retinal sensory cells.
3 The presence of a medullary plate in forms which appear to be still
lower in organisation than the Cyclostomes is perhaps a difficulty, but the
very diversity of their specialisations, Urochorda, Cephalochorda, Hemichorda
(see Lankester’s article “ Vertebrata,” ‘Encyc. Brit.’), suggests that they have
wandered off in different directions—all, however, towards degeneration—
from some more typical Vertebrate form. The lowness of their organisation
may perhaps be paralleled by that of the Copepods, which are apparently
simpler than the form Apus, which has the best claim to rank as primitive
among the Crustacea. I would suggest a similar explanation for both cases,
viz. the fixation and subsequent specialisation of larval forms. I have suggested
the same also to account for the Acarina, which in many respects are simpler
than I believe the ancestral Arachnid could well have been.
366 H. M. BERNARD.
“primary optic vesicle ” reproduces the retina and specialised
chromatophoral layer of the eye when it was at the stage shown
in Diagram II. This would account for the fact that the
lens, though it appeared, as I think, in the epidermal stage of
the eye, does not appear ontogenetically with the primary
optic vesicle, but has to develop separately.
Some confirmation of this view will be found if we compare
the shape of the optic cup with the retina at the first hypo-
thetical stage. The primary optic vesicle never apparently
forms a true cup; it is far more spoon- or ladle-shaped, the
sides gradually folding up over the handle or stalk. The
stalk itself tends to fold, and where the sides of the spoon meet,
just over the handle, the choroid fissure arises. This is an
almost exact reproduction of the retina shown in Diagram II.
As the primitive retina expanded, the optic nerve entering from
the edge may have been wide and fan-like. As then the retina
bulged deeper into the body, and its sides folded steeply up,
the optic nerve would also fold with it near its junction with
the spoon-shaped retina. This folding would be more marked
if the point where the nerve joined the retina left the level of
the skin and travelled inwards towards the base of the in-
vagination. I claim, therefore, that we have here direct em-
bryological evidence of the former existence of a ladle-shaped
retina almost exactly corresponding with that sketched in
Diagram II, i.e. without a lens in organic connection with it.
The ontogenetic development of the lens takes place, as is
well known, by an invagination of the ectoderm ; this obviously
does not and cannot repeat the evolutionary process as depicted
in the diagrams. It is, perhaps, probable that the embryo-
logical invagination which forms the lens may be some faint
attempt to repeat the invagination shown in Diagram III.
Any way this method of lens formation would clearly have to
be adaptive if the account above given of the origin of the
“primary optic vesicle”’ is approximately correct. As far as
I can remember, nothing similar to this suggested complica-
tion has been recorded, although if we knew enough we should
probably find it to have been very common. We have the
ATTEMPT TO DEDUCE VERTEBRATE EYES FROM THE SKIN. 367
slow and uniform development of an organ, viz. the eye,
through the lapse of time. Its ontogenetic development
cannot possibly repeat the evolutionary process, because, when
the eye was in the earliest stages, the embryo early acquired
the habit of producing the organ in that primitive stage by a
short cut. This short cut the embryo has never given up, and
it has thus lost all chance of being able to repeat the subse-
quent evolutionary processes. The development, therefore, of
all these later specialisations must necessarily be almost purely
adaptive.
With regard to the pineal eye, somewhat the same explana-
tion of its ontogeny is suggested. The rudimentary optic in-
vaginations, situated close to the median line, early became
involved in the developing medullary axis. They were,
however, not adaptive structures, but the actual optic pits.
Again, in this case also, the connection between the developing
optic pit and the medullary axis or groove would be found in
the optic nerve. The most serious objection to this suggestion
is probably to be found in the fact that the epiphysis seems to
appear later than the first rudiments of the Vertebrate eyes,
whereas it ought, perhaps, to appear earlicr. As a matter of
fact, very little reliance can be laid upon the order of appear-
ance. The functional eyes, whether pineal or definitive, seem
to be hurried on the scene almost before anything else. As
soon as the pineal eye became functionless or of less importance
than the definitive eyes, there would be no hurry in its develop-
ment, and the definitive eyes might appear first, although
historically much later developments than the pineal eyes.
It may be objected that there is not the same difficulty in
supposing the pineal eye to have developed from the brain, it
being but a single vesicle with a simple nerve-stalk, so that
its ontogeny might repeat its historical evolution. Spencer,
who assumes the embryology of the eye to be in the main
recapitulatory, compares the epiphysis with the tunicate eye,
but does not believe that it began as an eye. In order
to get over the difficulties in the way of deducing the pineal eye
structurally from the tunicate eye, the possibility of which
368 H. M. BERNARD.
had been assumed by Lankester, he suggests that the same
structure which, at an early stage of its development, formed
the tunicate eye developed into the pineal eye only after it
had grown out into a long stalked vesicle. The inherent
objections to this supposition seem to me to be as great as
those which it was intended to avoid. On the other hand,
however, the facts appear to me to be all in favour of the
account given above, the parallel between the retina with the
retinal pigment and the palisade layer with its subjacent
accumulation of pigment cells, between the connective-tissue
capsule of the eye, with its chromatophores and other wan-
dering cells, and the cutis—the fact that the eye is still em-
bedded in the thickened cutis in many forms, and that the
skin above the eye is frequently found transparent,—all seem
to indicate the skin, and not the brain, as the real mother-
tissue of the eye.
Given, then, the possibility of explaining the embryology
of the pineal eye as adaptive, there seems to me to be no valid
reason why we should not accept the teachings of comparative
morphology, and recognise the pineal eye as a relic of the
simpler type of skin of our Invertebrate ancestors.
The suggestion that the pineal eye arose out of detached
portions of the paired primary optic vesicles, I think, defies all
attempts to work out in detail. Beard’s explanatory diagrams,
it seems to me, cease to be manageable if, instead of single
layers of epithelial cells, we try to marshal functional eyes in
the same way.
In the foregoing pages I have endeavoured to show how the
Vertebrate eyes admit of being deduced directly from the skin,
as my theory of the origin of sight requires. My object has
been not only to remove a possible objection to the theory
arising from the embryological development of these eyes, but
also to show how they might be made even to bear testimony in
itsfavour. It must be borne in mind that the arguments here
adopted are not so strong as they might be. They would have
1 Beard, “‘The Parietal Eye of Cyclostomes,” ‘Quart. Journ. Micros.
Sci.,’ vol. xxix, pl. vi, figs. A, B, C. ;
ATTEMPT TO DEDUCE VERTEBRATE EYES FROM THE SKIN. 369
been greatly strengthened if it had been possible to incorpo-
rate some of the more direct evidence in favour of the theory
from which they start. This evidence is, however, cumulative ;
and in endeavouring, first, to show that the chief types of eye
are both morphologically and physiologically explicable by this
theory, I merely begin with that portion of the evidence
which came readiest to hand.
In selecting eyes for analysis in subsequent contributions to
this subject I shall not follow any systematic order; my
primary object not being any comparative study of eyes, but
merely the detailed working out of the observations which I
have made, which seem to me to afford evidence of the truth
of my theory of the origin of light sensations.
EXPLANATION OF PLATE 22,
Illustrating Mr. H. M. Bernard’s paper, “ An Attempt to
deduce the Vertebrate Eyes from the Skin.”
Fic. 1.—Diagrammatic representation of a pineal eye regarded as an
invagination of a simple palisade epithelium. The pigmented contents of the
chromatophores of the cutis find their way in the skin to the surface, where
they form a layer of slime (yellow), and in the eye into the cavity of the
invagination, where they form a slimy vitreous humour.
Fic. 2.—Hypothetical early stage in the evolution of the Vertebrate eye
proper. Externally the palisade layer of the skin has budded off layers of
cells, which as “ prickle-cells”” absorb the pigment and other matter passing
through the palisade cells, and, dying, become horny scales. The nerves of
the skin (.) no longer end in the palisade layer, but among the prickle-cells.
Hence, when the palisade cells become retinal cells stimulated by the passage
of the pigmented granules, the nerves have to bend back. The retinal pali-
sade cells early become cuticularised, apparently in order to oppose more
effectively the outward streaming of the pigmented granules. The nerves
are already grouped into a solid strand, and form a kind of handle to the
spoon-shaped retina which is being bulged in by the accumulation of slimy
maiter in the epidermal cleft which has arisen above the retina. Into the
370 H. M. BERNARD.
cavity thus formed, blood-vessels and connective tissue conveying chromato-
phores intrude, and help to supply the matter for the vitreous humour. The
horny covering of the retina is supplied by specially long “ prickle-cells,”
which form a kind of surface lens.
Fic. 3.—Hypothetical stage following that shown in Fig. 2, the retina
having bulged still further into the body, and the optic nerve (0. 2.) having
sunk more towards the base of the invagination, the whole eye, including the
surface lens, has been invaginated below the skin. In the stalk of the invagi-
nation, composed of cutis, lymph-spaces commence to appear in the optic axis
of the eye. ‘The intruding blood-vessels and connective tissue are shown
entering the cavity of the eye with the optic nerve, and radiating strands
towards the lens for the movements of the same are diagrammatically
indicated.
Fic. 4.—Shows the changes which would be required to complete the
sensory and dioptric apparatus of the Vertebrate eye. The lymph-spaces are
specialised into the aqueous chamber; the stalk of invagination persists as the
ligamentum annulare (or pectinatum). The radial contraction of the tissue
to form the iris gave rise to a fold—the rudiment of the ciliary processes.
The lens has become isolated by the rolling round (as indicated by arrows) of
the palisade layer of the distal portion of the eye. The portion of the
palisade epithelium thus given up to the lens is represented as having been
regenerated (dotted lines).
REPRODUCTION AND METAMORPHOSIS OF COMMON EEL. 37]
The Reproduction and Metamorphosis of the
Common Eel (Anguilla vulgaris).'
By
G. B. Grassi,
Professor in Rome.
Four years of continual researches made by me in collabo-
ration with my pupil, Dr. Calandruccio, have been crowned at
last by a success beyond my expectations,—that is to say, have
enabled me to dispel in the most important points the great
mystery which has hitherto surrounded the reproduction and
the development of the Common Eel (Anguilla vulgaris).
When I reflect that this mystery has occupied the attention of
naturalists since the days of Aristotle, it seems to me that a
short extract of my work is perhaps not unworthy to be pre-
sented to the Royal Society of London, leaving aside, however,
for the present, the morphological part of my results.
The most salient fact discovered by me is that a fish, which
hitherto was known as Leptocephalus brevirostris, is the
larva of the Anguilla vulgaris.
Before giving the proofs of this conclusion I must premise
that the other Murenoids undergo a similar metamorphosis.
Thus I have been able to prove that the Leptocephalus
stenops (Bellotti), for the greatest part, and also the Lepto-
cephalus morrisii and punctatus belong to the cycle of
evolution of Conger vulgaris; that the Leptocephalus
haeckeli, yarrelli, bibroni, gegenbauri, kéllikeri, and
many other imperfectly described by Facciola, and a part of
the above-named Leptocephalus stenops of Bellotti,
belong to the cycle of evolution of Congromurena my stax;
that the Leptocephalus tenia, inornatus, and dia-
1 From the ‘ Proceedings of the Royal Society,’ November, 1896.
372 G. B. GRASSI.
phanus belong to that of Congromurena balearica; that
under the name of Leptocephalus kefersteini are con-
founded the larvee of various species of the genus Ophich-
thys; that the Leptocephalus longirostris and the
Hyoprorus messanensis are the larve of Nettastoma
melanurum, and that the Leptocephalus oxyrhynchus
and other new forms are larve of Saurenchelys cancri-
vora, and that finally a new little Leptocephalus is the larva
of Murena helena.
The form known as Tylurus belongs to Oxystoma, of which
we unfortunately know nothing more than a figure by Raffi-
nesque. I have not been able to find the Leptocephalus of
Myrus vulgaris, of which I have had only a single indi-
vidual, in which the transformation was already far advanced.
Neither have I found the Leptocephalus of Chlopsis bi-
color, a very rare form, which is related to Murena and to
Murenichthys. As the result of these observations, the
family of the Leptocephalide has been definitely suppressed
by me ; the various forms of that family are, in fact, the normal
larvee of the various Murznoids.
In regard to the greater part of the above-named species,
the control has been threefold, namely :
Firstly, anatomical. I have compared the various stages in
all their structures, and have made the due allowance for the
changes brought about by the metamorphosis at the close of
larval life.
Secondly, natural. I have found in nature all the required
transitional stages.
Thirdly, experimental. I have followed, step by step, the
metamorphosis in aquariums.
Therefore the hypothesis of Giinther that the Leptocephali
are abnormal larvee, incapable of further development, must be
rejected. All this is related by myself at length, with all
historical details which concern the question, in a_ large
memoir which is about to appear in the journal edited by
Professor Todaro.
Until now all these facts have been unknown, because nor-
REPRODUCTION AND METAMORPHOSIS OF COMMON EEL. 373
mally they can only be observed in the abysses of the sea at a
depth of at least 500 metres. Fortunately, along a part of the
coast of Sicily strong currents occur, which must be ascribed
to the tide, producing very large displacements of the water in
the narrow Strait of Messina. I shall give further details con-
cerning these currents in my large memoir. In consequence
of the strong currents sometimes—I say sometimes because
there is no regularity, and one may have to wait for.a year with-
out obtaining any material—not only many deep-sea fishes, but
also all stages of the development of the Murzenoids are met
with in the surface-water. To these currents we owe all the
captures of Murzna helena with ripe eggs, which is in ac-
cordance with what I had already argued from other facts,
namely, that the reproduction of the Murznoids takes place at
great depths of the sea.
Before I proceed to speak of the common eel, I must pre-
mise that Dr. Raffaele has described certain pelagic eggs as
belonging to an undetermined species, putting forward the
suggestion that these eggs belong to some Murenoid. This
matter has been investigated by myself, and I have shown that
the newly hatched larve (called “ pre-larve” by me) derived
from these eggs have essentially the character of Leptocephali.
The life history of the Murzenoids, leaving aside for the
present the common eel, is as follows:—Females can only
mature in very profound depths of the sea, that is to say, at
least a depth of 500 metres. This fact I established by find-
ing well-known deep-sea fishes together with Leptocephali,
ripe Murzene, and quite ripe eels (see below). The females of
those species which do not live at this depth must therefore
migrate to it. The male, however, can mature at a smaller
depth, and therefore they migrate into the greater depth when
they are already mature. Fertilisation takes place at great
depths: the eggs float in the water; nevertheless they remain
at a great depth in the sea, and only exceptionally, for un-
‘known reasons, some of them mount to the surface.
From the egg issues rapidly a pre-larva, which becomes a
larva (Leptocephalus) with the anus and the urinary opening
VoL. 39, PART 3.—NEW SER. BB
374 GB; GRASSI.
near the tip of the tail. The larva then becomes a hemi-larva,
the two apertures just named moving their position towards
the anterior part of the body, which becomes thickened and
nearly round. By further change the hemi-larva assumes the
definitive or adult form. The larva, as well as the hemi-larva,
shows a length of body much greater than that exhibited by
the young Murzenoid of adult form into which they are trans-
formed. By keeping specimens in an aquarium I was able to
establish a diminution of more than 4 cm. during the meta-
morphosis. With regard to the greatest length which the larva
can attain in a given species, and the amount of diminution
which accompanies metamorphosis, there are great individual
variations.
The history of the common eel, to which I am now about
to refer, is very similar to that given above for the other
Murenoids. The common eel (Anguilla vulgaris) under-
goes a metamorphosis, and before it assumes the definitive
adult form it presents itself as a Leptocephalus, which is known
as Leptocephalus brevirostris. This Leptocephalus was
discovered in the Straits of Messina many years ago. A speci-
men was also captured by the ‘‘ Challenger,” and another
specimen was taken by the Zoological Station of Naples in the
Straits of Messina. This form is occasionally carried to the
surface by currents. By exception, in the month of March, in
the year 1895, we captured several thousands of them in one
day ; but the best way to secure this Leptocephalus (and a very
easy one) is to open the intestine of the Orthagoriscus
mola, a fish which is common in the Straits of Messina, and
in it one is certain to find a very large number of specimens.
It must be observed that Orthagoriscus mola is a deep-sea
fish. The specimens of Leptocephalus brevirostris found
in the intestine of Orthagoriscus are more or less altered by
digestion. Those specimens of Leptocephalus brevirostris
which are taken near the surface in the open sea are in a better
state of preservation, but, unfortunately, these also frequently
have the epidermis injured so that they cannot maintain their
life in an aquarium for more than a few days; they live long
REPRODUCTION AND METAMORPHOSIS OF COMMON EEL. 375
enough, however, to allow us to observe that it is their habit
to conceal themselves in the sand or in the mud as the common
eel (Anguilla) does. Here it is to be noted that the various
forms of Leptocephali have habits resembling those of the
Murenoids to which they belong, i.e. they dig into the sand
or abstain from doing so according as the adult form has or has
not this habit.
I now pass on to the characters of Leptocephalus brevi-
rostris. I give them here in the same order as I shall use in
my larger memoir. The length varies from 77 to 60 mm., the
same extent of variation as observed in other Murenoids.
The caudal fin tends to assume the form which it has in the
Elver! or young Anguilla. It is to be noted that in other
Leptocephali the caudal fin also tends always to exhibit the
adult form. The lower jaw projects sometimes more than the
upper jaw, asin Anguilla. The margin of the mouth is wide,
as in Anguilla. The tongue is free, as in Anguilla. On the
other hand, the youngest elvers which I have observed have
smaller eyes than Leptocephalus brevirostris, and this
need not surprise us, since we know that in other species of
Murenoids the diminution of the eyes occurs during the
metamorphosis. The nostrils are separated from one another,
the anterior tubes are relatively at a considerable distance from
the tip of the snout and from the rim of the mouth. They
are in a position in which they are observed in many other
Leptocephali, which are destined to transform themselves into
adult forms having the anterior nostrils in nearly the same
position as in the commoneel. The posterior nostrils, on the
contrary, are not tube-like, and are in the same position as
those occupied in the adult Anguilla. It is worth remarking
that in other Leptocephali also the posterior nostrils have
already assumed the adult position when the anterior ones are
still far removed from it. In L. brevirostris I find a larval
dentition which resembles that of the other Leptoce-
1 The word “ Elver” is used in this paper in its strict sense, viz. for the
young form of Anguilla vulgaris as taken when ascending rivers in vast
numbers.
376 G. B. GRASSI.
phali. In’ correspondence with the small size of Lepto-
cephalus brevirostris the number of larval teeth is
small. Researches founded, firstly, on the enumeration of
the myomeres; secondly, upon the enumeration of the dorsal
and ventral arches of the vertebrae of the caudal extremity
(hypurals) ; and thirdly, upon the enumeration of the posterior
spinal ganglia, lead with great certainty to the conclusion that
the Leptocephalus brevirostris is the larva of a Murenoid,
the number of whose vertebree must lie between 112 and 117,
most probably 114 or 115. Such a Murenoid is the Anguilla
vulgaris. The Murenoid indicated cannot be any other of
those occurring in the Mediterranean, because they all have a
number of vertebre higher than 124.! Counting the myomeres
in Leptocephalus brevirostris, one finds generally only 105
complete, five others incomplete, and all the others in a state
of transparency and incomplete formation. These latter are
fortunately at the posterior extremity, where other criteria
come to our assistance, namely, the spinal ganglia and the
vertebral arches. To show how I arrive at the number of
vertebree which must be possessed by the adult Individual,
corresponding to a given Leptocephalus brevirostris, I
quote the following example :—I assume that three vertebree
develop themselves in correspondence to the first four incom-
plete myomeres, and that 105 must develop themselves in
relation to the 105 complete myomeres,—that is to say, between
the fourth and fifth myomeres, between the fifth and sixth, and
so on until we reach the 105th vertebra lying between the
104th and 105th myomeres. I further conclude that seven
other vertebra are developed at the caudal extremity, as indi-
cated by the number of vertebral arches and the spinal ganglia
in that region. We count, therefore, in all 115 vertebra, and
this is the number which can be easily seen in many specimens
of Anguilla vulgaris.
1 Murenesox savanna is said to have 109 vertebre, but it is
doubtful whether it really occurs in the Mediterranean. The position of
its nostrils and the number of its branchiostegal rays render its association
with Leptocephalus brevirostris impossible.
REPRODUCTION AND METAMORPHOSIS OF COMMON BEL. 377
Here I must particularly insist that I have ascertained in
an absolute manner that during the metamorphosis of the
Murenoids the number neither of the myomeres, nor of the
vertebral arches, nor of the spinal ganglia is subjected to any
change. The hypurals of Leptocephalus brevirostris are
precisely the same as in the elver of Anguilla vulgaris,
The last hypural which is fused with the urostyle may pre-
sent itself as a single piece, or may be more or less cleft,
These are variations which are met with also in the elver.
Just as in the elver, the last hypural but one is always ex-
tensively cleft, or, if the expression is preferred, doubled. To
the last hypural correspond five rays, whilst four correspond to
the last but one, and one to the last but two, the whole
structure being identical with that found in the elvers of
Anguilla vulgaris. Of these ten rays, the eighth, seventh,
and sixth are bifid, both in Leptocephalus brevirostris and
in the elvers of Anguilla vulgaris. In the pectoral fin of
Leptocephalus brevirostris the definitive rays can be ob-
served, and these are of the same number asin the elvers of An-
guilla vulgaris. Leptocephalus brevirostris is transpa-
rent, and has colourless blood. The red corpuscles are wanting,
but there are present so-called ‘ blood-plates” (“Blutplattchen”
in German) similar to those of the inferior Vertebrates. The
bile is also colourless. This fact is observed in all the other
Leptocephali. Leptocephalus brevirostris is, however,
the only one which is free from all pigmentation. Corre-
spondingly, the common eel is the only species of Murznoid
which at the close of metamorphosis (i.e. in the youngest
elvers) is devoid of all trace of larval pigmentation. It was
this observation which first led us to the discovery of the
relations between Leptocephalus brevirostris and An-
guilla vulgaris.
In making transverse sections of Leptocephalus brevi-
rostris I found other characters which confirm the relation
between it and the common eel; for instance, the branchio-
stegal rays are ten to eleven in number, as is also observed in
the elvers of Anguilla vulgaris. In the common eel the
378 G. B. GRASSI.
well-known lateral branch of the fifth pair of the cranial
nerves exists. It is also found in Leptocephalus brevi-
rostris. This lateral branch could not be found by Dr.
Calandruccio in the other common Murenoids of Sicily, and
is wanting also in the other Leptocephali.
The mucous-canal-system (sensory canals) in the head are
already developed, partially in Leptocephalus_ brevi-
rostris, and are incompletely developed in the elver. As in
the elver, so in Leptocephalus brevirostris, the pyloric
ceca are wanting. ‘The blind extremity of the stomach and
the incompletely developed swim-bladder, which is as yet free
from contained gas, are present both in Leptocephalus
brevirostris and in the elver of Anguilla vulgaris. The
pronephros is in active function, as in the other Leptocephali.
The Malpighian glomerules of the kidney (mesonephros) are
lobed as in the eel, and their number corresponds with that
observed in the Helmichthys stage, of which I will speak
further on. The genital gland, not yet sexually differentiated,
is almost identical with that of the same stage. In short, it
may be said that the whole organisation of Leptocephalus
brevirostris corresponds with the organisation of the common
eel, if we make allowance for those changes which are
observed in the metamorphosis of the other species of Mure-
noids, such as reduction of the pancreas and of the liver, dis-
appearance of the protoskeleton, complication of the muscu-
lature, increase in size of the cerebellum, loss of the larval
teeth, development of the definitive teeth, &c.
From the description of these Leptocephali I must pass on
briefly to speak of the stages nearer to the condition of the
elver. I am, however, obliged to leave a break in the series,
which, however little its significance, yet certainly will make
some impression on the minds of those who do not realise with
what caution I have formed my conclusions. I must confess
that since I have learnt how difficult it is to procure an entire
series of the development of a Murznoid, I am more astonished
at being able to recognise a single stage in the development of
a given species than at not finding the whole series. I must
REPRODUCTION AND METAMORPHOSIS OF COMMON EEL. 379
point out that the break in my series of the development of
Anguilla vulgaris would have been much smaller if I could
have persuaded myself to kill and preserve one of the hemi-
larve which I happened to meet with at the end of the year
1892. They were really transitional stages between Lepto-
WN AIAG <<
\ LE a=
ae
SS
/,
Fie. 1.
Gener
rae MMMM EM A gaa
IIL
Fie. 1.—Leptocephalus brevirostris. Natural size.
Fie. 2.—Leptocephalus brevirostris. Later stage. Natural size.
Fie. 3.—Anguilla vulgaris. ‘Transition stage. Natural size.
Fie. 4.—Anguilla vulgaris. Definitive habit (Elver). Natural size.
cephalus brevirostris and that stage which I shall describe
further on. I published this fact in a preliminary note in the
month of May, 1893. They were transparent with almost
colourless blood, without any trace of pigmentation except at
the eyes, and had lost all the larval teeth, whilst they possessed
already very few and very minute teeth of the definitive series.
380 G. B. GRASSI.
The body was thickened, and already showed the cylindrical
form. They measured little less than 8 cm. In short, they
were Leptocephalus brevirostris on the way to trans-
formation into Anguilla vulgaris. Asa matter of history
they actually did transform themselves in my aquarium with
the usual diminution in their dimensions, and subsequently
proceeded to increase in bulk.' The metamorphosis took
place, as usual, without the animal taking in any nourish-
ment whatever. The resumption of growth was accompanied
by a resumption of feeding. Unfortunately I had no other
individuals of this stage.
The stage which I now pass on to describe (fig. 3) can
be obtained during the winter in the sea. I have never
found them at the mouths of rivers. The length varies
from 54 to 73 mm. Most individuals measured about 65 mm.
The body is relatively longer than in the elver. It is also
relatively deeper, as in Leptocephalus. We are reminded
of Leptocephalus also by the pigment of the eye, the vitreous
transparency of the body, the swim-bladder being indis-
tinguishable in the living animal, and the absence of all
larval pigmentation. The blood is slightly coloured, and the
bile is already green. Slight pigmentation can be seen along
the central nervous system, and at the middle part of the
caudal fin. This commencement of the definitive or adult
pigmentation in the regions named before it occurs in any
other part is also seen in other Murenoids. The definitive
teeth are very minute, and few in number. The intestine
contains no food. After what I had observed in the other
Murenoids, the simple observation of the barely indicated
teeth, and of the absence of aliment in the gut, would have
been sufficient to convince me that the stage now under notice
must be preceded by a Leptocephalus phase. Indeed, if we
did not admit such a preceding history, we could not under-
stand how this little fish could have attained such a size with-
1 The fact that I actually have obtained in an aquarium the transforma-
tion of L. brevirostris into Anguilla vulgaris is of prime importance.
The time occupied was one month.
REPRODUCTION AND METAMORPHOSIS OF COMMON EEL. 381
out acquiring well-developed teeth, and without nourishing
itself.
In conclusion, no one would hesitate, even not knowing
Leptocephalus brevirostris, to refer the stage now under
discussion to a Murenoid about to complete its Leptocephalus
metamorphosis, were it not for the fact that there has been so
much question concerning the reproduction of the common
eel, and that so many capable observers have failed in dealing
with it, that every new observation is received with scepticism.
The stage of which I am now speaking, in the hands of a
pure systematist, would probably be described as a Helmichthys,
a genus established for certain forms of Leptocephali far
advanced in transformation.
The next forms to which I have to refer are captured in the
course of migration from the sea into fresh water. When kept
in an aquarium they assume the characters of the elver, dimin-
ishing more or less in volume, and without nourishing them-
selves. The elvers of the common eel can present themselves
in stages differing little from that last described, as well as in
a form which has already developed the full pigmentation of
the adult. Even those which most resemble the preceding
stage always have a character which distinguishes them easily,
namely, the presence of a definitive pigment, more or less
superficially placed on the head, and not to be confounded
with the pigment round the posterior extremity of the brain,
which latter is already present in the preceding stage. In
specimens taken at the mouths of rivers this more or less
superficial pigment was, so far as I could ascertain, always
present.
As the pigmentation develops itself, the little eel gradually
undergoes a diminution in all its dimensions. It results from
my measurements, that the fully pigmented elver has an
average length of 61 mm., while for the more or less colourless
elver the average length is 67 mm. I found pigmented elvers
which were reduced in length to 51 mm.,a size which I never
observed in those elvers in which the development of pig-
ment had not taken place.
VoL. 39, PART 3,—NEW SER. cc
382 G. B. GRASSI.
The facts which I have stated demonstrate that the eel goes
through a metamorphosis, and that Leptocephalus brevi-
rostris is its larva. Some further considerations remain to
be given, although I believe that zoologists will not consider
the question still an open one after the record of facts given
above—facts which any one may verify by examining the
material which is preserved in my hands. Many to whom I
have related my discovery of the history of the common eel
have objected that eels are found almost everywhere, whilst
Leptocephalus brevirostris is limited to Messina. In
reply, I must say that, first of all, it is not true that Lepto-
cephalus brevirostris is limited to Messina; secondly, that
at Messina there are special currents, which tear up the deep-
sea bottom which everywhere else is inaccessible; thirdly,
although it is true that on the coasts of many countries where
Anguilla vulgaris is found, no one has ever seen a Lepto-
cephalus brevirostris; it is also true that in no country,
not even in those where eels are abundant, has anyone ever
seen an eel of less than 5 cm. in length. Since it has to be
admitted that no one knows the eel before it arrives at the
length of 5 cm., there is no greater difficulty in supposing that
during this unknown period the eel passes through a Lepto-
cephalus stage than in supposing that it does not do so. The
critical study of the literature of this subject, and a great many
continued observations, have occupied me for many years, and
have been undertaken just in those places where young eels
are to be found. They enable me, from my own studies, to
affirm with assurance that young eels with the definitive adult
form do not exist of less than 5 cm. in length.
From the study of the memoir of Raffaele on pelagic eggs, I
have come to the conclusion that the eggs of his undetermined
species No. 10, having a diameter of 2:7 mm., and differing
from all the others in the absence of oil-globules,! must belong
to the Anguilla vulgaris, because from them Dr. Raffaele
1 Renewed researches have convinced me that this egg is that of An-
guilla vulgaris. There is, however, another egg belonging to an unde-
termined Mureenoid which is devoid of oil-drops, and can easily be confused
with the true eggs of Anguilla.
REPRODUCTION AND METAMORPHOSIS OF COMMON EEL. 3883
obtained pre-larvee which had only forty-four abdominal
myomeres. I endeavoured for two years in vain to study
these eggs at the Zoological Station of Naples. I found only
a few of them, and these died prematurely.
In another point my researches have yielded a very inte-
resting result. As a result of the observations of Petersen,
we know now that the common eel develops a bridal colora-
tion or “ mating habit,” which is chiefly characterised by the
silver pigment without trace of yellow, and by the more or
less black colour of the pectoral fin, and finally by the large
eyes. Petersen inferred that this was the bridal coloration
from the circumstance that the individuals exhibiting it
had the genital organs largely developed, had ceased to take
nourishment, and were migrating to the sea. Here Petersen’s
observations cease and mine begin. The same currents at
Messina which bring us the Leptocephali bring us also many
specimens of the common eel, all of which exhibit the silver
coloration. Not a few of them present the characters de-
scribed by Petersen in an exaggerated condition,—that is to
say, the eyes are larger and nearly round instead of elliptical,
whilst the pectoral fins are of an intense black. It is worth
noting that in a certain number of them the anterior margin
of the gill-slit is intensely black, a character which I have
never observed in eels which had not yet migrated to the sea,
and which is wanting in the figures and in the originals sent to
me by Petersen himself. Undoubtedly the most important of
these changes is that of the increase of the diameter of the eye,
because it finds its physiological explanation in the circumstance
that the eel matures in the depths of the sea. That, as a
matter of fact, eels dredged from the bottom of the sea have
larger eyes than one ever finds in fresh-water eels I have
proved by many comparative measurements, made between eels
dredged from the sea bottom and others which had not yet
passed into the deep waters of the sea. Thus, for instance, in
a male eel taken from the Messina currents, and having a total
length of 343 cm., the eye had a diameter, both vertical and
transversal, of 9 mm. and in another eel of 334 cm. the same
384 G. B. GRASSI.
measurement was recorded. Ina female eel, derived from the
same source and purchased in the market, whose length was
483 cm., the vertical diameter of the eye was 10 mm., and
the transversal diameter rather more than 10 mm. These
are not the greatest dimensions which I observed, and I con-
clude from these facts that the bridal habit described by Petersen
was not quite completed in his specimens, and that it becomes
so only in the sea and at a great depth. In relation to these
observations of mine stands the fact that the genital organs in
the eel taken in the Messina currents are sometimes more
developed than in eels which have not yet entered the deep
water. Thus it has happened that male individuals have
occurred showing in the testes here and there knots of sper-
matozoa. These spermatozoa are similar to those of the
Conger vulgaris, and must be considered as ripe. As is
well known, so advanced a stage of sexual maturity has never
before been observed in the common eel. This appears to be
due to the fact that the males hitherto examined had not yet
migrated into the deep water of the sea.
Eels with big eyes taken from the depths of the sea were,
before the above facts were known, described as a distinct
species under the name of Anguilla bibroni (Kaup) and of
Anguilla kieneri (Kaup), not to be confounded with An-
guilla kieneri (Gunther), which is a synonym of Lycodes
kieneri.
In certain cloace of ancient Rome which to-day are disused
and contain pure water, remarkable eels are found of a length
of from 20 to 30 em., of a grey colour, without trace of yellow,
of male and female sex, with enormous eyes, and with more or
less rudimentary genital organs. They are individuals which,
confined in a place without light, have acquired prematurely
one of the characters of the bridal habit without a correspond-
ing development of the genital organs. These individuals are
probably incapable of ulterior development, as the condition
of their genital organs seems to demonstrate.
Under the name Anguilla kieneri (Kaup) there have
probably been included some individuals which had acquired
REPRODUCTION AND METAMORPHOSIS OF COMMON EEL. 385
big eyes under conditions similar to those described for the
eels of these Roman cloace. From these and similar observa-
tions it clearly results that all the European eels must be in-
cluded under-a single species; and this is an important fact
from another point of view, namely, that it destroys an
objection which might be raised against my conclusion with
regard to the development of Anguilla vulgaris from Lepto-
cephalus brevirostris, namely, the objection that Lepto-
cephalus brevirostris belongs not to Anguilla vulgaris,
but to Anguilla kieneri, or to Anguilla bibroni.
To sum up, Anguilla vulgaris, the common eel, matures
in the depths of the sea, where it acquires larger eyes than are
ever observed in individuals which have not yet migrated to
deep water, with the exception of the eels of the Roman
cloace. The abysses of the sea are the spawning-places of the
common eel: its eggs float in the sea water. In developing
from the egg it undergoes a metamorphosis, that is to say,
passes through a larval form denominated Leptocephalus
brevirostris. What length of time this development re-
quires is very difficult to establish. So far we have only the
following data:—First, Anguilla vulgaris migrates to the
sea from the month of October to the month of January ;
second, the currents, such as those of Messina, throw up, from
tle abysses of the sea, specimens which, from the commence-
ment of November to the end of July, are observed to be more
advanced in development than at other times, but not yet
arrived at total maturity; third, eggs, which according to
every probability belong to the common eel, are found in the
sea from the month of August to that of January inclusive;
fourth, the Leptocephalus brevirostris abounds from
February to September. As to the other months, we are in
some uncertainty, because during them our only natural fisher-
man, the Orthagoriscus mola, appears very rarely ; fifth,
I am inclined to believe that the elvers ascending our rivers
are already one year of age, and I have observed that in an
aquarium specimens of L. brevirostris can transform them-
selves into young elvers in one month’s time.
VOL. 39, PART 3,—NEW SER. DD
am
2
au
i.
=<.
5
|
a
7
-
«
:,
a
- v
. - 7m : _
7 - -* be : om :
& ae “yt ©
; ay HAE ROUTED 1 A7ROw
: - hea
et Py) Tits ee 5 ‘Gnit a an Le f on
prone aie a a Reena ts
ei eLwie: yell a ts G2 “unary
book aimee gee nee ei Pre aiJés us Die}
ote id me: oft ‘ ets < Tit Fi? ee 7 ius ‘a
shiver pope Tubs a foes Wl ih Ghee Hf 7
asta Lye Palsawth. er altt gmc de? é Peas et
4 eer Toomer Feat aor wy. Th Peat iit) 82
ae% - ine we " fought £2 o Ob SIN
-— a Su 7. te oe ee ay f s i
6 ‘ ’ 4 ' t 7 ' = :
2 r is fia i? &
; >. Si
i ? - 4
3. | bs
» - > iP a
3 bg t 4 7% rs
: (eeu pe ee 5G
- : chia Tsyg Viv aaa), |) Sage aay
aru oe ue , =) , r 502% oid ra
~~ em 7 ae by es * ages re, Shy i. » ea
me F es =
oy: ey BUN, 20 4 yi aie
= 1 Fr = Vash ir eae Be eee
® a = * >:
i
ee ans
ee Fa ie a oe,
FEB 16 1897
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 387
Changes in the Cell-organs of Drosera rotun-
difolia, produced by Feeding with Egg-
albumen.
By
Lily Huie,
Physiological Laboratory, Oxford.
With Plates 23 and 24.
ContTENTS.
PAGE PAGE
Literature : ‘ : . 888 | Results of the Present Research 400
Experiments. : ‘ . 393 | Summary and Conclusion . « 423
Methods . P : 394 | Description of Plates : ~ 424
General Structure of the Glands 396
INTRODUCTION.
Tue subject of this research was suggested to me by Dr.
Gustav Mann, with the view of throwing some further light
upon the functions of the nucleus and nucleolus, and to test
whether his views were correct,—namely, that the chromatin
segments are organs for the elaboration of food; that the
nuclear chromatin is not the essential element, inasmuch as
it only represents partly elaborated material; that the nu-
cleolus is a storehouse for some elaborated matter, &c.1 By
1 Gustav Mann, “ The Functions, Staining Reactions, and Structure of
Nuclei,” ‘ Proc. British Association,’ Section D, 1892, Edinburgh meeting.
VOL, 39, PART 4,—NEW SER, EE
888 LILY HUIE.
the kindness of Professor Gotch I have been permitted to
work in the Histological Department of the Physiological
Laboratory at Oxford, under the guidance of Dr. Mann.
LITERATURE.
Numerous as have been the papers published on the Dro-
seracee since Charles Darwin’s classical treatise upon ‘ In-
sectivorous Plants,’ comparatively few have dealt with
cytological changes induced by functional activity of the
glands; and these few have been mainly devoted to the obser-
vation and elucidation of the “Phenomenon of Aggregation
of Protoplasm,” a term applied by Darwin! to the following
phenomena :
“Tf a tentacle is examined some hours after the gland has
been excited, . . . . the cells instead of being filled with
homogeneous purple fluid, now contain variously shaped
masses of purple matter, suspended in a colourless or almost
colourless fluid. The little masses of aggregated matter are
of the most diversified shapes, often spherical or oval, some-
times much elongated,&c. . . . These little masses inces-
santly change their forms and positions, being never at rest.
A single mass will often separate into two, which afterwards
reunite. Their movements are rather slow, and resemble
those of amcebe or of the white corpuscles of the blood. We
may, therefore, conclude that they consist of protoplasm.”
Cohn, in his review of ‘ Insectivorous Plants, 2 threw
doubts upon Darwin’s inference that the aggregated masses
consisted of protoplasm, and called them ‘ Zusammenbal-
lungen” of the red-cell sap.
Francis Darwin,® who studied the phenomenon in the cells
of the stalks of the tentacles, reasserted the view of his father
1 *Tnsectivorous Plants,’ London, 1875, p. 39.
2 * Deutsche Rundschau,’ 1876.
3 Darwin, F., “On the Process of Aggregation in the Tentacles of
Drosera rotundifolia,” ‘Quart. Journ. Micr, Sci,’
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 389
that the aggregated masses were “protoplasm.” He states
the two opposing theories thus :
I, “ My father’s view, that the aggregated masses consist of
protoplasm, and that their movements are simply due to their
own contractility, excited by various external agencies.”’
II. “ Professor Cohn’s view, which appears to be that the
ageregated masses consist of condensations of cell-sap, and as
a necessary corollary that the movements are impressed on the
masses by some kind of protoplasmic action external to the
masses.”
That the latter view is the correct one was proved by
Schimper, Gardiner, and De Vries, all of whom subsequently
studied the phenomenon, and came to the conclusion that it is
the contents of the vacuoles, and not the protoplasm, which
give rise to the phenomenon described by Darwin.
Schimper! compared the condition of the cells before and
after stimulation in Sarracenia, Drosera, and Utricularia, and
found that the protoplasm under the influence of the stimula-
tion gains an increased power of imbibition. Water is with-
drawn by it from the cell-sap, and in Sarracenia and Drosera
the tannin with which the cell-sap is laden becomes concen-
trated, and gives rise to the appearance Darwin called
“ Aggregation.”
De Vries? describes the movements thus:
1. An increase in the rapidity and vigour of circulation in
the peripheral layer of protoplasm.
2. A division of the vacuole into several smaller vacuoles,
which are each surrounded by part of the original vacuole
wall.
3. A very marked diminution of the volume of the vacuole,
because some of its original constituents are expelled through
the vacuole wall, and collect between this and the circulating
protoplasm. The expelled fluid has, at least approximately,
1 Schimper, A. J. W., “ Notizen iiber insektfressende Pflanzen,” ‘ Bot.
Zeit.,’ 1882, Nos. 14 and 15.
2 De Vries, H., “ Ueber die Aggregation im Protoplasma von Drosera
rotundifolia,” ‘ Botanische Zeitung,’ 1886.
390 LILY HUIE.
the same attraction for water as the remaining contents of the
vacuole, but possesses a different chemical composition. The
colouring-matter, the tannin, and certain unknown albumins,
precipitable by ammonium salts, are retained inside the vacuole.
The grape-sugar and vegetable acids are probably, in part at
least, expelled. ‘l'urgor is almost the same in stimulated as
in unstimulated tentacles; it takes, e.g., a solution of between
2 per cent. and 3 per cent. KNO, to produce plasmolysis in
unstimulated tentacles. The author finds that aggregation
takes place first in the cells nearest the glands, and that divi-
sion of the vacuoles occurs also in the gland-cells themselves,
although no protoplasmic circulation is to be seen. When
stimulation ceases the cells return gradually to their original
condition, while the vacuoles increase in size, and ultimately
fuse again into a single one.
This author describes further a similar but distinct pheno-
menon. Under certain conditions, especially when stimulated
by salts of ammonia, by free ammonia in weak solution, by
iodine, by osmic acid, and even sometimes by stimulating a
fresh leaf with albumin, or by slowly killing a tentacle , by
drying, the albumins dissolved in the sap of the vacuole are
precipitated; first in the form of a fine granulation, afterwards
fusing into larger globules equalling in size the little vacuoles
formed by aggregation, when it is not easy to distinguish be-
tween this appearance and that resulting from aggregation.
But that they are different phenomena may be shown by first
inducing normal aggregation, and then treating the aggregated
cells with 1 per cent. ammonium carbonate.
The ammonia salts did not penetrate the thick cuticle of the
tentacle stalks, but entered at the cut places and also by the
gland-cells of the tentacle, and the little mammilla-shaped
glands situated on the stalk of the tentacle.
A year previous to the appearance of De Vries’ papers,
Gardiner! published an account of “The Phenomena accom-
1 Gardiner, W., “On the Phenomena accompanying Stimulation in the
Gland-cells of Drosera dichotoma,” ‘Proc. Roy. Soc, London,’ 1885,
p. 229.
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 391
panying Stimulation in the Gland-cells of Drosera di-
chotoma,” in which he described the phenomenon of aggre-
gation in living stalk cells, much as Darwin had done. He
also noticed the phenomenon produced by the action of car-
bonate of ammonia, and called it “ passive aggregation.” A
spindle-shaped or acicular body in the stalk cells and epidermal
cells of the leaf called a “rhabdoid” is also described, which,
as well as the elongated nuclei of the stalk cells, tends to be-
come spherical after long stimulation.
Gardiner studied also the changes in the appearance of the
gland-cells themselves after stimulation. His methods were
as follows :
Leaves stimulated from five minutes to seventy-two hours
by feeding with flies, or small pieces of frog’s muscle, were
examined fresh, or after fixing in alcohol, picric acid, or osmic
acid. The best results were got with specimens treated for
twelve hours with 1 per cent. or 2 per cent. solution of chromic
acid. He describes a typical resting gland-cell thus :—‘“ The
protoplasm is arranged in a network or reticulum. The meshes
of this reticulum are excessively close around the nucleus,
which is situated at the base of the cell; but towards the free
surface they are much more open, the close and more open
arrangement merging the one into the other. The meshwork
extends through the whole cell-cavity, and the interstices be-
tween the meshes are occupied by a pink cell-sap, and the
whole is bounded by a structureless ectoplasm. In neither of
the three layers of cells covering the tracheidal cells of the
head could any obvious movement of the protoplasm be de-
tected.”
After stimulation for twenty-four hours the following histo-
logical changes were observed :
“ A gland mounted in water exhibited a mottled appearance,
such mottling being caused by a vacuolation of the most peri-
pheral portions of the protoplasm of the gland-cells. In
section, such a cell showed that in the course of secretion
there had been a using up of the cell contents, and instead of
the meshwork occupying the whole of the peripheral portion
392 LILY HUIE.
of the cell, so as to give a fairly homogeneous appearance,
large spherical cavities had appeared in the reticulum here and
there, such cavities being occupied by the cell-sap. The sap
had, moreover, a much darker pink tint. Thus a breaking
down or destruction of some part of the reticulum has taken
place. After some seventy-two hours’ stimulation this break-
ing down of the reticulum had reached to such an extent that
in the peripheral portion before referred to, all the central
core of the meshwork had for the most part disappeared, and
replacing it was a single large vacuole filled with cell-sap.
The cytoplasm had, moreover, contracted from the upper or
free surface of the cell-wall. In no case does the destruction
and consequent vacuolation extend to the base of the cell,
where the nucleus is situated. The nucleus is always sur-
rounded by dense protoplasm; and there are grounds for
believing that after very long stimulation, when all the secre-
tion has been poured out, and before absorption begins, an
active growth of the protopiasm takes place around the nucleus,
and in the more basal portion of the cell.”
«The view here taken (which is supported by certain of the
staining reactions) with regard to secretion is that in the
gland-cells the more peripheral network consists of protoplasm,
together with a formed substance derived from it, and that the
outpouring of the secretion is caused by the repeated breaking
down (owing to stimulation) of the protoplasm into this formed
substance, which is of a mucous nature, and which rapidly
attracts water and escapes as the secretion to the external
surface.”
Gardiner does not mention what the staining reactions are.
The paper I quote from is called a ‘‘ Preliminary” one. I be-
lieve the author has never published a more complete account ;
and, as far as I know, no one has with modern histological
methods worked out the minute cytological changes which
occur in the gland-cells, especially with regard to the nucleus
and nucleolus and the staining reactions of the protoplasm.
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 3893
EXPERIMENTS.
With the view of observing the minute histological changes
induced by functional activity in the gland-cells of Drosera
rotundifolia experiments were made as follows.
I. In Perthshire, Scotland, in the beginning of September,
1895, leaves were fed with the white of a soft-boiled egg, cut
into pieces approximately 5 mm. in length by 1 mm. in breadth
and thickness. The leaves were fixed after the following
intervals of time :—one hour, twenty-four hours, forty-eight
hours, seventy-two hours. These intervals were chosen in
order that the leaves might be fixed at the same time of day
in each case to avoid all variations due to diurnal changes.
Young vigorous leaves, which showed no trace of having had
a previous meal, were likewise fixed at the same times, and the
whole experiment was conducted while the plants were growing
in their native habitat, an open moor.
II. Another series of experiments was made in May, 1896,
in Berkshire, England, where the leaves were again fed without
being removed from their natural habitat, and fixed at intervals
of from one minute up to the time when the leaves had again
fully expanded and appeared to have regained their normal
condition, which varied with the vigour of the leaf from four
to seven days.
III. Repetition of experiment in August, 1896, in Scotland,
to determine changes induced by an insoluble tissue, namely,
boiled pieces of cork.
IV. A number of plants were removed for convenience to a
cool greenhouse, and their leaves were fixed at various intervals
after feeding.
In each set of experiments 15—20 leaves were used for
each period of time, e. g. twenty leaves to determine changes
after one minute, twenty leaves to show alterations after two
minutes, &c.
The results of the four sets of experiments coincided on the
394 LILY HUIE.
tissues being examined by the methods about to be detailed ;
but none of my drawings have been taken from material of the
fourth series, because plants kept for a few days in a green-
house frequently show, as Darwin pointed out, a loss of the
bright red colour peculiar to them under natural conditions,
and they are therefore affected to some extent by their
artificial surroundings.
Meruops.
I used the four following fixing fluids:
1. One per cent. chromic acid solution, as recommended
by Gardiner.
2. Absolute alcohol.
3. Mann’s picro-corrosive alcohol.?
4. Mann’s weak watery picro-corrosive fluid.
The effects of these fluids were very different.
The chromic acid caused the controls (unfed leaves) to close
up as if powerfully stimulated ; and on subsequent histological
examination the glands presented the appearance of having
been stimulated for a short time,—for example, the nuclei of
the third layer of gland cells had become ameeboid, or irregular
in form, The whole tissue was cloudy, and stained with no
precision.
The two alcoholic fixing agents preserved the control leaves
in a beautifully expanded condition, but instantly made the
glands white and transparent-looking, as if some substance
had been extracted. Sections of such glands showed more or
less distortion or slight collapse of the cell walls, and great
vacuolation of the contents of the apical gland-cells. By the
picro-corrosive alcohol, however, the large chromatin bodies
1 Gardiner, W., “On the Phenomena accompanying Stimulation of the
Gland-cells of Drosera dichotoma,” ‘Proc. Roy. Soc. London,’ 1885.
2 Mann, G., “On a Method of preparing Vegetable and Animal Tissues
for Paraffin Embedding, with a few Remarks as to Mounting Sections,’
‘Trans. Bot. Soc. Edin.,’ vol. xviii.
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 395
of the later stages of stimulation were specially well fixed, and
they stained with great precision.
Manu’s watery picro-corrosive fluid is in every way satisfac-
tory, especially in its action upon control leaves. It both keeps
the leaves expanded and preserves the dark colour of the glands.
Though it causes some shrinkage of the protoplasmic body
from the cell wall, it does not destroy any of its characters,
and the points of attachment that remain in spite of the
shrinkage serve to bring into evidence the fact that the proto-
plasmic bodies of neighbouring cells are intimately connected
with the cell wall at corresponding places, and so are probably
in connection with each other. Further, material fixed in
the watery fluid takes the stains with most satisfactory
precision.
The formula for this fixing fluid is as follows :—Saturated
HgCl, in 3 per cent. NaCl, 1 part. Saturated solution of
picric acid in Aq. Dest., 3 parts. This fluid has a specific
gravity of 1020.
The material was left in this for twelve hours, then trans-
ferred for twelve hours to a saturated solution of corrosive
sublimate in normal saline, and afterwards dehydrated in
alcohol of gradually increasing strength, being placed in 50
per cent. alcohol for four hours; 60 per cent., four hours; 70
per cent., about ten hours (i.e. all night) ; 80 per cent., five
hours; 90 per cent., five hours ; absolute alcohol, ten hours ;
then next day into two further changes of absolute alcohol.
Chloroform was next introduced into the bottom of the vessel
with a pipette. An hour was allowed to elapse after the tissue
had sunk in the chloroform ; the fluid was then all poured off,
and fresh chloroform substituted, and this was changed once
more after six hours. Paraffin of 52° melting-point was next
added in small pieces, until saturation at a temperature of
30° C. was reached. The tissue was then placed in a warm
chamber heated to the melting-point of the paraffin, a little
more melted paraffin added, and then the chloroform was
allowed to evaporate slowly. Sections were cut 5:08 yu thick
by a Cambridge rocking microtome, fixed on glass slips in
396 LILY HUIE.
ribbons by Mann’s! egg-albumen method, and stained either
by M. Heidenhain’s iron-alum hematoxylin (with or without
previously staining in Bordeaux red), which gives very good
results for the merely morphological aspects; or by Mann’s?
eosin and toluidin blue method, for the study of the alkaline
and acid reactions. In every stage of stimulation the staining
reaction was controlled by having on the same slide a row of
sections of unfed material. The method of staining with
eosin and toluidin blue is as follows :—Free the sections from
paraffin by xylol; remove the xylol by alcohol; place the
sections in Gram’s iodine solution (double strength) for
five minutes; wash out the greater part of iodine with
alcohol.
Wash in water till the sections are quite white; place in 1
per cent. watery solution of water-soluble eosin (Gribler) for
fifteen minutes ; rinse in water; place in 1 per cent. solution
of toluidin blue for five minutes; rinse in water ; decolourise
in absolute alcohol till the control sections upon the slide
appear as on fig. 1, or pale blue to the naked eye. It is very
essential that the absolute alcohol is pure, or otherwise the
blue colour will be extracted too rapidly by the lime which is
used in the distillations of alcohol. Further, do not allow the
alcohol to drop on the slide, because this causes unequal
washing out; but always immerse the whole slide in a vessel
filled with absolute alcohol; clear in xylol by immersing the
slide, and mount in turpentine or xylol balsam.
GENERAL STRUCTURE OF THE GLANDS.
In the heads of the tentacles of Drosera rotundifolia,
D. dichotoma, &c., we may distinguish between glandular
and non-glandular elements (see the accompanying figure).
The latter consist of a group of large tracheids forming the
1 Mann, G., “ A New Fixing Fluid for Animal Tissues,” ‘ Anat. Anzeiger,’
vii Jahrg. (1893), Nos. 12 u. 13, p. 442.
? Mann, G., ‘ Zeitschrift f. wiss. Mikrose.,’ xi, 1895, p. 489.
397
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA.
* il 7 TU |
MDT TATRA AN AAR NU VV VY ey /} we ViVi
WON aA VAAN ANA ARR RRR ROMY
WITTE) 5 rns ee AKANAAAAN
A SM se
<\
<
2
te) TI
Tentacle of Drosera rotundifolia constructed from serial sections :—
cells; C = second layer of
apical gland-cells; B = lateral gland
AC
gland-cells; D
matogen; G
der-
tracheids; F =
third layer of gland-cells; H
periblem; H = plerome of stalk of tentacle.
398 LILY HUIE.
core of the head, and varying in number with the size and
vigour of the leaf, and in the centre of which terminates a
fibro-vascular bundle (consisting of one or two spiral vessels,
of small lumen, and an accompanying sheath of long narrow
starch-containing cells) which springs from the vascular
system of the leaf, and traverses the stalk of the tentacle.
These non-glandular elements presumably function as a
channel for the transference of fluids.
The glandular elements consist of three bell-shaped layers
of gland-cells covering the tracheids. The external layer is
composed of cells of two kinds. Those at the apex are thin-
walled elongated cells, numbering in a longitudinal section
12—14. The lateral cells of the outer layer are shorter, and
possess thick walls. Their external walls have tooth-like in-
ternal projections of cellulose.
The cells of the second glandular layer resemble closely the
lateral cells of the superficial layer, but their walls have no
projections.
The third layer is composed of comparatively few, but very
long cells, which are remarkable for the large amount of
tannin in solution contained in their cell-sap, and for the
strong cuticularisation of those walls, both longitudinal and
transverse, which separate them from one another. The
transverse walls are worthy of special attention, because they
are composed of two projections, one from each side. The
projection from the outer side is strongly cuticularised ; that
from the inner side (next the tracheids) is lignified. Between
the two there is a minute space through which intercellular
communication is permitted. This layer of cells is also epi-
dermal in its origin, and arises from a zone of epidermal cells
immediately below that forming the initial cells of the lateral
external layer. Hach cell of this zone elongates inwards and
then subdivides.
The entire tentacle head is cuticularised superficially. I do
not understand how Gardiner’ comes to describe the glands
1 Gardiner, W., ‘On the Phenomena accompanying Stimulation in the
Gland-cells of Drosera dichotoma,” ‘Proc. Roy. Soc. London,’ 1885.
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 399
of Drosera dichotoma as without cuticle, and Goebel! to
say that the apical gland-cells are not cuticularised. Like
those of Drosera rotundifolia, the gland is covered with
cuticle. After prolonged treatment by iodine and sulphuric
acid all that is left undestroyed of the median longitudinal
section of a gland of either of these species is the continuous
outline of cuticle and the cuticularised walls of the third layer
of gland-cells.
A method for readily seeing the cuticular envelope in com-
paratively large pieces is to take an entire leaf that has been
blanched in alcohol, to soak it in water, then to leave it for
some minutes in strong iodine solution, to lift it out and wash
off the superfluous iodine in water. A few of the tentacles
should then be cut off and laid upon a glass slip; a drop of
concentrated sulphuric acid be placed upon them, and a cover-
glass laid over them; and when the reagent is seen through the
microscope to have taken effect by the usual coloration being
produced, the cover-glass is to be pressed gently so as to rup-
ture the apices of the gland and squeeze out the contents.
The cuticular envelope of the gland usually remains attached
at its base, and floats out as a semi-transparent membrane
stained yellow by the iodine.
Gardiner? describes the gland-cells as “ remarkably pitted
on their upper or free surface; and Goebel® states that the
outer walls of the apical gland-cells are dotted. I have exa-
mined pieces of cuticle obtained in the manner just described,
with a 1/12 apochromatic oil immersion object-glass by Zeiss
with oculars Nos. 8 and 12, but have found no well-defined
pores. I have equally failed on examining sections. The
internal cellulose protuberances of the thickened outer walls
of the lateral cells of the external glandular layer often produce
the appearance of enclosing large deep pits; but it is a decep-
tive appearance, as a little careful study shows. There is no
1 Goebel, ‘ Pflanzenbiologische Schilderungen,’ Theil i, p. 198.
2 Gardiner, W., “On the Phenomena accompanying Stimulation of the
Gland-cells of Drosera dichotoma,” ‘ Proc. Roy. Soc. London,’ 1885,
3 Goebel, ‘ Pflanzenbiologische Schilderungen,’ Theil i.
400 LILY HUIE.
such appearance in the apical cells. The secretion must, of
course, leave the apical gland-cells somehow, and the albu-
men pass somehow into the tentacle, but how it is impossible
as yet to say quite definitely. Silver nitrate enters the tenta-
cles most readily through the lateral gland-cells, and clefts
appear between the apices of the apical gland-cells a few
minutes after stimulation, and the albumen seems to pass in
through these clefts, as will be shown later on.
As Charles Darwin pointed out,! the extreme marginal ten-
tacles of some leaves of D. rotundifolia differ from the
other tentacles in exhibiting a unilateral development. Their
heads are very much elongated, and bear the glandular surface
on the ventral side only. The plan of structure is, however,
the same as in ordinary tentacles, all the elements being
present, and the unilateral development appears to be merely
the result of pressure during early growth in the bud.
Results of Histological Examination of Material
fixed in Watery Picro-corrosive Sublimate, and
stained with Eosin and Toluidin Blue.
Before describing these results, it may be well to explain the
names applied to the nuclear organs.
Chromosomes are the organs which show the well-known
affinity for alkaline dyes, and which are situated at the peri-
phery of the nucleus.
Nuclear Plasm is the material which in the resting state
forms the main bulk of the nucleus, and is neutrophile in
character; it corresponds to what is frequently called the
nuclear sap.
Nuclear Sap is the more watery fluid inside the nucleus, in
which is suspended the nuclear plasm, the latter being precipi-
table by HgCl, and other reagents.
Nucleolar Chromosomes or Nucleoli are the spherical
organs which show a special affinity for acid dyes, and have a
more or less central position.
Endonucleoli are spaces inside the nucleolus.
1 Darwin, C., ‘Insectivorous Plants,’ p. 7.
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 401
APICAL GLAND-CELLS.
The Resting State as seen in Unfed Leaves used
as Controls. Fig. 1, Plate 23.
The Cell Wall stains pale blue. The free wall of the cell by
shrinkage has sometimes become slightly concave, but other-
wise appears turgid.
The Cytoplasm has retracted from the cell walls except in
the lower basal third, where it either remains entirely in con-
tact with the cell wall, or is attached to it by strands, which
are in line with similar strands in neighbouring cells. The
general appearance is suggestive of communication in this
region between the apical gland-cells themselves, and also
between them and the cells of the second layer. The cyto-
plasm is apparently homogeneous, and arranged in a foam-like
manner. It stains pure blue, and contains a coarse granula-
tion embedded in it which stains a deeper blue, and represents
some zymogen. The constant position of the nucleus is just
below the middle of the cell, and the cytoplasm between it and
the outer wall is generally densest in the middle third, being
more vacuolated towards the free surface and in the imme-
diate vicinity of the nucleus. Surrounding the nucleus, and
partly hiding this looser arrangement, is a more finely granular
mass of protoplasm which often shows a slight affinity for the
red stain. This zone gradually merges over into the general
plasm, but shows a number of distinct radiations passing both
outwards from the nucleus and towards the basal attachment
of the cell. The plasm, where retracted from the cell wall,
sometimes shows minute dark blue granules, which frequently
also occur on the margins of the vacuoles. Compared with
the lateral cells of the outer layer and with the cells of the
second layer, the cytoplasm of the apical gland-cells is much
deeper in colour, because of its greater abundance in these
cells.
The Nucleus is situated just below the middle of the cell.
It may be spherical or oval.
The nuclei vary in size, being generally smaller in the cells
402 LILY HUIE.
nearest to the apex. They are remarkably plump, exhibiting
a full and rounded contour. They stain of a purplish tint.
Whether a definite nuclear membrane exists is doubtful, but
in mid-focus the nucleus is sharply defined against the cyto-
plasm, partly because of the difference in tint, and also because
of the peripherally placed dark blue chromatin granules.
The Nuclear Chromosomes are very minute granules, appa-
rently shaped like diplococci, each coccus being slightly flat-
tened, and attached to its neighbour by its broad side. They
stain an intense blue without any reddish tint.
The Nuclear Plasm is granular and always very dense in
controls. Sometimes the granules appear to radiate from the
nucleolus, and in other cases their arrangement produces a
sponge-like appearance; or, again, the granules are too closely
packed to show any definite arrangement. They stain of a
purplish tint.
The Nucleolus is embedded in the nuclear plasm. One,
sometimes more, very distinct deep red nucleolus is always
present, surrounded by a narrow Frommann’s clear zone. If
there is only one nucleolus, it is large; it may be in any part
of the interior of the nucleus, but is generally about one
quarter to one-third of the diameter of the nucleus, removed
from its periphery.
It always presents a perfectly smooth spherical outline, and
contains one or more distinct endonucleoli.
Effects of Stimulation for One Minute on the
Apical Gland-cells. Type 1. Fig. 2:
The Cell Wall shows no evidence of diminished turgor, there
being no collapse. It stains purplish. Between the apical
thirds of neighbouring cells a very distinct reddish material is
to be noted, staining, in colour and intensity, exactly like the
semi-fluid egg-albumen in contact with the head. In tentacles
stimulated with pieces of boiled cork this red material is
always absent. We are dealing, therefore, with a substance
(albumen ?) which is passing in between the cells, and not with
a substance which is being excreted by the cell.
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 403
The cytoplasm in this specimen has retracted from all the
walls equally.
It is very difficult to distinguish in colour between the
cytoplasmic granulation and its matrix. The general tone is
of a deep-blue purple, and the matrix seems to be paler than
the granulation. Fully half the cell is occupied by very large
vacuoles filled with a fluid which stains pale blue. The walls
of the vacuoles are lined, to a great extent, with deep blue
granules which show no trace of purple or red. The vacuoles
are largest in the upper third of the cell, but usually one or two
large ones lie also near the base. The cytoplasmic granulation
is extremely dense round the nucleus, and stretches between
the vacuoles in radiating ameba-like processes.
Characteristic of this stage is the presence in the cyto-
plasm of one (occasionally two) very apparent blue “ corpus-
cles,” about twice the size of the nucleolus, and having a dark
blue granular periphery. The explanation I should like to
offer for this appearance is as follows :—In the unstimulated
cell some substance is freely distributed, which helps to give
rise to the affinity for the blue stain. On stimulation this
substance is withdrawn from the cytoplasm (which now takes
a purplish stain), and is collected in vacuoles of various sizes.
By mordanting the cytoplasm surrounding the vacuoles an
affinity for the blue dye is established, and the deeply stained
cytoplasm will give rise to the appearance of a distinct wall to
the vacuole. The smaller vacuoles, because seen in their
entirety, appear of a deeper colour than the large ones which
are seen in section, and could easily be mistaken for definite
corpuscles. The substance acting as a mordant is in this case
probably tannic acid.
The Nucleus maintains its normal position in the cell, its
normal size, shape, and outline. There is no sign of a collapse,
it being still plump and well filled out. It stains exactly the
same depth and tint as the cytoplasm, but is distinctly marked
off from it by its dark blue periphery.
The Nuclear Chromosomes are in no way different from
those of unfed specimens.
VOL. 39, PART 4.—NEW SER. FE
404 LILY HUIE.
The Nuclear Plasm appears very indistinctly granular, as
if the granules had swelled and fused, but is still extremely
dense.
The Nucleolus stains a purple-red. It is still full and
spherical, and frequently appears even enlarged. It is sur-
rounded by a broad Frommann’s circle, and does not show
the endonucleoli as distinctly as in the resting condition.
Effect of Stimulation for One Minute on the Apical
Gland-cells. Type 2. Fig. 3 (slightly more ad-
vanced than Fig. 2).
The Cell Wall is pinkish purple, as in the last specimen,
and there is the same appearance of some reddish material
between the apices of the cells.
The Cytoplasm is shrunken from the walls, but attached in
one place to the basal wall, and in two places to one lateral
wall. Its matrix and granulation are indistinguishable from
each other. The plasm is very red round the nucleus for a
distance equal to two thirds of the diameter of the latter.
Beyond this it is deep blue-purple, as in the last specimen, and
the apical fourth of the cell is entirely occupied by a network
of vacuoles, the reticulum between them containing masses of
the deep blue granules which line the vacuoles as in the last
specimen. The same granules are distributed, apparently in
strings, through the thick granulation occupying the middle
of the cell. A pseudo-corpuscle is present, as in the last
specimen.
The Nucleus is normal as to position, size, and shape. It is
of the same reddish tint as the cell-plasm surrounding it, but
is clearly defined by its dark blue periphery, and still maintains
its plumpness.
The Nuclear Chromosomes are like those of unfed specimens.
The Nuclear Plasm is of the same tint and general appear-
ance as the granulation surrounding the nucleus.
The Nucleolus is dark reddish purple, and is surrounded by
Frommann’s clear zone. The endonucleoli in all the cells are
much less distinct than in unfed specimens,
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 405
Effect of Stimulation for Five Minutes on Apical
Gland-cells. Fig. 4.
The Cell Wall is coloured as in specimens fed for one minute.
The specimen from which fig. 4 was taken had been only
lightly touched by the egg-albumen, but in specimens in eon-
tact with large masses of the albumen the same appearance of
a pink substance between the apices of the apical gland-cells is
seen as in the former specimen.
The Cytoplasm has retracted from the walls, but remains in
contact at one or two points with the basal wall and the lower
parts of the lateral walls as usual. The coloration generally
resembles that of fig. 2, though there is less of the scattered
dark blue granulation. The blue corpuscle-like vacuoles are
generally present in cells of this stage, though absent in the
one from which the drawing was taken, which, in fact, shows
scarcely any vacuolation, perhaps because its contact with the
albumen used as food was comparatively slight.
The Nucleus is, with regard to position, size, and shape,
normal. It is of the same red tint or slightly redder than the
surrounding cytoplasm, and its dark blue periphery is less
definite and regular than formerly, as if broken at intervals,
which may perhaps be due to a change in the nuclear mem-
brane, the latter no longer staining with the blue.
The Nuclear Chromosomes.—The individual granules have
become larger and more conspicuous. The gland from which
the drawing was taken was, however, the only one in the
material examined at this stage which showed this development.
The Nuclear Plasm is as dense as in unfed specimens, but
the granular structure has fused, as already described in
material after one minute’s stimulation.
The Nucleolus is as in specimens fed for one minute.
Effect of Stimulation for Ten to Twenty Minutes
on the Apical Gland-cells. Fig. 5.
The Cell Wall is in some specimens the same as in the last
description, while in others no penetration of albumen between
the cells can be made out,
406 LILY HUIE.
The Cytoplasm shows, as regards vacuolation, great variety
in material of this stage. The specimen from which the figure
(5) was drawn is more vacuolated than most of the cells, which
latter rather correspond in this respect to earlier stages already
described. All other details remain as in the earlier stages
(figs. 4 and 2), except that there are no blue granules, and the
blue “ corpuscles” are rare.
The Nucleus becomes less distinctly defined in outline.
Other details remain as in the last specimen described.
The Nuclear Chromosomes are distinctly larger and more
conspicuous than in unfed specimens.
The Nuclear Plasm shows enlarged pale spaces which, com-
bined with the less full and rounded contour, give to the
nucleus a slightly attenuated appearance as compared with
unfed specimens. In colour it resembles the earlier spe-
cimens.
The Nucleolus is unchanged from earlier stimulated spe-
cimens.
Effect of Stimulation for Twenty Minutes to One
Hour on the Apical Gland-cells. Fig. 6.
Notr.—It is extremely difficult to decide which is the
sequence of the three types next to be considered. They are
all characteristic of glands which have been active for periods
intermediate between twenty minutes and four hours, and it is
by no means certain that the order here adopted is the correct
one. The type shown in fig. 6, which was taken from a -ten-
tacle stimulated for one hour, was seen also in a tentacle fed
for only twenty minutes. I have, therefore, placed it before
the other two.
The Cell Wall is pale blue-purple. In many cells the walls
appear less turgid than in controls, being often somewhat
crooked or uneven; this is especially the case in the lateral
walls.
The Cytoplasm shows considerable vacuolation, especially
in the outer third of the cells, but not more so than in some
previously described earlier stages. Round the nucleus the
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 407
cytoplasm is dense. The entire cytoplasm stains red, except
that at the apex of the cell there is a trace of blue. A few
deep blue granules adhere to the exterior of the apical wall, a
condition which I leave unexplained. Whether in figs. 6, 7,
and 8 there is not more protoplasm than in cells stimulated
1—5 minutes is difficult to say. Probably, as the result of
stimulation, a temporary increase in the amount of cytoplasm
occurs.
The Nucleus is shrunken, leaving a clear white zone be-
tween it and the surrounding cytoplasm.
Nuclear Chromosomes appear as large granules, and are
brought into close proximity to each other by the shrinkage of
the nucleus.
The Nuclear Plasm is the same colour as the surrounding
cytoplasm, but appears deeper because of its greater density.
The Nucleolus stains the same tint as the surrounding
nuclear plasm, and therefore is difficult to see. In the speci-
men figured it is invisible, hidden probably by the enlarged
nuclear chromosomes.
Effect of Stimulation for One to Four Hours onthe
Apical Gland-cells. Fig. 7.
The Cell Wall is very pale blue. The apical cells appear
frequently to overlap each other, an appearance due to loss of
turgor.
The Cytoplasm is stained of a purple-red, becoming bluer in
the apical fourth, where it is greatly vacuolated. There are
very marked red granular aggregations in contact with the
nucleus on its basal aspect.
The Nucleus is somewhat swollen. Its position is normal.
The Nuclear Chromosomes are larger than in unfed leaves.
The Nuclear Plasm is shrunken to form a dense central
mass, attached at fairly regular intervals to the nuclear peri-
phery, so as to present a stellate outline. The vacuoles thus
formed appear slightly blue. The nuclear plasm, because of
its density, stains more deeply than the cytoplasm, but of the
same tone of colour.
408 LILY HUIE.
The Nucleolus is undiscernible. It is, however, probably
present, as in other material, also, showing this retracted con-
dition of the nuclear-plasm, I have been able to reveal it by
Heidenhain’s iron-alum hematoxylin.
Effect of Stimulation for One to Four Hours on the
Apical Gland-cells. Fig. 8.
The Cell Wall is very pale blue.
The Cytoplasm is greatly vacuolated. The main body of
the plasm is red, shading to blue round the vacuoles.
The Nucleus is spherical and swollen. Its outline is very
indistinct, and its periphery undefined.
The Nuclear Chromosomes are deep blue. ‘They show a
tendency to be displaced or arranged irregularly.
The Nuclear Plasm is purple. It appears thin because spread
out over a larger space than before. White spaces occur in it
here and there.
The Nucleolus is small and pale red.
Effect of Stimulation for Twenty to Thirty Hours
on the Apical Gland-cells. Figs. 9a and 10.
Note.—The state about to be described is characteristic of
glands that have been active for twenty to thirty hours, but is
sometimes reached in twelve hours.
The Cell Wall is pale blue. It presents the same appearance
of loss of turgor as noticed in specimens fed for one to four
hours.
The Cytoplasm is an extremely scanty network staining red,
or with a trace of blue here and there in the parts furthest
from the nucleus.
The Nucleus is normal as to size and outline, or is shrunken
so as to form an oblong body with its long axis at right angles
to that of the cell.
Nuclear Chromosomes now consist of eight large dark blue
segments. They are remarkably distinct and conspicuous in
uushrunken nuclei, and appear to be V- or U-shaped, reminding
one of the well-known stages in karyokinesis. In shrunken
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 409
nuclei they are brought very close together, and help to give
the nucleus a dark purple appearance.
The Nuclear Plasm is red. It is vacuolated in the un-
shrunken nuclei. It is dense purple-red in the shrunken
nuclei, which are therefore much darker than the cytoplasm.
The Nucleolus is pale red, inconspicuous, sometimes very
small in the unshrunken nuclei. It is undiscernible in the
nuclei that have shrunken. Endonucleoli are indistinguish-
able, owing to the paleness and transparent appearance of the
nucleoli. = -
Fig. 98 is a nucleus from an apical gland-cell of a leaf that
had been fed with yolk of boiled egg, and which after thirty
hours was fixed in Mann’s picro-corrosive alcohol. It is re-
markable for the large size of the chromosomes, which seem
to be approximately V-shaped. It is also remarkable for the
absence of nuclear plasm. The cytoplasm is in the same con-
dition as is shown in figs. 9 and 10.
In those experiments where pieces of cork were laid on the
leaves to produce stimulation, complete exhaustion of the cell,
similar to fig. 10, resulted as regards the cytoplasm; while the
nucleus, although shrivelled and shrunken, stained a pure red,
there being not a trace of blue colour in the chromosomes.
This peculiar behaviour of the chromatin segments I have also
seen, though not frequently, in leaves which had been fed
twenty to thirty hours previously with egg-albumen. I men-
tion this fact because there may be some analogy between this
affinity for red dyes as seen here, and also met with during the
middle period of mitosis—a view which seems to be supported
by figs. 9a and B, where one of the features of mitosis is re-
produced.
The Apical Gland-cells Two to Three Days after
Feeding (the Leaf beginning to reopen). Fig. 11.
The Cell Wall is pale blue.
The Cytoplasm shows the usual general retraction and basal
attachments. It is extremely dense and granular round the
nucleus, forming a well-defined zone which occupies the middle
410 LILY HOIE.
third of the cell, so that cells of this stage can be easily re-
cognised with low powers. The thick granulation stains a
deep purple. The matrix appears to stain the same colour,
but fainter. Below the nucleus are one or two very large
vacuoles, Above it the space is either occupied by one or two
large vacuoles, or, as in the cell figured, a protoplasmic net-
work traverses the upper third of the cell, enclosing numerous
smaller vacuoles in its meshes.
The Nucleus is in position, size, and shape as in unfed
specimens. It stains of the same tint as the cytoplasm, and
therefore lacks the clear definition which is characteristic of
the resting state.
The Nuclear Chromosomes are in the form of numerous
granules of slightly larger size than in the resting state. They
have a somewhat undefined appearance, and want of clearness
of outline.
The Nuclear Plasm is granular. It resembles very closely
the cytoplasmic granulation in every way ; so that the nucleus
appears to be only a portion of the cell-plasm enclosed by a
broken ring of chromatin granules.
The Nucleolus is either very pale and transparent, and
usually very small, or is quite undiscernible.
Apical Gland-cells, Seven Days after feeding (the
Leaf thoroughly opened up, and glistening with
exuded drops). Fig. 12.
The Cell Wall is pale blue.
The Cytoplasm has retracted as a whole from the cell walls,
but is attached in the basal third at various points, as is the
general rule, both in controls and fed leaves. It is densest
near the nucleus and below it. In the rest of the cell it is
largely vacuolated. In structure and colour it is exactly like
that of unstimulated glands. The vacuoles are large and
numerous, and their contents are perfectly homogeneous, and
are unstained ; though usually the vacuoles appear to have a
bluish tinge in mid-focus, caused by underlying or overlying
plasm.
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 411
The Nucleus is in position, size, and shape as in unfed speci-
mens. They also agree with the resting nuclei in their general
appearance of plumpness and turgor.
The Nuclear Chromosomes are peripheral granules which
are either equally small or very slightly larger than those of
specimens that have never been fed. They stain the same
deep blue, and are similar in shape and arrangement.
The Nuclear Plasm is granular. Its general arrangement
produces a sponge-like appearance, and is therefore similar to
the least dense specimens among controls. It shows affinity
for both dyes, as does the nuclear plasm of controls.
The Nucleolus is exactly as in controls,—that is, very evident,
with distinct endonucleoli.
LATERAL SUPERFICIAL GLAND-CELLS.
The Resting State, seen in Unfed Leaves used as
Controls. Fig. 13, Plate 24.
The Cell Wall is turgid, and stains pale blue.
The Cytoplasm is either not retracted at all from the cell
wall, or is retracted as in the apical cells, except at certain
points where it is applied to the basal wall and to the lower
parts of the lateral walls. It is always attached by strands to
the projections from the external wall.
The Cytoplasm is exactly similar to that of the apical cells
in its structure and staining properties, but frequently
appears paler because it is less dense. It is very dense round
the nucleus, but the more peripheral parts are largely vacuo-
lated. A reddish tint such as sometimes occurs in apical cells
round the nucleus is hardly ever seen in lateral ceils.
The Nucleus is placed in the centre of the cell or slightly
below it. It is similar to the nuclei of the apical gland-cells,
though sometimes slightly paler owing to the nuclear plasm
being somewhat less dense.
Nuclear Chromosomes are like those of the apical cells, or
sometimes the granules are slightly larger.
412 LILY HUIE.
The Nuclear Plasm resembles that in the apical cells, or
occasionally is less dense.
The Nucleolus is as in the nuclei of the apical cells.
The Effect of Stimulation for One Minute on the
Lateral Superficial Gland-cells. Type 1. Fig. 14.
The Cell Wall is pale purplish pink. Its turgor is un-
changed.
The Cytoplasm is attached to the basal and lateral walls, but
shrunken towards the apex. The distribution is like that of
unfed specimens. Its coloration is pinkish purple, and fairly
uniform throughout the cell. A blue ‘ corpuscle” is usually
present in the cell-plasm.
The Nucleus is normal as to position, size, and shape. It is
slightly deeper in colour than the cytoplasm surrounding it.
The Nuclear Chromosomes are as in unfed specimens.
The Nuclear Plasm is unchanged in structure, but shows
slightly more affinity for red than in unfed specimens.
The Nucleolus is deep purple-red, and is surrounded by
Frommann’s zone. The endonucleoli are less distinct than in
unfed specimens.
Nore.—The lateral superficial gland-cells at the base of this
tentacle-head show a much redder coloration than the other
cells of the layer, and therefore resemble fig. 15.
Effect of Stimulation for One Minute on the Lateral
Superficial Gland-cells. Type 2. Fig. 15.
The Cell Wall is extremely pale purple. There is no red
coloration between the cells.
The Cytoplasm is retracted from the basal and lateral walls,
but not from the apical to which it is closely applied. The
general distribution of the cytoplasm is like that of controls.
In colour it is like that of the apical cells,—that is, red
round the nucleus, becoming bluer towards the periphery.
The blue “ corpuscles” are present as a rule.
The Nucleus is normal as to position, size, and shape. It
is of the same shade of colour as the cytoplasm surrounding
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 413
it, or sometimes rather bluer, but is sharply defined by its
dark blue periphery.
The Nuclear Chromosomes are as in unfed specimens.
The Nuclear Plasm is less distinctly granular, though as
dense as in unfed specimens.
The Nucleolus is deep purple-red, and is surrounded by
Frommann’s zone. The endonucleoli are not distinct.
Effect of Stimulation for Five Minutes on the
Lateral Superficial Gland-cells. Fig. 16.
The Cell Wall is pale pinkish purple in the specimen figured.
In other specimens more densely surrounded by the albumen
used as food, the external walls, and the upper part of the
lateral walls are very red.
The Cytoplasm has apparently not diminished in mass, or
only slightly so. Its coloration is as in apical cells of this
stage. A blue ‘‘ corpuscle ” is frequently present.
The Nucleus is normal as to position, size, and shape; but
its blue periphery is more or less irregular.
The Nuclear Chromosomes are like those of the apical cells,
appearing as large, conspicuous, generally distinctly double
granules.
The Nuclear Plasm in general arrangement is as in unfed
specimens. In some cases it is slightly less dense. Its colour
is like that of the surrounding cytoplasm.
The Nucleolus is unchanged from the one-minute stage.
Effect of Stimulations for Ten to Twenty Minutes on
the Lateral Superficial Gland-cells. Fig. 17.
The Cell Wall, as in the last case, stains either pale purple
or red. In the specimen chosen the walls are pale purple.
The Cytoplasm is in general aspect, bulk, distribution, and
coloration similar to that seen in specimens fed for shorter
periods,
The Nucleus and Nuclear Chromosomes show no change
since the stage last described.
AL 4 LILY HUi#.
The Nuclear Plasm is less dense than formerly. Its colour
is like that of the surrounding cytoplasm.
The Nucleolus is unaltered from the last stage described.
Effect of Stimulation for Twenty Minutes to One
Hour on the Lateral Superficial Gland-cells.
Fig. 18.
Cell Wall is pale purplish blue. The walls of these cells do
not show the same loss of turgor that is seen in the apical cells.
Owing to their greater thickness they are more resistant.
The Cytoplasm is reddish purple. It seems to cover as
great an area as in control leaves, but is everywhere much
less dense. There is no shrinkage from the cell wall. The
peripheral layer is very blue. A blue corpuscle-like vacuole
is sometimes present.
The Nucleus is swollen, and its outline is very indistinct.
The periphery is apparently broken, and there is no trace of a
nuclear membrane.
The Nuclear Chromosomes are in the form of a great number
of irregular granules larger than in unfed specimens, but
somewhat indistinct in outline, and staining a less intense
blue.
The Nuclear Plasm is less dense than formerly, because of
the increased size of the nucleus. In it there are large white
spaces. It stains a bluish purple, and so gives the nucleus a
bluer tint than the surrounding cytoplasm.
The Nucleolus is pale red. Frommann’s clear zone is not
seen. The endonucleoli are not well defined, or are absent.
Effect of Stimulation for One to Four Hours on the
Lateral Superficial Gland-cell. Fig. 19.
The Cell Wali is very pale blue.
The Cytoplasm is red or reddish purple.
The Nucleus may be said to be perfectly wanting in outline.
It is chiefly defined from the cytoplasm by its distinct blue
chromosomes upon their light background of sparse nuclear
plasm, and large white spaces.
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 415
The Nuclear Chromosomes appear as a few large segments.
They are fewer and larger than those of the apical gland-cells.
They stain a pure deep blue.
The Nuclear Plasm is very scanty, and of the same tint as
the cytoplasm, but large white spaces occupy most of the
interior of the nucleus, which has therefore a washed-out,
empty appearance.
The Nucleolus is small.
Notr.—It seems as if the lateral superficial cells and the
cells of the second layer in this case were in advance of the
apical cells; for the collection of the chromatin into a few
large segments, and the vacuolation of the nucleus, are cha-
racteristic of more advanced stages of glandular activity.
Effect of Stimulation for One to Four Hours on the
Lateral Superficial Gland-cells. Fig. 20.
The Cell Wall is pale blue.
The Cytoplasm is red, shading to a bluer tint towards the
periphery of the cell. The red colour is deepest close to the
nucleus, but there are no such well-defined masses of red
material as are seen in the apical cells of the same gland
(fig. 7). A blue “corpuscle” is sometimes present.
The Nucleus, Nuclear Chromatin, and Nuclear Plasm are
the same as in the apical cells.
The Nucleolus is srnall, and also indistinct because it re-
sembles in colour the surrounding nuclear-plasm. Fig. 20
may be an earlier stage than fig. 19, because the nuclear
chromosomes have not as yet aggregated into eight big seg-
ments.
Effect of Stimulation for Twenty to Thirty Hours
on the Lateral Superficial Gland-cells. Figs. 21
and 22.
The Cell Wall is pale blue.
The Cytoplasm is as in the apical Cells.
The Nucleus is represented by the two types found in the
416 LILY HUIRE.
apical cells. The unshrunken nuclei are in all respects similar
to those of the apical cells of a corresponding period. The
shrunken nuclei differ somewhat in the two layers. In the
lateral cells they seldom, if ever, become elongated, but remain
approximately spherical.
The Nuclear Chromosomes, the Nuclear Plasm, and the
Nucleolus resemble those of the apical cells of same period,
the formation of eight chromosomes being again very evident.
The Lateral Superficial Gland-cells Two to Three
Days after Feeding. Fig. 28.
The Cell Wall is pale blue.
The Cytoplasm exhibits a granulation resembling that of the
apical cells (fig. 11), but is scarcer, and therefore paler; it
surrounds the nucleus, and stretches away from it, especially
towards the base of the cells. Cytoplasmic strands also run
out from the nucleus towards the apical peripheral layer, in-
tersecting the cell and dividing the space into large vacuoles.
The Nucleus resembles those of the apical cells, but is here
somewhat darker, because relatively denser than the surround-
ing cytoplasm.
The Nuclear Chromosomes resemble those of the apical
cells, or are slightly larger; they are apparently spreading
out to form less compact bodies than in figs. 21 and 22.
The Nuclear Plasm and Nucleolus resemble those of the
apical cells of the same period.
The Lateral Superficial Gland-cells Seven Days
after Feeding. Fig. 24.
The Cell Wall is pale blue.
The Cytoplasm is exactly like that of controls, or is some-
what less abundant, with larger vacuolated spaces.
The Nucleus, Nuclear Chromosomes, Nuclear Plasm, and
Nucleolus resemble those of the apical cells of same gland and
those of controls.
OHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 417
Tue Sreconp LAYER or GLAND-CELLS.
The Resting State, seen in Unfed Leaves used as
Controls.
The Cell Wall is pale blue.
The Cytoplasm is very scanty compared to cells of the
outer layer.
A thin peripheral plasmic layer always remains applied
to the wall of the cell. Cytoplasmic strands suspend the
nucleus in the middle of the cell, and radiate from it to the
peripheral layer lining the wall. Large vacuoles occupy the
spaces between these radiating strands. The structure and
staining properties of the cytoplasm are exactly the same as in
the apical and lateral gland-cells of the superficial layer. A
reddish tint round the nucleus is sometimes (though rarely)
seen.
The Nucleus, Nuclear Chromosomes, Nuclear Plasm, and
Nucleolus are like those of the lateral superficial gland-cells.
The Effects of Stimulation for One Minute to
Twenty Minutes upon the Gland-cells of the
Second Layer.
All the cytological changes correspond to those of the
lateral superficial cells, except that the cell walls are always
pale purple, never red.
Effect of Stimulation for Twenty Minutes to One
Hour on the Gland-cells of the Second Layer.
The Cell Wall and Cytoplasm are like those of the superficial
cells stimulated for a slightly longer period.
The Nuclei of the cells below the apical gland-cells have
shrunken in the same way as the apical nuclei, while the lateral
cells of the second layer have their nuclei in the same con-
dition as the lateral superficial cells, or only very slightly
shrunken.
The subsequent changes in the cells of the second glandular
layer agree in all respects with those of the lateral superficial
gland-cells.
418 LILY HUIE.
GLAND-CELLS OF THE THIRD LAYER.
The Resting State, seen in Unfed Leaves used as
Controls. ‘Fig: 254, Nueleus seen in Pomer
tudinal section; 258, the same in transverse
section.
I have only figured the nuclei of this layer, because the cells
are very large, and the cytoplasm is too scanty to show any
characteristic changes, except those of the staining reaction.
In tentacles fixed in absolute alcohol, however, a dense plasm
is seen filling the cell, and resembling mucus.
The Cell Wall.—The outer wall of a longitudinal section
stains pale blue. The inner wall next the tracheids stains
deep blue. The cuticularised parts do not stain at all.
The Cytoplasm is merely a thin peripheral layer, slightly
granular, staining pure blue. The layer lining the wall next
the tracheids generally shrinks away from it to the centre of
the cell, which it traverses like a cord, attached at its ends to
the point in the transverse wall between the cuticularised por-
tion and the lignified projection. As this is the case in all the
cells of the layer, a cord, or rather a lamella, appears to stretch
continuously from cell to cell.
The Nucleus is spindle-shaped. Its position is peripheral.
It lies either in the layer of cytoplasm lining the external wall,
or in that which lines the internal wall. Should the cyto-
plasm shrink to the centre of the cell, the nucleus is taken
with it. Its long axis corresponds to the long axis of the cell.
The Nuclear Chromosomes are minute granules which gene-
rally appear to be double. They stain deep blue. They are
peripherally placed, and their arrangement frequently suggests
that they are disposed like beads upon invisible threads, which
describe a wide-meshed reticulum.
The Nuclear Plasm is granular; and the granules appear
more or less distinctly to be arranged in chains, or to form a
kind of open network with the spaces occupied by a homo-
geneous substance. Both granules and spaces stain blue,
without any reddish or purple tinge, thus differing from the
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 419
nuclei of the other gland-cells. This is the general rule; but
in leaves taken in autumn the nuclei of this layer occasionally
correspond in tint with those of the other gland-cells.
The Nucleolus.:—There is one large nucleolus, or two smaller
ones. They stain deep red, and contain one or more distinct
endonucleoli.
Effect of Stimulation for One Minute on the Cells of
the Third Glandular Layer. Type 1. Fig. 26a,
Nucleus seen in longitudinal section; Fig. 26B,
in transverse section.
The Cell Wall is pale purple; deeper in tint and bluer on the
side next the tracheids.
The Cytoplasm is as in unfed specimens, except that it stains
pinkish purple. The occasional presence of a blue ‘‘corpuscle”
is doubtful.
The Nucleus is normal as to position and size. It shows no
alteration in shape, except that it is slightly less pointed at the
ends. Its general tint is pinkish purple; and it shows more
affinity for the red stain than the majority of the nuclei of the
other gland-cells.
Nuclear Chromosomes are as in unfed specimens.
The Nuclear Plasm is of a reddish-purple tint. It is dis-
tinctly granular, and is arranged in an open network in the
meshes of which is a pale purple homogeneous substance.
The Nucleolus is dark purple-red. The endonucleoli are
indistinct or absent.
Effect of Stimulation for One Minute on the Cells
of the Third Glandular Layer. Type2. Fig. 27.
The Cell Wall stains pale blue-purple, showing a much bluer
tint than the walls of the other gland-cells.
The Cytoplasm is pale pink, but otherwise is like that of
unfed specimens. There is no blue “corpuscle” present.
The Nucleus is normal as to position and size. It is less
finely pointed at the extremities than in controls. It exhibits
VOL. 389, PART 4.—NEW SER. aa
420 LILY HUIE.
a strong affinity for the red stain. This affinity is markedly
greater than in the nuclei of the other gland-cells.
The Nuclear Chromosomes are apparently as in controls.
The Nuclear Plasm is distributed as in unfed specimens, but
the granules stain red. The substance which occupies the
meshes of the reticulum is also red.
The Nucleolus stains dark crimson. The endonucleoli are
badly defined.
Effect of Stimulation for Five Minutes on the Cells
of the Third Glandular Layer. Fig. 28.
The Cell Wall and the Cytoplasm resemble those last
described.
The Nucleus has become much broader and shorter, and
shows an intense affinity for the red stain.
The Nuclear Chromosomes are very distinctly double. They
stain very dark blue.
The Nuclear Plasm stains red. It is a very open reticulum,
composed distinctly of granules strung together in a moniliform
manner.
The Nucleolus resembles those of specimens stimulated for
one minute.
Effect of Stimulation for Ten to Twenty Minutes
on the Cells of the Third Glandular Layer.
Fig, 29.
The Cell Wall and Cytoplasm resemble those of specimens
fed for one minute.
The Nucleus has become still shorter and broader. It
shows the same affinity for the red stain as the last type.
The Nuclear Chromosomes have slightly increased in size,
but in other respects they are like the Nuclear Plasm and
Nucleolus as in the last type.
Effect of Stimulation for Twenty Minutes to One
Hour on the Cells of the Third Glandular
Layer. Fig. 30.
The Cell Wall.—The external longitudinal wall stains pale
blue; the wall next the tracheids dark blue.
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 421
The Cytoplasm stains very pale bluish purple.
The Nucleus is almost spherical, and stains very pale blue-
purple.
The Nuclear Chromosomes are conspicuous deep blue
granules of various sizes, and considerably larger than in the
control.
The Nuclear Plasm stains reddish purple, and is gathered
together in the centre of the nucleus, instead of being distri-
buted in chains. It is still apparently granular.
The Nucleolus is pale red, and much smaller than in the
control (fig. 25).
Effect on the Cells of the Third Glandular Layer
of Stimulation for One to Four Hours. Fig. 31.
The Cell Wall and Cytoplasm are as in the last type.
The Nucleus is oval, spherical, or irregular in shape. It
stains a more or less intense blue, sometimes very deeply.
The Nuclear Chromosomes are conspicuous dark blue
granules.
The Nuclear Plasm is scanty and granular, staining reddish
purple. It is collected to form a little cloud round the nu-
cleolus.
The Nucleolus stains purplish red. It has a transparent
appearance, and contains one or more endonucleoli.
Effect of Stimulation for One to Four Hours on the
Cells of the Third Glandular Layer. Fig. 82.
The Cell Wall, Cytoplasm, Nucleus, Nuclear Chromosomes,
aud Nuclear Plasm as in the last description.
The Nucleolus is purple-red, and is indistinct because it
agrees in colour with the granules which closely surround it.
One or more endonucleoli are present. In the figured speci-
men the endonucleolus is large.
422 LILY HUIE.
Effect of Stimulation for Twenty to Thirty Hours
on the Cells of the Third Glandular Layer. Fig. 33.
The Cell Wall.—The external wall is pale blue; the internal
wall is deep blue.
The Cytoplasm is very scanty. It stains a very pale blue,
thus differing in a very marked manner from that of the other
gland-cells. Sometimes there is a trace of red, or an appear-
ance of a few red granules here and there.
The Nucleus is irregular in shape, and stains pale pinkish
purple, and apparently quite homogeneous.
The Nuclear Chromosomes consist of a number of dark blue
granules of varying size which are peripherally placed.
The Nuclear Plasm is altogether absent, at least no granu-
lation can be made out, only a diffuse stain.
The Nucleolus is pale red and transparent-looking. It
possesses one or more endonucleoli.
Cells of the Third Glandular Layer Two to Three
Days after Feeding. Fig. 34.
The Cell Wall.—The external wall is pale blue, the internal
wall deep blue.
The Cytoplasm stains pale blue.
The Nucleus is irregular in shape, slightly oval, and stains
blue.
The Nuclear Chromosomes occur as a number of scattered
granules, each apparently double.
The Nuclear Plasm consists of blue granules, arranged in
rows. It is rather scanty, and stains much less deeply than
the nuclear chromosomes.
The Nucleolus is large, stains purplish red, and contains one
or more endonucleoli.
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 423
Cells of the Third Glandular Layer Seven Days
after Feeding. Fig. 35.
The Cell Wall is pale blue.
The Cytoplasm is pale blue and somewhat granular, as it is
in leaves that have never been fed.
The Nucleus is spindle-shaped in some tentacles, but not
drawn out at the ends into such fine acicular points as in con-
trols. The spindle-shaped nuclei are coloured blue like those
of controls. Nuclei like fig. 34 are also common.
The Nuclear Chromosomes are exactly as in specimens that
have never been fed.
The Nuclear Plasm consists of blue granules like those of
control specimens, arranged in moniliform rows. In some
nuclei the appearance is most suggestive of a spiral arrange-
ment of these chains of granules round the nucleolus.
The Nucleolus is large and deep red as in controls, with
clear endonucleoli, and surrounded by Frommann’s clear
zone.
In general appearance the cells, as a whole, are very lke
those in the control specimens.
SUMMARY AND ConcLUDING REMARKS.
As the result of an appropriate stimulation of the gland-
cells of Drosera by food, it is seen that the basophile cytoplasm
becomes used up, and is represented ultimately by a very
scanty eosinophilous plasm. The restoration of the cytoplasm
is brought about by the nucleus absorbing food material,
metabolising it, and then excreting it into the cytoplasm, as is
proved by—
(1) The great increase in the bulk of the basophile nuclear
chromosomes during a period preceding the restoration of the
cytoplasm.
(2) The scanty amount of eosinophilous nucleolar chro-
matin, and small size of the nucleolar organ during the same
period.
(3) By the neutrophile cytoplasm first appearing immedi-
424 LILY HUIE.
ately round the nucleus; and this extra-nuclear plasm cor-
responding in colour reactions (chemically) and in size of
granules (morphologically) with the intra-nuclear plasm.
(4) By the diminution of the nuclear chromatin, and in-
crease in the amount of nucleolar chromatin, whenever the
cytoplasm has been restored.
That the extra-nuclear plasm undergoes a further change on
its own account seems to be proved by the purple colour
gradually being changed into blue.
Whether the nuclear or the nucleolar chromatin is the
primary product of metabolism; whether they are formed
simultaneously by the nuclear or nucleolar organs respectively ;
whether the one or the other plays the part of a ferment-
secreting organ in relation to nuclear metabolism, are questions
only to be answered after further investigation.
The aggregation of the nuclear chromatin into a definite
number of V-shaped segments—eight in Drosera—proves that
such a change is not a feature characteristic of mitosis, but
simply a sign of great activity in the nuclear organs.
I intend continuing my research on Drosera with the view
of determining whether food materials which differ chemically
will produce characteristic changes in the gland-cells.
DESCRIPTION OF PLATES 23 and 24,
Illustrating Lily Huie’s paper, “ Changes in the Cell-organs
of Drosera rotundifolia produced by feeding with
Egg-albumen.”’
All the figures were drawn with Zeiss’s camera lucida, Zeiss’s 3, apochro-
matic oil immersion objective, and No. 8 compensating ocular. The tube
of the microscope was drawn out to its full extent, and the drawing-paper was
laid on the table. The figures of the lateral superficial cells, and those of
the cells of the third glandular layer, were almost invariably taken from the
same gland, as the corresponding figures of the apical gland.-cells,
CHANGES IN CELL-ORGANS OF DROSERA ROTUNDIFOLIA. 425
Fries. 1—12 show the cytological changes in the apical gland-cells, pro-
duced by feeding with egg-albumen :
Fig. 1, The resting condition.
Fig. 2. One minute after feeding.
Fig. 3. One minute after feeding. Slightly more advanced.
Fig. 3a. Appearance of semi-boiled white of egg, fixed in picro-corrosive
and stained exactly like the sections. No blue colour visible. Granu-
lation much finer than in the cells.
Fig. 9B. Nucleus from cell, fed for thirty hours on yolk of egg.
Fics. 13—24 show the cytological changes in the lateral superficial gland-
cells.
Fig. 13. The resting condition.
Fig. 14, One minute after feeding.
Fig. 15. One minute after feeding. Somewhat more advanced.
Fics. 25—35 show the changes in the nucleus of the cells of the third
glandular layer.
Fig. 25a. Nucleus cut longitudinally.
Fig. 258. Nucleus cut transversely.
Fig. 26a. Nucleus cut longitudinally.
Fig. 268. Nucleus cut transversely.
eae
\ ae ane
i a ayy re PA itt; ae
pees a4 ithe os np ae fa. Col Seo Jae
¥ APs ~~ ay aa) », a oe a at, ah 4% an “a _ Vaary
/ ge Sh Sas Set Bega
< '
A ; a ’ a > vi ry 4 Thy ih ar eye? ru ‘aie
Potted! at rulad®
(7%) ie jae. ep Tf) 1? 7 oul ay
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 427
Observations upon the Development and Suc-
cession of the Teeth in Perameles; together
with a Contribution to the Discussion of
the Homologies of the Teeth in Marsupial
Animals.
By
J. T. Wilson, M.B.,
Professor of Anatomy,
and
4
J. P. Hill, F.LS.,
Demonstrator of Biology in the University of Sydney, New South Wales.
With Plates 25—32.
Part I.—IntrRopvuction.
REcENT years have witnessed a noteworthy revivai of interest
in the morphological interpretation of the facts of mammalian
dentition. This renewed interest is traceable to the more
systematic and thorough-going application of the methods of
embryological study to the distinctively morphological, and no
longer merely to the histogenetic aspects of the phenomena of
tooth development. These methods have yielded a valuable
supplement to the work of the systematist and paleontologist,
and have provided a very necessary basis for a critical estimate
of views arrived at by investigators working along such dif-
ferent lines of research.
As a result of the fresh activity to which we have alluded
there has sprung up a pretty copious literature, which, added
to that previously existing, renders the study of the problems
of dentition at the present day by no means a light one,
42.8 J. ZT: WILSON. AND J, P. HIth.
It is not, however, the object of the writers to attempt to
give any comprehensive sketch of the evolution of the problems
met with in such a study. Nor do they profess to supply any-
thing lke a complete and adequate discussion, even of the
present position of these problems. Their aim has from the
outset been a much more limited one, and they feel absolved
from making any such attempt all the more that general
literary surveys of the kind referred to, along with sufficient
bibliographical data, have been published comparatively re-
cently by more than one distinguished authority in this
department of science. Special reference may here be made
to the address by Professor H. F. Osborn at the American
Association in August, 1893, specially important from the
palzontological side, to the masterly résume given by
Professor Gustav Schwalbe in an address to the “ Ana-
tomische Gesellschaft” at Strassburg in May, 1894; to
the brief but lucid and interesting account by Mr. M. F.
Woodward in ‘Science Progress’ for July, 18945; and, lastly,
to that given by Professor W. Leche in his recent important
monograph on “ mammalian tooth-development in the ‘ Bib-
liotheca Zoologica,’ 1894-5.
But although in the following pages no effort is made after
historical completeness, abundant reference must necessarily
be made to the various phases of scientific opinion upon the
most important dentitional questions, and we trust that our
statement of these questions will be found to be not only accu-
rate, but sufficiently detailed for the purposes of discussion.
In spite of the substantial enlargement of our knowledge of
the mammalian dentition attained of late years, it must be
admitted that no final settlement of some of the more im-
portant issues has yet been arrived at. On the contrary,
certain of the more novel phenomena recently brought to light
have so far served rather to complicate than to simplify the
problems involved.
No mammalian order has called forth more discussion in
respect of its tooth equipment than the Marsupialia. Pri-
marily, no doubt, this is owing to the inherent peculiarities of
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 429
the dentition in this order. But in part also it is due to the
fact that the more or less peculiar characteristics of the denti-
tion in the Metatheria (and, a fortiori, in the Proto-
theria) have, on account of the usually accepted zoological
position of the group, been approached with a quite special
interest, as likely to afford important clues to the phylogenesis
of the teeth in Mammalia generally. A consideration of the
literature of this branch of the subject might indeed suggest
that something like unanimity has been attained upon several
important matters, e.g. the serial homology of the deciduous
premolar of Marsupials with the more anterior premolars. As
will appear in the sequel, however, it is the opinion of the
present writers that, even on the points referred to, the dis-
cussion is very far from being finally or satisfactorily con-
cluded. As regards several other weighty questions, no
pretence of agreement is made by various investigators; and
in not a few instances the accounts of what are obviously
identical phenomena reveal actual discrepancies of observation,
as well as very diverse interpretations of those facts in regard
to which their observations do coincide. It has accordingly
appeared to the authors highly desirable that the body of fact
and observation available for the construction of adequate
hypotheses concerning the marsupial dentition should be
further supplemented from the developmental side ; and, being
comparatively favorably placed for obtaining suitable ma-
terial, they have been induced to undertake the task of, at
least, providing further reliable records of observations upon
tooth development in Marsupials. With this object the poly-
protodont genus Perameles was chosen in the first instance,
and P. nasuta has throughout furnished the major part of
our material, and formed the basis of most of the descriptive
part of this paper. Other forms, however, have been examined
with reference to particular points, e.g. Dasyurus viver-
rinus, Phascologale cristicauda, Trichosurus vulpe-
cula, and Macropus ruficollis.
Perameles was specially chosen as our type for study and
description as representing a fairly generalised marsupial type
430 J. T. WILSON AND J. P. HILL.
which has not hitherto been made the subject of any complete
or extended investigation, though several authors (Rése [1],
Woodward [2], and Leche [3] ) have made reference to obser-
vations made by them on isolated individual stages of its
development.
The series of specimens at our disposal has been a fairly
complete and satisfactory one, and in its collection we are
indebted to the following HERLESET for material generously
placed at our disposal :
The Trustees of the Australian Museum, Sydney; Pro-
fessor W. A. Haswell, University of Sydney; Professor W.
Baldwin Spencer, University of Melbourne; Messrs. C. W.
De Vis, Brisbane Museum; A. G. Hamilton, George Masters,
A. M. lea, and Dr. R. Broom.
Review of Past and Present Opinion.—Whilst we
have disclaimed any attempt after historical completeness, we
nevertheless deem it expedient, or even imperative, before
attempting to set forth our own observations, to make such
reference to past and present views as may be necessary to pro-
vide a basis for the future discussion of our own observations.
In 1867 the law of the succession of the teeth in the Mar-
supialia was definitely formulated by Flower (4).
This author claimed to show by means of his own researches
and those of others that, wherever a tooth-change can’ be
shown to occur in marsupial animals, such tooth-change in-
volves only one tooth—the last of the premolar series.
This view being accepted, the question naturally arose—
“to which set of teeth in non-marsupial Mammals do the
non-changing antemolar teeth of Marsupials correspond ? ”
Flower concluded that they answered to the permanent or
successional teeth of higher Mammalia, and that thus only
one true milk-tooth was present in any Marsupial. This in-
terpretation, taken along with the more primitive character of
the marsupial organisation in other respects, led him to suggest
that “ the milk or deciduous teeth may rather be a set super-
added to supply the temporary needs of Mammals of more
complex dental organisation ” (4, p. 639).
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 431
The views of Flower were subsequently adopted and ex-
tended by Oldfield Thomas (5), who adduced in their support
certain important paleontological facts.
Upon the relation of the Eutherian to the Metatherian
stocks, Thomas is notably explicit. He considers that “the
change from Metatherian to Eutherian” occurred, “in all
probability,’ within the limits of the family Dasyuride.
His conclusions, therefore, that “the rudimentary tooth-
change of Marsupials represents an early stage in the first
formation ” of a complete change, and that the Marsupials
are “still in a backward condition out of which the Kutheria
have long since passed,” are of quite fundamental significance
in his interpretation of the phenomena of the dentition of
Mammals generally.
The general argument in favour of Flower’s view of the
primitive character of the single tooth-change in Marsupials,
derived from the zoological position of the order, is expressed
thus by Thomas :—“ When we consider that in every character
of their organisation the Marsupials are infinitely behind and
at a lower stage of evolution than the placental Mammals, it
would appear to be a total subversion of all the ordinary rules
to suppose that in this one character of their dentition they
should have passed on in advance of all the other Mammals,
and, having gone through the condition in which the latter
now are, should have again nearly evolved away that process
of tooth-change which is to its placental possessors so evidently
advantageous ”’ (5, p. 450).!
A more special argument is also derived from the fact “ that
five out of the six families of Marsupials, natives both of
Australia and America, have, with the comparatively unim-
portant exceptions already noted as occurring among the
Dasyuride, arrived at precisely the same stage of tooth-
change ”—a circumstance which would be unlikely to occur if
1 It may be remarked that this is precisely what has actually occurred in
Dasyurus and Sarcophilus in relation to other Dasyuride, according
to Thomas himself, i.e. after first evolving a tooth-change (single, of course),
they have next proceeded to get rid of it.
432 J. I. WILSON AND J. P. tbh,
the modern tooth-change were a remnant of a fuller one, for
then “ we should naturally expect that, under the very various
conditions of the struggle for existence, equally various
degrees of reduction would have been attained to.” It may
be pointed out that the cogency of this argument would
entirely disappear if it could be shown to be probable that the
(hypothetical) reduction was dependent upon conditions of
life common to, and peculiar to, the entire group.
In professing his firm adherence to Flower’s view that the
teeth of Marsupials in front of the last premolar represent the
permanent teeth of other Mammals, Thomas states that he
was led to that opinion by “ finding the impossibility of work-
ing out the general homologies of the teeth on the basis of
the opposite view,” and by very extended observations of
specimens.
It is perhaps unnecessary to follow Mr. Thomas through
his somewhat elaborate and rather far-fetched doctrine of a
retardation which ought to have occurred in Marsupials
if their teeth had formerly possessed milk predecessors which
were subsequently lost,—a retardation parallel to that which
occurs in the case of the last permanent premolar; and which
he alleges as also occurring in the case of the first incisor in a
number of Marsupials, probably by way of preparation for the
acquirement of a milk-tooth. Thomas himself remarks upon
‘the difficulties in the way of understanding how the ordinary
processes of evolution” could first have “ brought about such
a preliminary retardation,’—a remark with which his readers
will readily agree.
The above views have now an interest for us which is largely
historical, since Thomas has explicitly surrendered the main
point of his position in view of later discoveries.
The views of Flower, prior at least to Thomas’s advocacy of
them, had not as a whole received general approval. His
identification of the non-changing antemolar teeth of Marsu-
pials with those of the permanent series of Placentals was,
indeed, practically universally received. The view, however,
that the seemingly almost complete monophyodontism of
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 433
Marsupials was a primitive condition, and that the original
mammalian stock was actually monophyodont, had not met
with general acceptance previous to the publication of Thomas’s
paper in 1887. Notably Winge (6)! maintained, as against
Flower, that in all probability the condition in Marsupials was
secondary, and due to an almost complete loss of the milk
dentition in that group; and this opinion of Winge possessed
many adherents.
The emphatic reiteration of the views of Flower by so emi-
nent an investigator as Oldfield Thomas, upon the grounds of
his own extensive researches, seems to have served as a stimulus
to renewed consideration of the subject of marsupial dentition,
especially from the developmental point of view. The first-
fruits of the latter research appeared in December, 1891, in
the form of a paper by Kiukenthal (7) on tooth development
in Didelphys.
In his introductory remarks Kikenthal indicates the Flower-
Thomas theory of the primitive character of the single tooth-
change in Marsupials as the prime consideration with refer-
ence to which his investigations were directed. Thomas had
admitted that a discovery of successional germs accessory to the
non-changing marsupial (antemolar) teeth would be fatal to
the theory in question, and the result of Kiikenthal’s exami-
nation of Didelphys was to reveal the presence, at the lingual
sides of nearly all the developing teeth in that animal, of
epithelial ingrowths of the primitive dental lamina. These
ingrowths he naturally interpreted as rudimentary enamel-
germs of successional teeth, in series with the undoubted
enamel-germ of the actual successional premolar (p 3, i.e.
pm. 4 of Thomas).
Kukenthal thus believed that he had established on a per-
fectly secure basis, the view that the existing teeth of Marsu-
pials in front of the last premolar are in reality milk-teeth in
series, not with the persisting last premolar, but with its milk
1 We have been unable to consult Winge’s original paper, and have had to
rely for our knowledge of his views upon the accounts of Kiikenthal and
Leche.
434 J- T. WILSON AND J. P. HUGE.
predecessor. Kiikenthal’s claim was promptly admitted by
Thomas himself (8) as apparently incontrovertible, and he
therefore no longer refuses to subscribe to the view of a primi-
tive diphyodontism of mammals, ‘‘ probably in direct suc-
cession to the irregular polyphyodontism of their reptilian
ancestors,”—a diphyodontism which “ may even have existed
in what were in other respects members of the latter class.”
Nevertheless Thomas is far from admitting that the whole
problem of the phylogeny of the dentition in Mammalia
is thereby cleared up. He regards, e.g., the fact that the
Mesozoic Triconodon changed a single tooth only, as ren-
dered inexplicable by the newer theory. Still he definitely
adopts the latter, merely passing on to a criticism of Kiken-
thal’s views upon other issues, i.e. the homologies of the
individual premolars, and the questions of the nature and mode
of origin of the molars.
An important and elaborate memoir on the development of
the human teeth by Carl Rose was published in 1891 (9), and
apparently independently of Kiukenthal, and contempora-
neously with him, Rése had also worked out the tooth develop-
ment of Didelphys as well as of several other marsupial
forms. The results of this work were published in September,
18921) 2
As regards the question of the presence of rudimentary
enamel-germs of successional teeth, Rose largely confirms the
observations of Kiikenthal upon Didelphys,” and he showed
that lingual ingrowths of the dental lamina, similar to those
found in Didelphys, were present in all other Marsupials he
examined.
During 1892-3 Rose continued to publish the results of an
active investigation into the phenomena of tooth development
in a number of other forms, both mammalian and reptilian.
1 Rése’s views upon the marsupial dentition were, however, outlined in his
paper on tooth development in Edentata, published in July, 1892 (34).
2 An unfortunate error in Rose’s paper, subsequently acknowledged by
him, was his mistaking the enamel-organ of the deciduous premolar for that
of the first molar.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 435
Rose has also pushed boldly into the realm of theory, and
has presented hypotheses to account not only for the more
divergent types of mammalian dentition, but for the derivation
of the latter from that of polyphyodont and homodont
ancestors. We cannot avoid the feeling that in his views ~
upon these matters the quantity of theorising is somewhat
disproportionate to the extent of the basis of fact and ob-
servation. His schematic representation (29), for example,
of the mode of derivation of the human dentition from a lowly
reptilian-like type appears to us to be a rather striking
instance of a highly ingenious but a somewhat uncritical
exercise of the scientific imagination.
Rose’s contributions to our knowledge of tooth development
are, however, both extensive and of very great interest and
importance, and his interpretations are, as a rule, entitled to
the utmost respect.
As regards the mammalian dentition, Rése is a powerful
advocate (a) of the primitive diphyodontism of the mam-
malian stock, (0) of the doctrine that the multituberculate
teeth of the heterodont dentition originated by a process of
fusion of several individually distinct and simple conical
members of a homodont dentition.
In these views Rose has, of course, had many predecessors,
Kikenthal among others, but Rose’s attempts to further estab-
lish the validity of these propositions deserve special recognition.
As regards the marsupial dentition in particular, Rése’s
general conclusions may be thus summed up :—That dentition,
like that of mammals generally, was originally diphyodont:
** Allein die Thatsache, dass hinter allen Zahnen zeitweilig eine
zusammenhingende, bandférmige Zahnleiste verlauft, beweist
uns, dass die Beuteltiere von diphyodonten Saéugern und
weiterhin von polyphyodonten reptilienihnlichen Vertebraten
abstammen” (34, p. 508). And again: ‘Das Milchgebiss
der Siugetiere ist nicht eine Neuerwerbung, sondern eine
phyletische Vererbung und ist entstanden durch zusammen-
draugen mehreren reptilienahnlichen Zahnserien in eine
Einzige” (ib., p.509). Thus it is held that during phylogeny
VOL. 39, PART 4,—NEW SER. HH
436 Ju T. WILSON AND J.P. Hib,
the Marsupials have suffered loss of all the teeth of the “ per-
mauent” series in front of the last premolar, with, indeed, the
doubtful exception, according to Rose, of the last upper incisor
in some forms.
In this connection Rése remarks upon the “ noteworthy con-
stancy of the group with reference to the single tooth-change
from Mesozoic up to modern times—a constancy which had
seemed to Thomas so inexplicable,—and he offers the following
explanation :—“ Die Beuteltiere sind aber bei der Reduktion des
vielfachen Zahnwechsels der reptilienahnlichen Vorfahren der
heutigen Sauger gleichsam iiber das Ziel hinausgeschossen und
haben sich in eine Sackgasse verrannt, aus der kein Ruickweg
moglich ist” (1, p. 705). He also suggests that such a reduc-
tion could only be serviceable if at the same time the milk-
teeth came to grow from persistent pulps,—a condition which,
he remarks, is attained only by the wombat.
Rose agrees with Kikenthal in referring the molars to the
same series as the other teeth, i.e. to the milk or first denti-
tion, if they are to be referred specifically to one or other of
the two series represented by the antemolar teeth. Into his
later view respecting the possibility of the molars representing
*end-members of separate dentitions’’ we need not now enter.
In 1893 a highly interesting and important paper was pub-
lished by Rose (11) on the subject of tooth development in
Phascolomys. We are of opinion that the full significance
of the observations therein recorded has been apprehended
neither by subsequent writers nor by the author himself, though
the latter undoubtedly attached considerable importance to
them.
The results of Roése’s research (for which unfortunately only
one pouch specimen, of 19 mm. body-length, was available)
would seem to tend towards a confirmation of a long discredited
statement of Owen’s (15), to the effect that the incisor teeth
and the first molar are changed in the young animal. It
appears from Rose’s investigation that at a comparatively early
1 This view regarding the nature of the last polyprotodont incisor has
received no confirmation whatever from any later observer.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 437
stage in its tooth development the wombat possesses rudiments
of teeth entirely unrepresented in the adult dentition; and
that, further, certain of these teeth are precociously calcified,
and, according to the author, obviously belong to a generation
antecedent to that to which the permanent teeth of the animal
belong. In addition to this he finds, in a position in the jaw
corresponding to that occupied later by the single adult pre-
molar, a single pointed, precociously calcified tooth, provided
with a large and “ bulbous” lingual germ (“Ersatzleiste”),
which latter he regards as the true germ of the permanent
premolar; the rudimentary but well-formed calcified tooth
beside it being the deciduous predecessor to which Owen
apparently referred.
In summing up his results Rose first emphasises the fact of
the possession by the wombat of two typically distinct den-
titions “ ganz iihnlich wie die placental en Sauger.”” Of these
two incomplete series he points out that all the members of
the first, and several of those of the second, must be resorbed
in early life.
He then proceeds to put the important question, ‘ which of
the two tooth-series of wombat corresponds to the milk series
of other Marsupials?”? And just here he finds himself in-
volved in some little difficulty. For, while he finds it easy to
homologise the posterior elements of the series with those
found in other Marsupials by reckoning the “ Ersatzleiste”
of the small calcified “‘ premolar” as the germ of the persist-
ing one, he feels almost compelled to identify the calcified and
precocious germs of the first generation in the anterior seg-
ment of the jaw with the “milk” (persisting) dentition of
other Marsupials. He is thus led to postulate a fundamental
difference in homology between the teeth of the anterior and
posterior segments of the jaw respectively, and thus also be-
tween the anterior persisting teeth of wombat and those of
other Marsupials. “ Wahrend die Schneidezahne der polypro-
todont Beutler zur ersten oder Milchzahnserie gehoren, rechnen
diejenigen vom Wombat zur zweiten oder bleibenden.” He
adds, however, that it is not improbable that the condition
438 J. T. WILSON AND J. P. HILL.
found in wombat may ultimately be found characteristic also
of some other diprotodont Marsupials.
It will be noted that if the view for which Rose indicates
his preference be correct, the acceptance of it implies at least
a partial surrender of what we may call the Kiikenthal-Rose
position in relation to the marsupial dentition generally.
In his recent work (8, pp. 100, 101), Leche has definitely
adopted that interpretation of Rése’s observations which Rose
himself somewhat reluctantly put aside, and has thus attempted
to conserve in its integrity the ruling modern theory of the
marsupial dentition as a true ‘ milk ”’ series.
The solution which Leche confidently advocates involves the
interpretation of the prematurely calcified teeth of wombat,
not as homologues of milk-teeth, but as vestigial remains
of a “prelacteal” series inherited from the polyphyodont
ancestors of the Mammalia.
It will appear in the sequel that the present writers are
very strongly inclined to adhere to that interpretation of these
rudimentary teeth which Rose decides to accept in the paper
under consideration. But in so doing it will appear that they
also advocate a much wider extension of the hypothesis in-
volved in that interpretation than was contemplated by Rose.
In the ‘ Proc. Zool. Soc.,’ May 2nd, 1893, there appeared a
paper by Mr. M. F. Woodward (2) on ‘“ The Development of
the Teeth of the Macropodide,” in which the author gave
an account of his discovery in certain members of that family
of a number of small calcified teeth supplementary to the
proper rudiments of the adult dentition.
Thus in the upper jaw of Petrogale he found, in addition
to the germs of the three adult incisors, three “ minute
calcified rudimentary (or rather vestigial)” teeth. So, again,
in the case of the lower jaw he found two vestigial teeth
in addition to the enamel-germ of the permanent lower
incisor. From his facts Mr. Woodward draws certain
conclusions regarding the relations of the dentition of
Macropods to that of Polyprotodonts and primitive Mar-
supials, with which we are not here concerned. It is to be
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 439
noted that the author refers the adult incisors of Macropods
to the first dentition, accepting, as he does, the criterion
which had seemed sufficient to Kiikenthal and Rose. He
says, ‘‘If these various and often minute cord-like down-
growths of the dental lamina (lingually to the enamel-germs
of the adult incisors) are to be interpreted as representing
rudiments of teeth, as seems probable from comparison with
the known rudiments of the first or second dentition in other
mammals, then we find that in the kangaroos the incisor
teeth all belong to the first dentition, that the relations of the
canines are uncertain,” &c. &c. Again, on account of the
presence of a “distinct but small” lingual downgrowth by
the side of the enamel-germ of the foremost upper vestigial
tooth, Woodward referred that tooth to the first dentition
equally with the germs of the adult teeth; and he extends the
same determination to the other two, partly from their
general analogy to the first, and partly on account of certain
observations as to their relative position in reference to the
dental lamina, and to the enamel-germs of the neighbouring
adult teeth.
Similar reasonings prevailed with him in regard to the
lower incisors, and accordingly he is led to interpret the
vestigial incisors simply as decadent members of the same
series to which the adult teeth belong.
Mr. Woodward’s paper constitutes the earliest publication
in which were recorded observations upon the presence of
undoubted embryonic vestigial teeth in Marsupials. Rdése’s
paper, dealing with such structures in wombat, was not pub-
lished till 4th August, 1893.
On the same day there appeared in the ‘ Morphologisches
Jahrbuch,’ Bd. xx, a paper (12) by Professor W. Leche, which
bore the date January, 1893. This contribution was supple-
mentary to his paper (13) in the preceding volume on the
subject of mammalian tooth-development, and it contains an
account of his researches into the development of the teeth of
Myrmecobius, to which in his previous paper he has
referred as still in progress,
440 J. T. WILSON AND J. P. HILL.
In Myrmecobius Leche had discovered the presence of
calcified structures connected with the dental lamina, and
placed labially with reference to it and to the enamel-germs
of the adult teeth. These calcified tooth-remains, though less
perfectly formed, or perhaps we should say more completely
degenerated, than those described by Mr. Woodward, bear a
striking general resemblance to his, both in respect of their
topographical relationships and of their structural features.
Leche’s interpretation of them was, however, widely different
from that given by Woodward in the case of the Macropod
vestigial teeth, for he viewed the calcified structures as the
sole remains of an entire “ prelacteal” dentition which had
otherwise become suppressed. To this view Leche still
adheres in his latest work (3), where he claims that both
Woodward’s and Rose’s observations just referred to are to be
explained along similar lines.
Woodward himself (14) has accepted Leche’s notion
of a “prelacteal” dentition, so far at least as Myrmeco-
bius is concerned, and he believes that he has obtained
confirmation of the view through a similar discovery in
Phascologale.
By far the most important recent contribution to the lite-
rature of this subject is the comprehensive monograph by
Professor W. Leche already referred to (3).
In addition to a rich collection of observations copiously
illustrated, this book contains a systematic discussion of all
the more important issues raised by a study of the ontogeny of
the mammalian dentition. In the latter portion of this paper
Leche’s views are dealt with in detail in so far as they have a
bearing upon the subject-matter of this work. Meanwhile we
may remark that one of the most important features of Leche’s
contribution is his criticism of the commonly prevalent but
loose and unreliable notion as to what really constitutes a
morphological tooth-germ.
Such a critical determination is urgently needed if any
further progress is to be made in investigation of the dentition
along the lines of embryological research. In the development
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 4A1
of our own views we have been largely stimulated and aided by
Leche’s masterly essay.
GENERAL ACCOUNT OF THE LEADING CONCLUSIONS FROM OUR
OWN INVESTIGATION.
In approaching the subject of our own researches one or
two matters call for special remark.
In the first place it is necessary or desirable that we should
indicate the general grounds upon which we have felt compelled
to depart from the now usually accepted views of the marsupial
dentition. The attitude we have felt bound to adopt practi-
cally implies a return to Winge’s views on the main question
involved; but with the notable corroboration of that theory
derived from our interpretation of the so-called ‘ prelacteal”’
vestigial teeth as remains of the true milk series, which,
according to Winge, had been entirely lost from the marsupial
dentition, except in the case of the deciduous last premolar.
At the time our research was begun, and for some time
afterwards, we were fully persuaded of the truth of the
Kiikenthal-Rése theory, and found no reason to question
Leche’s identification of his vestigial teeth in Myrmecobius
as ‘“prelacteal,’ to which category we felt, with Leche,
strongly disposed to refer the vestigial calcified teeth dis-
covered by Woodward in the Macropodide.
We were accordingly highly delighted to (apparently) con-
firm the accepted views by the further discovery in Pera-
meles of a number of quite similar “ prelacteal”? rudimentary
teeth.
The examination of the earlier stages of tooth development
in Perameles, however, forced upon us irresistibly the con-
viction that the deciduous premolar in that animal
must belong to the same series as the so-called
* prelacteals.”
For in Perameles the deciduous premolar is differentiated
from the dental lamina contemporaneously with the so-called
prelacteal”’ teeth, and at a period prior to the differentiation
442 J. T. WILSON AND J. P. HILL.
of any other tooth of the adult dentition. This early forma-
tion of the deciduous premolar is not peculiar to Perameles.
It is important to note that it occurs also in Didelphys.
But its precocity cannot be explained in Perameles, as it
usually has been in Didelphys, by a reference to the future
size of the tooth, since, as is well known, it isin Perameles
by far the smallest of the whole cheek-tooth series, and, in-
deed, is absolutely a small-sized tooth. Neither can the early
development be explained by its position in the jaw. Indeed,
from its relatively posterior position in the young and develop-
ing jaw, quite the contrary, i.e. a retardation, might have
been expected, such as affects the hinder (molar) elements of
the dental series. The discussion of this important matter
will be resumed further on in this paper. So much it was
essential to state at the outset to render our position in-
telligible.
It was only with great reluctance that we arrived at this
apparently inevitable conclusion that the deciduous premolar
was a member of the “ prelacteal” series, and although it
seemed impossible to interpret the facts in the case of Pera-
meles in any other way, it seemed doubtful whether we
should be able to reconcile this view satisfactorily with the
facts observed in the case of other Marsupials, more particu-
larly in Didelphys. At least, it seemed difficult to do so in
a way that would carry conviction to those whose views on the
marsupial dentition were moulded upon the conditions repre-
sented in Didelphys. In that animal it has indeed been
shown (cf. Kikenthal [7] ) that the deciduous premolar is, at
the first, the most advanced of all the teeth; but the subse-
quent course of its development, resulting in the production
of a relatively large molariform tooth, seemed anything but
favorable to a theory according to which it must be regarded
as belonging to an otherwise degenerated “ prelacteal ” order
of teeth.
It may now appear somewhat curious that the very obvious
possible explanation of the whole difficulty did not dawn upon
us at the very outset. But so completely were we under the
DEVELOPMENT AND SUCOERSSION OF TEETH IN PERAMELES. 448
influence of the ruling theory of the homology of the perma-
nent marsupial dentition to the milk dentition of higher
Mammals, that it was some little time ere it occurred to us
even to call in question the grounds of this identification, and
at the same time the “ prelacteal”’ character of the vestigial
teeth so frequently met with in Marsupials. But we very soon
saw that, could the lacteal theory of the permanent mar-
supial dentition be successfully impeached, all difficulties
would be at once removed from the theory we felt bound to
advocate from a consideration of the condition in Perameles,
where the deciduous premolar and the so-called “ prelacteals ”
evidently belong to one and the same dentitional series. This
series would of course no longer be regarded as “ prelacteal,”
but as the true homologue of the milk series of the higher
Mammalia.
It was while in this mental attitude to the question that we
were first appealed to by Leche’s sceptical critique (3, pp. 132,
136, &c.) of the criteria usually relied upon as sufficient to deter-
mine whether or not any given lingual downgrowth of the
dental lamina is morphologically the enamel-germ of a
replacing tooth. For the fuller elaboration of our final
attitude to the questions thus suggested, and for the evidence
upon which our conclusions are based, we must refer to the
subsequent sections of this paper. Here it is only necessary
to say that we have been led to reject the presently prevailing
Opinion regarding the more or less swollen downgrowths
formed at the lingual sides of the developing marsupial teeth.
It is our emphatic opinion that these are not the repre-
sentatives of the replacing teeth of higher mammals, nor,
indeed, are they actual individual tooth-germs at all, but mere
residual appendages of the lamina, though they may, in a
sense well recognised by Leche, contain the “ promise and the
potency” of a possible tooth-generation. But the adult teeth
in front of the last premolar are the genuine homologues of
the replacing teeth of higher mammals, and the milk-teeth of
the latter are represented, in front of the deciduous premolar,
by the so-called “ prelacteal” vestiges, of which, as we shall
444 J. T. WILSON AND J. P. HILL.
have occasion to notice, there are abundant remains in
Perameles.
We are convinced that this view is amply justified by the
considerations set forth in this paper, and that, when
deliberately weighed, it will at once be found agreeable to all
the facts of development of the marsupial dentition, and to
afford a far more natural and unstrained explanation of these
facts than the hypotheses heretofore dominant. We cannot
but believe that the enormous simplification which would be
effected by the adoption (if satisfactorily demonstrated) of the
view we now advocate must be appreciated on every hand. In
its light the peculiarities of the marsupial dentition may be
rendered easily intelligible, and the difficulties of interpreta-
tion hitherto associated with these peculiarities forthwith
disappear.
In particular the new view would seem to afford the only
adequate explanation of Rése’s highly interesting discoveries
in the wombat. We have already seen how this observer felt
bound to interpret the rudimentary teeth he found in his
foetal specimen as belonging to a milk series; and although
Schwalbe and Leche have both subsequently expressed the
opinion that these rudiments are to be interpreted as ‘‘ pre-
lacteal,’’? we are of opinion that this view presents many
difficulties which are obviated by the simpler and more natural
view we feel bound to advocate on the grounds of our own
observations. We at least find it now impossible to adopt any
different conclusions on the points in question, respecting the
morphology of the marsupial dentition, than the following :
1. The permanent teeth of Marsupials are the homologues
of the permanent or replacing teeth of higher mammals.
2. The deciduous premolar is a true milk-tooth, and it is
not the sole representative of the series to which it belongs,
since the so-called ‘‘ prelacteal ” teeth are in reality milk-teeth
which have undergone reduction, and have well-nigh wholly
disappeared, under the operation of influences unfavorable to
their development.
3. The lingually situated downgrowths of the dental lamina
DEVELOPMENT AND SUCOESSION OF TEETH IN PERAMELES. 445
by the sides of the developing teeth are no rudimentary
enamel-germs. They are merely portions of a quite indifferent
“yesidual dental lamina” becoming liberated on differ-
entiation of the permanent teeth from the parent lamina. The
swelling of the distal portion of the downgrowth is a mere
thickening of the free border of the residual lamina, and exhibits
no differentiation which is really characteristic of the pro-
duction of actual enamel-organs.
It is possible that its presence is to be explained in terms of
a continued formative activity on the part of an unexhausted
dental lamina, an activity which might conceivably issue in
the production of enamel-germs of a third series, such as
Leche has shown exceptionally to occur amongst other
mammals. We are, however, unaware of any single instance
in which a differentiation genuinely characteristic of the for-
mation of an enamel-organ of a third dentition has been
observed among the Marsupialia.
We may here touch upon one other strong point in the
prima facie case we have been attempting to set forth in
favour of the serial homology of the deciduous premolar with
the so-called “ prelacteal” teeth. This is brought to light in
the investigation of those marsupial forms in which the de-
ciduous premolar is inconstant in its presence or altogether
absent from the normal dentition. Thus in Dasyurus viver-
rinus, in spite of the absence in the adult of any representative
of the last premolar, either deciduous or successional, we find
the deciduous premolar to be constantly present in the young
mammary foetus as a small and precociously calcified vestigial
tooth (figs. 80 and 81). It is indeed considerably larger than
the vestigial ‘‘ prelacteal”? teeth met with elsewhere, as in
Perameles, or evenin Dasyurus itself (see di 3, fig. 23).
But in all other essentials it agrees with the “ prelacteals,”
e.g. in time of differentiation, relative situation, and in the
fact of its absorption during foetal life.
An essentially similar condition we find in Phascologale
cristicauda, in which also no trace of any milk premolar
appears in the adult, This form, however, differs from
446 J. T. WILSON AND J. P. HILL.
Dasyurus in the presence in the adult upper jaw (though
only occasionally, according to Baldwin Spencer, 23, pp. 23
and 26) of the successional last premolar as a “ minute and
tubercular ” tooth (i.e. ‘‘p. 4,” cf. 22, p. 276). This tooth is,
however, absent from the adult lower jaw.
Thylacinus offers a still more advanced phase of develop-
ment, for in it the successional premolar is now a well-
developed tooth, larger than the second premolar (“p. 3” of
Thomas), and is preceded by a milk premolar, though that is
still only a minute and rudimentary one, and is shed during
infancy (22, p. 255).
In these respects Perameles may be regarded as standing
just above Thylacinus, for while the last permanent pre-
molar is well developed in both forms, its deciduous predecessor
in Perameles is a well-formed though still relatively a small
tooth, and is not extruded by the eruption of p. 3 (‘Thomas’s
“py, 4”) until the animal has attained about three fourths
adult size. The condition in Perameles may thus be re-
garded as itself intermediate between that in Thylacinus
and that in Didelphys, where the deciduous premolar is a
very large and multicuspidate molariform tooth.
It is beyond question that these several conditions repre-
sent stages in a process of reduction affecting first the
deciduous, and then the successional premolar, until the
Dasyurinz condition is reached.
Is it not in the highest degree probable that in this series
we have to recognise, as it were going on under our eyes, the
same process of reduction which, at an earlier epoch, has
brought the other antemolar milk-teeth down to the condition
of mere calcified vestigial structures—the so-called “ prelac-
teal”? rudiments? In other words, the deciduous premolar is
simply the hindmost member (so far as we know) of a dental
series exhibiting various stages of a retrogressive process, some
of the more anterior members of the series being represented
by various ‘ prelacteal ” rudiments,
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES, 447
Tue Question or NoMENCLATURE OF THE PREMOLARS.
After careful consideration we have not thought it well to
adhere to Thomas’s (5) numerical determination of the teeth in
the premolar region. That is doubtless an attractive theory,
which aims at enabling us to homologise the premolars of
modern Marsupials with those of the higher mammals; but
we are as yet unconvinced that the case has been sufficiently
made out.
Admitting the probability that one of the premolar series
has been lost in the ancestors of modern Marsupials, we
cannot regard Thomas’s contention that it is p. 2 which has
disappeared as placed beyond all reasonable doubt. Thomas’s
case for the homology of the last premolar of modern Mar-
supials to p. 4 of other mammals is made up of the following
factors :
(a) The existence of variations in the way of an occurrence
of the hypothetically missing p. 2; to which may be added—
(6) The alleged occurrence in ontogenetic development of
a possible rudiment of an enamel-germ in the position of
Thomas’s “ p. 2.”
(c) The probable phylogenetic relation of Triconodon to
modern Marsupials.
The present writers are, however, of the opinion that there
are too many uncertainties connected with each of these
factors to warrant our basing any system of nomenclature
upon a theory so conditioned.
With regard to the first of these, Bateson (21) has well
shown how indecisive is the evidence derived from a study of
tooth-variations in determining individual homologies of teeth.
In Bateson’s judgment ‘the system elaborated by Thomas
breaks down ; not because there is any other system which can
claim to supersede it, but because the phenomena of variation
are not capable of this kind of treatment,” because “it is not
possible to apply any scheme based on the conception that
each tooth has an individual homology which is consistently
44.8 J. T. WILSON AND J. P. HILL.
respected in variation.”” It will be noted that these conten-
tions do not necessarily affect our belief that the normal teeth
are individually homologous in different forms; but only
lead to the belief that the facts of variation do not yield reli-
able evidence in favour of such an homology in any given case,
If this be so, then the first and a very important support for
Thomas’s theory gives way.
Even less weight can be attached to the second considera-
tion adduced in support of Thomas’stheory. Rose does indeed
mention a possible rudiment of Thomas’s “ pm. 2” in Didel-
phys. But for this Kikenthal sought in vain, and, as a
result of a later and more extended research on Didelphys
and other marsupial forms, Leche formulates his judgment
upon the matter in the following passage:
“‘TIn diesem Zusammenhange mochte ich ausdricklich her-
vorheben, dass die ontogenetischen Untersuchungen bisher
keinen Aufschluss tiber die Homologien der einzelnen Zahne
der Beuteltiere und derjenigen der Placentalier gegeben habe.
Auch die von Thomas versuchte Homologisirung der Pramo-
laren der Beuteltiere gewinnt durch die ontogenetischen
Befunde keine Stiitze ” (8, p. 107).
Woodward, however, has recorded his discovery in Ma-
cropus giganteus (2, p. 463) of “numerous small enlarge-
ments and irregularities ” in the diastema between the canine
and the “third” premolar of the upper jaw, “some of which
may possibly represent the missing premolars.” This is vague
enough. But he goes on to say, ‘In the lower jaw, however,
there is a very distinct vestige of a tooth in the form of an
irregular enamel-organ with enamel epithelium and pulp (fig.
14). This, from its proximity to the third premolar, must re-
present pm. 2.” Since, however, there is another typical
marsupial premolar to be accounted for somewhere in this
situation, it seems to us not a little rash to pronounce the
enamel-organ in question—merely on account of its proximity
to p. 83—to be “pm. 2,” a tooth admitted to be otherwise
entirely absent from the marsupial dentition.
Like Kiikenthal and Leche, we have sought in vain for any
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 449
confirmation (in Perameles) of the view that Thomas’s
“pm. 2” is present in the developing marsupial jaw.
We must therefore hold with Leche that embryological in-
vestigation yields no support to the theory of the presence of
this hypothetical element of the marsupial dentition; and
though its absence may not be conclusive against that theory,
it forms an additional difficulty in the way of entertaining a
system of homologies based upon the ideal existence of such
an element. :
Only the third consideration quoted above in support of
Thomas’s nomenclature remains for consideration, viz. that
derived from the dental formula of Triconodon. This would
certainly seem to render it highly probable that the ancestors
of Marsupials were possessed of four premolars. Thomas’s
argument rests upon the usually accepted close phyletic or
even phylogenetic relationship between Triconodon and
modern Marsupials, but it also requires the identification of
the last premolar in the latter with “pm. 4” in the former.
Now both of these propositions may doubtless be justified as
highly probable, but neither, surely, can so far be regarded as
scientifically certain. And, in particular, we hold that the
evidence so far brought forward is insufficient to determine
which premolar, if any, has disappeared in the course of
evolution.
On the whole we believe it to be safer in the meantime to
designate the premolars simply numerically in the order of
their occurrence in the modern marsupial type, instead of
founding a system of nomenclature upon the condition of
Mesozoic forms whose precise zoological relationships can only
be inadequately determined.
And, in adopting in this paper the older method of numera-
tion, we do so the more readily that Thomas’s system has not
as yet gained general acceptance among Continental writers.
450 J. T. WILSON AND J. P. HILL.
MatveritaAL AnD Meruops.
For the purposes of our research we have had at our com-
mand fourteen different stages of Perameles young. Among
these stages P. obesula was represented by three specimens,
while the remainder of the material was furnished by P.
nasuta.
The youngest was an intra-uterine specimen of P. obesula,
rand from this early condition onwards we possess a pretty
complete series of stages.
The material representative of the more important stages
has been sufficiently abundant to enable us to obtain complete
series of sections in different planes.
Decalcification was effected chiefly by means of nitric
alcohol ; sections were cut in paraffin except in a few cases
where celloidin was preferred. The stains used were Gre-
nacher’s borax-carmine, hematoxylin, hematoxylin and eosin,
hematoxylin and picric acid; the hematoxylin stains were
chiefly used, and are greatly to be preferred, especially the
double stains named. We find a very dilute solution (very
faintly claret-coloured) of Renaut’s hzmatoxylic glycerine to
give the best results. The stain should be light if to be fol-
lowed by eosin, and deep (i.e. of considerably longer duration)
if to be followed by picric acid. The latter (as also the eosin)
is used in a weak solution in 90 per cent. alcohol. Sections
were generally stained on the slide, being fixed by Mayer’s,
or latterly by Maun’s albumen method.
1 Owing to the enormous number of serial slides we had to work with, each
slide containing very numerous sections, it was rather important to have a
short and convenient method of designating any particular section in our
notes and descriptions. We have found it exceedingly useful to employ a
18
~ 6-IN=E'
number of the section in any given row, counting from the left-hand end.
The Arabic numeral below the line refers to the number of the particular row
of sections on the slide, counting from the top. The Roman numeral indi-
cates the number of the slide in the series ; and the capital letter specifies
the series itself. For ordinary work the last may often be omitted. The
formula thus: Here the figure above the line indicates the
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES, 401
Part II.—Description oF Stages,
Stage I.—Perameles obesula: intra-uterine embryo.
Length from anterior convexity of head to curved hinder
extremity of body ' : ; . 875 mm.
Coronal and sagittal series examined.
This stage is that in which an allantoic placenta has been
described by one of us (H., 30).
The margins of the gap are not yet soldered together to
form the “ Saugmund.”
As regards its tooth development, this stage appears to
correspond pretty closely to that described and figured by
Rose as his earliest stage in Didelphys (1, p. 641, fig. 1).
The structure referred to by Rose as “ einer leicht ins Kie-
fermesoderm eingesunkenen Epithelwucherung,”’ and labelled
“7.1.” in his fig. 1, is present in our specimens in both upper
and lower jaws. In the upper jaw a very shallow groove like
that shown in Rose’s figure appears abruptly a little behind
the anterior end of the jaw. ‘This faintly indents the surface
of a slight thickened ingrowth of the oral epithelium into the
mesoderm. Both groove and solid cellular ingrowth when
traced backwards gradually fade away, the superficial groove
disappearing sooner than thesolid proliferating band of epithelial
cells.
In coronal sections of the lower jaw a lens-shaped thicken-
ing of the oral epithelium is first met with anteriorly. Soon
a shallow groove appears over this, and the thickening then
dips more deeply into the mesoderm, becoming almost tri-
angular in cross-section. Further behind, the groove becomes
description of a slide is thus easily condensed into a series of notes, each pre-
faced by a formula expressive of the individual section to which the note
refers. Frequently one wishes to refer to a number or groups of sections at
one time. ‘This is obviously to be accomplished by extending the upper line
18-26
6—-lI-E
sixth row of the third slide of series E is conveniently indicated. This
method is specially useful for reference from drawings to the sections
figured.
VOL. 39, PART 4.—NEW SER. 11
of the formula thus :— Here a group of eight sections in the
4.52 J. T. WILSON AND J. P. HILL.
slightly deeper and opener, the ingrowth at the same time
becoming thicker. Finally the groove opens out and dis-
appears, leaving only the thickened ingrowth, again lenticular
on cross-section, which in turn disappears a few sections further
back.
Rose’s fig. 1 (1) sufficiently illustrates the condition found
by us in this early stage.
Stage IJ.—P. nasuta: mammary fetus.
Length from vertex to root of tail . : . 17 mn.
Head length . : ; ‘ Tor as
Transverse (coronal) and sagittal series studied.
The dental lamina is already pretty fully developed, though
the differentiation from it of enamel-organs has not progressed
very far, so that only a very few enamel-organs can be recog-
nised as such, and there are relatively long stretches of quite
undifferentiated lamina. The latter is relatively thick (fig.
1, dl.), especially at its so-called “neck.” In other words,
the groove in the mesoderm occupied by the elongated mass
of ectodermal cells forming the lamina is a widely open one,
its lips not yet being approximated so as to thin out the
lamina towards its connection with the deep surface of the
oral epithelium.
Upper Jaw.—In the upper jaw the dental lamina extends
anteriorly almost, but not quite, to the mesial plane: mesially
it is separated by an interval from the lamina of the opposite
side. In this anterior incisor region the structure is best
studied in sagittal sections, for in the sagittal plane we obtain
nearly true transverse sections of the lamina for some distance
outwards. Fig. 6 shows a sagittal section parallel with the
mesial plane which just shaves the margin of the orifice of
Stenson’s duct (S.d.), and cuts the dental lamina (di/.) trans-
versely to its axis some distance in front of this. The lamina
is seen to be surrounded by young connective tissue, and the
whole is capped by an imperfect dome-shaped roof of bone—
the commencing alveolar portion of the premaxilla (pmz.).
This dome-shaped roof in turn causes an upward inflection of
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 453
the cartilage of the nasal floor (a/.) in this its most anterior
portion. The upward inflection of the cartilage of the nasal
floor is visible also in coronal sections some distance in front
of the naso-palatine foramen. In serial sagittal sections, as
one traces them outwards from the mesial plane, the dental
lamina soon after its commencement is seen to become dis-
tinctly thickened and enlarged. This enlargement extends
outwards for some distance, and occupies a considerable por-
tion of the anterior transversely lying portion of the lamina.
This we take to be the representative of the first upper incisor,
although there is as yet no differentiation of an enamel-organ
distinct from the dental lamina, and therefore the tooth, as
a distinct organ, cannot yet be said to have come into ex-
istence.
(It may not be out of place here at the outset to remark
that, where we have given interpretations of the facts and
phenomena observed [as distinguished from simple records of
observations], these interpretations are expressions of judgments
arrived at after careful collation of the facts, not only in the
stage under notice, but in those both preceding and succeeding.
An explanation may thus occur which would not be warranted
by the facts noted in the immediate context alone.)
Laterally to the swelling mentioned the lamina is again
shallower and less bulky for a short distance. Soon, however,
it again swells out to form an enlargement of considerably less
transverse extension than the first. This we believe to be the
representative of the second incisor. It is developed about
the point where the lamina begins to bend backwards in a
sagittal direction, as a comparison of coronal and sagittal sec-
tions indicates. Immediately external to it (or behind it) the
lamina bends much more acutely, and assumes definitely the
general sagittal direction which it maintains during the re-
mainder of its course backwards. And just at this acute
portion of its curve the lamina expands rather suddenly into a
large swelling, oval on cross-section, which is evidently repre-
sentative of the third incisor tooth. This swelling is of con-
siderable interest, because in connection with it we have a
454 J. T. WILSON AND J. P. HILL.
most important and striking feature present. This consists in
the differentiation, towards its anterior end, of a shallow but
quite definite mesodermal papilla, which indents its labial side,
forming the rudiment of the true milk predecessor of the per-
manent third incisor, whose Anlage forms the bulk of the
swelling.
The series of drawings (figs. 1—5) completely illustrates the
characters of this milk rudiment (di) and its relations to
the swollen portion of the dental lamina representing the
Anlage of the permanent tooth (¢3).
Thus fig. 1 shows the lamina (dl.) in advance of the 73
region destitute of any differentiation. In the next succeeding
section, fig. 2, the labial outgrowth (d23) has appeared very
abruptly growing out from the “ neck” of the lamina. In the
following section (fig. 3) the milk enamel-organ (di3) is
broader, and its relation to the lamina is slightly modified.
The connective tissue on its labial aspect promises to form a
papilla (mp.), and there is a slight tendency to cupping of the
enamel-organ itself. As yet the main portion of the dental
lamina exhibits very little modification, but from this point
backwards it swells considerably, so that in fig. 4, three sec-
tions further back, it is decidedly more massive, whilst at the
same time its differentiation from the milk enamel-organ is
hardly visible, being indicated only by the arrangement of the
nuclei of the constituent cells. Still the milk papilla (mp. di 2)
is distinct though very shallow. Three sections still further
back (fig. 5) the milk enamel-organ has wholly disappeared,
leaving, however, a much-thickened dental lamina, representing
the hinder and main portion of the Anlage of the permanent
third incisor (72).
The precise position of the milk-tooth rudiment on the
antero-lateral aspect of the entire third incisor Anlage is defi-
nitely established, not merely by an examination of the serial
sections figured, but by comparative study of sagittal series,
which affords complete corroboration of the statements just
made.
It will be shown in connection with the description of the
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 455
next stage that the milk enamel-organ just described under-
goes a rapid but most imperfect differentiation, and that in
later stages (from Stage rv onwards) it has entirely disap-
peared. At no time is there any formation of hard dental
tissues in connection with it. Comparison of the present
stage (11) with the next (111) proves that the labially placed
papillary indentation now described has nothing to do with the
formation of the papilla of the permanent third incisor.
Behind the site of 22 the dental lamina is again much
reduced in size for a few sections, but presently it enlarges
slightly in the succeeding ones and again decreases. This
almost certainly indicates the site of the future 74+. Behind
this the lamina again enlarges, its cross-section becoming
swollen and rounded, and after continuing thus through a
number of sections it changes its shape, but without decreas-
ing in sectional area, so as to form a figure elongated in cross-
section. This gradually undergoes still further enlargement
and elongation by a markedly deeper ingrowth of the lamina:
this enlargement we have definitely ascertained to be the
canine rudiment. Whether the shallower thickening of the
lamina continuous with it in front has any significance as the
Anlage of 72 we cannot be certain, though probably it may be
thus interpreted.
The shape of the canine rudiment is quite distinctive in
early stages, for already the part of the dental lamina answer-
ing to it manifests that tendency to disproportionately deep
ingrowth into the surrounding connective tissue which is indi-
cative of the future deep-seated position of the differentiated
tooth. Its development may be followed with ease through
all the stages from this onwards.
But, as now differentiated, the markedly enlarged canine
moiety of the dental lamina is not to be regarded as merely
the rudiment of the permanent canine. Subsequent stages,
especially that immediately succeeding, prove that from the
present Anlage a rudimentary deciduous (milk) tooth (d*) is
also formed. Indeed, the formation of a d* is already fore-
shadowed by the existence of a labially directed outgrowth
456 Jw. T. “WIESON- AND rd. “PL HERE.
(vide fig. 7) from near the basal portion (i. e. the “ neck”) of
the elongated laminar ingrowth. A comparison with fig. 13,
which shows the exactly.corresponding structure in the case of
the lower canine rudiment, will sufficiently demonstrate the
first steps in that evolution which results in the interesting
condition so beautifully shown in the next stage, where we
have the well-formed miniature enamel-organ of a deciduous
canine tooth above and below (figs. 20 and 37).
On account of the deep extension of the canine Anlage it
forms a prominent feature in sagittal sections which pass
through its plane, all the more that a long extent of the
dental lamina behind it, though quite continuous throughout,
is relatively very shallow and insignificant. This latter por-
tion of the dental lamina, representing the whole of the pre-
molar region between the canine and the deciduous premolar,
is at the present stage absolutely void of any special structural
differentiatiou. It is, as already said, comparatively shallow
(though a “neck”’ is slightly indicated), and extends back-
wards through very many sections to the region of dp3.
Here it again deepens considerably, its deeper part or fundus
becomes swollen, and ere long it is seen to be deeply indented
on its labial aspect by a large papilla, giving it the typical form
of a cup-shaped enamel-organ. This is represented in fig. 8,
in which it is to be noted that the process of formation of a
cup-shaped enamel-organ has gone on without any trace, so
far, of the emancipation of the developing organ from the
parent lamina. An early stage of the latter process may,
however, be recognised in fig. 9 (rdl.), which represents the
corresponding organ in another embryo from the same pouch.
(For the more advanced stage of the same process compare the
next stage as illustrated in fig. 24.)
Special attention may once more be called to the fact that
although dp= in Perameles is never other than a small and
insignificant tooth compared with the other functional members
of the dental series, yet nevertheless at this early stage it is
already large and well advanced in its development as an
enamel-organ; and it is the only antemolar tooth-rudiment,
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 457
amongst those which subsequently attain maturity, which as
yet exhibits the least trace of papillation. The precocious
character of its development is well illustrated in fig. 10
(dp*), showing in sagittal section its relation to the oral
epithelium (0.e.) and to the molar portion of the dental lamina
(dl.) behind it, which is cut vertically in the sagittal plane of
the section.
The molar portion of the dental lamina is massive and well
developed for quite a considerable distance behind dp3, as
may be seen in fig. 10. At first it is merely flask-shaped on
transverse section, but soon its cross-section alters in shape
by its distal margin bulging out obliquely in a labial direction.
This constitutes the Anlage of m+, and is seen in fig. 1] to
exhibit a very slight depression at its fundus. . . . Opposite
this depression the connective-tissue cells are aggregated to
form the very earliest Anlage of a dermal papilla (mp.m+).
Both in front and behind, the swollen Anlage gradually
diminishes in thickness. It is worthy of special remark that
here we have the very youngest stage in the differentiation of
a separate molar enamel-organ, and that in this there is not
the slightest indication of a composite character of the dermal
papilla whose earliest stage of evolution is here exemplified.
It is, from the first, single and individual. In the next
succeeding stage we shall find that this incipient enamel-organ
has rapidly attained a typical cup-shaped or even “ bell-
shaped” (‘‘glockenformig”’) character, and has “caught up
with” dp2 in its progress towards full differentiation.
Behind the Anlage of m+ the molar lamina gradually
diminishes in size, and finally comes to a rather abrupt termi-
nation in a somewhat truncated hinder extremity.
It will be evident from fig. 10 that at least the major portion
of the dental lamina in the molar region is directly continuous
with the deep surface of the oral epithelium. If the third suc-
ceeding section of this series were also figured, it would be seen
that this continuity of the lamina with the oral epithelium
extends right up to the very hindmost end of the molar lamina.
It is thus evident that the dental lamina in the molar region
458 J. T. WILSON AND J. P. HIM.
at the present stage has arisen by direct proliferation and in-
growth of the cells of the Malpighian layer of the oral epithe-
lium, as is the case in front of the molar region.
An examination of later stages proves that the line of conti-
nuity between the oral epithelium and the molar lamina grows
backwards, for some time, pari passu with that rearward
progress of the lamina itself which accompanies the gradual
elongation of the jaw.
At the present stage the entire dental lamina has retained
its connection with the oral epithelium.
In a very young mammary foetus of Dasyurus the molar
dental lamina does not end abruptly, but, after somewhat
suddenly becoming shallower, undergoes a further very gradual
diminution in depth, until it fades away into a thickening
of epithelium, lens-shaped in cross-section, like that which
has been described by Rose as constituting the earliest trace
of the dental lamina in the anterior region of the jaw.
Lower Jaw.—The armature of epithelium on the oral
surface of the lower jaw near its tip is only moderately thick,
and at first it thins slightly as it is traced backwards; then
suddenly on each side a localised thickening of epithelial cells
appears, which invades the underlying connective tissue, and
suggests the character of an extremely thick, broad, and some-
what shallow dental lamina. In reality, however, the definitive
dental lamina appears abruptly a short distance behind as a
further downgrowth from the thick cell-mass in question.
(The precise significance of the latter is doubtful. As it is
traced further back it becomes largely continuous with the
corresponding mass of the opposite side. Further back still,
the fused structure is seen to be absolutely continuous with
the epithelial cell-mass forming the tip of the tongue, which
is thus firmly glued to the upper surface of the lower jaw. As
the tongue is traced backwards it gradually frees itself from
the latter, and the thickening of the jaw epithelium is then
seen to have largely disappeared.) The dental lamina itself,
springing abruptly as aforesaid from the deep surface of the
epithelial armature of the jaw, becomes almost at once some-
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 459
what enlarged, and forms a swelling whose outer aspect 1s
somewhat flattened, while its fundus is more rounded and is
somewhat mesially inflected. The outer, more flattened por-
tion now comes to project labially, and is seen to consist
mainly of cells, the chromatin of whose nuclei is evidently in
the active phase. This feature causes this portion of the mass
to appear pretty definitely differentiated with a haematoxylin
stain. Some sections further back the labial projection of the
lamina is deeply indented by a dermal papilla (fig. 12), so as
to form a labially directed enamel-organ (d7;) in the cup-
shaped stage. This is as yet incompletely differentiated from
the common mass of the swollen dental lamina. The entire
enlargement of the lamina in this situation corresponds to the
region of the first incisor, and the labially cupped enamel-
organ, in process of differentiation, is the Anlage of the first
milk incisor (diz). This rudimentary tooth is shown in
subsequent stages to undergo a precocious though imperfect
development into a minute calcified tooth.
Opposite the papilla of di; the main mass of the dental
lamina has become considerably enlarged, and the increase in
bulk continues behind the rudiment of the milk-tooth. Part
of this posterior thickening doubtless represents the hinder part
of 7, but the more posterior portion must provide the material
for the development of tz. This is rendered certain by com-
parison with the succeeding stage (111), for in both stages the
dental lamina makes an abrupt turn outwards and runs trans-
versely for a very short distance, instead of antero-posteriorly.
Now in Stage 11 it becomes certain that 7, is developed at
the postero-external limit of this transversal part of the lamina,
and 7; at its antero-internal extremity. Further, in the
present stage the short transversely running part of the
lamina commences immediately posterior to the thick part of
the lamina which lies behind di; ; whilst at the outer end of the
same short transverse part of the lamina appears a new
thickening. This, as we shall see, is the only other swelling
in front of the canine, and it corresponds exactly in position
to ig in Stage 111.
460 J. T. WILSON AND J. P. HILL.
The posterior limit of this rudiment of 7; may be recognised
by a marked diminution in the sectional area of the lamina,
~ which continues back quite small until it again expands into
the typically elongated canine Anlage. The lamina in this
region is relatively much elongated on transverse section. It
does not, however, penetrate the deeper tissues in a vertical
direction, but is diverted rather mesially, parallel with, and
pretty closely adjacent to, the deep surface of the oral
epithelium. In it a very interesting step in differentiation
may be observed in the shape of avery slight but unmistakable
attempt at the formation of the enamel-organ of the milk
canine (fig. 18, d.;). This is just on the point of becoming
cup-shaped, and the dermal papilla (mp.d;) is beginning to
form. In the succeeding stages this Anlage -will be followed
through the course of its evolution. Here, as in the other
examples of milk Anlagen, the structure is first developed not
only on the labial aspect of the common canine enlargement
of the dental lamina, but opposite its anterior part, the
swelling attaining its maximum development just behind the
milk rudiment.
As is the case in the upper jaw, the premolar portion of the
lamina behind the canine Anlage is small and presents no
local differentiation until we reach the region of the last
premolar (dp;). Here the dental lamina becomes much
enlarged in all its dimensions, and the bulbous mass becomes
indented labially by the large dermal papilla of d.p.z (the
“milk molar”). Its characters are so closely similar to
those of the corresponding organ in the upper jaw that the
figures of the latter may suffice for both (figs. 8 and 9).
Fig. 14, however, illustrates the appearance seen in sagittal
section.
The molar region of the dental lamina is in a precisely
similar condition to that exhibited in the upper jaw. There is
only a slight indication of the differentiation of the first molar
enamel-organ. For the most part the molar lamina forms
simply a slab-like lamina of considerable dimensions, extend-
ing backwards, and continuous in front with the milk premolar.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 461
It is throughout in direct connection with the deep surface of
the oral epithelium, as serial sections, sagittal and transverse,
show most definitely. But as in the upper jaw, so also here,
there is found at a short interval behind dp.z an outward pro-
jection or bulging of the labial aspect of the lamina entirely
comparable to that which in the upper jaw serves to broaden
the lamina transversely. Here, however, it is placed much
nearer to the “ neck ” of the lamina, and throws a very slightly
cupped surface, with the rudimentary papilla in relation to
it (mp.m;z), more distinctly on to the labial aspect of the
Anlage. Fig. 15 illustrates the condition described, and may
be advantageously compared with the figure of the lower
canine Anlage (fig. 13).
Behind the Anlage of m; the labial projection disappears,
and the lamina is left as a simple but fairly thick band lying
parallel with the oral epithelium, to which it is attached by a
“neck” curved almost at right angles with the dorso-ventral
axis of the lamina. As in the upper jaw, this disappears alto-
gether with considerable abruptness.
Our readers will thus observe that in following the develop-
ment of the dental lamina in Stage 11 we have been able to
trace what to all appearances are the earliest steps in the
direction of individual tooth-differentiation. During this
period the dental lamina still includes the Anlagen of all the
future teeth; none of these have as yet been emancipated
from it, and the great majority of the Anlagen have as yet
little or no claim to be considered as individualised tooth-
germs.
Regarding the morphological interpretation of our observa-
tions, especially in the way of the determination of tooth
Anlagen, we believe that in the case of a number of these
there can be difference of opinion. It has already been indi-
cated that the justification of our interpretation in certain
other cases depends to some extent upon our observations in
subsequent stages of development. A fuller discussion of
these more disputable points must therefore be reserved till
later.
462 J. T. WILSON AND J. -P.- HILL.
Stage 11].—P. obesula: mammary feetus.
Length from vertex to root of tail . . . 22 mm.
Head length : : ‘ . ~ oe
Studied in coronal and horizontal series.
The examination of this stage reveals features of uncommon
interest and importance.
Although our specimen representing this stage is not speci-
fically identical with those of the stages immediately pre-
ceding and succeeding, we are confident that this difference is
of no consequence, and does not in the least prejudicially
affect the conclusions arrived at. So far as it is possible to
ascertain and to judge, the condition presented by the stage
answers exactly to what one would expect to find, intermediate
between Stages 11 and tv. The correspondences, indeed, are
throughout unmistakable.
Here for the first time the great majority of the permanent
teeth are sufficiently differentiated to be recognisable as dis-
tinct and individual enamel-organs. Most of these, however,
as we shall see, are in a very early stage of their individual
differentiation, and are not yet emancipated from the dental
lamina. The latter structure is relatively massive and abso-
lutely continuous throughout. Its line of attachment to the
oral epithelium (proximal margin or “ neck ”) is, however, in
parts considerably attenuated; and in a few places, for a
section or so, it is just possible to recognise a discontinuity
between these structures. This is doubtless due to the com-
mencement of the process of resorption which Rose has shown
(9) to produce fenestration of the lamina.
With these extremely rare and slight interruptions the at-
tachment to the Malpighian layer extends up to the posterior
end of the molar part of the lamina.
Upper Jaw.—In the incisor region the extreme anterior
end of the dental lamina (fig. 16, d/.) is directed somewhat
mesially, but it no longer lies so distinctly in a transverse
direction as in the preceding stage. *
The Anlage of the first permanent incisor (fig. 16) is indi-
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES, 463
cated close to the anterior extremity of the lamina by the
presence of a subglobular outgrowth (i+), bulging in an an-
tero-lateral direction, i.e. at right angles with the axis of the
lamina (d/’.)in this situation. The lamina itself is here plump
and swollen, so that on section the two together form a bilobed
mass. The labial lobe, however (“7+ ”’), is merely a localised
swelling, while the lingual (dl.), becoming more attenuated
behind, is continued for some distance onwards to connect
with the second incisor region.
It will be remembered that, in the preceding Stage 11, i+
was represented by a mere uniform enlargement of the dental
lamina. The outgrowth of the labial mass we have now de-
scribed constitutes therefore the next step in the differentia-
tion of the enamel-organ of the first incisor. And the growth
of this labially projecting mass involves the development,
between it and the main body of the lamina, of a depression.
This is seen both in front, behind, and especially above, where
it is somewhat more pronounced. The surrounding capsule of
connective tissue, where that is in relation with the depression
in question, exhibits distinct though slight evidences of cel-
lular differentiation. Fig. 16 shows a horizontal section of
the first incisor region with the sulcus extending on to the
anterior face of the Anlage. Here the connective tissue is
less fibrous, and its cells are more closely aggregated to-
gether to form what is in all probability the Anlage of the
dermal papilla. The latter indeed projects slightly into the
sulcus.
The formation of the labial lobe above referred to, and the
associated production of a depression between it and the main
portion of the dental lamina, are of great importance. We
find that when the earliest stage of the process of cupping of
an enamel-organ can be traced, that process is commonly, if
not invariably, initiated by an overgrowth of the cells of the
dental lamina so as to form a projection on the labial aspect ;
and that the future cup-like depression into which the dermal
papilla subsequently fits is foreshadowed by such a sulcus as
we have above described. And when the cupped enamel-
4.64 J. LT. WILSON AND <3. OP. ATE
organ is at length evolved by this process, the outer part
of the rim of the cup is constituted by the labially
outgrowing lobe, and the inner part of the rim of the
cup is formed by the main body of the dental lamina
itself. It will be observed that this explanation is essen-
tially in agreement with the idea expressed in Rése’s simile
regarding the relation of the cupped enamel-organs to the
dental lamina, when he says they are arranged “like swallow-
nests on a board.” Our present stage illustrates the condition
found just prior to the actual ‘“swallow-nest” stage of
differentiation of the enamel-organs. At this stage, indeed,
“sulcus” is the term which more adequately expresses the
character of the depression corresponding to the future cup of
the enamel-organ. It is open both fore and aft, because it
owes its first formation to a mere labial outgrowth of some-
what inchoate character. It may be noted further that this
sulcus may be very shallow, or even absent, if the outgrowth
of cells takes place near to the free margin of the lamina and
extends simply outwards. Thus in the figure given (fig. 11)
of the first upper molar Anlage in Stage 11 the labial promi-
nence is so wide, and extends so near to the free margin of the
lamina, that the intervening sulcus is broad and extremely
shallow. The cellular condensation indicative of the future
dermal papilla is, however, unmistakable. So also, with some
difference, in the Anlage of m4, Stage 11 (fig. 15).
The definitive cupping of such a bilobed enamel-organ of
course involves the subsequent appearance of a margin or rim
to the cup both in front and behind, in consequence of the
more pronounced local limitation of the processes of enamel-
organ formation,—processes which at first operate along a
stretch of dental lamina indefinitely limited fore and aft.
As may be gathered from Rése’s ‘ swallow-nest ” compa-
rison, the lingual wall of the cupped enamel-organ is formed at
first simply by the dental lamina,—the “ board ” against which
the “‘swallow-nests ” are placed. But by-and-by, as we shall see
hereafter, this dental lamina undergoes further differentiation,
whereby a proper and distinct inner wall is contributed to
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 465
complete the enamel cup, leaving a more or less liberated dental
lamina independently of the enamel-organ, and situated at its
inner or lingual side. This we term the “residual dental
lamina.”
After this somewhat necessary digression—partly antici-
pative of what we shall have to point out in detail later on—
we may return to the description of the lamina in Stage 111.
Behind the posterior limit of the Anlage of i+ the dental
lamina becomes itself progressively reduced in sectional area
until it approaches the region of the second incisor, its axis
being directed somewhat outwards as well as backwards to
reach the latter.
Here it quickly increases both in height and in thickness,
and is seen to form the lingual portion (d/2) of another
bilobed mass (fig. 17), comparable to that in the region of 7+.
To this bilobed swelling a similar interpretation must be given
as that employed in the case of the first; i.e. its lingual por-
tion is simply the enlarged dental lamina (d/?), while the
labial portion of the main mass projecting outwards (“72”)
is the outer or labial portion of the enamel-organ of 12. Here
_again the lobes are separated by a groove (fig. 17) passing
over the summit of the Anlage, and reaching its anterior and
posterior aspects.
In fig. 18 “II” represents a horizontal section of the whole
epithelial mass in the region of the second incisor. Here dl?
indicates the lingual lobe of the mass corresponding to the
swollen continuation of the dental lamina itself. This hori-
zontal section is taken above the level which would show the
continuity anteriorly of dl2 with dil+ in fig. 16, aud about
the level indicated by the line ad. in fig. 17, which shows a
coronal section through the same region. (It will be remem-
bered that between the developing teeth the dental lamina is
considerably lower than opposite them.) In this figure of a
horizontal section there is seen to project from the labial
aspect of i2 a small epithelial mass (di2), which we believe to
be the representative of the enamel-organ of the second milk
incisor. The same projection is visible in fig. 17, showing a
4.66 J. T. WILSON AND J. P. HILL.
transverse (coronal) section of the same region.! The back-
ward recurvature of this small labially projecting cell-mass,
seen in fig. 18, must not be mistaken for an indication of com-
mencing papillation from behind forwards. The series of sec-
tions proves that the projection in question is simply being
gradually pinched off from the main body (22). Higher up,
the horizontal sections show only the two lobes of the Anlage
of i2 (d/2 and ‘‘22”’), It will presently appear that all these
features are still better marked in connection with the region
of the third incisor, which is partly illustrated in the same
drawing (fig. 18, “ III’).
Horizontal sections taken at a lower level than fig. 18 show
the dental lamina (d/2) in the region of the second incisor to
be prolonged backwards into continuity with d/3 in the third
incisor region.
Horizontal sections passing through the upper part of the
third incisor-mass closely resemble those through the upper
part of the second—viz. in each case we have lingually the
lobe formed by swollen dental lamina, and labially the lobe
formed by the differentiating labial moiety of the incisor
Anlage. As in the cases of the two more anterior tooth-
germs, the two lobes are separated by a slight groove or
sulcus, and, corresponding to this, over the summit of the
Anlage the connective tissue shows a slight cellular aggrega-
tion indicating the commencement of the dermal papilla.
Horizonal sections taken a little lower down begin to show
a narrow labial projection in a position corresponding to that
of di2 in the second incisor region. Still lower there suddenly
appears, closely external to the last-named outgrowth, an
isolated mass of epithelial cells. In sections taken still more
inferiorly, as shown in fig. 18, “III,” this cell-mass (d73) is
found to be really connected with the labial aspect of the main
mass of 72 by a narrow neck (cb.). This labial cellular out-
1 The single coronal section here figured might seem to suggest that the
di% projection simply forms the lower limit of a papillated depression on the
labial aspect. This is a merely accidental resemblance, as a study of the
neighbouring sections in the series amply proves.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 467
growth is further seen to be the seat of retrogressive processes,
for in neighbouring sections it shows invasion by connective
tissue, partially breaking it up into smaller cell-groups.
This cell-mass is incontrovertibly the serial homologue to
the projection figured under “ di2” in the region of the second
incisor in figs. 17 and 18, though plainly it is in a more ad-
vanced phase of evolution, as its greater isolation from the
main mass and condition of degeneracy testify. We are
satisfied that this cell-mass can be regarded as none other
than the degenerating remains of that very early, labially
placed enamel-organ, which we have described and figured
above (p. 454, and figs. 2, 3, and 4) in connection with a pro-
minent swelling of the incisor region of the dental lamina in
the preceding Stage u. It will be remembered that the
localised thickening of the dental Jamina was set down as the
Anlage of 78, and the labially developing enamel-organ, im-
perfectly approximating to the cupped stage, was regarded as
that of di3. The judgment there expressed on the strength
of our observations upon the differentiation of the dental
lamina in that stage, is now seen to be borne out by the degree
of development subsequently attained by the presumably ho-
mologous structures.
It is interesting to note that in the present stage, as in the
preceding, the third incisor Anlage as a whole is still the most
bulky of the incisor series, although the disproportion between
it and the other incisor Anlagen is now greatly diminished.
It must be noted in regard to fig. 18 that, in a longitudinal
section passing so near to the base of the dental lamina in the
73 region as that figured, it is only the basal portion of the
Anlage of 73, and the vestigial remains of d2, which are
recognisable. Itis only at higher planes that the differentiation
of the Anlage of i2 into labial and lingual lobes is as recog-
nisable as it is, e.g. in the case of 72 in the same section.
Lower down in the case of the latter region, indeed, the labial
portion of the Anlage of 72 similarly disappears from the
section as a distinct lobe, leaving only the common lamina
plus the labial projection of di2.
VoL. 89, PART 4,
NEW S5ER. K K
468 Jj. La WALSON AND J. P: HILL;
The structure of the differentiating dental lamina in the
third incisor region will be rendered further intelligible by a
reference to fig. 19, which illustrates a coronal section through
this region. Here again the bilobed character is only ob-
scurely manifested, but in the next succeeding section it is as
plainly seen as in 72 in fig. 17, and the cellular aggregation
at the site of the future papilla is also clearly seen. The line
ab. shows the approximate level of the horizontal section in
fig. 18. Here the almost isolated remains of the enamel-organ
of di3 are well shown.
Tn this figure (19) it will be observed that the attacliment of
the whole mass to the oral epithelium is already narrowed.
The deep notch between oral epithelium and dental lamina on
the lingual side represents the result of the process of pinching
off from the oral epithelium which is now in progress. Origi-
nally the attachment no doubt corresponded to the whole
breadth of the base of the mass lingually. As a result of the
pinching off, the primitive continuity of the dental lamina on
its lingual side with the oral epithelium is now obscured.
The plane of the horizontal section through the region of 73
(fig. 18) is such that although it does not shave the upper free
border of the low dental lamina between the low-lying 72 and
72, it does cut the lamina between 72 and i4, obliquely for a
certain distance (fig. 18 d/4). The fourth incisor Anlage
itself, however, lies wholly at a higher level.
It is unnecessary to enter upon a detailed description of the
Anlagen of the fourth and fifth incisors. The following state-
meuts will suffice. The dental lamina is continuous throughout,
both in its own extent and with the oral epithelium. It has
undergone enlargement at two places corresponding respec-
tively to the fourth and fifth incisors. These localised
enlargements are for the most part of a general character,
though bulging mainly to the labial side, except that in each
case there is already to be found, projecting labially from near
the base of each rounded swelling, a somewhat attenuated
epithelial process, precisely comparable in general appearance,
structure, and relations to that marked ‘‘ di? ” in figs. 17 and
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 469
18. And as in the latter case, so here, we take these rudi-
mentary labial outgrowths of epithelial cells to represent the
abortive enamel-germs of true milk incisors, di4 and di.
It is, however, when we come to deal with the canine rudi-
ment in this stage that we really obtain the most emphatic
testimony to the validity of the interpretation which we have
given of these interesting vestiges.
From the fifth incisor the dental lamina, though small, is
continued back without interruption, and very soon it exhibits
that great vertical enlargement which we find to be character-
istic of even the earliest canine rudiments in both upper and
lower jaws.
In fig. 21 we show a horizontal section which in front
passes through the upper part of the fifth incisor mass (28).
This is seen to bulge out labially (but this is quite above the
level of rudimentary d=). From the inner or lingual part of
the Anlage of 22 the dental lamina is continued backwards for
a short distance, to lose itself, in this sectional plane, in a
broad area of epithelium (0.e.). This, as may be learnt from
the drawing, is simply the deep layer of buccal epithelium
sliced parallel with its free surface. A section or two higher
this epithelial patch disappears, giving place to the prolonga-
tion backwards of the narrow dental lamina. Still higher in
the series the lamina is seen to undergo enlargement so as to
form the most inferior portion of the canine swelling of the
lamina (ef. fig. 20). This enlargement markedly increases in
thickness as it is followed upwards through the horizontal
series. But although the lamina is so definitely enlarged, the
differentiation of the permanent tooth has hardly yet set in;
certainly no distinct outgrowing labial lobe is apparent, like
that observable in the anterior incisor germs. Itseems highly
probable, however, that the upper and labially deflected portion
(marked “c” in fig. 20) of the whole Anlage is really
equivalent to the labial lobe of the bilobed incisor masses, and
is thus about to form the labial portion of the cupped enamel-
organ of the permanent canine. If this be so, then the
proper morphologically distal margin of the dental lamina is
4.70 J. T. WILSON AND J. P. HILL.
formed by the elbow-lke projection marked “c.dl.”? in the
figure, and the flattened surface between c.d/. and the apical
point of the whole enlargement would represent the site of the
future enamel cup. However this may be, the one specially
striking and important feature of this general canine Anlage
is the presence, near its “ neck,” of a very perfect miniature
enamel-organ in the cup-shaped stage. ‘This structure
projects directly outwards in a labial direction from the
proximal portion of the dental lamina, and in it we
recognise, in accordance with the views we have set forth
in the Introduction, the true rudimentary “milk” canine
(fig. 20, d®).
In fig. 22 we have shown another horizontal section through
the upper jaw of this stage, through the same region as, but at
a higher level than, the section illustrated in fig. 21. In this
section we remark the canine thickening of the dental lamina
¢, and its more attenuated prolongation backwards into the
premolar region (p+di.). Beside the thickened canine portion,
the enamel organ of d£is seen. Further important reference
will be made to this figure later on. In the meantime we
would draw attention to the fact that the minute, and up
to this stage perfect, enamel-organ described above
is entirely homologous in position and relations
with those small lobular outgrowths of the epi-
thelium of the dental lamina situated labially from
the Anlage of the four hindmost incisors, which we
have interpreted as representatives of enamel-
organs of deciduous incisors. And it is not mate-
rially diverse from these in regard to its chrono-
logical position.
Now if the serial homology with the incisor vestiges be
granted—and we think it is impossible to avoid such a con-
cession—this at once establishes the title of these vestiges to
be regarded as true representatives of the enamel-organs of a
tooth-generation preceding the permanent teeth. And in our
opinion this generation is homologous to the milk series of
Eutheria.
DEVELOPMENT AND SUCCESSION OF 1 EETH IN PERAMELES. 471
So far as we can ascertain, there is not to be discovered
either in the present or in the preceding stages any such dif-
ferentiation in the first incisor region as might be taken to
represent there the rudiment of a true milk-tooth, i.e. one
which belongs to a tooth generation preceding that to which
the first permanent incisor belongs. Nevertheless there is
some reason for the belief that traces of an attempt at the pro-
duction of such a tooth do appear at some time during the
development of the teeth in Perameles. And if any such
rudiment does occur, then the cell material of which it consists
must be derived by outgrowth from the labial side of the base
of the mass we have described as constituting the Anlage of
the first incisor. It has been stated above that in its size and
degree of differentiation (including that of a “milk” rudi-
ment) the third incisor Anlage is the most developed member
of the upper incisor series at the present stage. The second
is less developed, but yet exhibits features similar to those in
the third. So also in the case of the fourth and fifth incisor
Anlagen. Although, therefore, no such labial remains occur
in connection with the first incisor Anlage in Stage 1ir as we
have been able to show associated with the incisors behind it,
and more especially in the canine region, we consider ourselves
justified in the belief that certain cellular processes situated
on the labial side of the root of the dental lamina in the first
incisor region of the next stages (1v and v) may possibly
represent the remains of the missing rudiment of a dit. In
the present stage we may suppose that the latter is not struc-
turally defined from the basal portion of the common first in-
cisor Anlage.
In further corroboration of our views upon the incisor
development in Perameles it may not be inappropriate to
refer to the condition we have found in another polyprotodont
Marsupial, Dasyurus viverrinus. Among our series of
stages of development of this form we possess one which, in
respect of the third incisor, exhibits a general stage of de-
velopment intermediate between the present stage (111) and
the following stage (1v) of Perameles.
472 J. 2. WILSON’ AND? 7, °P. Bibi:
As in Perameles, but in much more marked degree, the
third is the largest and most advanced in its development
among the upper incisors. Here (fig. 23) the definitive cup-
ping of the enamel-organ of 2%, has at last been effected, and
a well-marked dermal papilla (mp. i*) formed, while at the
same time the main body of the dental lamina has undergone
such differentiation that the mass of cells forming the lingual
portion of the cupped enamel-organ is now becoming distinct
from the rest of the dental lamina. The latter is, in fact, in
process of liberation as a “residual”? dental lamina (rdi.).
(In an antecedent stage the Anlage of 73 is found to be differ-
entiated to exactly the same extent as the corresponding
Anlage in Stage 111 of Perameles.) A remarkable feature
in this section is the rudimentary di3. Whatever may be
said in detraction of the claims of the structure we have
described in connection with 72 in Perameles, to figure as a
vestigial enamel-organ, it is impossible to deny that here in
Dasyurus we have an indubitable vestigial predecessor to
28, It seems to us impossible to gainsay the homology
testified to by this comparison between Perameles and
Dasyurus. And it is superfluous to point out that an ad-
mission as regards the identity of the structure labelled di2
in Perameles with the vestigial tooth in Dasyurus involves
the admission of a similar homology for the structures in
Perameles in series with “di3.” Here in Dasyurus
(fig. 23) di= is not simply a degenerate cell-mass, but has
assumed specific dental characters by the precocious evolution
of a small but thick dentine cap, which again is covered bya
hood of regularly arranged enamel-cells. Instead of forming
a projection from the rest of the incisor mass, the vestigial
tooth rather lies embedded in the labial aspect of the latter
near its base. Perhaps this may be explained by a reference
to the great labial outgrowth of the Anlage of the permanent
incisor which must have occurred to bring about the condition
figured, together with the later stationary and unprogressive
character of the calcified vestigial di%.
The examination of an earlier stage of Dasyurus has
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 4738
shown us the third milk incisor at a stage prior to its calcifica-
tion as a deeply cupped enamel-organ structurally comparable
to the enamel-organ of d‘, figured by us in Perameles
(fig. 20). Thus we have shown that, in Dasyurus, the rudi-
mentary enamel-organ of di2, does not undergo the rapid
degeneration which is exhibited by di3, in Stage 11 of
Perameles, but passes into the regular cup-shaped stage which
we have seen in the upper jaw of Perameles to be attained
only by d3.
A thoughtful comparison of the following series of figures
will render it easy to understand the derivation of the condi-
tions found in the third incisor region of Dasyurus from the
earlier condition illustrated in Stage 11 of Perameles:
Cf. Perameles, Stage II, 73 and di in figs. 3 and 4,
" epee LUN) Lares in figs. 17 and 18.
9 oo sae eo in figs. 18 and 19.
” ” » +(C and ds in fig. 20.
In fig. 21 it is seen that the dental lamina is continued
backwards, diminished but unbroken, from the region of the
canine (marked “0. e.”) to a point where it somewhat suddenly
expands to form the Anlage of the first premolar (p+). It
will be evident from the figure that, though this enlargement
appears as a general expansion of the lamina, the addition or
increase of cell material is towards the labial side, so that the
bulging is in that direction. On coronal section, indeed, one
obtains an appearance not far removed from that exhibited in
this same stage by 74, although it can hardly be said that a
definite furrow has appeared between the labial bulging and
the main body of the dental lamina. Both in figs. 21 and 22
the stretch of dental lamina extending from the canine to the
first premolar is recognisable.
In the first-named figure, however, the section passes along
the more constricted basal portion or ‘ neck” of the lamina.
In the second figure the section passes through the lamina
near its free margin or “ fundus,” where that reaches a high
vertical extension in the neighbourhood of the canine. Further
back the lamina disappears from the section by becoming ver-
A474 J. 1. WILSON: AND. J. P, HELL.
tically shortened, the horizontal section thus cutting it obliquely.
The hinder part of this obliquely cut portion of the lamina, as
shown in fig. 22, does to some extent coincide with the anterior
part of the first premolar swelling in fig. 21. It constitutes,
indeed, as coronal sections prove, the lingual portion of the
latter mass. The labial bulging of the Anlage is lower down,
and is not cut through at this level (cf. fig. 21).
The interval between the tapering ends of the Anlagen of
the first and second premolars is a comparatively short one.
Here the dental lamina, although quite continuous, appears
low and relatively insignificant.1_ This is especially interesting,
as it is just here that one might expect some trace of Thomas’s
hypothetical ‘ pm2.”
A glance at fig. 21 is sufficient to show how far forward the
anterior tapering end of the second premolar enlargement of
the lamina extends. This enlargement is, indeed, of very
considerable antero-posterior extent. Anteriorly its cross-
section is rounded, but posteriorly it is much more vertically
extended, gradually narrowing at the same time, to be con-
tinued on as the vertically elongated but otherwise unmodified
lamina, which is found in the region between the second and
third premolars (cf. text Fig. 1, a—e).
Fig. 22, representing a horizontal section taken at a higher
level than fig. 21, shows the higher hindmost portion of p2
cut somewhat obliquely, and gradually passing back into con-
tinuity with the freed dental lamina on the lingual side of the
well-developed milk premolar, dp.
On the labial side of the basal margin or “neck” of the
1 The lamina looks lower than it really is because it is somewhat folded in
its length, so that in cross-section the long axis (being attached to free
margin) is not straight, but crooked. Such bending of the lamina near the
neck is a common enough phenomenon, and here it is strongly marked. An
outwardly projecting elbow in cross-section thus produced may easily enough
be mistaken for a true outgrowth of cells in some cases, and such an error
must be carefully guarded against. The separation of the lamina from the
oral epithelium frequently leaves such an elbow projecting freely, and this
may form another source of error in interpretation,
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 475
(G)
Fie. 1.—a—c. Series of coronal sections (each three sections apart) from
premolar region of upper jaw of Stage 111, and passing through dental
lamina (d./.) between p2 and dp3. (x 90 diameters.)
476 J. tT; WILSON AND JS. P.-Hiih.
dental lamina, opposite the anterior half of p2, there is to be
seen through several sections a small mass of cells projecting.
It is possible that this is representative of dp2. If so, it is
very degenerate, and there is no similar structure beside the
Anlage of p+, though a similar small cell-mass is present in
the lower p2 (see below).
The epithelial projection in question is at all events not due
to a bending of the lamina, whose “neck” is short, straight,
and perfectly preserved in the region referred to.
It will be remembered that in the preceding Stage 11 no
trace of localised thickening or other differentiation was ob-
servable in the premolar region of the dental lamina in front
of the site of the third premolar, but that here a tooth Anlage
in the cup-shaped stage was already present. And whilst in
the present stage the rudiments of the two anterior premolars
(p+ and p2) are sufficiently defined as such by localised en-
largement of the dental lamina, the Anlage of that premolar
tooth which was already cup-shaped in the preceding stage is
now still further advanced, showing differentiation of the
stellate reticular tissue of the enamel-organ (figs. 22 and 24).
Furthermore it is to be remarked that the process of libera-
tion of dp® from the lingually-placed dental lamina, which was
stated to have just set in the preceding stage, has progressed
considerably, so that the freed dental lamina (fig. 24, rdJ.)
now forms a partly independent epithelial ingrowth by the
lingual side of the tooth rudiment. The latter is still seen
(fig. 24) to be somewhat broadly attached to the labial aspect
of the liberated lamina at some distance from its apex, or,
strictly speaking, from its free distal margin, and extending
nearly to its attached proximal margin or “ neck.”
A careful investigation of the earliest stages of
premolar tooth-differentiation has thus served to
show that this third in order among the premolars is
not differentiated at all contemporaneously with the
1 The question of the significance of certain outgrowths of the dental
lamina on its labial side is discussed below in connection with the molar
problem.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 477
other premolar germs. On the contrary, it is com-
paratively far advanced in structural differentiation
by the time there appears any trace at all of the
anterior premolars, or more than very slight traces
of the formation of the enamel-organs of any perma-
nent tooth whatever.
The vertical elongation of the cross-section of the dental
lamina just anterior to dp? has been referred to. Just in
front of the plane at which the enamel-organ of the latter
tooth comes into view the dental lamina is higher than at
any more anterior part of the premolar region, and as high as
it is at the canine thickening. This height is retained by the
residual lamina opposite to dp®. Towards the hinder part of
the tooth the lamina becomes separated off from the oral epi-
thelium.
A couple of sections or so behind the disappearance of the
last trace of the enamel-organ of dp3, the free margin of the
dental lamina undergoes an abrupt further elongation, i.e. it
suddenly penetrates considerably deeper into the connective
tissue to be prolonged into the rudiment of m+, which forms
immediately behind as a large cup-shaped enamel-organ.
This is not visible in fig. 22, since the plane of that section
passes altogether beneath m+. Fig. 25, however, represents a
section taken at a higher plane, where the extreme upper or
basal portion of dp2 is cut through, along with its associated
residual lamina (rdl.). The latter is seen to be continued
backwards into the first molar, and the lower face of the
enamel-organ of this tooth (m+) is seen to be shaved through.
Fig. 26 shows a still higher plane, where therefore the con-
tinuity with the lamina in the premolar region is not visible,
but where the lamina (m2 dl.), ascending posteriorly, and
becoming freed from the hinder part of the developing tooth,
is continued backwards into the region of the future second
molar. Here the first molar is seen to be a deeply indented
(“glockenformig”) enamel-organ (cf. fig. 27), which in its
degree of maturity now rivals dp. It will be remembered
that in the preceding stage the enamel-organ of m+ had only
478 J. LT. WILSON AND J. P. HILL.
just become recognisable as a distinct Anlage, though dp2 was
already a cup-like enamel-organ.
Figs. 28 and 29 represent coronal sections through m4,
rather behind the middle of the tooth, and through the
posterior part of the tooth, respectively. In these the process
of differentiation of the tooth Anlage from the parent lamina
is seen to be well advanced, so that the residual dental lamina
(rdl). which remains, after separation from it of a tooth
rudiment, appears with its free distal margin or fundus some-
what swollen, and freely projecting by the side of the first
molar Anlage.
As the residual lamina is traced backwards from the region
of m+ (ef. figs. 25 and 26) it retains for a time on its labial
face some irregular projections over the area corresponding to
that occupied further forwards by the differentiated enamel-
organ of m1. A comparison of figs. 29 and 30 (the latter
being the tenth section behind the former) will serve to
explain this statement. It will also appear that the residual
dental lamina of the first molar region has very slightly
increased in thickness. Here, of course, it simply constitutes
the primitive and undifferentiated lamina intermediate between
the region of the first and second molars.
Fig. 31, taken some little distance further back, shows con-
siderable broadening of the lamina at its free margin, after
somewhat the same fashion as we have seen to characterise the
Anlage of the first molar in the preceding Stage 11.
Here, again, the thickening of the fundus appears to be
chiefly due to a proliferation of the cells at the labial lip of the
free margin of the lamina.
Fig. 32 shows a section immediately in front of the plane at
which the molar lamina abruptly terminates.
In the last three sections (cf. especially figs. 80 and 31) the
prominent labially directed process (/.0.) near the base or
attached margin of the dental lamina is the cross-section of a
continuous secondary lamina in this region (m+ to m2),
whose significance will be discussed below in connection with
the questions of molar homology (vide Part II).
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 479
Lower Jawof Stage III.—The phase of tooth develop-
ment exemplified in the lower jaw is on the whole a little in
advance of that found in the upper. This is more notable in
some regions than in others. During the period of development
which intervenes between Stages 11 and 111, rapid advancement
has been made in the differentiation of the first lower incisor
Anlage. An attentive comparison of our figures of sections
through this region in the two stages (figs. 12 and 388) will
chiefly show that the early papillated Anlage of di, in Stage 1
has rapidly passed through the succeeding stages of its
development up to and including the development of a perfect,
though thin, dentine cap for the papilla. No enamel has been
formed, but the enamel-organ—apparently arrested in its
evolution—has become separated from the main mass of the
dental lamina by a considerable interval, remaining connected
with it only by a delicate and disintegrating strand of
epithelial cells forming a “ Verbindungsbriicke” (fig.
33, cb.).
It may be noted that the position of diz relative to the
Anlage of 7+ appears to have altered during the transition from
Stage 11 to Stage 111. In the former the rudiment of di>
occupied the labial aspect of the more anterior portion of the
swelling corresponding to 77. Here, however, the differentia-
tion and segregation of diz have allowed of a growth forward
of the bulky Anlage of i;, unhindered by the presence of any
more anterior dental element. Hence, in the present stage,
the vestigial dit has come to lie opposite the more posterior
moiety of its large morphological successor. The relative
forward extension of the latter appears to continue at least up
to a period represented by our Stage 1v (see fig. 56).
We have in the next place to point out that, in the present
stage (111) the dental lamina, which was almost wholly
undifferentiated from the enamel-organ of diz in Stage 11, has
gone on to form the large papillated dental germ of 7; (fig. 33).
But as yet there is no trace of differentiation of a residual
dental lamina from the Anlage of 71. In the succeeding Stage
tv, however, a residual lamina, with its free marginal portion
4.80 J. T. WILSON AND J. P. HILL.
distinctly swollen, is exceedingly well developed (fig. 55)
along the whole extent of the tooth (¢;) so that the latter
has come to appear as a huge appendage of the liberated
lamina.
In the present stage no such differentiation has occurred,
and the whole of the large dental lamina remaining over after
the differentiation of diz appears for the present totally con-
verted into an enamel-organ for 77. Thus the sequence of
phenomena observable in this series of three stages (11, III,
and iv) in the case of this particular region goes far to illus-
trate and support our contentions as to the nature and con-
ditions of occurrence of a residual dental lamina, the so-called
“ Krsatzleiste ” of Rose.
As the papillated Anlage of 7; is traced backwards it
gradually diminishes in sectional area, becoming at the same
time more consolidated by loss or marked reduction of the
looser epithelial cells in its interior, which become the steilate
tissue. When the last trace of papillation has disappeared
posteriorly, the structure is seen to be again a mere undif-
ferentiated dental lamina transitional from the first to the
second incisor region. In the latter situation it swells out
again, and at the same time becomes slightly indented below.
There is also a somewhat denser aggregation of connective-
tissue cells in relation to this indentation than elsewhere in
the fibrous capsule, and the cellular accumulation forms a
faint elevation, corresponding to the indentation of the enamel-
germ.
This condition of affairs corresponds to what we have
already noted in connection with the upper incisors. Beyond
the Anlage of 73 the dental lamina runs for a short distance
almost transversely outwards, and then swells into the Anlage
ofiz. This relationship is very striking. At first we were
puzzled by the appearance of the coronal sections in this region,
for both 7; and zz appear in the same section (fig. 34), and
the relation of the intervening part of the dental lamina seems
obscure. (his p+)
is morphologically in front of dp> (his dp*) is at all so in-
evitable as he regards it. The proofs drawn from the mere
appearance of the dental lamina in transverse sections are, in
our opinion, of very little value. A quite similar development
of the deeper part of the dental lamina, and an identical
development of loose and stellate-like epithelial tissue in its
interior, may frequently be met with where there is no
question of the presence of a tooth at all.
It is only the definite localised elevation, sharply limited
fore and aft, figured in the drawing of a wax model of this
region of the lamina in Petrogale which seems to us atall a
formidable evidence of the view put forward. And even this
DEVELOPMENT AND SUOOESSION OF TEETH IN PERAMELES, 493
we cannot regard as conclusive. For, according to Mr. Wood-
ward’s statement and figures, the enamel-organ of dp2 (his dp*)
at the stage referred to has not yet differentiated itself as an
organ distinct from the dental lamina. The whole lamina in
the dp* region appears as if converted into the enamel-organ,
as is commonly the case in the earlier stages of tooth develop-
ment. This phase of development is, however, only a passing
one. By-and-by the separation of the enamel-organ from the
lamina which gives it birth is indicated, and the resulting
appearance of a residual lamina or “ Hrsatzleiste” we
regard as the first condition of the appearance of a succes-
sional tooth.
This first condition is as yet absent in the early stage of
Petrogale under notice. Mr. Woodward in effect maintains
that in its absence we must forthwith accept the presence of
an enlargement of the lamina in front as the true Anlage of
p* (his p*). We hold, on the contrary, that until we are in
a position to trace the fate of the residual lamina, which will
assuredly manifest itself by the lingual side of dp, it is pre-
mature to decide upon what constitutes the earliest germ of
p® (“pt”). If the evolution of the swelling figured by
Woodward can be clearly traced up to the formation of the
enamel-organ of p= (“p+”), without suspicion of a confusion
with the subsequent evolution of a true “Ersatzleiste” of
dp2, then Mr. Woodward’s case for Petrogale will have been
made out.
We therefore insist that it is necessary to have recourse to
somewhat later stages for corroboration, or otherwise, of Wood-
ward’s view. And what is the result of such a reference? We
shall quote Mr. Woodward’s own words. “In a slightly older
embryo the same condition was observed, but owing to the de-
velopment of the third and fourth premolars, especially of the
latter, they somewhat overlap the lamina connecting the two,
and it in consequence becomes displaced to the inner side of
these structures. So much does the fourth premolar grow
forward with age that it appears as if this lamina was
a downgrowth from the inner side of the enamel-
494, J. T. WILSON AND J. P. HILL.
organ of pm [the spaced type is ours]; this, however, 1s
really not the case; the lamina is morphologically in front of
that tooth, and only attains a secondary connection with it”
(2, p. 458). Now, if we disentangle the statement of facts
from the accompanying explanatory hypothesis, we recognise
that in the later stage, i.e. when dp? has become thoroughly
differentiated, a free dental lamina is found by its lingual side
appearing “as if this lamina was a downgrowth from the inner
side’ of its enamel-organ.
But this is precisely the condition we find in Perameles,
where the appearance of such a laminar “down-
growth” is most assuredly not due to any process
of overlapping or disproportionate growth of any
kind. Nor is there in that form any antecedent enlargement
of the lamina in front of the dp3 region. We infer from his
language that Mr. Woodward found some slight difficulty in
explaining the fact of an actual connection between dp3
and the lamina by its lingual side, on his theory of a forward
growth producing overlapping, since he denominates such a
connection as “secondary.” Now we are firmly convinced
that no truly secondary connection is ever de-
veloped between a dental lamina and an enamel-
organ which has previously been differentiated off
from the lamina. Itis,a priori, most unlikely that any
such secondary connection should be established. And we
believe that, wherever a connecting bridge is present, it will
be found necessary, and not very difficult, to interpret it as a
remnant of the original primary connection between enamel-
organ and parent lamina. We must believe that Mr. Wood-
ward’s method of explaining away the importance of the con-
nection he found in his later stage of Petrogale rests upon
an erroneous interpretation of the facts. We hold that in all
probability he is wrong in supposing that the part figured in
the earlier stage in front of dp> is the same as the part found
in the later stage by the side of that tooth. The latter part is
simply the free dental lamina of the region of dp® itself, now
liberated by a more complete differentiation of the enamel-
DEVELOPMENT AND SUCCESSION OF THETH IN PERAMELES. 495
organ of the latter; and it is from it, and not from the lamina
morphologically in front of dp, that p? is subsequently de-
veloped. We have put our re-interpretation of Mr. Wood-
ward’s facts in frankly dogmatic form for the sake of clearness.
There is too much unavoidable uncertainty about any question
of this kind to encourage a genuinely dogmatic attitude in
regard to it. And we may, indeed, with Leche, admit after
all the possibility of dp® arising in front of p32 without
concluding, with Woodward, that the teeth so arising belong
to the same series. At the same time the facts in Perameles
do not permit us to entertain any doubt with regard to the
origin of dp*, however it may be in the case of Petrogale.
Nor do our investigations in Marsupials lead us to side with
Leche in his view that the germs of the permanent teeth do
arise in front of those of the milk-teeth. Judging from the
observations already detailed with regard to the position of
the early rudiments of the vestigial milk incisors, we must be-
lieve that the Anlagen of the permanent teeth arise side by
side with those of the milk-teeth, though the latter are placed
rather opposite the anterior portions of the former. This we
have already seen to be originally the case in the region of the
first lower incisor of Perameles. Nevertheless at a com-
paratively early period the vestigial milk enamel-organ comes
to lie opposite the hinder end of the permanent tooth, so that
here the primitive relation comes to appear reversed.
Woodward calls attention to the anomalous eruption of p3
in Perameles rather in front of its “ supposed” pre-
decessor dp.
But, after all, the point of eruption of a tooth is of
secondary importance, and we find, on the other hand, that
not only at its first appearance, but during the very long
latent period which precedes its elaboration as a cupped
enamel-organ, the germ of p3 is placed, not in front of, but
opposite to dp3,
it would be difficult to draw any other conclusion from our
fig. 48, and here we are in a position to state that there is
absolutely not the slightest scrap of evidence tending to prove
496 Je Lf. WILSON AND J; PP: “HU:
that any such “ secondary attachment” has been entered upon
as Woodward has imagined to take place in his parallel case.
We are perfectly confident that our interpretation is the
correct one for Perameles.
From the description of dp2 given above, it will be observed
that it is still in a more advanced phase of evolution than
the two anterior premolars. But the difference is rapidly
becoming less marked. The comparatively rudimentary
character of the deciduous premolar of Perameles is now
expressing itself in a much less rapid increase in size, so that
it is now being rapidly overhauled in its progress by the
enamel-organs of p+ and p2. In consequence of this fact, the
very striking discrepancy in the periods of appearance of the
enamel-organs of dp> and of the anterior premolars respectively,
is, from now on, less and less apparent. It is thus quite
easy to understand how observations upon a series of stages
which did not include any younger than the present (Stage rv)
might lead to the conclusion that these teeth were serially
homologous.
The appearance of the upper first molar in Stage tv is
represented in fig. 51. Here, as in dp3, a thin dentine cap is
present over the summit of the protocone. A well-marked
residual dental lamina (rdl.) is present, whose significance will
be discussed in connection with the general question of molar
homology. ‘This figure alone is sufficient to disprove Wood-
ward’s view (2) that “lingual downgrowths of the dental
lamina” do not occur in connection with the molar teeth.
Two drawings (figs. 52 and 53) are given of the enamel-
organ of m2 from distinct series of sections (and not quite at
corresponding points). In each case a well-marked swollen
(“knospenférmig ”) “Ersatzleiste” or residual dental
lamina (rdJ.) is visible.
There is also a small labial “ sprout ” or “ outgrowth ” (/.0.)
opposite a knee-like bend or “ Knickung” of the lamina, the
significance of which is elsewhere discussed.
1 Woodward has apparently given up this view, so far at least as certain
rodents are concerned (cf. 18).
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 497
After the enamel-organ of m2 disappears from the serial
sections, the dental lamina continues on, still for the most
part in connection with the oral epithelium, and by-and-by it
becomes swollen to form the undifferentiated Anlage of m3
(fig. 54). In this its hinder portion the lamina preserves the
kink or flexure above referred to, as well as the (laminar) labial
outgrowth (/.0.) from this. The latter has, we are convinced,
nothing to do with tooth production (cf. p. 564 et seq.).
Lower Jaw.—In the lower jaw the dental lamina arises
somewhat abruptly a little distance in front of the first incisor
region. Here it is fairly thick and plump. This port on
must be regarded as having been produced by a slight exten-
sion forwards of the lamina since the period represented by
Stage 111, occurring pari passu with the continued elonga-
tion of the jaw. When the region of 7; is reached the lamina
is seen to be continued on into a thickened residual dental
lamina by the lingual side of 7; (fig. 55, rd/.). The enamel-
organ of the latter is large and well developed, and is con-
nected with the lamina by a thin epithelial connecting bridge.
Further back the latter is lost, while the lamina itself has
become much elongated and thinned, and here (fig. 56) is
found attached to its neck a delicate strand of cells (cd.)
forming a “ Verbindungsbriicke” for the small vestigial
ealcified dt;. Only a faint trace of a dentine cap is found at
the apex of 7, in spite of the large size of the tooth.
Behind the first incisor region the very deep and thin lamina
is continued on, and soon the enamel-organs of iz and i;
appear almost simultaneously in the coronal series as a result
of the peculiarity in the topographical relations of these teeth
referred to in connection with the earlier stages. 7; is placed
more superficially and labially, and it would have been impos-
sible from an examination of this stage alone to have rightly
interpreted the appearance presented (fig. 57). In this figure
there will be observed a distinct labial outgrowth (diz) from
the root of the dental lamina. Adjacent sections show it
better as a thickened cell-mass attached to the root of the
lamina by a thinner connecting bridge, and they also prove
498 J. T. WILSON AND J. BP. HILL.
that it is related more intimately to 7, than to 7,, as might be
suggested by fig. 57. We can only interpret this as a vestige
of a diz, of which a trace was found in Stage 111 in the form
of a flattened mass of epithelial cells labially to the Anlage
of 73.
As iz and is disappear posteriorly, the dental lamina is
greatly reduced in depth for some distance, and then gradually
deepens again, becoming more and more elongated and swollen
as the canine region is approached. Here (fig. 58), but still
some distance in front of the actual canine enamel-organ, we
find the vestigial d; attached to the lamina by a delicate epi-
thelial strand (cd.).
It is of interest to note that in one of our series of sections
of this stage this is a precociously calcified tooth lke di;
(fig. 58), while in another series it exists in the form only of
a loose and degenerating epithelial mass.
The identification of this vestigial tooth as d; is rendered
easy by a comparison of its relations in the earlier stages
where no possible doubt could be felt, and where its present
relative dislocation is evidently in progress.
The lower permanent canine has not progressed very rapidly
during the period intervening between Stages 111 and Iv.
Although it has attained considerable dimensions, its papilla is
still very rudimentary, and its enamel-organ is only slightly
cupped. Furthermore it is not yet differentiated off from the
lamina, so that the residual lamina has not yet come into
existence. Only a short stretch of lamina intervenes between
the last trace of the canine and the commencement of the
first premolar (p;). The latter tooth is more advanced in
development than the canine, though not greatly so. It is
deeply cupped by a prominent papilla, and the differentiation
of the residual dental lamina has begun.
The second premolar is in a precisely similar condition to
that of the first, tapering away at its posterior end into a fairly
deep and thick dental lamina, which is continued back into the
third premolar region. In this situation dp; comes into view,
placed away to the labial side of the lamina. Since the
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 499
appearance of this tooth (except that it as yet shows no calci-
fication) and its relations to the liberated lamina are exactly
as described and figured in the upper jaw, no further par-
ticulars need be given.
Behind the third premolar region the lamina (still con-
tinuous throughout with the oral epithelium) undergoes
further deepening and becomes continuous with the large
enamel-organ of the first molar, which is larger and more
complicated than the corresponding upper tooth. Here the
dental lamina is for some distance separated off from the oral
epithelium. A residual dental lamina is well developed, as in
the upper jaw, and its freely projecting margin is moderately
swollen. Towards the posterior end of m; the dental lamina,
now unconnected with the tooth, is also for the most part
separated off from the oral epithelium, and is continued into
the region of the second molar. Here again it shows a labial
connecting strand (“ Verbindungsbriicke”) joining it to
the enamel-organ of mz; whilst its free marginal portion
forms a large residual dental lamina, just as in the upper jaw.
Towards the hinder end of mz the dental lamina again becomes
connected with the oral epithelium, exhibiting near its point
of connection a somewhat elongated labial laminar outgrowth.
Further back the lamina entirely loses its connection with the
oral epithelium, becoming more deeply placed among the con-
nective tissue of the jaw. Its distal portion is converted into
the papillated enamel-organ of mz, whose posterior end forms
the present termination of the lamina. The liberation of a
residual dental lamina by the side of mz has already been par-
tially effected.
Stage V.—P. nasuta: mammary foetus.
Length from vertex to root of tail . é . 37mm.
Head length ‘ : : ass
Coronal and horizontal series examined.
This stage represents only a slight advance on the preced-
ing.
Upper Jaw.—All three first incisors are now well developed,
voL. 389, PARY 4.—NEW SER. MM
500 J. T. WILSON AND J. P. HILL.
the third especially showing a distinct advance from the pre-
ceding stage. It is now triangular and pointed in coronal
section, and possesses a distinct dentine cap, as also do the first
and second.
The dental lamina is definitely interrupted between i+ and
12, but otherwise it is continuous, though often greatly re-
duced in size, throughout the rest of the incisor region, and it
retains its connection with the oral epithelium.
Towards the anterior part of the first incisor region irre-
gular epithelial cell-processes spring from the oral epithelium
close to the root of the dental lamina, similar to those men-
tioned in connection with the corresponding region in Stage
Iv (figs. 43—46).
The residual dental lamina is still clearly marked by the
sides of 74 and 72, but it is now considerably reduced in size,
absolutely as well as relatively to the tooth-germs. The resi-
dual dental lamina in the region of i has become definitely
established, and indeed it now closely resembles that of 7+
and i2, In each case, though forming a distinct and freely
projecting laminar downgrowth beyond the connecting stalk
of the tooth-germ, it is already a somewhat attenuated and
insignificant structure. We should certainly not be war-
ranted in following Leche by reading any special importance
into its persistent presence at this stage. It is plainly steadily
disappearing.
The fourth and fifth incisor enamel-organs are in a much
less advanced stage than those in front. They are still rela-
tively shallower and cup-like, with rounded papillz, much as
they were in Stage tv. They still appear as modifications of
the dental lamina which is directly continuous with the lin-
gual lip of their cup-like body, and no residual lamina has as
yet differentiated itself from them. As one might expect,
the portion of the lamina with which they are continuous is
bulkier, i.e. less reduced, than it is in regions where tooth-
germs have attained a more advanced stage of development.
The lamina behind 72 passes directly and with a somewhat
abrupt deepening into the canine lamina. This, with the
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES, 501
associated tooth-germ, is shown in fig. 59. Behind the canine
the lamina is abruptly shortened, and is continued after a
short interval into the bell-shaped enamel-organ of p+, whose
papilla is destitute of any dentine cap.
The differentiation of the residual dental lamina has set in
by the lingual side of the enamel-organ.
The second premolar exhibits a condition precisely similar
to that of the first, except that the faintest trace of a dentinal
cap to the pointed papilla may be observed. The residual
dental lamina is fairly developed.
Behind p2 the dental lamina becomes gradually larger on
cross-section, the germ of dp= soon appearing in the coronal
series. By the side of the latter tooth the dental lamina
exhibits precisely the same kind of modification as was seen in
the preceding stage in the corresponding situation (cf. fig. 49).
Fig. 50 shows a horizontal and fig. 60 a coronal section of
the structures concerned. It will be noted that in this stage
also the relative development of dp* is considerably retarded
as compared with that exhibited in earlier stages. Its dental
cap is only a trifle in advance of that of p2. From its present
condition alone it would be quite impossible to arrive at that
conclusion with regard to its serial homology which we find
it necessary to adopt in view of the earlier characters, both of
the tooth itself and of the rest of the dentition.
In the molar region the connection of the dental lamina
with the oral epithelium is frequently interrupted.
The first two molars are large and well developed. Lingually
from each there is a large and well-developed residual lamina
comparable to that figured for the last stage. The Anlage of
the third molar is slightly more bulky than in Stage rv, but
is still represented by a mere distal swelling of the lamina, as
represented in fig. 61. In this figure the projection marked
‘cb. m2” is seen by comparison with the anterior sections to
correspond to the root of the laminar connection (“ Ver-
bindungsbriicke”’?) with m2.
Lower Jaw.—The dental lamina in front of the region of
the first incisor shows some further antero-posterior extension
502 J. I. WILSON AND J. P: HILL.
since the period represented by Stage 1V. This anterior seg-
ment of the lamina now exhibits an attenuation of its “ neck,”
while its fundus shows a very definite oval enlargement, which
disappears as the anterior plane of 7; is reached. Here, as in
the preceding stage, the lamina is continuous with the “ re-
sidual” lamina by the lingual side of that tooth.
The first incisor is a large pointed germ, with a fairly thick
dentine sheath to the papilla. The milk incisor, di; (fig. 66 a)
is present as a strongly calcified miniature tooth, placed at
the labial side of the apex of 7. It has a small but distinct
dermal papilla, and the dentine shell is surrounded by a sheath
of small epithelial cells, which is attached by a laminar con-
nection to the root or attached border of the dental lamina
beside 23.
As in the preceding stages, 73 occupies a markedly labial
position, and owing to the kink of the lamina, whose occur-
rence has been already described, its anterior portion has come
to lie opposite the posterior end of the large iz. Posteriorly
it occupies a similar position with reference to nearly the
whole of tz. Indeed, in a few sections all three germs are
visible in the same coronal plane. Both iz and 7z possess
dentine caps, and end behind in nearly the same coronal plane,
though iz extends a trifle more posteriorly than 73. The con-
tinued dental lamina is, however, placed beside zz, the mor-
phologically posterior member of the series.
A long though attenuated residual lamina appears by the
lingual side of 7;, projecting far beyond the connecting bridge
between the latter and the dental lamina (fig. 66, rd/.). The
residual lamina is also present beside zg and iz, but, owing
to the dislocation of iz, the lamina has undergone corresponding
dislocation and is partly interrupted. As already indicated,
it is resumed beside the posterior end of 73, from which it
extends backwards.
A delicate epithelial lamina is to be seen springing from
the root of the dental lamina (i.e. from its attachment to the
oral epithelium) opposite 7g and 73, and extending labially to
i, This ends in a minute mass of cells, and it persists through
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 503
a considerable number of sections, and evidently answers to
the labial outgrowth figured in connection with Stage iv
(fig. 57) as a vestigial diz. It is now in process of absorption,
and apparently has never shown any sign of calcification.
From its first appearance in Stage 11 it has never attained
the dignity of tooth-like differentiation of structure.
Behind the third incisor region the dental lamina again
deepens, though its connection with the oral epithelium is
interrupted for a time. Its free margin thickens, and is con-
tinued directly into the canine Anlage, but very soon is
partially freed from it, being prolonged as a residual dental
lamina. The canine enamel-organ is now bell-shaped, and as
yet is uncalcified.
In the anterior premolar region the oral connection of the
dental lamina is frequently interrupted ; p; is slightly calcified,
while pz has as yet developed no dentine.
The deciduous premolar and its relations to the Anlage of
pz resemble those in the upper jaw, except that the pz thick-
ening does not exhibit the peculiar indentations present in the
upper jaw, and that dp is distinctly less advanced and is still
uncalcified ; in fact, it appears slightly less mature than in the
preceding stage.
The molar region shows an advance upon the condition in
the lower jaw of Stage rv, in that mz is now a fairly large
and deeply papillated tooth showing traces of differentiation
of cusps. A well-developed residual dental lamina is present
on the lingual side of each molar, and extends far beyond the
epithelial connection with the enamel-organ.
Stage VI.—P. nasuta: mammary foetus.
Length from vertex to root of tail . : . 44mm.
Head length : ‘ ; . 24
= ”
Lips now separated to form the ‘‘definitive Mundspalte.”
This stage exhibits a marked advance upon the last in
regard to tooth development. In it we may be said to en-
counter for the first time throughout the adult form of the
various teeth, except in the most posterior molar region all
504 J. T. WILSON AND J. P. HILL.
the teeth are well calcified, and have now attained such size
relatively to the jaws that, especially in the incisor region, a
good deal of overlapping occurs in both jaws.
Further, in Stage v the dental lamina, though frequently
interrupted and generally exhibiting signs of reduction and
commencing resorption, was, on the whole, fairly complete.
In the present stage (v1), on the other hand, the lamina is
exceedingly imperfect and scrappy ; and even where it appears
it usually displays much attenuation and irregularity of form
from the operation of the degenerative process.
Upper Jaw.—Beside the first incisor of the upper jaw
only comparatively slight traces of the dental lamina are
visible, and these are now quite unconnected with either the
tooth or the oral epithelium. It is impossible to distinguish
any specific remains of the residual dental lamina as such.
Beside i2 it is just possible to recognise portions of lamina
corresponding to ‘ Ersatzleiste” and “Verbindungs-
briicke” in a few places ; while in the region of 23 no resi-
dual lamina can be identified.
The fourth and fifth incisors are now well calcified, though
smaller than the anterior ones. Their more juvenile character
is perhaps evidenced by the persistence, in more regular form,
of the dental lamina with its residual prolongation. But these
latter also persist to a large extent in connection with the
canine, as shown in fig. 62 (rdl.), which represents the anterior
portion of that tooth (and therefore does not show the pointed
and strongly calcified character of its apex).
In the region of the anterior premolars the dental lamina is
scrappy, but vestiges of the free residual lamina remain here
and there, as well as portions of the rest of the lamina and of
the epithelial connecting bridge.
The lamina becomes more regular and complete behind p2,
and in the region of dp* its marginal portion is swollen, form-
ing the Anlage of p2 as shown in fig. 63. In adjacent sections
here and there it exhibits some indentations, but of irregular
character. In this its swollen portion it is surrounded by
markedly condensed connective tissue.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 505
The thickening of the residual lamina becomes even more
sharply accentuated than in fig. 63 in sections further back.
The deciduous premolar is a well-calcified tooth having a
sharply pointed chief cusp, while the anterior portion through
which the section passes in fig. 63 shows a low and flattened
crown.
Behind dp® the continued dental lamina becomes more uni-
form in thickness. In the region of m+ traces of the connec-
tion with the latter tooth are discernible, as also in region of
m2, has totally disappeared.
The character of m2 is shown in fig. 70. It exhibits only a
very slight advance on the preceding stages of the same tooth.
Lower Jaw.—In sections passing through the anterior
tip of 7; the same rudimentary enamel-organ is encountered
as was noted in Stages 1v—vit. It is, however, more irre-
gular than in Stage vir, and is evidently undergoing resorp-
tion, without having shown the least trace of calcification.
The tip of 7; has now grown far forward by the labial side
of z;, and has insinuated itself between the latter tooth and
its vestigial milk predecessor diz. This is a clear case of
relative displacement of ¢; similar to that which we have
shown to occur in the case of 2; with reference to the rudi-
mentary enamel-organ morphologically in front of it. 7, at
this stage shows the now calcified posterior secondary cusp
very distinctly. di; is in statu quo as regards its struc-
tural characters, only its relative topographical position is
slightly altered by forward growth of iz. 7; appears further
back by the labial side of 7,. It bears the same topographical
relation to 7; as the latter does further forward to 7;, so that
all trace of the early peculiarities in the relative situation of
these teeth is now lost.
The basal part of 7; may be traced far back, lingually to,
and in a deeper plane than, 7; and iz, all three appearing to-
gether in the sections throughout a considerable extent of the
anterior part of the jaw.
VOL. 89, PART 4,—NEW SER. NN
516 J. T. WILSON AND J. P. HILL.
dp is now a narrow tooth much elongated vertically, which
has cut the gum. Its residual lamina first becomes discernible
opposite its anterior end.
The Anlage of pz; is now more bulky than in preceding
stages. It is placed opposite the hinder part of dpz, and
though of larger size exhibits only a very slight depression
answering to the cupping of the enamel-organ ; and here its
papilla can hardly be said to exist as such, though the con-
nective tissue is evidently tending towards its formation by
local cellular accumulation opposite the depression.
In the posterior molar region the dental lamina ends as a
free laminar mass by the lingual side of mz, and in front of
the middle of that tooth. The Anlage of mz does not seem to
have made progress in the interval which separates this from
the preceding stage.
Stage X.—P. nasuta.
Length from tip of snout to root of tail ; . 118 mm.
Head length . : Se SON:
This stage is distinctly eamtior naan on the whole than
the preceding one, but the deciduous premolars are only just
cutting the gum, while in Stage 1x they had already protruded
somewhat.
In the upper jaw no teeth have undergone eruption save dp3,
but in the lower jaw the lower incisors have broken through in
addition to the deciduous premolar.
In the region of the canine there are still to be found abun-
dant remains of the residual dental lamina in the shape of
detached cell-masses, embedded in the fibrous connective tissue
lingually from the well-developed canine. These disappear
and reappear frequently in the serial sections, and, in various
sections widely separated from one another, one of the detached
cell-masses assumes a very definite pear-shaped form, and is
surrounded by a distinct connective-tissue capsule (fig. 72).
This pear-shaped mass is always the deepest mass present, and
obviously corresponds to the fundus-portion of the residual
lamina. Its appearance is highly suggestive of the “ knospen-
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 517
formig” tooth-germs of other writers. But the fact that
such appearances occur intermittently throughout nearly
the whole extent of the lengthy canine region seems sufficient
to negative this idea. We can point to quite a number, no
less than seven, absolutely separate and distinct “bud-
hike enamel-germs” of this sort, existing in antero-pos-
terior series by the side of the canine of this stage. Each of
these is fairly bulky, pear-shaped, and concentrically encap-
sulated. It is beyond doubt that they are nothing more than
portions of the swollen free distal margin of the residual lamina
which persists for some time after the antero-posterior con-
tinuity of that margin has been interrupted by regressive pro-
cesses. It is equally difficult to imagine a production of several
rudimentary “ bud-like enamel-germs of replacing teeth” by
the side of the single canine, or to imagine a primitively single
“bud-like”’ germ subdivided at intervals into a lengthy longi-
tudinal series of bud-like masses.
Let it be specially noted that in the case before us we have
presented the phenomenon of “knospenférmig” cell-masses,
long persistent, and distinctly encapsulated by the surrounding
connective tissue, i.e. offering the only credentials to which
Leche appeals for the determination of such structures as true
tooth-germs. The demonstration of their multiplicity in such
a case as the present seems to us to be the reductio ad ab-
surdum of the replacing-enamel-germ theory.
It seems to us that the presence of these structures at a
comparatively advanced stage of the canine tooth development
is largely explicable by reason of the considerable bulk of the
entire canine Anlage, to which attention has frequently been
called in these pages. The notable size of the canine lamina
is doubtless associated with a higher degree of formative
activity, which is longer in being exhausted here than else-
where. In any case we can attach no special morphological
significance to the continued presence at this period of even
such definite epithelial structures as those described.
The two anterior premolars are large pointed teeth, and cause
marked projection of the gum. Vestiges of the residual lamina
518 J. T. WILSON AND J. ‘P. HILL.
are here and there discernible by the lingual side of the teeth
as in the canine region, but on a smaller and less impressive
scale. Here and there the form of the residual lamina (pear-
shaped, with long stalk) is preserved, but for the most part the
vestiges—generally absent altogether—are in the form of scat-
tered minute epithelial cell-groups.
The deciduous premolar is in the act of cutting the gum with
the most anterior part of its crown.
The process of eruption is very clearly seen to be attended
by the flattening out of the enamel epithelium covering the
tooth crown and its conversion into squamous epithelium with
concomitant formation in it of epithelial “ nests’ or “pearls”
close to the tooth cusps. This nest-formation also proceeds in
the oral epithelium over the tooth. Eventually the epithelial
pearls become placed quite near the surface, and they then
undergo rupture and disintegration, their cores of concentric
epithelial cells being lost and the tooth crown exposed. These
different stages can be plainly followed in this one case by
tracing dp? backwards from its point of actual eruption to-
wards the posterior more deeply placed part of the tooth, where
the preparations for eruption are proceeding. The cell-nests
here noted as forming in the enamel epithelium are in all pro-
bability similar to those epithelial nodules noted by Poulton
(25) as appearing in the “ most superficial part of the middle
membrane of the enamel-organ immediately over the apex of
each chief cusp of the large broad posterior tooth” of Orni-
thorhynchus (see his pl. ii, figs. 4,5, 11, and 12, and pl.
ili, fig. 6). It may be noted, however, that in this stage of
Perameles the tooth is very much more advanced than those
figured by Poulton in Ornithorhynchus, and the middle
stratum of the enamel epithelium can no longer be distin-
guished as such. All we can say is that in the midst of the
epithelial investment of the tooth-crown, formed by the now
flattened enamel epithelium (as well as in the neighbouring
oral epithelium), these cell-nests are formed, and that the
process of eruption of the tooth under notice is accompanied
and so far conditioned by their denudation and disintegration.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 519
It is noteworthy that, in respect of its enamel formation,
dp in Perameles displays a striking degeneracy. At no
time do the cells of its inner enamel layer attain any great
degree of elongation as compared with those of other teeth,
and the layer of enamel actually formed is a very thin lamella
indeed.
The Anlage of p2 with its distinct though small papilla is
well represented in fig. 73.
In the molar region the residual dental lamina has prac-
tically disappeared from the first molar, but it is present by
the sides of m2 and of m2, in which latter situation it is shown
in fig. 71.
This section passes through the plane of the hinder end of
m3, which lies considerably to the labial side of the part
figured. It will be observed that the connective tissue shows
a definite concentric arrangement around the residual dental
lamina in this figure. Such an arrangement persists through-
out the whole extent of the residual lamina, though it is only
considerably behind the plane figured that it constitutes the
Anlage of m4.
Lower Jaw.—The first two incisors have already completely
broken through the epithelium at their apices. The anterior
primary cusp of 72 is just in process of eruption, and here there
may be observed at its tip a follicular or nest-like arrangement
of epithelial cells which is in the act of opening out on the
surface by disintegration, as was noted in the case of dp> in
the upper jaw.
di; is still observable, but it is now more degenerate in
character, merely consisting of asolid dentinal nodule, in which
no trace of a papilla or cellular core can be recognised.
dp; is in process of eruption, and this is associated with cell-
nest formation between the surface and the erupting tooth
point. The Anlage of p; resembles that of the upper jaw
pretty closely.
The fourth molar is a simply but deeply papillated enamel-
organ with the middle layer of stellate tissue well developed.
A residual dental lamina has already appeared lingually as a
520 J. T. WILSON AND J. P. HILL.
result of the differentiation of the anterior part of mz; the
hinder part of the tooth is not yet constricted off from the
parent lamina.
Stage XI.—P. nasuta: pouch specimen.
Length from tip of snout to root of tail : . 127 mm.
Length of skull . : ‘ ‘ ~ AS Ss
This stage is only slightly, if at all, older than Stage x.
Indeed, in certain respects development is less advanced, for in
the upper jaw no teeth have yet broken through the gum,
though the canine and premolars are causing very marked
projection of the latter.
Both the deciduous premolar and the Anlage of p3 closely
resemble those of the preceding stage. The proximal part
(or “‘neck”’) of the residual lamina, where that forms the
Anlage of p3, offers a very striking example of cell-nest for-
mation as a comcomitant of the process of disintegration of
the lamina. Here we have the papillated Anlage of p2 borne
upon an attenuated lamina, whose proximal border towards
the oral epithelium exhibits a number of large globular epithe-
lial swellings showing the concentric arrangement of the con-
stituent cells characteristic of epithelial “ pearl ” formation.
The molars resemble those in Stage x.
In the lower jaw diz is in the same degenerate condition as
in Stage x, but the dentinal nodule representing it is even
smaller. Otherwise the lower jaw presents no noteworthy
features.
Stage XII.—P. obesula: pouch specimen.
Length from tip of snout to root of tail ‘ . 120 mm.
Length of skull . : : 5 s Agen.
The lower jaw only of this specimen was studied. Though
its dimensions do not exceed those of Stage x1, it represents
a decidedly older stage. This is explainable by difference in
specific characters.
The Anlage of pz still shows a rudimentary papilla project-
ing into the slightly cupped enamel-germ. The whole Anlage
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 521
has an irregular appearance in many sections, owing to the
presence of irregular appendages derived from the “ neck”
portion of the dental lamina.
Stage XIII—P. nasuta: pouch specimen.
Length from tip of snout to root of tail ‘ - 180 mm.
Length of skull . A ; : Gnas
It will be seen from fig. 75, representing the upper jaw of
this stage, that all the antemolar teeth (except p®) have cut
the gum, and most of the general characters of the adult den-
tition are recognisable. Of the molars two in the upper jaw
and three in the lower have broken through. The very small
dp is seen occupying the slight interval between p2 and m1.
The Anlage of pz in the lower jaw is shown in section in
fig. 74. It is now a well-developed cupped enamel-organ,
and, in one of the two series cut, the dermal papilla is much
larger than in the section figured, filling up the whole of the
somewhat deeper cup of the enamel-germ.
In the latter the stellate tissue of the middle layer is now
well developed. In fig. 74 it will be observed that the enamel-
organ is now in process of constriction off from the lamina,
whose free end is already beginning to project as a new resi-
dual lamina (rdl.). That this interpretation is the correct one
is proved by reference to Leche’s figs. 140 and 142 (here-
with reproduced for comparison as figs. 76 and 77), showing
the corresponding enamel-organs of p? in Phascolarctus.
In connection with the latter a very evident and swollen
residual dental lamina has anew developed itself by the side
of p3. This, according to Leche, provides for the possibility
of the production of a tooth of the third (his “ fourth ”’) denti-
tion. We believe that this possibility must not be lost sight
of, but in the meantime we prefer to point out that the regular
development of such an Anlage as has been discovered and
figured, both by Leche and by ourselves, by the lingual side
of p&, serves in the most striking way to confirm our view of
the homology of p? to the other adult antemolar teeth, whose
residual dental lamine (‘‘Hrsatzleisten ”’—auct.) are there-
522 J. TIT. WILSON AND J. P. HILL.
fore merely equivalent to that which later develops
beside the third permanent premolar.
It may be noted that in Phascolarctus p? appears to
differentiate at a much earlier period than does the corre-
sponding tooth in Perameles; for though the stage figured
by Leche is considerably less advanced in its general develop-
ment, and as regards the rest of its dentinal characters, than
the present stage of Perameles, yet the Anlage of pz has
attained a somewhat higher degree of organisation.
Stage XIV.—P. nasuta: dried skull of a young adult
about three quarters grown. (See figs. 78 and 79.)
Skull length : 5
Here (fig. 78) in the upper jaw the fourth molar is in course
of eruption, and the premolar tooth-change is now almost
completed. The third permanent premolar, p2, has broken
through, and its crown has nearly attained the level of the
anterior premolars. By the eruption of its posterior portion
it is displacing the small dp®. On the right side the latter
tooth is missing, but may have been accidentally lost. On the
left side dp2 is still present, lodged in the slight hollow over
the posterior end of p3. In the lower jaw (fig. 79) the erup-
tion of pz is not so far advanced, but the relations are other-
wise entirely similar to those on the upper jaw. The crown
of dp? is worn and flattish or slightly hollowed. In this con-
nection we may note that Flower (4, p. 635, and Pl. XXX,
fig. 1) describes and figures the tooth-change in Perameles
in an animal not quite full grown. His description of the
condition is as follows :—‘‘ The permanent incisors, canines,
and two anterior premolars are in place. Behind these in each
jaw is a very minute, rather compressed tuberculated tooth,
succeeded posteriorly by the true molars of the permanent
series. In the alveolus above this minute tooth, which is the
temporary or deciduous molar, is lodged the germ of the
posterior permanent premolar, a tooth having a large com-
pressed, pointed, triangular crown, with small anterior and
posterior basal tubercles.” (PI. XXX, fig. 1.)
70 mm.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 523
PART III.
Discussion oF GENERAL PROBLEMS OF THE MARSUPIAL
DENTITION IN THE LIGHT OF THE FOREGOING OBSER-
VATIONS.
Our own conclusions may, perhaps, be most conveniently
introduced in the course of a discussion of the views of pre-
vious writers—already partly outlined in the introductory
portion of this paper. And in order to focus the discussion,
and to set forth its issues more clearly and definitely, we have
thought it useful to formulate certain leading questions, and
to define the attitude which has been assumed towards these
by other observers whose views are related to our own either
by way of agreement or contrast. In the formulation of these
questions we have pretty closely followed the lines, and even
in part the language, of writers who, like Thomas, Schwalbe,
and Leche, have so greatly advanced the intelligent discussion
of the subject of mammalian dentition.
I. “Do the existing teeth of Marsupials in front of
the last premolar correspond to those of the milk or
of the successional dentition of higher mammals?”
It will presently appear that the discussion of this question
involves a consideration of those criteria which may enable us
to determine a given embryonic structure as the germ of a
successional tooth.
II. How are we to explain the single tooth-change
or succession (i.e. in the case of the last premolar),
hitherto almost universally recognised as the sole
instance of tooth-change among Marsupials?
III. To which dentition belong the molars of Mar-
supials and of mammals generally? and—
IV. What is the nature of that generally admitted
process of evolution by which multicuspidate teeth
have been derived from a primitively simple uni-
cuspidate type?
024 J. T. WILSON AND J. P. HILL.
I. Homology of Antemolar Teeth of Marsupials.
In reference to the first question, it has already been pointed
out that Flower (and formerly Thomas) held that the non-
changing antemolar teeth of Marsupials answered to the
permanent or successional teeth of the Eutherian orders. This
determination was adopted with great unanimity up till the
period of publication of Kiikenthal’s discoveries, even by those
who dissented on other fundamental points from the Flower-
Thomas hypothesis regarding the dentition of Metatheria.
But from the time that Kiikenthal first recognised the per-
sistence of somewhat swollen (‘‘kolbig ”) downgrowths of the
dental lamina at the lingual sides of the developing teeth of
Marsupials, these downgrowths have on every hand been re-
garded as the true equivalents of the Eutherian successional
series, and the entire theory of the marsupial dentition has
accordingly been remodelled to suit the newer interpretation.
Up to the present, so far as we know, not a single investigator
has ventured to proclaim his dissent from the latter-day creed
in this respect.1
If, then, we proceed to inquire upon what foundation this
belief is based, we find that it rests solely on the occurrence,
in Marsupials, of the epithelial downgrowths referred to, taken
along with the similarity of the latter (@) to the earliest known
rudiments of successional enamel-germs in other mammals,
and (4) to the earliest stage of the one successional tooth
admittedly existing in the marsupial jaw. But the question
cannot, we believe, be thus summarily disposed of. It must
be subjected to a much more thoroughgoing criticism, such
as is embodied in the very weighty and thoughtful discus-
sion of the entire subject by Professor Leche in his recent
monograph (8).
In that work Leche has pointed out that, in order to decide
in any given case whether a tooth belongs to one or other
dentition, certain criteria have been employed whose worth
and validity are of very different degrees. He attempts to
1 But see addendum, p. 581.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 525
determine which of these may legitimately be utilised in
arriving at such a decision.
(1) The Criteria of Serial Tooth Homology as set
forth by Leche.
We propose first of all to summarise Leche’s treatment of
this fundamental problem. After reviewing his position the
extent of our agreement and difference with him will become
apparent in the course of discussion.
In the first place Leche discusses (a) the significance of
the occurrence of epithelial downgrowths of the
dental lamina lingually to developing enamel-organs.
That the presence of swollen epithelial downgrowths in-
ternal to the enamel-organs of mammalian teeth was sufficient
to stamp the latter as belonging to the milk dentition, no one
apparently, until Leche, had thought of questioning.
Woodward (2), indeed, is somewhat guarded in his language
when he writes, “If these various and often minute down-
growths of the dental lamina are to be interpreted as repre-
senting rudiments of teeth, as seems probable from comparison
with known rudiments of the first or second dentition in other
mammals,” &c.; but, nevertheless, he appears to accept in
their entirety those dentitional theories which depend wholly
upon the determination in question.
Leche, however, enters fully into the question of the validity
of the assumption that Kiikenthal’s swollen downgrowths
actually represent successional enamel-germs; and his con-
clusions are by no means decisive in favour of that assumption.
The following passages will serve to illustrate his attitude :—
‘“Da, wie bereits erwahnt, die Differenzirung des Schmelz-
keimes ausschliesslich oder doch vorzugsweise an der labialen
Flache der Schmelzleiste erfolgt, so ist auch von vorneherein
zu erwarten, dass das tiefe Ende der Schmelzleiste lingualwarts
vom Schmelzkeim auftritt. Ist also jene ‘ Knospe,’ welche
von demselben verdichteten Mesodermgewebe welches das
Zahnsackchen bildet, umgeben ist (figs. 9, 10), nichts anderes
526 J. T. WILSON AND J. P. HILL.
als das zuerst sichtbare Product des Abschniirungsprocesses
des Schmelzkeimes von der Schmelzleiste, so legt schon diese
Thatsache den Schluss nahe, dass die ‘Knospe’ nicht, wie
noch mehrfach auch von denneuesten Autoren angegeben
wird, an und fiir sich identisch mit einem Schmelzkeim, re-
spective einer Zahnanlage sein kann. Dieses giebt auch
daraus hervor, dass, wie die vorstehenden Untersuchungen
lehren, die Entstehung einer ‘ Knospe’ nicht an eine be-
stimmte Dentitionsreihe gebunden ist: sie tritt nicht nur neben
den typischen Milchzaihnen sondern auch neben solchen Zahnen
auf, die in der Regel ohne Nachfolger sind wie die Ersatzzahne
(figs. 55, 78, 79, u.a.) und die Molaren (text—fig. 2).”
The substance of this criticism is frequently repeated by
Leche, thus (8, p. 183):—**Steht es somit fest, dass das
Auftreten einer ‘ Knospe’ zunachst nur den beginnenden
abschniirungsprocess des Schmelzkeimes von der Leiste kenn-
zeichnet, ohne dass dadurch unbedingt ein neuer Zahn zu
Stande kommt,” &c.; and again (3, p. 1386), “ Das Vor-
kommen einer ‘ Knospe’d. h. das mehr oder weniger frei
hervortretende Schmelzleistenende neben einen Schmelz-
keime keineswegs beweist, dass der letztere zur ersten Den-
tition gehort. Ja, wir konnen weiter gehennicht einmal die
weiterbildung dieser ‘ Knospe’ zu einem wirklichen Schmelz-
keime berechtigt zu dem Schlusse, dass der mit einer solchen
Zahnanlage ausgestattete altere Schmelzkeim unbedingt der
ersten Dentition angehort, da, wie wir aus den obigen Unter-
suchungen wissen, auch lingualwarts von typischen und
unbestrittenen Reprasentanten der zweiten Dentition solche
knospenformige Schmelzkeime vorkommen ké6nnen, welche
sich in einigen Fallen zu vollstandigen Zahnen ausbilden.”
Having thus provisionally disposed of the assumption that
the mere presence even of a ‘ bud-like” lingual downgrowth
of the dental lamina is sufficient to establish the “milk”
homology of the tooth beside which it is found, Leche goes on
to pass in review other criteria which might be taken as afford-
ing the necessary test of a supposed serial homology. He
dismisses (b) “simultaneousness of function” as afford-
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELBS. 527
ing quite incompetent evidence of serial homology between
individual teeth, on the ground that in some mammals (notably
in Erinaceus) certain of the milk-teeth function along with
successional teeth.
Again, (c) the different depth at which one enamel-
germ springs from the dental lamina compared with
another can, in Leche’s view, afford no reliable criterion of
serial homology. Thus in Erinaceus the dental lamina in
the situation of several of the antemolar teeth is markedly
shallowed, and is wholly transformed into an enamel-organ
very superficially placed. But it is found that the dental
lamina, at the places referred to, has to produce only one
tooth, and the anomalous position of the enamel-germ is held
to be explained merely by the absence of tooth-change in that
particular segment of the jaw.
While agreeing in general with the judgment expressed
above, we are bound to add that the reasoning appended to
it is not conclusive to us, for the absence of a tooth-change
will not account for discrepancies of situation in many cases,
e.g. the striking depth of the upper canine Anlage frequently
referred to in this paper may be contrasted with the relative
shallowness of the enamel-germs of the two anterior premolars
which are without doubt serially homologous with the canine.
On the other hand, we have found that enamel-organs which
at first are close under the oral epithelium, and which have
(as in Leche’s description) no dental lamina or only a very
short one, superficial to them, do afterwards exhibit quite a
well-marked dental lamina, and from this by-and-by a residual
lamina is differentiated. This is what happens in the case of
the anterior premolars of Perameles, though they never attain
the great depth of the canine Anlage.
We are of opinion that the relative depth of the various
enamel-organs depends upon several factors, partly, perhaps,
intrinsic to the lamina itself, and dependent upon the degree
of its formative activity as an epithelial mass; and partly
extrinsic, and dependent upon its topographical relations to
various other structures.
528 J. T. WILSON AND J. P. HILL.
(d) The time of development of a tooth, in so far as
this takes account merely of the attainment of maturity and of
the period of eruption of a tooth, fails, in Leche’s judgment,
to provide the necessary evidence, and constitutes a quite un-
reliable test of the serial identity of different teeth.
This test is, however, constantly relied upon by Rose, appa-
rently without any misgivings as to its validity. Thus, con-
cerning the last premolar and third incisor of Phalangista
Cookii, he says (1, p. 702), ‘‘ Diese Zihne sassen noch tief in
ihren Alveolen, die Farbe und Dichtigkeit ihrer Schmelzkappe
wies darauf hin, dass sie von bedeutend jiingerer Bildung
sind als ihre Nachbarn. Von einem Zahnwechsel, d.h. der
Resorption eines Zahnes der ersten Reihe konnte auch hier
keine Andeutung angefunden werden. Es scheint mir sehr
wahrscheinlich, dass auch bei Phalangista Cookii nicht
allein der letzte Pramolar, sondern auch der dritte Incisivus
des Oberkiefers aus der zweiten Zahnreihe entsteht,” &c.
(e) But, on the other hand, the time of development of
a tooth takenin a different sense furnishes a test which
Leche recognises as of the highest importance, if not of para-
mount authority. Its value depends upon the validity of the
following generalisation :— Die Anlagen der zu derselben Den-
tition (Zahngeneration) gehohrigen Zahne differenziren sich
gleichzeitig oder nahezu gleichzeitig an der Schmelzleiste”
(3, p. 137).:
To this law of contemporaneousness of origin of the mem-
bers of the same dentition Leche does indeed attach certain
qualifications, but it nevertheless remains for him the most
valid criterion in the determination of the serial homologies
of the different teeth.
(2) The Serial Homology of the Marsupial Antemolar
Teeth as tested by Appropriate Criteria.
We may now turn to the results of Leche’s application of
his critical method just expounded to the general question
under discussion, viz.: Are the existing teeth of Marsupials
in front of the last premolar homologous to the first or to the
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES, 529
second dentition of higher mammals? And it is noteworthy
that Leche’s conclusions are to a very large extent negative as
regards the value of just those developmental features upon
which alone Kiikenthal and Rose relied to establish the “milk”
homology of the non-changing antemolars of Marsupials.
It had simply been assumed by Kiikenthal and Rése that
the presence of a residual lamina (‘‘ Ersatzleiste” of Rise),
more or less swollen, by the lingual side of an enamel-organ
gave sufficient assurance of the ‘ lacteal” character of the
latter. Leche’s weighty and, as we believe, entirely valid con-
tentions have, at any rate, opened the door to scepticism with
regard to this doctrine. It remains to be seen upon what
grounds Leche after all gives in his adhesion to a view which
has been rendered assailable with the weapons he has himself
provided.
The Nature and Morpologieal Value of a
“Residual Lamina.”
In one of the numerous passages in which he discusses the
morphological value attachable to the liberated residual dental
lamina, Leche utters his opinion as follows :—‘* Das durch die
Emancipation des Schmelzkeimes freigewordene Schmelz-
leistenende bildet an und fir sich nur die Voraussetzung fiir die
Entstehung einer Zahnanlage : je bedeutender dieser Schmelz-
leistentheil ist, desto grésser ist die Pradisposition fiir die
Bildung eines neuen Zahnes. Und wir k6onnen hinzufigen:
ist dieses Schmelzleistenende wirklich kolbenoder knospen-
formig angeschwollen und von einem Zahnsackchen umgeben
erst dann ist diese Moglichkeit als realisirt zu betrachten, erst
dann konnen wir von einer (knospenformigen) Schmelzkei-
manlage reden, einerlei ob diese Anlage sich spater weiter
entwickelt oder nicht ” (8, p. 134).
It will be remarked that this passage is by itself quite inde-
cisive. It merely expresses the author’s view that any free
end of the dental lamina swollen (i.e. “kolbig” on section),
and surrounded by an investment of condensed connective
tissue, has a right to be regarded as an enamel-germ, though
530 J. T. WILSON AND J. P. HILL.
it may be an abortive one. And, since the author holds that
such abortive germs are at least occasionally met with beside
the true successional teeth of higher mammals, it is evident
that we are not enabled by their mere presence to
determine which tooth generation, if any, they re-
present.
Elsewhere, however, in a very important passage Leche
explicitly sums up the argument in favour of the true “suc-
cessional” character of the lingual downgrowths beside the
various marsupial teeth in the following words:
*“‘ Charakteristisch fiir die Beutelthiere sind also sowohl das
constante Vorkommen und die scharfe Auspragung dieser
Schmelzkeime und ihre Uebereinstimmung mit dem Schmelz-
keim des p. 3, als auch ihre lange Permanenz, welche Higen-
schaften diese Gebilde nicht unwesentlich von den lediglich
durch die Emancipation der Zahnanlagen von der Schmelzleiste
entstandenen ‘ Knospen’ unterscheiden, ein Punkt, den ich
hier ganz besonders betonen méchte. An einigen dieser Schmelz-
keime sind deutliche Zahnsackchen vorhanden” (38, p. 103).
Now we cannot but feel that if Leche’s criticisms detailed
above be well grounded, and if the sceptical lesson he repeatedly
inculcates have taken hold upon the reader, the latter will find
it difficult to discover in the summary quoted anything at all
sufficient to reassure him of the certainty of that identification
which the author advocates.
It is difficult to understand how “ constancy of occurrence ”
—‘‘a sharply stamped character ’’—and “long persistence ”
can possibly constitute a ‘not unessential” difference. To
us there appears nothing at all “essential”? about such dis-
tinctions, and thus far we must hold that Leche’s own verdict
ought to have been, at most, “ not proven.”
In one of the passages just quoted from Leche (on p. 108)
we find him asserting that whenever the free end of the
(residual) dental lamina is actually swollen in a “kolben”
or “ knospenformig” manner, and when in addition we can
recognise the presence of an investment of condensed connective
tissue to form a “‘ tooth-sac,” then we have a right to speak of
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES, 531
the structure as an “‘ enamel-germ,” whether it develops further
or not.
With regard to the first of these two alleged guarantees
of identity as genuine enamel-germs, it is held by the present
writers that undue importance has been attached to the dis-
tinction between the free end of the dental (residual) lamina
while in the condition of a mere prolongation beyond the de-
veloping enamel-organ, and the same prolongation when it has
undergone some slight enlargement. With most writers, it is
true, that distinction is not a very prominent one. Indeed, it
would appear that, to the majority of them, practically any
lingual prolongation of the dental lamina beyond its point of
connection with a developing tooth is to be regarded as an
“Krsatzleiste,” and therefore as the homologue of a suc-
cessional tooth. In Kikenthal’s earliest reference to the
marsupial condition (24) we find him basing his view of the
“milk ” homology of the marsupial incisors upon the possession
by them of an “Ersatzleiste” with a swollen (“kolbig’’)
free end. On the other hand, neither Rose nor Woodward lay
any particular stress upon the enlargement of the free end of the
residual laminal downgrowth. They do not ignore its frequent
occurrence, but neither do they seem to regard that feature as
asine qua non of its title to consideration as an enamel-
organ. The passage already quoted from Woodward (p. 525)
will sufficiently bear out this statement as far as that writer is
concerned.
But according to Leche a mere lingual prolongation of the
dental lamina beyond the point of attachment of the enamel-
organ of a tooth gives of itself no certain promise of future
tooth formation, or perhaps one should say, no reliable evi-
dence of its morphological value as an enamel-organ of a
succeeding tooth generation. Leche is thus forced to base a
fundamental morphological distinction in large part upon that
swollen (“kolben” or “knospenférmig”) character
which is held to distinguish the residual dental lamina beside
the teeth of Marsupials. It may be quite true that such
thickening of the free end of the residual lamina is an invari-
VoL. 39, PART 4,—NEW SER. 00
532 J. T, WILSON AND J. P. HILL.
able antecedent to the first formation of a true successional
tooth. But it is not equally permissible to assert, conversely,
that wherever such a feature is perceptible there we must
recognise the presence of a rudiment of an enamel-organ
belonging to a definite tooth generation. The fact is that
under nearly all the discussions touching the ‘‘ swollen ” resi-
dual lamina or “Ersatzleiste”’ of Marsupials—not even
excluding Leche’s more critical remarks—there appears to us
to lurk an assumption which is traceable to the use of some-
what ambiguous descriptive terms. Thus the term “ bud”
(“ Knospe’’) seems to us wholly misleading as descriptive of
the somewhat thickened downgrowth (our residual dental
lamina) so frequently met with by the sides of the developing
enamel-organs in the marsupial jaw. The word ‘ bud-like ”
(““knospenférmig”) may indeed describe the outline ap-
pearance of the lingually prolonged lamina as seen in a
cross-section, but it is thoroughly inapplicable to the actual
solid form which that downgrowth really possesses,
We must therefore raise a most emphatic protest against
the use of a phraseology such as that employed by Leche in
the following sentence :—‘‘ Zuniachst ist zu betonen, dass bei
den Beutelthieren die Schmelzkeime der zweiten Dentition
sich meist langer als die Schmelzleiste erhalten, also ganz wie
bei einer Anlage, aus der ein Zahn sich wirklich entwickelt :
hatte die Zahnanlage jede Bedeutung eingebusst, so ist
schwer einzusehen, wesshalb sie als knospenformiger Schmelz-
keim sich langer als ihr Mutterboden, die Schmelzleiste erhalten
sollte” (8, p. 105). Here the distinction assumed between
‘“enamel-germs of the second dentition” and the ‘ dental
lamina” is a most unjustifiable one. The idea that, during
disintegration of the latter, we can distinguish “ bud-like
enamel-germs ”’ which are preserved longer than “the parent
structure—the dental lamina” is a wholly fanciful one, de-
pendent upon the confusion of the outlines of cross-sections
with those of solid extended structures.
In his recent criticism of Leche’s general attitude towards
the question of marsupial teeth (27) Kiikenthal employs the
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 533
sane objectionable form of expression— bud-like enamel-
germs”: “Es sind nun, wohlgemerkt, nicht etwa die Enden
der frei endigenden, einfachen Zahnleiste, die nach innen von
den durchbrechenden Zahananlagen der Beuteltiere liegen ”
[but they are really nothing more], “ sondern knospenformige,
teilweise von verdichtetem Bindegewebe umgebene Schmelz-
keimstadien,” &c. (27, p. 655).
Localised serial swellings of the continuous
dental lamina in its liberated or residual phase,
such as we can justifiably speak of as separate
enamel-germs, do not exist, either distinguished
from, or as integral parts of, the residual lamina
aforesaid.
An exception to this statement must of course be made in
favour of p. 3, which arises in the first place as a localised
thickening of the residual lamina by the side of dp. 3. But
p. 8, as we shall attempt to prove more conclusively below, is
serially homologous with the enamel-organs of the other ante-
molar teeth of the adult, and not with the residual lamina
developed by their sides.
We cannot insist too strongly on the laminar character of
the residual lamina. As we have again and again shown in
the course of our description of stages, it arises as a result of
the constriction process whereby the permanent enamel-organs
are differentiated from the main dental lamina. And, when
thoroughly established, it is in perfect morphological con-
tinuity with the undifferentiated dental lamina fore and aft,
intermediate between the region of one tooth and those of the
next in front and behind. This is well brought out in our
fig. 82, which is a combination drawing from several suc-
cessive horizontal sections of the upper jaw of our Stage v.
The complete antero-posterior continuity of the dental lamina
in the anterior premolar region could not be demonstrated in a
single section, but in the combination figure no violence has been
done to any essential relationship. And the figure will be suffi-
cient to illustrate the manner in which the free residual lamina by
the lingual side, say, of p+, simply forms the direct continuation
5384 J. T. WILSON AND J. P. HILL.
of the undifferentiated dental lamina before and behind, which,
in turn, is continued into the residual lamin of £ and of p?.
(We have carefully verified the planes of section, and have
definitely determined that they correspond to the free portion
of the epithelial [residual laminar] downgrowth. But, indeed,
the view here expressed is supported by the whole body of our
observations.)
The merest glance at our figure will serve to show that, at
least in this region, the term ‘‘ bud-like ” applied to the resi-
dual laminar downgrowth is ludicrously inappropriate. The
swollen “ bud-like ” outline seen on cross-section is due to a
general and continuous thickening of the free marginal portion
of the lamina.
It may with justice be contended that a perfectly similar
marginal thickening of the primary dental lamina may be
noticed prior to the differentiation from it of the earliest
enamel-organs (though then the whole structure is much fuller
and plumper than is the residual lamina of later stages), and
that the later recurrence of such a condition may well be taken
as heralding the advent of a new dentitional series. To this
we reply that that general and continuous marginal thickening
of the lamina, primary or residual, of itself signifies
nothing at all in the way of tooth differentiation.
It is possible, indeed, to affirm—but this is equally true of a
non-swollen residual dental lamina destitute of any tooth-sac—
that the epithelial residuum is the potential equivalent of
further successional teeth. But in our view there is no more
reason to identify the residual dental lamina with any one tooth
generation than there is to identify the primary and undifferen-
tiated dental lamina of a higher mammal specifically with the
“milk,” to the exclusion of the “‘ permanent ”’ series of mam-
malian teeth.
We regard the lingually placed ‘epithelial downgrowths ”
beside the normal adult enamel-organs merely as portions of a
dental lamina not yet wholly exhausted of its formative acti-
vity, and therefore possibly still capable, exceptionally, of
providing material for the production of teeth homolo-
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 535
gous with a third Eutherian dentition (fourth dentition of
Leche).
The hypothetical product of the residual dental lamina may
perhaps be represented as tooth-series to the mth power,
rather than as any definite series. Kiikenthal (24) aptly
remarks—and in this he is following the lead of Baume (26)—
that the two mammalian dentitions are sisters whose mother
is the undifferentiated dental lamina. This is true on the
whole, but according to our theory it requires the further
comment that the “Ersatzleiste” in Marsupials is analo-
gous to the mother after the second parturition (the
product of the first pregnancy having been early
aborted), and ought not to be mistaken, as heretofore, for
the younger daughter. The theoretical possibility of the
occurrence of a third pregnancy need not be excluded. Leche
has shown that occasionally in mammals such an event does
actually occur; but so far as our experience goes the formal
possibility or promise of a third dentition has been absolutely
unfulfilled in Marsupials. In this connection we may ask the
reader to compare Leche’s figures (e.g. his figs. 106, 113, 115,
116, 118, 121, 124, and 137) of residual laminar downgrowths
in various marsupials, which he labels as “ knospenformig,”
or even actually as ““Schmelzkeime,” with his figures (e. g.
figs. 28—380, 74, 84, 94, and 95) of residual downgrowths of
the lamina by the lingual sides of the true permanent or resi-
dual teeth of certain other mammals, and also with the two
figures (our figs. 76 and 77) we have reproduced from his figs.
140 and 142, showing the similar appendage to p3 in Phasco-
larctus. It will be seen that no structural differences serve
to distinguish the downgrowths in these different cases.
In view of the foregoing discussion the term “ knospen-
formig ” will be seen to be thoroughly out of place as applied
to the slightly thickened laminar downgrowths figured in
cross-section in Leche’s figs. 187 and 140.
It has been shown in connection with our Stages 11 and 111
that the first rudiments of the enamel-organs of the individual
teeth arise as localised enlargements of the primary lamina
536 J. T. WILSON AND J. P. HILL.
especially in the form of labial outgrowths therefrom. And
if it could be shown that the continuous residual lamina of
Marsupials does actually develop regular localised enlarge-
ments, bud-like or other, by the lingual sides of the enamel-
organs of the adult teeth,—of similar character to that which
we have seen to constitute the first rudiment of p3, then the
modified views of Leche would receive some support. But
such is not the case. Of definite and regular “ bud-like ”
swellings, other than expressions of that general and conti-
nuous marginal enlargement to which we have already alluded,
we find no trace, unless indeed the production of isolated
groups of cells and the formation of epithelial “ pearls’? and
*‘cell-nests”’? during the final disintegration of the residual
lamina were to be reckoned as such. And in the latter event
we have shown that no less than seven rudimentary succes-
sional germs must be held to be present in the region of the
upper canine of Perameles.
It is true that the residual lamella is not absolutely even
thoughout its entire extent, but presents various slight irregu-
larities and inequalities, both in contour and in the thickness
to which its free marginal portion attains here and there.
These, however, never amount to localised swellings, which
could possibly be spoken of or regarded as in any sense
differentiated from the lamina. (The mode of origin of p.3
will receive special attention.) Where local variations may be
recognised they are extremely indefinite, and their only sem-
blance of regularity of arrangement results from the fact that
the level of the free distal edge of the residual lamina is to
some extent affected by the depths of the enamel-organ, beside
which it lies ; hence, e. g., it is much deeper opposite the canine
than beside the last incisor. These differences have an obvious
and easy mechanical explanation. Nothing further issues
from the structural condition thus brought about, except
perhaps a slightly longer persistence of degenerating remains
of the deeper lamina.
The view here set forth is not only borne out by our own
pretty extensive observations, directed with special reference to
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES, 537
this point, but also by examination of Rése’s figures of his models
of the developing dentition in Didelphys (11; cf. especially
his fig. 12). In none of the stages there represented do we
get any indication of “ bud-like” or partially isolated struc-
tures arising from differentiation of the residual lamina, which
is there quite clearly shown (cf. his figs. 12 and 15) as a con-
tinuous band. It is true that the contour of the latter, whilst
plain and fairly even in parts, is in other places crooked and
irregular; but the latter condition is obviously attributable,
chiefly if not entirely, to the overcrowding of the developing
teeth (cf. especially the incisor region of his fig. 12).
In Rése’s models of the early stages of the human dentition
(9) there may be followed the process of formation of a residual
dental lamina, which up to a certain point exhibits characters
very similar to those we find in the case of the residual dental
lamina of the marsupial jaw. But the structural similarity
between the indifferent stage of the human residual lamina
and that of Marsupials really yields no evidence either for or
against their actual and strict homology. For Leche has
clearly proved! that residual laminze may be developed by the
side of the enamel-organs of the second (permanent) dentition
of higher mammals. And indeed, as already pointed out, he has
figured a residual laminar downgrowth by the side of the one
undoubted successional marsupial tooth in Phascolarctus
(figs. 76 and 77), an observation which we have to a large
extent been able to confirm for Perameles (cf. p. 521 and
fig. 74).
It is, to say the least of it, very suggestive that just in
Phascolarctus, where the deciduous premolar is very rudi-
mentary, and where, therefore, we may suppose the formative
activity of the dental lamina to be less exhausted than in other
1 See especially his notes on the third incisor and canine of Hrinaceus (6)
on pp. 25, 26 of his monograph (8), and his illustrative figures, 28—30. Cf.
also fig. 74 showing residual lamina (“ Schmelzkeimahnliche”) beside
p. 1 in Phoca; fig. 94 showing same in Desmodus; and fig. 97 showing a
similar condition of residual lamina beside i2 of the same animal. Cf. also
Marett Tims’ researches on the dentition of Canis (87).
538 J. IT. WILSON AND J. P. HILL.
forms, we find, by the lingual side of the early maturing
p. 38, a well-formed residual lamina as figured by Leche. So
also the preservation of the dental lamina in Desmodus and
Phoca, referred to and figured by the same author, is doubtless
due to the weaker development of both tooth series, i.e. there
is more formative material left over than in other cases.
Connective-tissue Condensation around a Supposed
Enamel-organ.
Leche, as we have seen, attached considerable importance to
the occurrence of a “rudimentary tooth-sac,” in the shape of
a condensation of the connective tissue around a supposed
enamel-germ. But while he utilises it for diagnostic purposes,
he is yet by no means blind to its equivocal character. ‘ Dass
jene Verdichtung des Mesoderms an sich durchaus nicht immer
die Anlage einer Zahnpapille oder eines Zahnsackchens zo
sein braucht, dass sie vielmehr das rein mechanische Produkt
des Eindringens des Ektoderms ist, geht ausser aus den
obigen Thatsachen auch aus dem Umstande hervor, dass, wie
schon Baume (p. 66) beobachtet hat nicht nur der Schmelz-
keim sondern auch die Schmelzleiste, falls sie geniigend tief
in das Mesoderm eindringt, von verdichteten Mesoderm-gewebe
umgeben ist’ (3, p.15). And again :—~ Sowohl aus Baume’s
als meinen Untersuchungen geht hervor, dass die Schmelz-
leiste tiberall da, wo sie geniigend tief in das Me-
soderm eindringt, eine Verdichtung in diesem her-
vorruft. Wie ich oben (p. 15, figs. 4 and 5) naher ausge-
fihrt habe, ist also diese Verdichtung und Abplattung der
Mesodermzellen durchaus nicht immer die Anlage eines Zahn-
sickchens oder einer Zahnpapille, sondern vielmehr als das rein
mechanische Product des Eindringens der Ectodermleiste
aufzufassen. An den Stellen, wo die Schmelzkeime entstehen,
schreitet durch den verstarkten Druck, welchen diese auf die
umgebenden Mesodermzellen ausiiben, die Verdichtung und
Abplattung der letztern weiter zur Bildung von Zahnsickchen
und Zahnpapille, wahrend durch die Riuckbildung der
Schmelzleiste in den Zwischenréumen zwischen den Schmelz-
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 539
keimen die von jener hervorgerufene Differenzirung im Meso-
derm wieder ausgeglichen wird ” (8, p. 131).
These quotations from Leche are sufficient to dispose of the
claim that a mere dermal condensation around an epithelial
downgrowth can give any reliable guarantee of its nature.
Our own observations have fully convinced us of the worthless-
ness of this feature as a test character. Connective-tissue
condensation, in which the cells become flattened and con-
centrically arranged around an epithelial downgrowth or
enamel-organ, is beyond doubt simply the response to the
stimulus supplied by the continued proliferative activity of
the epithelium. Naturally such condensation will precede
tooth formation, but it will also accompany every other con-
tinued growth of the epithelial lamina to whatever cause it is
due; and, as a matter of fact, it is most markedly seen around
those proliferating epithelial masses or cell-nests which so
frequently appear during the absorptive and retrogressive
metamorphosis of the dental lamina.
Even on the ground of his own admission, then, we may put
aside as incompetent that test of the genuineness of an enamel-
germ which Leche bases upon the presence or absence of a
rudimentary tooth-sac. And there remains only the swollen
(“kolbig” or “knospenférmig”’) character of the epi-
thelial downgrowths of the dental lamina, already criticised
at some length.
That Leche should find it necessary to appeal to the “ con-
stancy of occurrence,’ “sharply stamped character,” and
* long persistence ” of the supposed successional germs seems
to us to indicate some mistrust of the judgment founded upon
their structural features. But if the latter are insufficient to
suggest that theory of their homology which he espouses, we
are at a loss to see how the theory can be strengthened by the
appeal to the vague characteristics referred to. Nor have the
characteristics themselves greatly impressed us. The struc-
tures in question do, doubtless, “‘ constantly occur,” but they
are no more “sharply stamped” in character than the fact of
their mere existence would demand ; nor do they “ persist ”
540 J. T. WILSON AND J. P. HILL.
beyond a period consistent with the general process of dis-
integration of the dental lamina of which they form parts.
Worthlessness of Proof from Agreement of Develop-
ment of p. 83 with that of Supposed Enamel-germs.
In the passage quoted above from Leche (p. 198) it will be
found that in discussing the residual laminar downgrowths
he mentions their agreement with the enamel-germ of p. 3 as
tending to establish their claim to be regarded as themselves
rudimentary enamel-organs.
It is doubtless true that in its evolution p. 3 arises out of
the residual dental lamina of dp. 3, and that that lamina at an
early period exhibits the same developmental phase exhibited
by the residual lamina subsequently developed by the lingual
sides of the other enamel-organs. But in view of our previous
contentions the fact of this general agreement can carry no
weight as a proof of true serial homology.
Application of the Criterion of “ Contemporaneous-
ness of Origin.”
Leche in the next place passes on to apply in this connec-
tion his favourite criterion of serial homology, viz. that of
Contemporaneousness of origin of the different An-
lagen from the dental lamina (cf. p. 528).
Remarking that the centre of gravity of the whole question
of the homology of the persisting antemolar teeth of Mar-
supials lies in that of the relations of dp. 3 and p. 3 to the
other antemolar teeth, Leche proceeds to apply to this question
the above-mentioned test. And instead of obtaining a clear
and unquestioned verdict in his favour on this head he finds
himself from the first involved in a difficulty, which has to
be obviated by an important qualification of his doctrine of
*contemporaneousness.” He finds, in fact, that the Anlage
of dp. 8 is, after all, not contemporaneous in its first differ-
entiation with the Anlagen of the other antemolar teeth with
which he is attempting to demonstrate its serial homology,
but is really in advance of the latter. The further hypothesis
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 541
invoked in explanation of this is that the stage of differentia-
tion of the Anlagen is conditioned by the future size and grade
of development of the teeth concerned. And just in so far as
this contention is valid, the value of Leche’s paramount test
is of course discounted. The author’s own words in this im-
portant connection may be quoted: “ Nun ergiebt sich sowohl
aus dem von Rose (vi, fig. 4) abgebildeten Modelle eines
153 mm. langen Didelphys-Jungen als auch aus der
Untersuchung meines Stadium B (17 mm. lang), dass alle
Zahnanlagen dieselben Beziehungen zur Schmelzleiste zeigen,
dass aber die Entwicklungsstufe der verschiedenen Anlagen—
und dies geht auch aus Kiikenthal’s Mittheilungen (I, pag.
662) hervor—schon auf diesen zeitigen Stadien der Kiinftigen
Grosse und Ausbildung des betreffenden Zahnes entspricht.
So ist der grosste von ihnen, namlich Pd. 3,—ich sehe natur-
lich von den Molaren ab,—auch Zugleich der am weitesten
entwickelte, dann kommt der nachstgrdsste (Cd.), welcher
weiter entwickelt ist als die schwacheren Pd. 1, Pd. 2, und
die Schneidezahne. Man hat daher ebensowenig Recht, dem
Pd. 8 seiner hoheren Entwicklungsstufe halber, einer alteren
Dentition Zuzuzahlen, wie wenn man aus demselben Grunde
in Frage Stellen wollte, dass der Kckzahn zu derselben Denti-
tion gehort wie die tibrigen Zahne (Pd. 1, 2, und Schneide-
zahne). Bei dem etwas alteren Stadium C von Didelphys
(siehe oben pag. 88) finden wir sogar, dass Pd. 3 weniger
entwickelt ist als z. B, Pd. 2. Es bildet also dieser Umstand
eine Sttitze fiir die Zurechnung des Pd. 3 zu derselben Denti-
tion wie die persistirenden Ante-Molaren ” (3, p. 102).
Now it may be quite true that the future size of a tooth
does, to some extent, influence the period of the differentiation
of its Anlage from the dental lamina, i.e. the mere bulk of
the lamina in a given position may determine the more hasty
evolution of a tooth-Anlage in that situation. We may
further point out that yet another factor does, at least in the
case of the molars, determine the order of appearance of the
Anlagen of teeth of the same series, viz. the relative
position in the jaw. And it is at least possible that this
542 J. T. WILSON AND J. P. HILL.
factor may have some influence in the case of the antemolar
teeth, seeing that the first formation of the dental lamina
itself progresses from before backwards. But the significantly
early differentiation of the enamel-organ of dp. 3, which we
have shown to occur in the case of Perameles, obviously
cannot be explained either by position in the jaw or by future
size of the tooth. It will be remembered that dp. 3 in
Parameles is the smallest of all the (functional)
teeth in that animal. And yet it is differentiated from
the dental lamina before any of the enamel-germs of the adult
dentition have taken origin. Leche has been misled in the
case of dp. 3 through the accident of its large size in Didel-
phys. Had he been conversant with the facts of develop-
ment as we have ascertained them in Perameles, he could
not have rested satisfied with his suggested explanation.
The relatively early differentiation of the adult canine
Anlage, to which Leche refers in Didelphys in order to
illustrate the law that future size influences period of differen-
tiation, is not to be remarked in Perameles in our, for this
purpose, most critical Stage 111.1 On the other hand, it is a
fortunate circumstance that just in this stage we possess a
perfect example of the differentiation, in connection with the
canine Anlage of an enamel-organ which is truly synchronous
with the enamel-organ of dp. 38, and which cannot, in our
judgment, be regarded as other than that of a genuine canine
“milk” tooth. This fact will, however, be more fully dis-
cussed in the sequel.
The fact, to which Leche several times refers, that in some-
what later stages of Didelphys (e. g. his Stage C) dp. 3 is
less advanced than, e. g., the second premolar (p. 2), is of no
avail in diminishing the importance of the fact of its early
appearance.
So also in Perameles we have found that dp. 8 in its
1 In our Stage 11 the common canine Anlage is certainly very large, but it
is yet in an indifferent phase. In the lower jaw it may be seen that an
enamel-organ is in process of differentiation from its labial surface, but it is
not that of <, but of d:.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 543
subsequent evolution and maturation lags behind the neigh-
bouring and more juvenile teeth, and is finally outstripped by
them in its developmental progress. But Leche has himself,
as we have seen, shown the worthlessness of the criterion of
period of maturity, and we cannot follow him in his virtual
application of that criterion in the passage above quoted.
The subsequent behaviour of a developing tooth does not
cancel, or otherwise alter, the significance of its first appear-
ance, and we have shown that the latter cannot be disposed of
in the way Leche suggests. We are far from affirming the all-
sufficiency in every case of Leche’s test of synchronicity as
applied to the earliest differentiation of enamel-organs. No
doubt other considerations, such as those above referred to,
must in many cases be taken account of. Here, however, we
believe that we have a case in which the value of the criterion
of contemporaneousness is clearly illustrated; and it teaches,
in our view unmistakably, the doctrine that dp. 3 is not
serially homologous with the other antemolar members of the
adult dentition.
This decision is of course the key of the whole position in-
volved in the question under consideration, viz.:—‘‘ Do the
existing teeth of Marsupials in front of the last premolar cor-
respond to those of the milk or of the successional dentition of
higher mammals ?”
It can hardly be denied by anyone that the deciduous pre-
molar of Marsupials is homologous to the milk-teeth of other
Mammalia.’ A recognition that dp.3 is not homologous with
the other antemolar teeth, but belongs to an “ earlier” cate-
gory, practically implies the serial homology of these teeth
with the successional p.3, and therefore their true homology
with the successional teeth of higher mammals.
The view that p.3 is in series with the teeth in front of it
may at first appear, when viewed in isolation, an improbable
one. But we think it may be taken as established by the
facts before us, and at the very least estimate it agrees better
with the facts in Perameles than the alternative theory.
1 But see above, p. 441.
54.4, Je TL. WILSON AND J. P. HILL.
Kiikenthal, who laid special emphasis upon the precocious
differentiation of dp.8 in Didelphys, held that the early
liberation of the residual dental lamina entailed by the con-
striction off of the enamel-organ of dp.3 could only be con-
ceived as the differentiation of the first Anlage of the succes-
sional tooth (7). It will be plain from the views we have
enunciated throughout this paper that we cannot assent to
this interpretation. The residual lamina beside dp. 31s not to
be conceived as actually at its first appearance a rudimentary
enamel-organ. It is at first indifferent. But it is not long
before a definite and progressive localised thickening and en-
largement sets in (cf. Stage 111) similar to that which constitutes
the earliest rudiments, say, of the other premolars. With the
latter, indeed, the rudiment of p. 3 is in direct serial continuity.
We may therefore very well consider the rudiment of p.3 as
appearing tolerably early in development, though its further
structural evolution is greatly delayed by the development
beside it of its predecessor dp. 3.
II. Interpretation of the Tooth-change in
Marsupials.
In reference to the second question formulated, it will be
borne in mind that all recent authors are agreed that in those
Marsupials whose dentition has been satisfactorily investigated,
one tooth only is replaced by a successor, and that this tooth
is invariably the last premolar.
The statement of Owen (15) that in Macropus, and also in
wombat, the milk incisors are shed during the mammary
foetal stage, has never been confirmed by any subsequent
observer, and has been universally discredited, the more readily
that the statement is made without any definite indication of
the details of such a process having been actually observed.
It is, nevertheless, hard to believe, as Woodward (14) has
pointed out, that such an accurate observer should hazard a
statement of this kind absolutely without evidence. And in
view of the discoveries of Woodward and Rise regarding the
presence of vestigial teeth in the forms named, it may yet be
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES, 545
found that phenomena in the way of shedding or absorption of
rudimentary teeth actually do occur, such as might naturally
be interpreted by Owen as cases of ordinary, if precocious,
tooth-succession.
As yet, however, we can only deal with the one undoubted
manifestation of tooth-change admittedly occurring in Marsu-
pials. Is such tooth-change to be regarded as the survival of
an originally more complete succession? With the single ex-
ception of Professor Leche all the recent authorities would
seem to adopt the affirmative answer to this question, thus
taking their stand upon the primitive diphyodontism of Marsu-
pials, and inferentially of mammals generally. We have
already seen how Mr. Oldfield Thomas (8) definitely, if rather
reluctantly, gave in his adherence to this theory as a conse-
quence of Kukenthal’s researches. Any hesitation he exhibits
is confessedly due to the difficulty he finds in explaining upon
this hypothesis the striking fact that Triconodon already in
Mesozoic times exhibited just the same fulness (or meagre-
ness) of tooth-change which to-day characterises the marsupial
group. But after all it does not appear that the peculiar fact
mentioned (even if Triconodon is admitted as really of
marsupial organisation, as seems probable) is easier of explana-
tion on any alternative hypothesis. Surely itis just as difficult
to account for the fact that a new acquisition of milk-teeth
(as supposed by the older Flower-Thomas hypothesis) should
have stopped short with Triconodon in the marsupial group,
as to believe that the process of degeneration (supposed by the
newer hypothesis) should have been suspended as far back as
the Mesozoic age. For any secondary condition capable of
accounting for the former suspension, such as the peculiar
nutritional conditions of Marsupials, may without much diffi-
culty be applied so as to account for the latter.
Leche has more recently come forward as an advocate of the
alternative hypothesis that the tooth-change in the case of
the last premolar is not the remnant of a more complete change,
but the first appearance of a new dentition foreign to the
primitive marsupial organisation. His theory, however, is
546 J. T. WILSON AND J. P. HILL.
markedly distinct in other respects from the original view of
Thomas and Flower, for it is framed in full view of the pre-
sence of those ingrowths of the dental lamina which, in
common with Kikenthal, Rose, and others, he regards as rudi-
mentary enamel-germs. These, however, he interprets not as
the retrogressing vestiges of an older, but as the products of a
progressive evolution heralding the advent of an entire new
dentition, which among the Eutheria has arrived at a con-
dition of greater or less completeness. We have seen, too, that
Leche’s theory is widely different from that of Flower and
Thomas in that he holds fast the homology of the marsupial
teeth to the milk dentition of Eutheria; in his opinion it is
the “second” dentition, not the “ first,” which is in process
of acquisition.
In supporting his theory Leche refers to the difficulty which,
as we have already seen, was experienced by Thomas in re-
conciling the condition observed in Triconodon with Kiken-
thal’s theory of the vestigial character of the second dentition
in modern Marsupials, and he claims that the difficulty vanishes
when his theory is adopted, since he is able to point to the
peculiar adaptation of the marsupial mouth to the sucking
function as a cause adequate to account for the prolonged failure
to develop a second dentition in the anterior region of the jaw.
Leche would thus figure as a supporter of the original
monophyodontism of Marsupials, were it not for the considera-
tions introduced by his highly significant discovery of calcified
vestigial teeth. These he interprets as belonging to a “ pre-
lacteal ” dentition, since they are evidently antecedent in de-
velopment to the existing adult teeth in these animals.
Leche’s Denial of the Vestigial Character of the
Supposed Second Dentition of Marsupials.
If the theory of the persisting teeth of Marsupials, which is
held by Leche in common with every recent investigator, be
conceded, it is not easy to follow Leche in his opposition to
the Kiikenthal-Roése doctrine of the vestigial character of
the supposed successional enamel-germs. It is true he simply
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 547
refuses his assent until further proof. But the reasons he
urges in support of the alternative theory do not appear to us
to be of a very convincing character. Those sceptical argu-
ments which alone appear to us to be specially cogent as applied
to the theory of Kiikenthal and Rése, and which we owe largely
to the logic of Leche himself, appear to us to be almost equally
destructive of the presuppositions of his own position.
Leche first calls attention to the long persistence of the
so-called “ enamel-germs of the second dentition.” We have
already adversely criticised the idea, common both to Leche and
his opponents, that genuine and actual successional enamel-
germs are present. But even. supposing that such are really
in evidence, we cannot see that their ‘‘ long persistence” is
any argument against their vestigial character. On the con-
trary, if they are really vestiges of teeth which have been lost,
it is precisely these remnants of structures which formerly
persisted throughout a considerable part of the life of the animal
which will retain the strongest tendency towards durability.
Again, the occasional occurrence in other mammals (e. g.
Erinaceus and Phoca) as discovered by Leche of teeth of a
third dentition, developed from outgrowths of a residual dental
lamina by the lingual sides of teeth of the second dentition,
does no doubt establish the capacity of the dental lamina—
after partial exhaustion by the production of a first and a
second dentition—to once more liberate itself as a free residual
lamina, from which in rare cases there may actually be gene-
rated members of a third and entirely new dentition.
But just in so far as this argument is relied upon for Leche’s
purpose his case is weakened for the establishment of the
homology of the supposed potential enamel-germs of Mar-
supials to the Eutherian second dentition, a case which is, at
its best, in our opinion an exceedingly weak one. In fact, we
may say that if the production in some mammals of teeth of a
third dentition, from “enamel-germs” similar to those of
Marsupials, be admitted, this may fairly be claimed as at least
establishing, on Leche’s own premises, the possibility ofa
true morphologicalcorrespondence between the similar structures
VOL. 39, PART 4,.—NEW SER. PP
548 ds T.. “WILSON AND 0, P. HELE.
in the forms compared, and thus as tending to overthrow that
homology of the adult marsupial teeth to the Eutherian
milk dentition which Leche accepts as definitely fixed.
To this deduction from Leche’s reasoning we have already
given in our complete adherence, and we are thus after all at
one with Leche in denying that the so-called “ enamel-germs ”
are vestigial remains of fully developed successional teeth,
since we deny that they are “ enamel-germs” at all.
In further opposition to the interpretation of the residual
laminar downgrowths as vestigial in character, we may cite the
fact of the non-occurrence, in any single case, of occasional
more advanced stages of tooth development, such as might
legitimately be expected to crop up as the result of atavistic
tendencies. That such an expectation is not without warrant
is shown by the fact that in cases where undoubted vestigial
teeth are found, as in the case of the so-called “ prelacteal ”
teeth of Leche (our “milk” teeth), these frequently, if not
generally, exhibit a prematurely advanced and “ abbreviated ”
developmental condition.
Denying then, as we do—with Leche—the vestigial cha-
racter of the supposed rudimentary enamel-germs in Marsupials,
and—against Leche—their character as successional enamel-
germs, we have to make the attempt to homologise the mar-
supial dentition with the typical mammalian upon totally
different lines. Such an attempt can, we believe, be quite
satisfactorily carried out, with the help of material in the way
of facts and observations, of which some of the earliest and
most significant were contributed by Leche himself. These
facts and observations concern the existence of representatives
of an undoubtedly vestigial tooth-series, whose imperfect de-
velopment precedes the evolution of the persisting adult teeth.
Concerning the So-called “Prelacteal” Teeth.
Vestigial representatives of this earliest “ prelacteal”’ tooth
generation have been described by Leche in Myrmecobius,
and, as indicated in our introductory sketch, he has interpreted
the discoveries by Woodward and Rése of similar vestigial
teeth as demanding an identical explanation.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 549
If now it be admitted, on the strength of our previous con-
tentions, that the doctrine of the homology of the existing
antemolar teeth of Marsupials to the Kutherian milk denti-
tion rests upon very slender grounds, surely the discovery in
Marsupials of vestiges of a dentition preceding the adult one
may well give us pause before adding, with Leche, a theore-
tical third to the two dentitions already recognised as the
typical mammalian heritage. We do not think that due con-
sideration has yet been given to the question whether these
discoveries alone do not enable us to decide upon the precise
value to be attached to the supposed rudimentary enamel-
germs, in whose determination as such Leche has at least
taught us to exercise a very cautious and even sceptical criti-
cism, and which he himself after all accepts rather as poten-
tial than as actual “ enamel-germs.”
If these problematical germs can be dispensed with, the
view of their successional homology proving to be ill-founded,
the theory of the marsupial dentition may be reduced to very
simple terms. The whole of the adult marsupial antemolar
teeth (except dp. 3) could then simply be recognised as homo-
logous to the permanent or successional teeth of other mam-
mals; whilst the homologues of the Eutherian milk-teeth
would be found in the deciduous premolar, together with any
vestigial so-called ‘ prelacteal”’ teeth which may be recog-
nisable.
This now proposed explanation of the marsupial dentition
has not been wholly ignored by Leche. Thus, in reference to
the supposed “ prelacteals”? of Myrmecobius, he says, ‘‘ Ent-
weder stellen die fraglichen rudimentéren Zihne die erste
(Milch) Dentition dar, welche bis auf diese Reste ver-
schwunden ist, wahrend die zweite Dentition, welche die erste
wahrend der Phylogenese ganzlich ihrer Funktion enthoben
und verdrangt hat, durch die persistirenden Zahne reprasen-
tirt wird. Oder, die persistirenden Zahne entsprechen bei
Myrmecobius wie bei den andern Beutelthieren der ersten
Dentition, so dass die erwaihnten rudimentaren Zahne nichts
anderes als Reste einer Dentition, welche der ersten Dentition
550 J. T. WILSON AND J. P. HILL.
vorangegangen ist, darstellen konnen” (3, p. 91). The argu-
ments used by Leche to dispose of the first view seem to us
exceeding weak. ‘“‘Gegen die erste alternative spricht nun
zunachst der Umstand dass dieselbe ohne jegliche Analogie
bei den iibrigen Beutelthieren ist, denn bei diesen entspricht
ja, wie die neusten Untersuchungen tibereinstimmend darthun,
das persistirende Gebiss der ersten Dentition der placentalen
Sdugethiere. Und da gerade Myrmecobius in Bezug auf die
Anzahl der Backenzihne die primitivste Form unter den
lebenden Beutelthieren ist, wurde, falls wir diese Alternative
acceptiren wollten, das Myrmecobius-Gebiss durch das Vor-
kommen einer ganzen Reihe von Zahnen der zweiten Denti-
tion zugleich hoher als die tibrigen Beutelthiere entwickelt
sein-eine Annahme, welche durch ihren Mangel an wahrschein-
lichkeit von selbst fallt ” (8, p. 91).
It is plain that the “general acceptance in Marsupials of
the homology of the persisting teeth to the milk-teeth of Pla-
centals’’ simply goes for nothing in the present connection,
seeing that it is just the validity of this “general acceptance ”’
that is challenged, partly on the grounds of the existence of
the “prelacteals.” To determine the latter as morphologically
‘‘ prelacteal”? on the grounds of the accepted “ milk ” homo-
logy of the persisting marsupial teeth appears to us like
reasoning in the very narrowest of circles.
And with regard to the second argument from the condition
in Myrmecobius we are at a loss to perceive its bearing upon
the question at issue. Granting for the present, at least, that
Myrmecobius is the most primitive marsupial form, are we
in the least bound to assume that even the most primitive
mammal should exhibit partial monophyodontism, or at least
incompleteness of its successional series? ‘To make such an
assumption is again to beg one of the most important questions
at issue. Why, we ask, should it not simply be held that the
most primitive mammals possessed both a milk and a succes-
sional tooth series, but that the former had undergone almost
total suppression amongst the ancestors of the modern Marsu-
pials (? Triconodon). We have already pointed out that the
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 951
retention of the single tooth-change from Mesozoic times to
the present offers neither more nor less difficulty than any
other mode of explaining the Triconodon condition (cf. also
Leche [28], p. 278).
The only argument used by Leche in this relation which
appears to us to require serious consideration is his statement
respecting traces of true “ prelacteal’’ tooth-rudiments labially
from the milk enamel-organs of higher mammals.) But when
we examine the descriptions and figures upon which Leche
bases this startling statement, we find that the indications
referred to are of the very feeblest and most inconclusive kind
(see his figs. 3 and 5’). We have ourselves met with struc-
tural indications in the pig (cf. figs. 83—85 ; text-figs. 2 and
3) which at first tend to suggest, even more strongly than
Leche’s, an interpretation such as he has given. But they are
really to be explained in a quite different manner, as we hope
to show incidentally later on when dealing with the question
of molar homology (cf. p. 567, e¢ seq.).
It is unnecessary here to recapitulate our own observations
upon the so-called “ prelacteal”’ vestigial teeth in Perameles,
which, in accordance with our theory, we have constantly
designated as “milk ” or deciduous teeth. It is sufficient to
say that we have recognised what we believe to be vestiges of
these teeth in connection with at least four, and perhaps all,
of the upper incisors, at least two lower incisors, and both
upper and lower canines. Of these, d22, di+, and d< have
been traced in the cupped and papillated stage, while di;
becomes strongly and d; weakly calcified.
In our examination of these interesting rudiments we very
early became convinced of their serial homology with the
tooth dp. 3. As we have stated in the introductory part of
this paper, this conviction at first compelled us to take up the
' The argument from the topographical relations between those portions of
the dental lamina connected with the rudimentary and persisting teeth,
respectively, may be passed over as unimportant. The various stages we
figure of vestigial teeth in Perameles quite sufficiently explain any pecu-
liarity in the late stages of the vestigial teeth, alone studied by Leche.
Bae J. T. WILSON AND J. P. HILL.
novel and revolutionary view that the deciduous tooth of
Perameles, and probably of other Marsupials, was a member
of a supposed degenerate “prelacteal” series. But ere long
we were led, first to call in question, and then confidently to
reject the “ prelacteal” theory of this earliest tooth-series, as
a violation of that principle of parsimony which should govern
our procedure in the construction of explanatory hypotheses.
For if dp. 3 and the supposed “ prelacteals”” be regarded as
representatives of an otherwise suppressed milk series of mar-
supial teeth, we shall have before us a workable hypothesis
which we believe to be free from the difficulties and complica-
tions which beset that more generally current, and which
offers no insuperable difficulties of its own.
In preceding pages we have dwelt upon the application of
Leche’s criterion of the contemporaneousness of origin
of the tooth-Anlagen from the dental lamina. We
have shown that, judged by this test, dp. 3 belongs to a
different category than that of the other persisting antemolar
teeth. The identification of dp. 3 and the “ prelacteals” as
members of one and the same tooth-series may be designated
as a positive result of the application of the same cri-
terion. The investigation of our Stages 11 and 111 has yielded
the most striking evidence of the validity of such a conclu-
sion.
Thus in Stage 11 the enamel-organs of the future permanent
antemolar teeth are in abeyance, or are represented only by
localised undifferentiated thickenings of the dental lamina,
But already dp? and di+ show well-marked cupping (figs. 8, 9,
14, and 12), while d¢3 shows a slighter degree of it (figs. 2,
8, and 4), and the enamel-organ of d£ is, at least, in process of
differentiation from the main mass of the canine Anlage
(fig. 13).
But it is in connection with Stage 111 that the most strik-
ing pictures are obtained. The horizontal series from this
stage was of especial use to us during our period of transition.
Specially would we refer to our fig. 22, as affording a most
instructive comparison between the canine ‘ prelacteal ”
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 553
enamel-organ, so-called (our de.), and the deciduous third pre-
molar, dp. 3.
The position of the latter with regard to the neighbouring
dental lamina is seen to correspond—not to that of the other
adult teeth—but most accurately with that of the canine
** prelacteal ” enamel-organ, which is well seen in the section ;
dp* and d? are both seen to project labially so as to lie ina
sagittal plane distinct from that occupied by the developing
enamel-germs of the permanent teeth. This relation is also
very evident, if slightly less striking, in coronal sections
through the same Anlagen ; we have only to imagine the bulk
proportions of the respective parts in figs. 20 and 24 to be
somewhat altered in order to realise the most complete corre-
spondence between the two.
It is of the greatest interest to compare Rése’s fig. 1,
representing the dental condition of a foetal Phascolomys
(11, p. 752) with that just described. There, in the wombat,
it appears that dc. is comparatively large and calcified, but is
evidently of precisely the same order as the smaller calcified
cheek-tooth marked D, whose homology with dp. 3 cannot, in
our opinion, be doubted (cf. also his fig. 3). Here, therefore,
the bulk proportions of de. and dp. 38 are altered from what
we find in Perameles, but otherwise the condition presents a
remarkable similarity.
The very marked labial situation occupied by dp3 in all the
entire series of our stages is most striking, and must not: be
confounded with such a lateral displacement of. enamel-organs
as may ensue as a result of over-crowding of the jaw. This
latter occurs more especially in the incisor region, and its
effects are well illustrated in the figures of Rése’s models of
Didelphys already referred to. But the distinction in the
orientation of the enamel-organ of dp. 3 is established from
its very first appearance. It is most striking when the germs
of the anterior premolars begin to appear, and long before any
displacement through encroachment on their part could be
imagined. ‘Thus both in serial position and in period of
differentiation the enamel-organ of dp. 3 relates itself to the
554 J. T. WILSON AND J. P. HILL.
so-called ‘ prelacteals,”
teeth.
Along with these considerations may be borne in mind the
tendency, so often illustrated amongst Marsupials, for this
same tooth to share the fate of its fellows and to become
vestigial. We have referred in the introduction to its con-
dition in Dasyurus and Phascologale, and here we need
do no more than refer to our figures of the tooth in the first-
named of these forms (figs. 80 and 81).
On the grounds, therefore, both here and elsewhere detailed,
we can have no hesitation in affirming that dp. 3 ought to be
regarded as a member of the same series which includes the
vestigial canines and incisors, and that this series corresponds
to nothing else than the normal mammalian milk-series, which
in Marsupials has been profoundly modified in the way of
suppression.
and not to the permanent antemolar
Primitive Diphyodontism and “Suppression.”
Along such lines of reasoning we thus reach the standpoint
of the primitive diphyodontism of mammals—a_ position
already occupied, though, as we believe, on erroneous grounds,
by Kikenthal, Rose, and others. We believe that we are
justified in seeking for the cause of the almost total suppres-
sion of the milk-teeth in front of the last premolar, in the
modified condition of the mouth in marsupial young in con-
sequence of its peculiar adaptation to the sucking function.
This would seem to be a more natural employment of this
factor than that of Leche, who has already suggested it. For
he seeks by it to explain the non-appearance of an entirely
new series of replacing teeth. Surely such an organic
modification would more immediately tell upon an existing
milk dentition in the way of suppression, rather than—
merely through delay of such dentition—affect the subsequent
development of successors.
In fact, Leche himself, in opposing the view that the second
dentition has actually once been present and has since dis-
appeared, contends in so many words that, according to the
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 905
principle of ‘‘ abbreviation,” the suppression should affect the
first rather than the second dentition. “‘ Ferner falls wirklich
jemals eine vollstandige zweite Dentition ausgebildet gewesen
ware, ist es kaum zu erklaren, wesshalb z.B. bei Didelphys,
wo durchaus keine Ursache zu einer Reduction oder retro-
graden Entwicklung des Zahnsystems als Ganzen vorliegt und
actisch auch keine Reduction eintritt, die zweite und nicht
vielmehr die erste, im allgemeinen schwachere Dentition un-
terdriickt wurde—etwas, das ja auch nach dem Princip der
Abkiirzung der Entwicklung zu erwarten gewesen ware”
(3, p. 105. See also 28, p. 372 [2)]).
We note also in this connection the admission of the same
author, ‘‘ Wo aber sonst innerhalb der Klasse der Siugethiere
Monophyodontismus auftritt, spricht der zur Zeit vorliegende
Thatsachenbestand zu Gunsten der Annahme, dass die erste
Dentition verschwunden ist und die zweite persistirt ” (3, p.143).
Leche finds in Erinaceus an illustration of this view. He
has come to the conclusion—we believe justly—that in this
form those antemolars which are subject to no tooth-change
are members of the second dentition which have hastened their
time of development, and that Krinaceus represents “a transi-
tion form from the diphyodont to the monophyodont stage.”
Thus suppression in Krinaceus is interpreted in a totally
different way from that occurring in Marsupials, so that three
different mammalian conditions may be recognised: (a) the
lowest or marsupial being characterised by an almost complete
monophyodontism, from failure to develop successional teeth ;
(6) the typical mammalian condition, where complete duplicate
sets of antemolar teeth—milk and successional—are developed ;
and (c) a still more advanced condition, attained in Erina-
ceus, Bradypus, and Pinnipedia, which again approaches
monophyodontism, but this time through defect of the first
(milk) or historically older dentition. An elaborate hypothesis
like this, to explain what are essentially very similar con-
ditions as regards suppression in Hrinaceus, &c., and Mar-
supials, seems to us to bear its condemnation upon its face.
We think the proofs are overwhelming that in Marsupials, as
556 J. T. WILSON AND J. P. HILM.
in the other forms referred to, but to an even greater extent,
the suppression has affected the milk dentition and not the
successional, and that we may extend to Marsupials generally
the conclusion of Rose, that “‘ Phascolomys Wombat ganz
ahnlich wie die placentalen Sauger zwei typische gesonderte
Dentitionen besitzt 7 (11, p. 745).
In a paper upon the milk dentition of the Rodentia (18)
Woodward has discussed the occurrence of vestigial teeth
in certain members of that order, and has shown that, in the
forms dealt with, these vestigial teeth are representatives of
the milk dentition. It may be noted that this conclusion is
reached in the case of the mouse in spite of the fact that
‘there is present at one stage in the development of the large
lower incisors a slight downward prolongation of the dental
lamina on the posterior side of the enamel-organ of that tooth.”
This is, of course, simply a repetition of the ordinary mar-
supial condition. Upon it he comments :—‘‘ This might be
regarded as presenting an indication of a permanent tooth, the
large incisor being referred to the milk dentition, while the
vestigial incisor in front might represent the pre-milk dentition.
On the other hand, if this very slight prolongation of the
dental lamina has any morphological value at all, I should be
inclined rather to consider that it represented that supposed
fourth dentition, as described by Kiikenthal and Leche for the
Seals” (p. 626). Here, then, we find Woodward applying
to the vestigial teeth of the mouse precisely that mode of
interpretation which is claimed by us to be alone capable of
affording a satisfactory solution of the marsupial problem.
In Kikenthal’s latest utterances (in reply to Leche) upon
the problems of the mammalian dentition (27) we have a con-
venient summary of his views, which have undergone little if
any change from those of his earlier papers.
He adheres firmly to the view that both the ordinary denti-
tions of mammals were inherited from sub-mammalian ances-
tors, but apparently admits as established the assertion that
“traces of two other extinct dentitions are occasionally still
present embryonically.”
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES, 557
But it is extremely difficult to follow Kikenthal in his ex-
planation of the manner in which suppression has_ been
brought about in different groups. He first formulates a
“law of reduction,” stipulating that the tendency to sup-
pression has constantly affected the first dentition most
strongly. Thus he explains the suppression of the supposed
“ prelacteal” series. So far all is intelligible. But when he
comes to explain suppression as witnessed within the mamma-
lian class itself the case is much ess clear. ‘ Bei den nie-
deren Sdugetieren dominirt die erste Dentition, die zweite
kann wohl ausgebildet sein (Edentaten),! oder aber sie fallt
aus besonderen Griinden-—secundaren Anpassungena—dem
Rudimentarwerden anheim (Beutler, Cetaceen).? Bei den
hdheren Saugetieren verliert die erste Dentition an Bedeutung,
und die zweite kommt mehr zur Geltung. Wenn jetzt gewisse
Lebensbedingungen einen Zahnwechsel ungiinstig sind, wird
nicht die zweite Dentition unterdritickt werden, sondern die
erste eine raschere Entwickelung und ein fritheres Ende
nehmen.” (The last sentence expresses the corollary from the
previously enunciated “law of reduction,” and formulates
precisely what we believe to have happened in the case of
Marsupials.)
In view of this startling exception of Marsupials from the
ordinary operation of the law of reduction, it is rather
strange to find Kukenthal going on at once to say that “ein
einheitliches Gesetz beherrscht also die Dentitionen aller
Wirbeltiere ” (p. 657). But why should Kikenthal ask us to
make a special exception of the Marsupials from the operation
of his own “law”? Why should he expect us to assent to
the idea that while ‘‘ conditions of life unfavorable to tooth-
change,” amongst mammals generally, should tend towards
suppression of the milk dentition, yet nevertheless the
? Upon this statement with regard to the Edentata, as well as upon the
entire passage, Leche’s rejoinder (28, p. 275) may be consulted.
2 We are constrained to avoid special discussion of the Cetacean condition
because we are disposed to hold with Leche that the homology of the persis-
tent teeth of Cetacea must still be left as an open question (28, pp. 274-6).
558 J; Tl WILSON AND J}; BP: HIGE.
‘secondary adaptations ” which have operated amongst the
marsupial group have resulted in the well-nigh complete ex-
tinction of the successional dentition ?
We have little doubt that the answer to these queries ulti-
mately resolves itself into this,—that Kiikenthal has not been
able to bring himself to question the validity of the prevalent
judgment which has stamped mere secondary or residual laminar
downgrowths—sometimes thickened marginally—as veritable
enamel-germs of successional teeth.
We are quite unable to sympathise with Kiikenthal’s final
protest against Leche’s views as calculated “to disturb anew
the now clear conception” of the dentitional problem. In
our opinion Leche’s systematic criticism has been most timely,
and if he has not gone the whole length in carrying out his
criticisms to what we believe to be their logical conclusions,
yet to him belongs the credit of having first introduced syste-
matic critical criteria into the discussion of the developmental
aspects of the problems of dentition.
III. Serial Homology of the Molars.
The question concerning the serial homology of the molars
is not altogether separable from the fourth question, which
deals with the origin of multicuspidate teeth in general.
For a connected historical sketch of the state of opinion on
these points we may simply refer the reader to Leche’s mono-
graph (3, pp. 145-8). A succinct account may also be found
in Woodward’s paper (14), and a more extended discussion of
the subject in Schwalbe’s address (16). The principal views
held may be shortly summarised as follows: :
(a) The molars belong to the first or milk dentition (Owen,
Beauregard, Kiikenthal [earlier view], Rose [earlier view],
Osborn, Hoffmann, Leche).
(2) The molars belong to the second or “replacing” denti-
tion (Lataste, Magitot, Woodward).
(c) The molars are not homologous to any single series, but
to two or more dentitional series, either by actual fusion
of the tooth-germs of the latter, or of the material from
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 509
which such tooth-germs would have been formed (Schwalbe,
Kikenthal).
(d) The molars are the lateral end-members of separate
dentitions: thus, in man, the dentition of which m. 1 is the
end-member lies between the first and second dentitions (Rése
[later view]).
The arguments mainly relied on by Leche for the establish-
ment of the “ milk” affinities of the molars largely hinge
upon that interpretation of the development of the antemolar
teeth with which we are now sufficiently familiar.
The presence of a residual dental lamina—when its distal
portion is somewhat swollen, and especially when the sur-
rounding connective tissue is somewhat condensed—has been
liberally drawn upon by him to furnish proofs of the milk
homology of the antemolar teeth. And, having been accepted
as practically sufficient for that purpose, it is now utilised in
establishing a like homology for the molars.
There can be no doubt that the molar segment of the dental
lamina in Marsupials does very constantly develop a well-
marked and often thickened residual laminar downgrowth
Opposite each tooth as a consequence of its differentiation from
the parent lamina. Here, as elsewhere, however, the residual
lamina is in each case directly and serially continuous, fore and
aft, with the intermediate undifferentiated parent lamina; in
other words, the residual laminar downgrowth does not ex-
hibit discrete swollen portions, but is an elongated continuous
band, more or less uniformly thickened towards its free distal
margin.
Value of the Molar Residual Lamina.
No secure basis for the establishment of the dentitional value
of the molars is afforded by the presence of this residual
lamina. It might be hastily concluded that, with our definite
opinion of the “successional” character of the persisting
antemolar teeth, we are bound to attach the same value to the
molars. But we are by no means compelled todo so. If
the molars are to be regarded as homologous with any one
560 J. T. WILSON AND J. P. HILL.
dentition, as is likely enough, they may very well be equivalent
to the milk series, as represented, e.g., by dp.3. For the pro-
duction of a residual dental lamina, during the differentiation
of the enamel-organ of a tooth from the parent lamina, for us
neither involves nor excludes the belief that successors
may or do arise from that residual lamina. In other words,
the production of a residual dental lamina by the lingual side
of a milk tooth-germ in itself does no more than guarantee the
possibility of the origin from it of a true successional tooth,
just as its presence by the side of the Anlage of a successional
tooth involves the potentiality of the origin from it of a tooth of
a post-successional dentition. .
We may, perhaps, advantageously refer once again to the
condition of the first molar in the early stages (11 and 111) of
Perameles. It will be remembered that in Stage 11 that
enamel-organ is already recognisable as such, and its stage of
development, though slightly less advanced than dp. 3, is fairly
on a par with that of the anterior milk-teeth (‘ prelacteals”’).
Its papilla is just indicated.
In Stage 111 the resemblance between dp. 3 and m.1, both
in their stage and manner of development, is worthy of note.
At the very least it may be urged that there is no prima facie
case for assigning m. 1 to a different series from dp. 3.
Woodwaid at first (2, p. 460) seems to have based his
belief in the successional character of the molars upon his
inability to discover in Didelphys, in the Macropodide,
and some other mammals “rudimentary enamel-germs” or
“HKrsatzleisten” at their lingual sides. And he attempted
to explain away what he regarded as a “suggestion ” of such
structures in Lepus and Talpa by a critical distinction
which, thoroughly carried out in reference to marsupial develop-
mental features, leads directly up to the main thesis of this
work. Woodward in a subsequent paper (18) has fully re-
cognised the existence of such lingual growths of the molar
dental lamina in certain Eutherian forms, and in that connec-
tion reconsiders the question of molar homology. He still
inclines strongly to the theory of their successional character,
DEVELOPMENT AND SUCCESSION OF TEBTH IN PERAMELES. 561
though, of course, no longer upon the grounds of the ab-
sence of what he regards as rudiments of possible successors.
But the latter he would prefer to relegate to the category of
Leche’s fourth (post-sucvessional) dentition. His reasons for
such a decision are largely of a speculative character.
But there is one feature of his case which is by no means
merely theoretical, but is based directly upon observation of
structure. We refer to the presence of alleged rudiments of
a molar dentition older than the existing molar series. We
shall recur to this subject presently.
We have examined several series of sections of early Ma-
cropus embryos, and, like Woodward, we there find that the
large and deeply cupped molar enamel-organs are as yet indis-
tinguishable from the dental lamina, the latter appearing for
the time being to have been wholly converted into the cup-like
enamel-organs. But this is precisely what we find to be the
case with the milk enamel-organs of the incisor and canine
teeth in embryo pigs of 20 mm. head-length, where of course
successional teeth arise later on. It is entirely a question
of the stage of advancement of the developmental processes.
And however it may be in Macropodidx, there is not the
slightest doubt that lingually placed “ residual” prolongations
of the dental lamina (and with swollen free distal margin) do
occur constantly in Pe rameles during the development of every
one of the molars, as shown in our descriptions and figures.
Thus fig, 27 shows the tolerably early appearance (Stage 111)
of the residual lamina by the side of m4. Here there can be
no possible confusion with the Anlage of m2 which is seen in
the horizontal sections (figs. 25 and 26) to be just developing
further back. That the evidence derived from these horizontal
sections is not to be gainsaid becomes apparent in the exami-
nation of Stage 1v. There m2 is a well-developed cupped
enamel-organ, and has itself developed a definite residual
lamina with a swollen fundus (fig. 52), while beside m+
the residual dental lamina is seen persisting as a typical
“ Ersatzleiste,” and appears on cross-section as a swollen
and “ bud-like” (“‘knospenférmig ”) downgrowth (fig. 51).
562 J. T.- WILSON AND J.-P. HILL.
And with the development of each successive molar a similar
condition is brought about.
It would certainly appear that the evolution of the macropod
molar proceeds less rapidly than it does in a polyprotodont
form like Perameles.! In any case it becomes quite certain,
from examination of our consecutive stages, that a residual
dental lamina with thickened margin comes to project freely
by the lingual side of the molars, just as in the case of the
antemolar teeth.
Concerning the Significance of Labial Outgrowths
of the Lamina in the Molar Region.
Holding such views as we do of the comparatively trifling
significance to be attached to the occurrence of a residual
dental lamina (even when “knob-like” or ‘ bud-like ”
[“kolbig” or “knospenférmig™”] on cross-section), it be-
comes the more important for us to come to a decision with
regard to the real character of those labially directed out-
growths of the dental lamina in the molar region noticed by
Woodward. These, as has been mentioned above, have been
interpreted by him as possible or even probable vestigial
remains of predecessors of the existing molar teeth.
In such vestiges he is inclined to believe (2, pp. 460 and
470, and cf. 18, p. 630) we have the true serial homologues of
the antemolar milk-teeth.
It is plain that, from our point of view, a discovery of
vestigial remains of teeth preceding the functional molars
would tend to place the latter in the same category with, say,
the permanent canine of Perameles, where, as has been
shown, a pretty large residual dental lamina is differentiated
alongside the Anlage of the permanent canine at a time when
a mere vestige of the milk canine (dc.) alone remains (fig. 47).
Woodward describes as “ present on the outer side of the
enamel-organ (in Petrogale) a conspicuous outgrowth of
its cells, extending down into the gum at right angles to the
' Probably the future large size of the macropod molar has something to
do with the relative delay in the process of its differentiation.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES, 563
swollen portion of the dental lamina.” He goes on to remark
that “‘ we can only interpret these appearances on the grounds
that the functional molars belong to the second or replacing
dentition, and that this downgrowth on their outer sides repre-
sents the rudimentary, or rather vestigial, milk or first
dentition.”
We can fully confirm the observations of Woodward in
regard to the existence of such labial projections as he has
thus described (and figured in his figs. 25, @ and 8, and 26).
The like are constantly seen in Perameles in certain regions,
and at first we were inclined to interpret them in the same
manner as Woodward has done.
We have, however, been led to entertain grave doubts as to
whether such a construction can legitimately be placed upon
the facts observed by Woodward and ourselves. In the first
place it may be observed that the whole labial surface of the
dental lamina often presents considerable irregularity as com-
pared with the lingual surface, and the irregularities not
infrequently constitute actual outgrowths or processes. This
condition is specially noticeable in those parts of the lamina in
front of or behind the enamel-organ of a molar tooth (cf. fig.
30, “d”). And even opposite the tooth-germ cellular projec-
tions may be observed upon the labial aspect of the “ neck ”
of the dental lamina, proximal to the place of attachment of
the enamel-organ, as in figs. 28 and 29 (“a”). This kind of
process appears to consist of cellular material, originally form-
ing part of the enamel-organ, which has been left attached to
the main stem or “neck ” of the lamina during the process of
constriction off by which the enamel-organ is liberated from
the lamina; and it is very probably comparable to those
villus-like projections which not infrequently occur upon
the outer surface of the developing enamel-organ itself. Indi-
cations are shown in fig. 29, “c,” and especially in fig. 28, “c,”
of such villous irregularities over the general surface of the
enamel-organ of m-t towards its hinder end. In some sections,
several (as many as three) labial projections from the neck of
the dental lamina of the character indicated in figs. 27, 28, and
VOL. 39, PART 4,—NEW SER, QQ
564 J. T. WILSON AND J. P. HILL.
29 (a”) may be recognised, one above the other. It is plain
that, whatever be the real meaning of these, it is impossible to
regard them as vestigial representatives of teeth. This form
of labial projection must therefore be excluded. Again, it may
be found that a projection on the labial side of the dental
lamina is traceable to the point of severance of the lamina
from the oral epithelium.
Thus in the section shown in fig. 27, an outgrowth of this
character (“ b”) appears to be the direct prolongation of the
proximal margin of the dental lamina, which has just been
separated from the oral epithelium though it is still in contact
with it. If a comparison of this figure be made with that of
dp2. in fig. 24, it will become evident that the condition in the
former is easily derived from that of the latter. We have only
to imagine that the freed margin of the lamina in the region
of m+ has become shifted, or has actually grown, labially from
its late point of attachment (or later point of severance). Such
an explanation may possibly apply not only to the case quoted,
but to that represented in Woodward’s fig. 26, m’ (cf. also his
fig. 20, d. 1.).
On the other hand, neither of the explanations suggested
above appears sufficient to account for the appearance shown
in his figs. 25 a, and 25 6. These represent a form of out-
growth or projection with which also we are familiar in Pe-
rameles. Thus in figs. 30 and 31, passing through the lamina
just behind m+ in Stage 111, and where the last trace of the
enamel-organ of that tooth is disappearing from the sections,
we find that a labial process (/. 0.) of very distinct and definite
character manifests itself, springing from the neck of the
dental lamina. This becomes established, not as a mere local-
ised projection, but as a continuous secondary lamina
extending right back to the place where the entire dental
lamina ends abruptly in the Anlage of m2 (figs. 25-6).
The structure described cannot, we believe, be explained by
either of the modes of interpretation already put forward as
applicable to some other cases. It cannot in any sense be a
consequence of disintegration or severance, and its constancy
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 565
and persistent continuity in the hindmost molar region forbid
our regarding it as a more or less accidental or capriciously
developed villous outgrowth or process. But at the same time
the fact that it does not constitute any mere knob-like excres-
cence, but that it really extends longitudinally back-
wards asa flattened laminar or ridge-like outgrowth,
seemed to us upon closer consideration to be unfavorable to
its claims to be regarded as of rudimentary or vestigial dental
character.
On examining the next succeeding stage (1v), where m2 is
a well-developed enamel-organ, and m3 has attained a degree
of development comparable to that of m2 in Stage 111, we
find that the secondary labial ridge or lamina aforesaid has
now disappeared opposite m+ and up to near the hinder part
of the now developed m2. Its disappearance is associated
with the disappearance in most of the sections of the connec-
tion between the dental lamina and the oral epithelium, i.e.
with the onset of disintegration of the proximal portion of the
dental lamina.
But towards the hinder end of the now well-developed m2
the secondary labial laminar outgrowth again comes into view
(figs. 52 and 53, /.0.), this time extending backwards from
near the hinder end of m2 to the posterior limit of the dental
lamina with its thickened Anlage of m3 (fig. 54, /.0.).
We see, therefore, that the conditions which formerly
obtained further forwards are exactly repeated here at the pro-
gressively growing segment of the dental lamina. Here
again, however, there is nothing to suggest a vestigial dental
character. The continuous laminar character, indeed, is an
indication to the contrary.
We may sum up these and further observations in this
direction by affirming that, towards the hinder end of the
dental lamina (which in the molar region is the growing end,
and is thus ontogenetically younger than the parts of the
lamina in front of it), there is constantly to be found, in addi-
tion to the main stem or axis, as it were, of the dental lamina
a secondary offshoot or lateral leaf-like outgrowth, directed
566 J. T. WILSON AND J. P. HILL.
labially, nearly at right angles with the stem, as shown in figs.
30 and 81 (/. 0.) and in Woodward’s figs. 25 aand 25 6.
The definiteness, constancy, and longitudinal extension of
this form of labial projection markedly distinguish it from
those other kinds of outgrowths on the labial side of the dental
lamina to which we have alluded above.
No satisfactory explanation of the meaning of this structure
seemed forthcoming until, after some incidental study of serial
sections of pig embryos, the conclusion was forced upon us
that this secondary labial laminar projection really represented
the continuation of the labio-dental, or, better, labio-alveolar
epithelial lamina (precursor of the lip-furrow or labio-alveolar
groove) which the researches of Rose (9) in the human teeth
have shown to arise from a common Anlage with the dental
lamina proper.
This identification seems to us so important and so unex-
pected, as applied to structures so far back in the molar region,
and especially in marsupial animals (in which, at least ante-
riorly, the lip-gum furrow is long suppressed), that we find it
necessary to exhibit in some detail the facts upon which our
conclusion is based.
It must be admitted that the explanation offered seems far-
fetched and improbable, nor do figs. 830 and 81 seem to lend
much countenance to it. Nevertheless we hope to show good
reason for the belief that the labial laminar outgrowth is, at
least, an epithelial lamina which is largely independent of the
dental lamina proper.
In the first place we must state that in respect of the mode
of extension backwards of the dental lamina in Marsupials,
our observations are so far at variance with those of Rése
upon human molar development. There can be no doubt that
in the animals examined by us the greater part of the molar
lamina grows posteriorly by extending backwards its con-
tinuity with the oral epithelium, and not merely by extending
back freely into the mesoderm beneath the epithelium as Rése
has described it. The anterior, greater part of it appears to
be as much the product of direct ingrowth from the oral
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 567
epithelium as is the dental lamina in the antemolar region.
That this is the case as far as the entire molar lamina in Stage 11
is concerned we have already pointed out, but we may here
at once state that we have observed the extension backwards
of the dental lamina in continuity with the oral epithelium in
other stages as far as the region of the third molar inclusive
in the upper jaw, and just short of this region in the lower.
In fact, it may be stated that this continuity with the oral
epithelium exists just as far back as the oral epithelium
overlies the molar Anlagen. The most posterior molars
appear to originate precociously from a backward free pro-
longation of the lamina before the lengthening of the jaw
permits a co-extensive prolongation backwards of the oral epi-
thelium.
In pig embryos of 20 mm. head-length, we also find the
connection of the dental lamina with the oral epithelium
maintained right up to the hinder end of the molar region, or
within a couple of sections of it.
We are, therefore, little surprised to find that in the pig the
labio-dental or labio-alveolar lamina (i. e. the epithelial Anlage
of the lip-furrow which is formed anteriorly) is continued back
into the molar region as the Anlage of what we may here call
the gum-cheek-furrow. And at the extreme posterior end of
the proper dental lamina the latter is seen to be fused near its
basal or attached margin with the labio-alveolar lamina, the
two structures forming parts of acommon Anlage, as Rése has
shown to be the case at first in the anterior region of the jaw.
Fig. 83 illustrates the condition met with close to the pos-
terior end of the molar lamina in the upper jaw of such a pig
embryo. The lamina stops rather abruptly a few sections
further back.
The section figured passes close behind the posterior end of
the molar Anlage,—in fact, partly shavesit. Projecting labially
from the neck of the lamina near the oral epithelium there is
seen in cross-section a laminar projection (d.0.) of the kind
whose significance we are discussing, and which is proved, by
examination of the series, to be neither more nor less than a
568 J; & “WILSON AND: J: BP. RI,
continuation backwards of the labio-alveolar lamina or Anlage
of the “ lip-furrow.”
The accompanying figures in the text are outlines of the
identical epithelial structures in the lower jaws of the same
D
Fic. 2, a—z.—Pig embryo of 20 mm. head-length. Four serial coronal
sections through the molar region of the lower jaw. a. Some distance
behind m;. 8B. Two sections in front of a. c. Through posterior part
of mz. D. Through middle region of mz. /.0,= labial laminar out-
growth. /.gr.=lip groove. x 45.
series. (The lower jaw was chosen as affording a better series
of gradations of otherwise exactly similar conditions.)
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 569
Text-fig. 2 (aA) shows the common laminar Anlage attached
to the oral epithelium (which exhibits on its oral surface a
trace of the actual furrow). This section corresponds exactly
Fic. 3, a—c.—Pig embryo of 20 mm. head-length. Three serial coronal
sections, upper canine region. a. Just in front of © 3B. Through
middle region of ¢. oc. Through posterior region of * x 45.
in its relation to the molar lamina with that shown in fig. 83
from the upper jaw, and shows only a very slight variation
therefrom in form. The next three outlines (text-fig. 2,
B, C, D,) are from sections taken at intervals of a few
sections further forwards. Of these, p shows the definitive
separation of the labio-alveolar Anlage from the proper dental
lamina by an increase of the connective-tissue basis of the
alveolus.
570 J. T. WILSON AND J. P. HILL.
The remaining figures (text-fig. 3, a, B, and c, and figs.
84 and 85) from the pig are taken, not from the molar
region, but from much more anterior parts of the jaws, in
order to show the striking appearances in the way of labial
processes which are produced at certain stages of the differ-
entiation of the common Anlage of labio-alveolar groove and
proper dental lamina. Attention may be especially directed
to the transitional forms of the labial projections in the upper
and lower jaws figured in figs. 84 and 85. It must be men-
tioned that every one of the labial projections figured has
actually been traced into continuity with the definite lip-
furrow Anlage as clearly as in the case of the series shown in
the first four outline figures (text-fig. 2) taken from the molar
region. With regard to the latter it may be remarked that,
as might be expected from the fact that the more immature
portion of the molar lamina is the posterior, the primitive
connection of the Anlagen of labio-alveolar and proper dental
lamina is recognisable behind, and in proceeding forwards in
the molar region a transition toa state of mutual independence
of the two structures occurs. The fact that again, anteriorly,
there is areturn to the status of a common Anlage is doubtless
conditioned by other developmental factors.
Comparison of fig. 83 and text-fig. 2 of the molar lamina in
the pig, with figs. 30 and 31 of Stage 111, and 52—54 of
Stage 1v of Perameles, and also with Woodward’s figs. 25a
and 25 6 of Petrogale, will hardly fail to carry conviction of
the essential identity of the structures therein represented.
From a careful investigatiou of these labial appendages in
the molar region we therefore conclude that while such ap-
pendages may not be uniformly homologous, yet those which
alone are definite, and irreducible to the results of unimportant
incidents of development, require a different mode of inter-
pretation from that suggested by Woodward. We must accord-
ingly dismiss the idea that any structural features occur in the
development of the marsupial molars which can fairly be
interpreted as vestiges of degenerate “‘ milk” predecessors.
We are thus thrown back upon the already noted resem-
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 571
blance between the development of dp? and m+ in the early
stages. Such resemblance as has been adverted to cannot, of
course, by itself prove that an actual serial homology exists
between the two teeth. It may indeed establish a prima facie
case for such an homology. But, however we may decide the
special problems of the distinctively marsupial dentition, the
serial homology of the molars must remain as a general question
of mammalian dentition, whose decision must yet be regarded
as a separate and largely independent issue.
The theories of molar homology set forth under (ce) and (d) on
page 558 still remain to be dealt with. Of these the latter
has not as yet been supported save by its author, and may, we
believe, be regarded as a mere speculation. The former, which
is supported by both Kiikenthal and Schwalbe, may best be
discussed in connection with the next question, which primarily
concerns the mode of origin of multicuspidate teeth, but which
raises the general question of fusion of originally distinct
dental units.
lV. TuHeoriEs oF DENTAL FUSION AND THE ORIGIN OF
MULTICUSPIDATE TEETH.
The results of the present research do not, any more than
Woodward’s, lend countenance to the view once advocated by
Kikenthal, that molars are a result of the fusion of germs of
more than one dentition, in the sense in which fusion has been
supposed to occur between tooth-germs of the same series.
But Kikenthal’s present position in regard to this question is
rather less simply expressed. He holds “dass die echten
Molaren im Wesentlichen zur ersten Dentition gehéren, dass
sie aber ein Verschmelzungsproduct der Anlagen erster Denti-
tion mit dem Material, aus dem sonst die zweite Dentition
entsteht, darstellen ” (20, p. 112, and 27, p. 659).
Against this view both Woodward (14, p. 447) and Leche
(3, p. 146) have pointed out that it ignores the fact, upon
which Kukenthal had himself insisted, that a residual dental
lamina does normally develop by the lingual side of the first
972 J. Us WILSON “AND oJ RP. HUGE.
molar. ‘To this criticism Kikenthal has omitted to reply
during the course of his criticism of Leche, embodied in his
latest publication on this subject—‘ Zur Dentitionenfrage ’ (27).
And we find him towards the close of this last paper simply
reiterating the opinion in the course of his remarks in reply to
Hoffmann.
We confess that Kiikenthal’s view is still very far from
being clear to us. When he speaks of the molars as repre-
senting “ein Verschmelzungsproduct der Anlagen erster Den-
tition mit dem Material, aus dem sonst die zweite Dentition
entsteht,” are we to understand that the enamel germs of a
first dentition are first of all differentiated from certain residual
“material ” representing enamel-germs of a second dentition,
and that, having been differentiated from one another, these
are subsequently fused together? If this be what is meant,
then certainly the facts of marsupial development give no
countenance to such anidea. Indeed, the mode of origin, pre-
servation, and gradual disintegration of the molar residual
lamina prove that certain “ material”? remains over which is
not fused or taken up during molar development.
But if Kikenthal’s words are not to be understood in the
sense referred to, and if we are to believe that he does not
conceive of a morphological separation or form-differentia-
tion ever having been effected between the material of the
first dentition and the material representing the second (as,
indeed, his employment of the term material would seem to
imply), then we can only say that, so far as we can see, his
view differs but slightly, if at all, from Hoffmann’s idea of
“ physiological ” fusion.
The opinions of the last-named author are thus approvingly
summed up by Leche in a passage which deserves special
comment :—‘* Will man aber den Begriff der Verschmelzung
umbedingt beibehalten, so kann man, wie Hoffmann richtig
bemerkt, und wie auch ich bereits oben (p. 142) hervorge-
hoben habe, sich vorstellen, dass das Schmelzleistenmaterial,
welches bei den niederen Wirbelthieren zur Ausbildung einer
Ganzen Anzahl von Zahnserien verwendet wird, bei den Sauge-
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 5738
thieren zur Ausbildung von bedeutend wenigeren, dafiir aber
komplicierteren Zahnen benutzt wird” (3, p. 155).
Concerning such a view we would remark that it amounts
simply to a surrender of the idea of the fusion of morphological
structures as such. The aggrandisement of one morphological
structure at the expense of the material required for the deve-
lopment of another, whereby the survivor is enabled to launch
out into a further structural differentiation, may be a justifiable
conception, but it ought not to be confused with a case of
morphological concrescence. True morphological homo-
logy deals with the identity of organic forms, not with that of
the mere “ material” organised under such forms. In fact,
it may be questioned whether the identity of organic
material, as distinguished from that of the form
under which it is organised, is at all a fruitful or even
valid biological concept.
In any case, until Kikenthal has disposed of the objection
based upon the presence by the side of the molars (e.g. of
Marsupials) of a residual lamina entirely comparable to that
found beside other admittedly simple teeth, his theory of
homology of the molars must be regarded as untenable.
A still more fundamental question underlies the preceding
discussion, i.e. whether the multicuspidate form of the modern
mammalian molar has originated by the concrescence of several
originally distinct and simple tooth-germs, or by the gradual
evolution of cusps as outgrowths of a single simple conical
tooth.
In favour of the latter view are ranged the majority of
paleontologists (cf. especially Cope [10] and Osborn [84]),
who have elaborated a most plausible scheme showing the
steps in the supposed evolution of the complex mammalian
molars from a primitive conical type of tooth, through a very
early tri-tubercular condition.
The fusion theory, on the other hand, has taken shape chiefly
at the hands of a few embryological investigators—notably
Kikenthal and Rose.
Rose has described the appearance, at a very early period
574 J. T. WILSON AND J. P. HILL.
of the development of a molar enamel-organ, of several meso-
dermal papillary upgrowths prior to any marked differentiation
of the enamel epithelial germ. These upgrowths he confi-
dently interprets as the homologues of the originally distinct
denticles of which, in his view, a complex tooth is really built
up (cf. 29, fig. 1).
Woodward has remarked upon Rése’s attempt to prove
his theory in the case of the chameleon, “ where the back teeth
are each composed of three cones, which, according to him
(Rése), arise independently of one another,” to the following
effect :—this is true to a certain extent, but these cusps develop
under a common enamel-organ, and there is no indication of
their ever having possessed independent organs, as would have
been the case if they were distinct teeth, there being merely a
differentiation of the cylindrical enamel-epithelium over each
cusp, which Rése considers sufficient evidence in favour of his
view’? (14, p. 446).
We have occasionally come across appearances which at
first seemed to suggest multiple papillary upgrowths in con-
nection with the developing molar lamina. Thus in fig. 69,
though the whole of the surrounding connective tissue has not
been represented, it is plain that the irregularities of the labial
surface of the dental lamina might (if the connective tissue
were all filled in) suggest multiple papillation. But, as we
have already shown in our previous references to this figure,
this view would be quite erroneous. The true and only papilla
of m2 is marked “mp.” Fig. 69 may be compared with figs.
68a and 70. So far as our observation and experience go a
truly multiple papillary character is never found in the
earliest stages of development of a marsupial molar. At its
first differentiation as such the molar enamel-organ is as
simple, single, and uncomplicated in form as any incisor
enamel-organ. And as certainly we find that we can watch
the ontogenetic evolution of the marsupial molar cusps by a
gradual differentiation of the primitively simple cup-shaped or
bell-shaped enamel-organ.
It would appear, however, that such a statement as the
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 575
latter is by no means regarded, at least by Kiikenthal, as in-
consistent with the doctrine of phylogenetic concrescence of
dental germs. Replying to Hoffmann, he observes, “ Der
Backzahn eines heutigen Saugers is ein so complicirtes plas-
tisches Gebilde, welches im Laufe seiner Stammesgeschichte
einer solchen Summe von Verinderungen unterworfen worden
ist, dass ich nie daran gedacht habe, dass seine Entwickelung
uns tiber die ersten Stammesgeschichtlichen Vorgange, die in
die Uebergangsperiode von Reptilien zu Saugern fallen, wiirde
Aufschluss geben kénnen, und ich habe schon friiher....
die wenig glickliche Idee Rése’s, dass die Ausbildung einzelner
kegelf6rmigen Schmelz- und Dentinkappen an der spitze von
Backenzahnanlagen fiir die Verschmelzung spricht, als nicht
beweiskraftig zuriickgewiesen” (27, p. 658).
Certainly the crude theory here disclaimed will not stand
even a very superficial criticism. But the differentiation of
the molar dental lamina in Marsupials does not in the least
suggest the occurrence of a fusion-process of any kind. And
we are in a position to add that even the teeth of Ornitho-
rhynchus! are to be found, at a stage considerably anterior to
that described by Poulton (25), still in the condition of quite
simple deeply-cupped enamel-organs, from which the advanc-
ing complication of form shown in Poulton’s figures, and well
known in the fully developed teeth, is derived by differentiation
in the way of mere inequality of growth.
There can be no doubt that the advocates of the fusion-
hypothesis, instead of being aided by ontogenetic considera-
tions, will have to face the fact—perhaps not necessarily fatal,
but at least disconcerting—that, in a series of comparatively
primitive mammalian forms, the cusp-development of a compli-
cated molar can be traced out as a mere growth-differentiation
of a primitively smooth and simple bowl-shaped enamel-organ,
just as, according to Osborn (84, p. 206), we can be “in at
the birth of every successive cusp” in phylogenetic develop-
ment.
1 We hope ere long to make our observations in this direction the subject
of a special paper.
576 J. T. WILSON AND J. P.. HILL.
It may be conceived, of course, that prior to the simple cup-
shaped enamel-organ, there is a stage like that alleged to exist
by Rése, in which several mesodermal papille beneath a
common enamel-germ represent the future tooth. But the
evidence for the existence of such a condition is as yet so
extremely meagre and inconclusive that we cannot admit that
a serious case for it has been made out.
Schwalbe’s observations in favour of fusion, referred to by
Kikenthal, cannot be regarded as necessarily possessing any
great phylogenetic significance. The fusion referred to (16,
p- 21) of the upper milk incisor with the precociously develop-
ing crown of the successional tooth need not be taken as
possessing either more or less morphological importance than,
e.g., syndactyly in the human subject. Like the latter, the
condition might even be to some extent hereditary, but its
interest and importance are probably pathological rather than
morphological. Were it not for the undeniable importance of
Kikenthal’s observations on the process of division of the
cheek-teeth of whales into single conical teeth (19),! and of
the further statement of the same author that he has been able
actually to detect a fusion as taking place between molar
enamel-germs in the case of the walrus (20), we should be
disposed to regard the fusion theory as a very shadowy hypo-
thesis indeed. And even with the support alluded to we
cannot regard its claim to acceptance as at all a strong one.
There is plenty of room for sceptical criticism. Thus, concerning
Kiikenthal’s fig. 89 (19) Woodward has remarked (14, p. 447)
that the condition illustrated and described (19, p. 411) appears
“rather indicative of the formation of a cusp by outgrowth
from a simple conical tooth” than of the fusion process
supposed.
1 Oshorn has pointed out (84, p. 199) that “even by Kikenthal’s hypo-
thesis the typical Mesozoic mammal could not furnish as many teeth as are
found in some of the dolphins;” and he suggests as a likelier explanation
that “‘as the jaws were elongated the dental fold was carried back and the
dental caps were multiplied.”
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 577
PART IV.
ConcLupING REMARKS.
Before concluding this discussion it may not be wholly out
of place to ask the question whether our observations and the
conclusions deducible therefrom can be made to throw any
light upon the more general problem of the affinities of Mar-
supials with other mammalian groups.
Hasty generalisation is certainly to be deprecated, and it is
only with the greatest diffidence that we even approach the
discussion of such a weighty question with the equipment of
our own partial and imperfect knowledge, more especially of
the paleontological aspects of the problem.
There are, however, one or two more or less obvious deduc-
tions from the views which have been advocated in the fore-
going pages, and partially summarised in the introductory
section of the paper (p. 441).
It is plain that, if our view of the “ milk” homology of the
so-called “ prelacteal” teeth be admitted, we are bound to
believe that the marsupial order as a whole—if not derived
from truly Eutherian ancestors, as seems unlikely from the
general type or organisation exemplified—is at least an off-
shoot from a diphyodont stock common to both Metatheria
and Eutheria.
Following upon the publication of Kiikenthal’s and Rése’s
researches upon marsupial teeth, Osborn (34, p. 204) has
remarked that “the discovery of the complete double series
seems to have removed the last straw from the theory of the
marsupial “ancestry of the Placentals.’ And the adoption
of the conclusions of the present paper in no whit weakens the
general purport of this criticism, since it still leaves un-
touched the important fact that one of the two typical mamma-
lian dentitions has been lost during the evolution of the Mar-
supialia. In respect of their tooth-equipment then, the
Metatheria are degenerate mammals.
With this view it is interesting to correlate such specula-
578 J. T. WILSON AND J. P. HILL.
tions as inevitably arise in consequence of the discovery by
one of us (H., 30) of a true allantoic placenta in Perameles.
It is obvious that this discovery admits of only a limited
number of alternative explanations. It implies either (@) that
an allantoic placenta has been quite independently evolved
within the limits of the Metatherian and Eutherian groups,
thus affording a quite remarkable instance of parallelism in
development ;! or (4) that the Eutheria have inherited their
allantoic placenta actually from polyprotodont marsupial
ancestors closely allied to the modern Peramelida; or (c)
that both the Metatheria and the Eutheria have sprung
from an earlier mammalian proto-placental stock.
If the latter alternative were adopted (and we believe that it
alone will in the end commend itself to the judgment of
morphologists), then we should be in a position to define the
common mammalian ancestors of the Metatheria and Eu-
theria as both placental and diphyodont.
The much-debated question as to whether or not this early
mammalian stem was also marsupial in its organisation may,
in view of developmental researches upon the mammary organs
during recent years, admit of a not unsatisfactory reply. For
the researches of Gegenbaur (86), Klaatsch (82), and others
appear to have established the fact of the possession of rudi-
ments of ‘‘mammary pouches” (“ Mammartaschen”’) by
embryo Kutherians. Such mammary pouches (to be distin-
guished from a true Marsupium) are, in ruminants, accord-
ing to Klaatsch, converted into the tubular teat-cavity.? Bonnet
expresses Klaatsch’s view in this connection as follows :—“ Wo
aber die Mammartasche unverdndert bestehen bleibt, wie bei
den Wiederkaiiern und den Equiden macht er den Anschluss
dieser Formen an ganz niedere Zustande, die gar kein Mar-
supium ausgebildet haben, wahrscheinlich ” (81, p. 628).
1 But in such case Perameles, contrary to all previous belief, would
exhibit a more advanced or specialised condition than any other known
Marsupial.
2 The contrary opinions of Rein, though accepted by Minot in his ‘ Human
Embryology,’ have been subjected to severe criticism by Klaatsch, and the
views of the latter observer have been adopted by Bonnet (31).
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 579
It would appear probable that we must be prepared to en-
tertain favorably Klaatsch’s suggestion that the “ Placentals ”
have originally all passed through a kind of marsupial stage,
“ein Form, die jederseitz zwar eine Mammartasche, aber
keinen Beutel besass.” He denominates the animals which
exhibited this form “ Taschentiere ” or “ Bursalia.”
Such a rudimentary or primitive condition must have been
a characteristic of the common stock of both Meta- and
Eutheria, and was only a marsupial one in the sense in
which Echidna is now marsupial.
Klaatsch has recently pointed out (83) what strong reason
there is for the surmise that the evolution of mammary pouches,
such as Echidna possesses, is probably an even older trait of
mammalian development than the evolution of the mammary
glands themselves. Its existence in the oviparous Echidna
proves that it isa much older mammalian characteristic than any
form of placental connection can possibly be. We cannot follow
Hubrecht (35) in his assumption that the Protamniote an-
cestors of the Promammalia (and of the Sauropsida) must
have been viviparous.
In our view the Monotremes must be taken to represent a
quite distinct offshoot of a common Promammalian (Hypo-
therian) stem, which was originally both oviparous and “bursal”’
(in the sense of Klaatsch), and in which both hairs and mam-
mary glands were gradually developed. We do not believe that
in the long run it will be found possible to maintain the essential
dissimilarity of the mammary glandular organ in Monotremes
(36, and cf. Minot, 1. c., pp. 565-6).
After the separation of the Monotreme phylum we must
suppose the placenta to have appeared, constituting the steck
protoplacental. At all events, the existence of a placenta in
the marsupial order seems to indicate that the primitive pouched
common ancestors (Bursalia) of the Meta- and Eutheria,
long after losing their oviparous character, developed an
allantoic placental connection between the embryo and the
uterine wall. At the same time the mammary function,
already evoked during the oviparous period for the nourish-
VOL. 39, PART 4,—NEW SER. RR
580 J. T. WILSON AND J. P. HILL.
ment of the immature young as in Monotremes, became further
and more perfectly developed. We may further suppose
that now one branch of the stock thus circumstanced, i.e. the
Eutherian, went on to further exhaust the nutritive possibilities
implied in the acquirement of a placental connection; retain-
ing, however, and at the same time elaborating, the function
of lactation in relation to more advanced stages of development.
Another branch,—the Metatherian,—neglecting, from some
cause or other, the opportunities offered by the original for-
mation of a placental connection, may be supposed to have gone
on developing the latent capabilities of the mammary function
for the provision of adequate means of nutrition for the imma-
ture young. These capabilities included the replacement of
the primitive mammary pouches by a true Marsupium, partly
at least derived from the latter. The original pouched condi-
tion is thus to be carefully distinguished from the derivative or
truly marsupliated condition of the Marsupialia, which is
as much a secondary development along one line as, e. g., the
ruminant mammary condition is along another.
And the character of the marsupial dentition as we interpret
it renders it very highly probable that the secondary evolution
of a true Marsupium has all along been accompanied by a
retrogressive development or degeneration from the normal
early mammalian diphyodont dentition. Indeed, as Leche has
pointed out, the functional adaptation of the mouth of the
marsupial young to the peculiar suckling conditions prevalent
in the order has no doubt conditioned the almost entire sup-
pression of one of the two dentitions. The dentition sup-
pressed is, according to Leche, the second (his third) ; but in
our view it is, beyond all doubt, the first or milk dentition
which has degenerated.
University oF Sypvey, N.S.W.;
July 4th, 1896.
DEVELOPMENT AND SUCOESSION OF TEETH IN PERAMELES. 581
ADDENDUM.
Reference has been made in a foot-note in the body of this
paper to a contribution by Dr. Marett Tims (87) of ‘‘ Notes on
the Dentition of the Dog,” published in February in the
‘ Anatomischer Anzeiger.’ Since our research was completed
and the bulk of our work written up before Dr. Tims’ paper
came into our hands, we have found it easier to leave its con-
sideration for this special note.
From his observations upon tooth-development in the dog,
Tims has arrived at certain conclusions with regard to the
mammalian dentition with which our work will be found in
general agreement. And although this writer has apparently
not carried on special researches on the teeth of Marsupials,
his own work upon the dog has led him to foreshadow in some
degree the views independently arrived at by the present
writers on the grounds of a somewhat extensive and prolonged
series of observations.
Dr. Tims’ further confirmation of the common occurrence of
a residual dental lamina beside the permanent teeth of
Eutheria is a very welcome one.
The conclusion from his paper with which we are most parti-
cularly concerned is the following :—‘“‘ That there is not suffi-
cient evidence for believing in the existence of a pre-milk
dentition even in a rudimentary condition in the Mammalia
of the present day.” And a corollary, drawn in part from
this proposition, is that if Tims’ view be accepted, “ the teeth
of Marsupials would once more appear to belong to the per-
manent series.”
582 J. T. WILSON AND J. P. HILL.
J
BIBLIOGRAPHY.
. Rost.—* Ueber die Zahnentwickelung der Beutelthiere,” ‘ Anatomischer
Anzeiger,’ Bd. vii, 1892.
. Woopwarp.—“ Contributions to the Study of Mammalian Dentition.
Part I. On the Development of the Teeth of the Macropodide.”
‘Proc. Zool. Soc.,’ London, 1893.
. Lecue.—“ Zur Entwickelungsgeschichte des Zahnsystems der Sauge-
thiere,” ‘ Bibliotheca Zoologica,’ Heft 17, Stuttgart, 1895.
. FtowEer.—“‘ On the Development and Succession of the Teeth in the
Marsupialia,” ‘ Phil. Trans.,’ London, vol. clvii, 1867.
. THomas.—“‘ On the Homologies and Succession of the Teeth in the
Dasyuride, with an Attempt to trace the History of the Evolution of
the Mammalian Teeth in General,”’ ‘ Phil. Trans.,’ London, vol. elxxviii,
1887.
. Winez.—‘ Om Pattedyrenes Tandskifte isaer med Hensyn til Taendernes
Former,” ‘ Vidensk. Meddel. fra d. Naturhistoriske Forenung i
Kjébenhavn,’ 1882.
. KixentHaL.—“ Das Gebiss von Didelphys,” ‘ Anat. Anz.,’ Bd. vi, 1891.
. Tuomas.— Notes on Dr. Kiikenthal’s Discoveries in Mammalian Denti-
tion,” ‘ Ann. and Mag. Nat. Hist.,’ series vi, vol. ix, 1892.
. Roést.—‘‘ Ueber die Entwicklung der Zalne des Menschen,” ‘ Archiv
f. mikros. Anat.,’ Bd. xxxviii, 1891.
. Cope.—“ The Mechanical Causes of the Development of the Hard Parts
of the Mammalia,” ‘ Journ. of Morphology,’ vol. iii, 1889.
. RosE.— Ueber die Zahnentwicklung von Phascolomys _Wombat,”
‘Sitzungsber. der k. preuss. Akad. d. Wissensch. zu Berlin,’ Bd.
xxxvili, 1893.
. Lecns.— Nachtrage zu Studien tiber die Entwicklung des Zahnsystems
bei den Saugethieren,” ‘ Morph. Jahrbuch,’ Bd. xx, 1893.
. Lecne.—“ Studien iiber die Entwicklung des Zahnsystems bei den
Saugethieren,” ‘ Morph. Jahrbuch,’ Bd. xix, 1892.
. Woopwarp.—* On the Succession and Genesis of Mammalian Teeth,”
‘Science Progress,’ July, 1894.
. OwEn.—* Odontography,”’ London, 1840-45, ‘ British Association Re-
ports,’ 1848, p. 93.
. SchwaLBe.— Ueber Theorien der Dentition,” ‘Verhandlungen der
Anatomischen Gesellschaft auf der 8te Versammlung in Strassburg,’
1894.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES, 583
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
Osporn.—‘‘The Rise of the Mammalia in North America,” ‘Proc.
American Assoc.,’ vol. xlii, 1898.
Woopwarp.—* On the Milk Dentition of the Rodentia, with a Descrip-
tion of a Vestigial Milk Incisor in the Mouse (Mus musculus),”
* Anat. Anz.,’ Bd. ix, 1894.
KUKentHat.— Vergleichend-anatomische und Entwicklungsgeschicht-
liche Untersuchungen an Walthieren,” ii Thiel, ‘ Denkschriften der
medic.-naturw. Gesellschaft zu Jena,” Bad. iii, 1893.
KixentHat —“ Entwicklungsgeschichtliche Untersuchungen am Pinni-
pediergebisse,” ‘ Jenaische Zeitschr. f. Naturwissensch.,’ Bd. xxviii,
1893.
BatEeson.—‘ Materials for the Study of Variation,’ London, 1894.
Tuomas.—‘ Catalogue of the Marsupialia and Monotremata,’
Brit. Mus., London, 1888.
Batpwin Spencer.—‘ Report of the Horn Expedition to Central
Australia,’ part ii, Zoology, February, 1896.
Kixentuat.— Hinige Bemerkungen tiber die Saugetierbezahnung,”’
* Anat. Anz.,’ Bd. vi, 1891.
Povtton.—‘ The True Teeth and Horny Plates of Ornithorhyuchus,”
‘Quart. Journ. Micros. Sce.,’ vol. xxix, 1889.
Baume.—‘ Versuch einer Entwicklungsgeschichte des Gebisses,’ Leipzig,
1882 (quoted by Leche and others).
Ktxentuat.— Zur Dentitionenfrage,” ‘ Anat. Anz.,’ Bd. x, 1895.
Lecur.—“ Zur Dentitionenfrage,”’ ‘ Anat. Anz.,’ Bd. xi, 1895.
Roéss.— Ueber die Entstehung und Formabanderungen der menschlichen
Molaren,” ‘ Anat. Anz.,’ Bd. vii, 1892.
Hitt.—* Preliminary Note on the Occurrence of a Placental Connection
in Perameles obesula and on the Foetal Membranes of certain
Macropods,” ‘ Proc. Linn. Soc.,’ New South Wales, vol. x (2nd series),
part 4, 1895.
Boynet.—“ Die Mammarorgane im Lichte der Ontogenie und Phylo-
genie,” ‘Ergeb. d. Anat. u. Entwick.,’ Bd. ii, 1892.
Kuaatscu.—‘“ Ueber die Beziehungen zwischen Mammartasche und
Marsupium,” ‘ Morph. Jahrb.,’ Bd. xvii, 1891.
Kiaatsco.—“ Studien zur Geschichte der Mammarorgane. 1 Thiel: Die
Taschen und Beutelbildungen am Driisenfeld der Monotremen,”
‘Zoologische Forschungsreisen in Australien,’ Bd. ii, Lieferung ii,
Jena, 1895.
Rosz.—“ Beitrage zur Zahnentwicklung der Hdentaten,” ‘Anat. Anz.,’
Bd. vii, 1892.
584, J. T. WILSON AND J. P. HILL.
35. Husrecut.— Die Phylogenese des Amnions und die Bedeutung des
Trophoblastes,” ‘Verhand. der K. Akad. v. Wetenschappen te
Amsterdam,’ Tweede Sectie Dl. iv, n. 5, 1895.
36. GecenBAUR.—‘ Bemerkungen iiber die Milchdriisen-papillen der Sauge-
tiere,” ‘ Jenaische Zeitschr.,’ Bd. vii, 1873. ‘Zur genaueren Kenntniss
der Zitzen der Saugetiere,” ‘Morph. Jahrb.,’ Bd. i, 1876. ‘ Zur
Kenntniss der Mammarorgane der Monotremen,’ Leipzig, 1886.
37. Maret? Tims.—‘ Notes on the Dentition of the Dog,” ‘Anat. Anz.,
Bd. xi, 1896.
EXPLANATION OF PLATES 25—82,
Illustrating Messrs. J. T. Wilson’s and J. P. Hill’s paper,
“Observations upon the Development and Succession of
the Teeth in Perameles ; together with a Contribution to
the Discussion of the Homologies of the Teeth in Mar-
supial Animals.”
All sections drawn were outlined by means of Zeiss’s camera lucida.
List oF Common REFERENCE LETTERS.
C. S. Coronal Section. H.§. Horizontal Section. %. S. Sagittal Section.
U. J. Upper Jaw. LZ. J. Lower Jaw.
cb. Connecting bridge. d/. Dental lamina. mp. Dermal papilla. dl.
Residual dental lamina. cm. Meckel’s cartilage. 7%. 0. Labial laminar out-
growth. o.¢. Oralepithelium. ¢. Tongue. 7, *, p?, &c. First upper incisor,
upper canine, second upper premolar, &. 73, z pg, &e. Second lower incisor,
lower canine, third lower premolar, &c. diz, d*, dp, &e. First lower milk
incisor, upper milk canine, third upper milk premolar, &c.
The lingual side of the figure of each coronal and horizontal section is
indicated by a cross.
Stace II.
Fics. 1—5.—Serial coronal sections, U. J.
Fig. 1. In front of third incisor region.
Figs. 2 and 3. Next succeeding sections through enamel-organ of di3,
attached to “neck” of dental lamina, and showing earliest formation
of dermal papilla.
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES, 585
Fig. 4. Third section behind Fig. 3, showing di merging posteriorly in
common dental lamina.
Fig. 5. Third section behind Fig. 4, through Anlage of 73.
All x 210 diameters.
Fic. 6.—S. §. Cutting the anterior portions of the upper and lower dental
lamine transversely, not far from the mesial plane. a. Alinasal cartilage.
ch. Choanal passage. cm. Meckel’s cartilage. J.O. Organ of Jacobson.
J. c. Cartilaginous capsule of Jacobson’s organ. ma. Maxillary palate. x. e.
Epithelium of nasal cavity. pmz. Premaxilla. rf. Olfactory bulb. 8. d.
Stenson’s duct. ¢. Tongue. x 25.
Fic. 7.—U. J., C. §. Canine Anlage. x 210.
Fig. 8.—U. J., C. S. Milk premolar (dp3). x 210.
Fic. 9.—U. J., C. S. Milk premolar (dp?) in another embryo from same
pouch, with commencing liberation of the residual dental lamina. x 130.
Fic. 10.—S. §. dp® and molar segment of dental lamina (d/.). x 50.
Fie. 11.—U. J., C. S. Anlage of m+ with commencing papilla-formation
(mp.m+). xX 180.
Fie. 12.—L. J., C. S. Anlage of diz. x 230.
Fie. 13.—Z. J., C. 8. Canine Anlage. x 210.
Fig. 14.—8S. S. dpz and molar dental lamina (d/.). x 50.
Fie. 15.—Z. J., C. 8. First molar Anlage. x 180.
Stace ITI.—P. obesula.
Fic. 16.—U. J., H. S. Anterior end of dental lamina, showing the labial
portion of the Anlage of 2+ as a lobe projecting at right angles from the
main dental lamina (d/.), and the continuation backwards of the latter (d/’.)
towards the region of 22. x 140.
Fie. 17.—U. J., C. S. Second incisor mass, showing the labial portion of
the Anlage of 22 projecting from the swollen lamina, and the representative
of di2. x 180.
Fic. 18.—U. J., H. S. Second and third incisor masses (II and III), with
rudiments of di® and di3. The level of the section is indicated by the dotted
linesa ....6in Figs.17 and 19. x 180.
Fic. 19.—U. J., C. 8. Third incisor swelling, showing the narrow ‘ neck”
attaching the mass to the oral epithelium and the remains of the enamel-organ
of a3. x 180.
Fic. 20.—U. J., C. S. Canine mass, showing the as yet undifferentiated
Anlage of the permanent canine, with the cup-shaped enamel-organ of the
milk canine (da), c.dl. The morphologically free edge of dental lamina.
c. The probable equivalent of the labial lobe of the incisor masses. x 210.
586 J. T. WILSON AND J. P. HILL.
Fie. 21.—U. J., H. S. Cutting the oral epithelium at the point marked
0.:6.- X90:
Fic, 22.—U. J., H. S. At a somewhat higher level than Fig. 21. x 90.
Fie. 23.—U. J., C. S. Through the third incisor (2) of an early fcetus of
Dasyurus viverrinus. x 330.
Fic. 24.—U. J., C. §. dp. x 180.
Fie. 25.—U. J., H. S. Molar region at a higher level than Fig. 22. x 90.
Fic. 26.—U. J., H. S. m+ at a higher plane than Fig. 25. x 90.
Fie. 27.—U. J., C. S. Middle region m2. x 130.
Fie. 28.—U. J., C. S. m+ in the third section behind Fig. 27. x 180.
Fic. 29.—U. J., C. §. Posterior region of m1. x 120.
Fig. 30.—U. J., C. S. Dental lamina just behind m2. x 120.
Fie. 31.—U. J., C. S. Dental lamina five sections behind Fig. 30. x 120.
Fie. 32.—U. J., C. 8. Dental lamina in region of future m2, and just in
front of its termination. x 120.
Fic. 33.—Z. J., C. 8. iz and diz. x 280.
Fie. 34.—L. J., C. §. Anlagen of tg andiz. x 120.
Fic. 35.—Z. J., H. 8. Anlage of tz and zz. x 90.
Fie. 36.—Z. J., H. S. Nearer the oral epithelium than Fig. 35. x 90.
Fic. 37.—L. J., C. S. Anlage of permanent <, with deeply cupped enamel-
organ of dz. x 210.
Fig. 38.—Z. J., C. 8. Anlage of pz. x 150.
Fie, 39.—L. J., C. 8. dpz. X 280.
Fie. 40.—Z. J., C. 8S. dpz, seven sections behind Fig. 39, showing the
differentiation of cusps. x 180.
Fic. 41.—L. J, C. 8S. mz. x 120.
Fic. 42.—Z, J., C. §. Dental lamina behind enamel-organ of mz. x 1920.
Stace IV.
Fic. 43.—U. J., €. S. Dental lamina in front of 22. ca. Common Anlage
of lip-groove and dental lamina proper. xX 120.
Fie. 44.—U. J., C. 8. i+. ep. Epithelial ingrowths (see text). x 120.
Fic. 45.—-U. J., C. S. i2 three sections behind Fig. 44. x 120.
Fic. 46.—U. J., C. 8. i+ five sections behind Fig. 45. x 120.
Fic. 47.—U. J., C. S. ¢ with remains of d. x 100.
Fic. 48.—U. J., C. 8. dp® passing through the protocone. x 130.
Fie. 49.—U. J., H. 8. Through the region between p2 and m2. For ex-
DEVELOPMENT AND SUCCESSION OF TEETH IN PERAMELES. 587
planation of the projections from the dental lamina marked “a” and “db”
see text. x 100.
Fig. 50.—(Stace V.) U.J., H.S. Region between p? and m2. xX 100.
Fie. 51.—U. J., C. S. m+ passing through its protocone. x 80.
Fie. 52.—U. J, C. 8. m2. x 80.
Fig. 53.—U. J., C. S. m2 in another series, and nearer its posterior end
than Fig. 52. x 80.
Fic. 54.—U. J., C. 8. Anlage of m3. x 80.
Fie, 55.—Z.J., 0. 8. iz. x 100.
Fic. 56.—L. J., C. 8. iz some distance behind Fig. 55, showing diz. x 100.
Fie. 57.—L. J., C. 8. tg andiz. x 50.
Fie. 58.—Z. J., C. S. Dental lamina in front of ; with d&. x 210.
Stace V.
Fie. 59.—U. J., C. 8S. £. x 50.
Fic. 59a.—(Stace VII.) U.S,C. 8.5. 50.
Fie. 598.—(Stace VIII.) U.d.,C.8.£ x 50.
Fie. 60.—U. J., C. S. dp® and dental lamina in region of future p2. x 165.
Fie. 61.—U. J., C. 8. Anlage of m3. x 220.
Stace VI.
Fie. 62.—U. J, 0. 8. *. =x 50.
Fic. 63.—U. J., C. S. dp® with Anlage of p®. x 50.
Fic. 64.—L. J., C. 8. dpz with Anlage of pz. x 120.
Stace VII.
Fic. 65.—ZL. J., C. §. Passing through the forwardly directed apex of #;,
and showing the enamel-germ of an incisor (r. e. 0.) which arises from the
dental lamina in front of 7; x 210.
Fie. 66.—L. J., C. S. Through iz, showing diz independently attached to
the oral epithelium. x 150.
Fie. 664.—(Stace V.) JZ. J., C. §. Through 15, showing di; becoming
attached to the oral epithelium labially from the neck of the dental lamina.
x 150.
Fic. 67.—(Stacz IX.) U.J., C0. 8. dp® with Anlage of p?. x 30.
Fie. 68.—(Stace VI.) U.J., C. S. Posterior portion of m2. x 120.
Fie. 68a.—(Stace VI.) U.J., C. 8. Some distance behind Fig. 68, show-
ing Anlage of m&. x 120.
Fic. 69.—(Stace VIIL) U.J., C. 8S. Enamel-germ of m2. x 120.
588 Ji Di! WILSON: AND! J: P. HELE.
Fie. 70.—(Stace IX.) U.S, C0. 8. m&. x 120.
Fig. 71.—(Stacre X.) U. J, C. S. Enamel-organ of m® with residual
dental lamina, surrounded by concentrically arranged connective tissue. x 80.
Fie. 72.—(Stace X.) U.J., C. 8. Passing through £, showing one of the
enlargements of the residual dental lamina surrounded by a connective-tissue
capsule. x 230.
Fic. 73.—(Stace X.) U.J., C. S. Hnamel-germ of p&. x 120.
Fic. 74.—(Stace XIII.) U.J.,C. 8. Enamel-organ of pz, with com-
mencing formation of a residual dental lamina. x 90.
Fic. 75.—(Stace XIII.) U.J. From below. x 23.
Figs. 76 and 77.—Coronal sections through the third premolars (lower and
upper) of Phascolarctus cinereus, showing residual dental lamine by
the sides of p2. S/’. Residual dental lamina by side of pz. x 30.
(After Leche, Taf. xvi, figs. 140 and 142.)
Fic. 78.—(Stace XIV.) U.J. Region of premolars and molars from
below. xX 33.
Fie. 79.—(Stace XIV.) JZ. J. Region of premolars and molars from
above. X 33.
Fie. 80.—Dasyurus viverrinus. ZL. J., C.S. dpz. X 180.
Fic. 81.—Dasyurus viverrinus, JZ. J.C. 8. dpz in an older fetus.
x 1380.
Fic. 82.—(Stace V.) JU. J. Combination-drawing from several hori-
zontal sections. x 55.
Fic. 83.—U. J. Pig embryo of 20 mm. head-length. @C. 8. Dental lamina
in posterior molar region (see text). X 75.
Fic. 84.—Z. J. Pig embryo of 20 mm. head-length. C. 8. Passing just
posterior to ¢z, showing labial laminar outgrowth (/. 0.) and lip-groove
(1. gr.). xX 30.
Fic. $5.—Z. J. Pig embryo of 20 mm. head-length. C. 8. passing through
iz, showing labial laminar outgrowth (/. 0.) and lip-groove (/. gr.). x 30.
INDEX TO VOL. 39,
NEW SERIES.
Acanthocephala, a new genus (Arhyn-
chus) of, by Shipley, 207
Allen,
nervous system of crustacea, 33
Amphioxus, a new species of, by
Willey, 219
Anguilla vulgaris, metamorphosis
of, by Grassi, 371
Arhynehus hemignathi, a new
Acanthocephalon, by Shipley, 207
Arthropod’s heart and vascular systein
compared with enlarged vessels of
Magelona, by Benham, 15
Ascidian, new form from New Guinea,
by Willey, 161
Asymmetron, new species of, by
Willey, 219 , ~
Benham, W. B., on fission in Nemer-
tines, 19
Benham, W. B., on the blood of |
Magelona, 1
Bernard, an attempt to deduce the
Vertebrate eyes from the skin, 343
Blandford, W. F. H., translation of
Grassi on Termites, 245
Blood colouring-matter of Magelona,
by Benham, 4
Blood-corpuscles of Magelona, and
effect of reagents thereon, by
Benham, 5
Callotermes, Grassi on, 249
Carinella linearis, Mont., identi-
fied, by Benham, 21
Carinella (Nemertine), histology of
the nerve-cord and muscular coats,
by Benham, 28
Cheetz (= sete) of common earth-
worm chiefly composed of chitin,
those of smaller Oligocheta not
so, Goodrich on, 66
Edgar J., studies on the |
Chitin and chitinoid substances pre-
sent in the cceelomic corpuscles and
epidermic products of Enchytreus
and other Oligocheta (micro-chem-
ical examination of), by Goodrich,
62
Chlamydomyxa montana, a new
Protozoon, by Lankester, 233
Ceelomic corpuscles of Euchytreus
hortensis described, by Goodrich,
55
Ceelomic corpuscles of Pachydrilus,
sp., described, by Goodrich, 61
| Ctenophora, systematic relations of,
by Willey, 3238
Ctenoplana, by Willey, 323
Cuticle of the common earthworm,
soluble in distilled water, Goodrich
on, 65
Degeneration of polypides and em-
bryos in Lichenopora verru-
caria, by Harmer, 97
Dentition of Perameles, and of Mar-
supials in general, by Wilson and
Hill, 427
Development of the teeth of Pera-
meles, by Wilson and Hill, 427
Digestion, changes during, in gland-
cells of Drosera, by Huie, 387
Drosera, changes in the gland-cells
of, during digestion, by Huie, 387
Kel, common, metamorphosis of, by
Grassi, 371
Egg, formation of, in Lichenopora,
by Harmer, 104
Enehytreus hortensis, n.
described, by Goodrich, 51
Eyes of Vertebrates, deduced from
the skin, by Bernard, 343
sp.
590
Fission of embryos observed in
Lichenopora, and compared with
that in Crisia, by Harmer, 119
Fission of posterior segments con-
taining ripe gonads in Carinella,
and histology of the process, by
Benham, 22
Gland-celis of Drosera, changes in,
during digestion, by Huie, 387
Goodrich, E.S., notes on Oligochetes,
with the description of a new
species, 51
Grassi on the reproduction and meta-
morphosis of the common eel, 371
Grassi on the Termites, 245
Harmer, 8S. F., on the development
of Lichenopora verrucaria,
Fabr., 71
Hill, see Wilson.
Huie, L., on the changes in the gland-
cells of Drosera during digestion,
387
Lankester, Prof., on Chlamydo-
myxa montana, 233
Leptocephalus, various forms of, by
Grassi, 371
Lichenopora verrucaria, Fabr.
(Polyzoon), right-handed and left-
handed colonies described, by Har-
mer, 76
Magelona, a marine Chetopod, its
mode of occurrence, characters of
its blood and vascular system, by
Benham, 1
Marsupials, dentition of, by Wilson
and Hill, 427
Methylene blue used for demon-
strating cells and fibres in the
nerve-ganglia of embryo lobster
(Homarus), by Allen, 33
Nautilus, nepionic shell of, by Willey,
222
Nautilus, photograph from life, by
Willey, 179
Nautilus, varieties of, by Willey, 227
Nautilus, Willey on, 145
INDEX.
Nephridia of Enchytreus hor-
tensis described, by Goodrich, 51
Nepionic shell of Nautilus, by Willey,
229
Nerve-fibres, course of, in nerve-
centres of embryo lobster, 33
Nerve-ganglion cells of lobster (Ho-
marus) described and classified, by
Allen, 35
Ornithorhynchus, brain of, by Elliot
Smith, 181
Perameles, developmental history of
the teeth, by Wilson and Hill, 427
Polyclades of New Guinea, Willey
on, 153
Protozoon, a new species of (Chla-
mydomyxa montana), by Lan-
kester, 233
Shipley on Arhynchus, a new genus
of Acanthocephala, 207
Siphuncle of Nautilus, Willey on, 168
Smith, Elhot, on brain of a foetal
Ornithorhynchus, 181
Spectrum of blood colouring-matter
of Magelona, 4
Styeloides eviscerans, an As-
aWuen from New Guinea, by Willey,
1
Teeth of Perameles and Marsupials,
by Wilson and Hill, 427
Termes lucifugus, Grassi on, 249
Termites, Grassi on, 245
Thread-cells of Myxine compared
with the cclomic ‘“ thread-cor-
puscles” of Enchytreus hor-
tensis, by Goodrich, 59
Vertebrates, eyes of, deduced from
the skin, by Bernard, 343
Willey, letters from New Guinea, 145
»» On anew Amphioxus from the
Louisiade Archipelago, 219
», on Ctenoplana, 323
», On the nepionic shell of Nau-
tilus, 222
»» On varieties of Nautilus, 227
Wilson and Hill on the dentition of
aria especially of Perameles,
PRINTED BY ADLARD AND SON,
BARTHOLOMEW CLOSE, E.C., AND 20, HANOVER SQUARE, W.
or 2 ft ny fu! :
Aas Aaa :
pes
% ie ,
7‘)
*
7 i
ys
Ef
‘J 7
a
oh
iy
ve) f =
: yy ar
i : 7 ia) A? a
: : ay
od
Li
f ‘ i 7 ue IF
: a hon |
7 Vu pe
f
i,
7 As
rer,
ut ey
(CP ab ao
yy
walt
ERNST MAYR LIBRARY
IAEA OAT
3 2044 110 319 829
Date Due
APR 1970