Professional Documents
Culture Documents
IV.—A Monograph on the general Morphology of the Myxinoid Fishes, based on a study
of Myxine. Part VI. The Morphology of the Vascular System. By F. J. Cole,
D.Sc. Oxon., Professor of Zoology, University College, Reading. Communicated by
Professor J. H. ASHWORTH, D . S C , F.R.S. (With Two Text-figures and Five Plates.)
(MS. received July 17, 1925. Read February 9, 1925. Issued separately November 6.)
CONTENTS.
PAGE PAGE
A. Thymus and Pancreas 310 10. Subcesophageal Sinus 327
B. Pericardial Cavity 310 11. Carotid Sinus . . . . 327
C. Arterial System 312 F. Red Lymphatics—Peribranchial Cycle-
D. Venous System 314 1. Cardiac Sinus . . . . 327
E. Lacunar Spaces. Red Lymphatics—Subcutaneous 2. Peribranchial Sinuses 328
Cycle— 3. Peribranchial Anastomosis 329
1. Subcutaneous Sinus 322 4. Subchordal Sinus . 329
2. Subcutaneous Anastomosis . . . . 323 G. White Lymphatics—
3. Rostral Sinus 324 1. Lateral Chordal Sinus 330
4. Dorsal Oral Sinu3 324 2. Segmental Lymphatics . 331
5. Hypophysio-velar Sinus * 324 3. Intestinal Lymphatic Trunk . 331
6. Subhypophysial Sinus 325 H. Circulation of the Blood . 332
7. Ventral Sinus 325 J. Summary 336
8. Dental Sinus 326 K. Literature 337
9. Lingual Sinus 326 L. Explanation of the Plates. 339
The first five parts of this work on the skeleton, muscles, vascular system, and visceral
organs were published in the Transactions of the Society in 1905, 1907, 1909, 1912, and 1913.
The present section deals with a difficult and puzzling system which adds to its funda-
mental incongruity a range of variation which reduces the normal status to a condition of
speculative uncertainty. The works of A. A. RETZITJS, J. MULLER, and JACKSON have resulted
in a fairly accurate knowledge of the arterial system ; but the venous system, which is much
more important as regards its morphological bearings, is only imperfectly known, and the
series of so-called veno-lymphatic spaces hardly known at all. Unhappily, material for a
study of the development of the vascular system of Myxine has still to be obtained. The
smallest individual out of many hundreds which I have collected was over 10 cm. long.
BEARD records a badly damaged 6'5 cm. specimen and MAAS one of 8 "5 cm. The structure
of BEARD'S example, which I have examined, does not differ in any noteworthy respect from
that of the adult, and hence throws no light on embryonic conditions. Consequently the
interpretation of the venous system which has been adopted in this paper awaits embryological
confirmation, and is to be regarded in the meantime as a working hypothesis only. The
spawning grounds of Myxine have not yet been discovered, and it seems likely that the
animal spawns in deep water in places remote from its summer resorts.
The terminology of the skeletal system is the same as that adopted in Parts I and III
of this work, which it would be premature to change at this stage, but it must not be regarded
as having any established morphological significance. Similarly, in the absence of develop-
mental data, it would be unwise to attempt any systematic comparison of the Myxinoid
blood-vessels with those of higher Vertebrates ; and even when this information is available,
as it is in the case of the allied Lampreys, the task is not an easy one (cf. D E BEER, Q.J.M.S.
TRANS. ROY. SOC. EDIN., VOL. LIV, PART II (NO. 4). 41
310 PROFESSOR FRANK J. COLE
68, p. 327). The terminology adopted in the following description is therefore subject to
revision.
I am indebted to my museum assistant, Mr W. E. STONEMAN, B.A., who has, in a series
of dissections and injections, checked many of the results here described.
The expenses of this research have been defrayed by a grant from the Government Grant
Committee of the Royal Society.
I have nothing of importance to add to the structure of the systemic heart, but the
relations of the abdominal and pericardial coeloms must be dealt with. These two cavities
overlap each other—the abdominal coelom on the right side extending almost as far forwards
as the pericardial. The latter lies entirely dorsal to the former, the anterior lobe of the
liver almost completely separating the two cavities, which are only in contact on the extreme
right where the abdominal coelom bends over the right edge of the liver on to its dorsal sur-
face, at this point being in close contact with the pericardial coelom. It is here that the
* GIACOMINI does not confirm SCHAFFBB'S work on the thymus of Ammocuetes.
ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. 311
posterior wall of the latter cavity becomes continuous with the wall of the abdominal coelom
to form the pericardio-peritoneal foramen (fig. 9, p.p.f.), which is thus situated on the right
side and was described for the first time by J. MULLER in 183G. Through this, foramen the
portal vein appears to pass, but of course is morphologically outside it. The relations of the
foramen only are indicated in fig. 9, the actual position being much further back. The pos-
terior boundary of the pericardial cavity is irregular, and on the left side extends beyond the
anterior extremity of the posterior lobe of the liver. On the right side it also extends far
back, and almost entirely encloses the anterior end of the portal vein. Anteriorly it is
bounded on the right side by the last gill sac and on the left by the ductus oesophago-cutaneus.
It will be noticed on examining fig. 9, in which the pericardial cavity is represented in
black, that the heart projects into it from the ventral side and the portal heart from the
dorso-median. In the former case a wide gap is left ventrally on the left side which is
occupied by the sinus venosus, so that a ventral cardiac ligament or mesocardium, described
by JACKSON in Bdellostoma, has only an indefinite existence in Myxine. On the other hand,
the portal heart (and also the anterior portion of the portal vein) is almost completely
enclosed, forming a portal septum on the dorso-median side of these structures (cf. fig. 9).
It should be noted that the heart is pushed into the pericardial cavity from in front and
below, and the anterior extremity of the ventricle projects slightly beyond it. Like the
portal heart and vein, the common portal vein is sunk into the pericardial ccelom from above.
The distinction between the cardiac sinus (cd.s.), which is a true blood space and actually
consists of a very coarse venous plexus almost entirely surrounding the ventricle, and the
true pericardial cavity is very obvious in fig. 9. MECKEL erroneously identified the cardiac
sinus as the pericardial cavity, but the mistake was corrected by J. MULLER. JACKSON does
not mention the cardiac sinus, which he appears to have confused partly with what he calls
the " inner chamber of the right pericardial cavity."
Whilst the pericardial space is a single continuous cavity, it may be conveniently divided
into three regions. The central one encloses the ventricle of the heart but does not pass
anteriorly on to the ventral aorta, which lies in a forward extension of the cardiac sinus. The
central chamber is situated between the anterior lobe of the liver and the gut, and behind
the ventricle it becomes greatly narrowed dorso-ventrally. The right chamber is related to,
and more or less encloses, the dorsal division of the anterior section of the right posterior
cardinal vein with its included pronephros (r.a.c.1), the portal heart, portal vein, common
portal vein and systemic aorta, and communicates with the abdominal ccelom by the large
pericardio-peritoneal foramen, which forms its posterior boundary {p.p.f.). The left chamber
commences in front at the left of the ductus oesophago-cutaneus. It is related to, and more
or less encloses, the auricle of the heart, the dorsal division of the anterior section of the left
posterior cardinal vein with its included pronephros (l.a.c.1), the ventral division of the same
(l.a.c"), and sinus venosus.
It is important to note the asymmetrical disposition of the Myxinoid heart, since this is
clearly associated with other and more striking anomalies of the vascular system. The
ventricle is almost median but slightly to the right, whilst the auricle and sinus venosus are
well to the left, the auricle being entirely dorsal to the sinus venosus, which latter is lateral
and slightly ventral to the ventricle. The auricle is dorso-lateral to the ventricle. It is
obvious that this arrangement is quite different from that of the Selachian heart, and the
developmental history of the Myxinoid heart should prove very interesting. Its asymmetry
is accompanied by an asymmetry in the respiratory organs—the ductus cesophago-cutaneus,
which represents presumably a seventh gill, being restricted to the left side.
312 PROFESSOR FRANK J. COLE
The arterial system of Myxinoids is well known from the publications of J. MILLER
and JACKSON, and I propose to refer to a few points only which have been overlooked.
The coeliac or gastric artery (coe.a., coe.a.') is paired, the left, the larger of the two,
usually arising behind the right from the anterior extremity of the systemic aorta, but
the left may be the more anterior of the pair. The left coeliac near its origin is generally
attached to the anterior border of the bile duct, and it gives off the hepatic artery when
close to the gall bladder, finally crossing the bile duct to be distributed to the gut and
occasionally to the posterior liver also. A somewhat different condition is figured in my
Part V, PL II, fig. 4, where the hepatic artery is given off earlier in the course of the coeliac,
and is the vessel which accompanies the bile duct. The two coeliacs form a ring round
the gut, and meet on its ventral wall. They may there fuse completely so as to form a
single vessel, but they may only be connected up by an anastomosis, or may even remain
completely distinct. The hepatic artery (h.a.) is a branch of the left coeliac, and normally
divides into three branches to the gall bladder and two lobes of the liver, but in one case
the hepatic supplied only the anterior liver and gall bladder, the posterior liver deriving its
artery from the right coeliac. The coeliacs supply the gut as far back as the level of the
posterior extremity of the liver.
The pronephros (pn.) usually receives its arterial blood from the coeliac. Generally only
one artery is present on each side, and this passes to the glomus (fig. 1, right side), but there
may be more than one, in which case the body of the pronephros may receive an artery.
The pronephros, however, may be supplied by a branch from a segmental artery arising
from the aorta behind the coeliac of its side (fig. 1, left side), or from the root of the latter.
In one case the glomus of the pronephros received branches both from the cceliac and from a
segmental artery, which latter arose from the aorta some little distance behind the coeliac of
its side. Several other variations might be recorded. The isolated Malpighian bodies with
attached fragments of pronephric duct (pn.'), which are commonly found between the
pronephros and the anterior extremity of the abdominal kidney, are always supplied by
branches from segmental arteries (fig. 1, both sides).
In Bdellostoma, according to JACKSON, the coeliac artery is not paired, does not normally
supply the pronephros. and lacks the posterior branch or branches to the wall of the gut.
J. MULLER mentions only one coeliac artery (the left), but states that the pronephros (also
presumably the left) derives its branches mostly from it, and further, that it supplies the
liver, gall bladder, and the " stomach."
The position of the heart coincides in the tranverse plane with the constitution of the
single systemic posterior aorta (fig. l).
The posterior or abdominal aorta is formed by the union of three main arteries—a
median and two lateral. The former is the anterior systemic aorta or median anterior aorta
or vertebralis impar, and lies over the gut, not necessarily in the median position, whilst the
latter are the common carotid arteries or paired anterior aortse, situated at the side of the
gut. The common carotids are sometimes regarded as commissural vessels, and in the gill
region they certainly have that appearance in Bdellostoma, as J. MULLER points out, but their
integrity is maintained for a considerable distance in front of the gill region. The median
and paired anterior aortae are connected up by four pairs of anastomoses, of which the last
pair are usually very small and may be absent. I have, however, never found less than
three. The position of the anastomoses also varies, and they are apparently not always
ON THE GENERAL MORPHOLOGY OF THE MYXINOLD FISHES. 313
homologous in different individuals. Each gill has two efferent branchial arteries coursing
respectively in front of and behind the corresponding efferent gill duct. In Bdellostoma
there are one or two efferent gill arteries according to species. The only arteries which
arise from the median aorta are segmental arteries, the gut and all muscles other than those
of the body wall being supplied from the common carotid. The segmental arteries have no
regular arrangement, and, as in the case of the segmental veins, a separate vessel is not
detached for each segment, there being about four arteries to every seven segments.
In the gill region the common carotids vary considerably. The artery widens and
narrows in a capricious manner, loops are formed, and the vascular papillse, described in my
fourth part, are more numerous and complex than elsewhere. Posteriorly the artery may
become so thin as to be almost imperforate, and the object of the anastomoses appears to be
to ensure a passage for the arterial blood to the median aorta. This applies particularly to
the posterior half of the gill region. In front the artery preserves an even calibre, and its
individuality cannot be questioned. The vascular papillae are confined to the common
carotid, and no trace of them appears on the median aorta. They occur along the whole
course of the common carotid from the point where the posterior abdominal aorta is con-
stituted to where the carotid divides to form the external and " internal" carotids. In one
specimen I counted 95 on the left carotid and 85 on the right. For further details of these
curious structures the reader is referred to my Part IV (4).
On approaching the skull, but whilst still well behind it, the median and lateral aortae
break up. The common carotid divides into external and " internal" carotids, but as the
latter amongst other things supply somatic musculature and the skin, they are obviously
something more than this name implies. These latter vessels (fig. 1, v.i.'") by their fusion
in the middle line above the gut constitute the factors of what J. MULLER calls the unpaired
vertebral artery of the head or the vertebralis impar capitis (v.i/). This vessel was first seen
by MtiLLER, and the fusion which forms it completes what he calls the circulus cephalicus of
Myxinoids. At the point where the fusion occurs an artery is despatched downwards into
the velum (vel.a.), and in this way red blood reaches the velar portion of the hypophysio-
velar sinus.
The median aorta, somewhat anterior to the point where the common carotid splits,
divides into three vessels, which, omitting certain minor complexities, behave as follows
(fig. 1) :—The median one unites with the left factor of the vertebralis impar capitis, and
the paired vessels pass upwards and forwards to the side of the notochord, the right one,
however, sending an anastomosis to the right factor of the vertebralis impar capitis as it
passes it. The paired vessels give off two segmental vessels, and also send branches to the
constrictor pharyngis, velo-spinalis, and cranio-hyoideus muscles.
The median portion of the vertebralis impar capitis, after a short course covering about
three segments over the roof of the gut behind and the hypophysial tube in front, splits into
two vessels (figs. 1 and 6, v.i.), which bend dorsally and follow the inner ventral diverging
margin of the auditory capsule dorsal to the hypophysial tube (fig. 6, h.c). They give off
the artery to the brain (cb.a.), which passes upwards median to the trabecula, and perhaps
represents the true internal carotid, and is the first conspicuous vessel to arise from the
vertebralis impar capitis counting from behind. This artery is difficult to find even in
sections, and was not seen by J. MTJLLER.* The branches of the median section, closely
attached but external to the membranous cranium, then pass upwards and outwards internal
to the trabecula, and hence do not pass through any of the lateral fenestrae of the skull.
* According to STERZI there are three encephalic arteries.
314 PROFFSSOR FRANK J. COLE
They continue their upward and outward course, bend externally over the trabecular and
palatine bar, and finally pass under the eye to their final distribution. They therefore course
obliquely forwards, upwards, and outwards through the big gap in the base of the skull
bounded by the notochord, auditory capsules, trabeculse, and hypophysial plate. They
supply the M.M. velo-quadratus, copulo-quadratus profundus, tentacularis posterior, palato-
ethmoidalis superficialis, and the segmental muscle (parietalis), and a prominent branch is also
despatched to the skin (v.i.").
The external carotids (figs. 1, 4, 5, 6, ex.c.) from their posterior origin pass forwards and
downwards round the side of the gut on to its ventro-lateral surface, where they take up a
position just external to the ventral or lower division of the second branchial arch and the
lateral margin of the sub-cesophageal sinus. They now travel straight forwards in this
position, passing internal to the first branchial arch, until they reach the lateral margin of
the middle segment of the basal plate, which they now accompany. In this way they reach
and thereafter follow the lateral margin of the external bars of the anterior segment of the
basal plate, which, owing to its greater width and spread, extends upwards at the side of the
gut, so that the arteries are now lateral to the gut. They continue to pass upwards and
forwards, and just in front of the anterior extremity of the basal plate they lie dorso-lateral
to the mouth between the copulo-ethmoidalis and palato-ethmoidalis profundus muscles.
From this point they course at first outwards and then inwards and upwards between the
lateral labial cartilage and the dorsal limb of the tentacularis posterior muscle to their final
distribution at the tip of the snout.
It is largely by the anterior terminal branches of the external carotid artery that blood
enters the subcutaneous sinus.
Posteriorly the external carotids give off segmental vessels, but in front the segmental
musculature does not exist, and hence there are no segmental arteries and veins anterior to
the nasal capsule. The external carotids otherwise supply the gut, the MM. hyo-copulo-
palatinus, copulo-palatinus, quadrato-palatinus, copulo-quadratus profundus, copulo-quadratus
superficialis, palato-ethmoidalis profundus, tentacularis posterior, nasalis, transversus oris and
ethmoideo-nasalis, and also the skin of the snout. Two important branches are those which
supply the tongue—the anterior (l.a.) and posterior (La.1) lingual arteries. The former is the
more important, and supplies the greater part of the tongue. The latter reaches the tongue
via the " tether " of the hyo-copulo-glossus muscle, and supplies the ventro-lateral region of
the posterior extremity of this organ. In this way red blood reaches the dental sinus, in all
parts of which it is found.
On the tail the aorta splits into two vessels somewhat in front of the anterior knob of
the median ventral bar of the caudal fin skeleton (fig. 10, r.c.a., l.c.a.). The right vessel
passes on to the ventro-lateral surface of the notochord on this side and becomes small and
inconspicuous (r.c.a.). Further back it enlarges again, and the two vessels balance both as
regards size and position. They are, of course, separated by the median ventral bar of the
caudal fin skeleton. Posteriorly the caudal arteries are connected by a capillary system with
the factors of the caudal vein.
The best description to date of the Myxinoid venous system is still that of A. A.
RETZITJS, published in 1822.
The venous system varies very greatly in its details and even in certain aspects of
ON THE GENERAL MOEPHOLOGY OF THE MYXTNOID FISHES. 315
its general arrangement, and I select for description an example which exhibited the venous
system of Myxine in its most interesting and significant condition, and was investigated,
as it happened, with particular care. I have, however, confirmed the facts as described on
many other examples. I believe the Myxinoid venous system has undergone at least one
fundamental change in comparatively recent times, since individuals may from time to time
be found which depart from what may be regarded as the normal existing status in a
direction which can only be interpreted as ancestral.
The cerebral veins (figs. 2 and 6, cb.v.) arise ventro-laterally from the olfactory bulbs
as a pair of vessels, which then bend upwards and backwards on to the dorsal surface of
the brain, where they fuse at or near the middle line, the result further back becoming
the left efferent vessel of the brain. Apart from their origin, all the cerebral vessels are
confined to the dorsal surface of the brain. They are not symmetrical, the above vessel,
which we may call the left, draining the whole of the left half of the brain and also the
right side of the anterior region, the right posterior part of the brain having a special
vessel which, however, forms with the posterior section of the left a symmetrical pair of
vessels. Posteriorly the pair anastomose (fig. 2), but soon separate out again. They then
bend outwards, downwards, and backwards, and pierce the cranium by a special foramen
dorsal to the posterior region of the auditory capsule and posterior and dorsal to the vagus
foramen (indicated by.circles in fig. 2). They may now be described as the right and left
superficial anterior cardinal veins (r.s.c). The vessels at once pass outwards dorsal and
external to all skeletal parts and take up a position internal to the parietal muscle but
external to the constrictor pharyngis (fig. 7, l.a.c.s.). The superficial anterior cardinal veins
usually receive the segmental vessels of the region which they serve.
At about the transverse level of the eighth spinal ganglion the cardinal veins exhibit a
very curious arrangement which, however, I believe to be normal (fig. 2). The right super-
ficial cardinal (r.s.c.) perforates the constrictor pharyngis muscle and fuses with the right
deep anterior cardinal (r.d.c), the resulting vessel (r.d.c.') therefore passing backwards
internal to the constrictor pharyngis muscle. On the left side it is the deep cardinal which
perforates the muscle, so that the composite vessel (l.a.c.) courses external to the constrictor
pharyngis. Therefore the superficial vessel disappears posteriorly on the right side and
the deep vessel on the left, This anomalous state of affairs, which according to JACKSON
does not exist in Bdellostoma* continues for about eleven segments, i.e. up to about the
level of the anterior margin of the first gill. Consequently in this region we have vessels
which are not homologous performing the same functions. Each receives segmental vessels,
and also factors from the gut and the constrictor pharyngis, copulo-copularis, and parietal
muscles. At the level of the nineteenth spinal ganglion the right vessel swerves inwards
and becomes the median inferior jugular vein (i.j.v.).
Apart from the cerebral and cutaneous veins, no definite veins exist in front of the
auditory capsule, and the venous system of the snout is represented by the irregular so-
called veno-lymphatic spaces, which must therefore be regarded as an integral part of the
blood vascular system. The deep anterior cardinal vein (fig. 6, r.d.c.) commences as a large
sac, narrow from side to side but extensive dorso-ventrally, which is wedged in between
the hyoid arch and the velar portion of the gut. It receives its blood anteriorly by means
of a valved aperture (figs. 2 and 6, cl.h., h.v.s.') from the hypophysio-velar sinus, and some-
what further back there are other valves (fig. 2). The anterior extremity of the vein,
* In one Bdellostoma stouti dissected, the disposition of these vessels as regards the constrictor pharyngis muscle was exactly
as in Myxine.
316 PROFESSOR FRANK J. COLE
therefore, is a large chamber guarded at each end by valves. This fact, and the disposition
of the valves, at once suggests that the chamber may act as a local heart, collecting the
blood from the large irregular spaces of the snout, where the blood-pressure may well be
a minus quantity, and driving it posteriorly towards the heart. 1 had already noticed
(Part II, pp. 686 and 721-2) that the velo-quadratus and velo-spinalis muscles closely resembled
the extrinsic muscles of the caudal heart, and thus differed from all other muscles of the
body, but was not then aware of the interesting condition of the vein. If a Myxine which
has been narcotised in warm sea water be dissected so as to expose the " hyoid" region,
the rhythmic action of the small-fibred velar muscles can be readily observed. The result
of this action is to squeeze the vein against the hyoid arch, which itself is kept stable
by the constricting action of the cranio-hyoideus muscle—a muscle the function of which
is otherwise difficult to explain (cf. Part II, PI. IV, fig. 11, c.h.). I therefore regard the enlarged
anterior extremity of the deep anterior cardinal vein as structurally and functionally similar
to the caudal heart, and have named it the cardinal heart (figs. 2 and 6, cl.h.).
Behind the cardinal heart the deep anterior cardinal vein passes outwards through the
fourth fenestra of the skull, i.e. internal to the first branchial arch and external to the
second (cf. Part III, fig. I,/*.4, br.a.1, br.a.2). The transverse level is indicated by the circles
in fig. 2 of the present communication. It now lies lateral to the gut and internal to the
constrictor pharyngis muscle (fig. 7, l.a.c.d.). Before fusing with the superficial cardinal
it may receive one or more ventral segmental vessels, and also factors from the rectus,
copulo-quadratus superficialis, copulo-glossus profundus, and constrictor pharyngis muscles.
To continue the description of the anterior cardinal system, the left anterior cardinal
is clearly nothing but the backward extension of the superficial cardinal, with which the
deep cardinal has fused (fig. 2, l.a.c). This vessel represents the anterior cardinal sinus
or jugular vein of Elasmobranchs. At first it lies external to the constrictor pharyngis
muscle, but soon bends downwards and takes up a position lateral to the copulo-copularis
muscle. It soon rises again and assumes its old position external to the constrictor pharyngis,
where it is to be found for the greater part of its course. Behind the latter muscle it
lies external to the dorsal anterior limb of the constrictor branchiarum et cardise muscle,
afterwards occupying a similar position in respect of the posterior external sheet of the
same muscle (fig. 8, l.a.c). By this time it has increased very considerably in size, and
finally, in the neighbourhood of the ductus oesophago-cutaneus, it perforates the constrictor
branchiarum muscle and fuses with the corresponding vessel of the right side by means
of the extensive and somewhat irregular cardinal anastomosis (fig. 2, <?.«.), which passes
between the systemic aorta and the dorsal surface of the gut. The position of this
anastomosis excludes the possibility of its forming any part of the systemic heart. It is
surprising that no previous observer has seen the anastomosis, the existence of which con-
siderably modifies the course of the circulation, and is possibly associated with the loss of
the right Cuvierian duct. If it were absent, for example, the blood in the right anterior
cardinal could only reach the heart by passing through the liver. The anastomosis is easily
found both in Myxine and Bdellostoma. At the point where the anterior cardinal joins
with the anastomosis there is a valved aperture on each side, by which the sixth peri-
branchial sinuses discharge their contents into the venous system (fig. 2, p.b.s.'). Posteriorly
the left anterior cardinal received a vessel formed by the union of the dorsal and ventral
divisions of the anterior section of the left posterior cardinal (fig. 2, l.a.c.', l.a.c"). The
left anterior cardinal receives the segmental veins of its region, and also vessels from the
gut and the constrictor pharyngis, copulo-copularis, and constrictor branchiarum et cardise
ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. 317
muscles. JACKSON described the left anterior cardinal of Bdellostoma as opening directly
into the sinus venosus. This is certainly not the case in Myxine nor was it the case in
one Bdellostoma which was dissected with special reference to this point.
On the right side the vessel which passes backwards and corresponds functionally with
the left anterior cardinal is obviously the direct continuation of the right deep cardinal
vein and is therefore not homologous with the vessel of the left side (fig. 2, r.d.c.') After
receiving the right superficial cardinal it courses backwards lateral to the gut and internal
to the constrictor pharyngis muscle. At the posterior end of the club muscle it passes
outwards and downwards under the constrictor pharyngis, and curving ventrally round the
club muscle it soon reaches the neighbourhood of the mid-ventral line, as first described
by A. A. RETZITTS. At this point the contents of the lingual sinus are discharged into it
by a single aperture with a double valve (fig. 2, l.s.'). From this point backwards the vessel
is known as the median inferior jugular vein (figs. 2 and 8, i.j.v.). It is situated internal
to the muscles of the body wall and the constrictor branchiarum et cardise somewhat to
the right side of the median plane, and is closely opposed to the ventral wall of the cardiac
sinus. Further back, owing to the crowding together of the surrounding organs, it becomes
small. In the region of the ductus oesophago-cutaneus it crosses over to the left side,
becomes greatly enlarged but retains its relations to the surrounding muscles, and passes
imperceptibly into the sinus venosus. During the lateral part of its course the vessel
receives the segmental veins of its region, and also factors from the constrictor pharyngis
and copulo-copularis muscles. I did not detect any important vessels opening into it during
the median part of its course. The inferior jugular vein may be formed by the fusion of
the deep cardinal of both sides, the anterior factor disappearing on the left side. Its
formation and variations suggest this. A small vessel is sometimes received from the
region of the first gill pouch of the left side which may represent the missing left
factor.
A vessel which is undoubtedly the morphological equivalent of the left anterior cardinal,
having precisely the same relations to the surrounding organs, arises far back at the level
of the anterior margin of the first gill, i.e. at the point where the right inferior jugular
swerves towards the mid-ventral line. It is constituted by the union of the dorsal and
ventral factors of a typical segmental vein, and is much smaller than the vessel of the left
side. It is, however, a typical right anterior cardinal (figs. 2 and 8, r.a.c), and represents
the posterior extremity of the right superficial cardinal, which on this side is interrupted
for about eleven segments (seventh to eighteenth spinal). It is a vessel named by JACKSON
the anterior portal vein on account of its posterior connection with the portal heart, but
JACKSON failed to recognise the identity of this vessel with the anterior cardinal. Posteriorly
it communicates with the left cardinal by the cardinal anastomosis (fig. 2, c.a.) as already
described. Immediately posterior to the anastomosis it receives a large vessel formed by
a factor from the portal heart fusing with the dorsal division of the anterior section of
the right posterior cardinal (fig. 2, r.a.c.'). The right anterior cardinal receives a few
segmental vessels but no others of any importance. The extraordinary fact that the right
anterior cardinal communicates with the portal heart was first observed by A. A. RETZIUS.
The caudal veins are constituted at the tip of the tail by capillary systems which
place them in direct communication with the paired caudal arteries. The vein is also
paired, the sagittal plane being occupied by the median ventral bar of the caudal fin
skeleton (figs. 2 and 11, c.v.', m.v.b.). I have seen three cases in which there was more than
one caudal vein on each side. The caudal veins pass forward and expand to form the
TRANS. ROY. SOC. EDIN., VOL. LIV, PART II (NO. 4). 42
318 PROFESSOR FRANK J. COLE
paired caudal hearts, of which they are the afferent vessels. Anterior to the median ventral bar
the efferent vessels of the caudal heart (fig. 2, c.v.") unite to form the median caudal vein
(c.v.), and at this point the valves are situated which prevent any posterior movement of the
blood-stream (figs. 2 and l l ) .
The posterior cardinal veins (fig. 2, r.p.c, l.p.c) are formed by the bifurcation of the
median caudal vein anterior to the cloaca. The left posterior cardinal is much larger
than the right, and the two are connected by transverse anastomoses which may be
simple or ampulliform and pass ventral to the aorta. The posterior cardinals lie
ventral to the systemic aorta, which itself is ventral to the lateral chordal sinuses. The
latter are situated immediately below the notochord. There is no renal portal system, but
the pronephros is very closely associated with (presumably) the posterior cardinal system.
In the neighbourhood of the 31st spinal segment and the posterior extremity of the
anterior lobe of the liver, the right posterior cardinal effects its last anastomosis with the
left, and the single vessel anterior to this point is known as the common posterior
cardinal vein (fig. 2, c.p.c). The fact, however, that in many, if not most examples, the
right posterior cardinal is actually continued forwards in front of this region (fig. 2, r.p.c.')
shows that the so-called common posterior cardinal is simply the anterior extension of the
left vessel. This forward continuation of the right posterior cardinal was evidently seen
(but misinterpreted) by J. MULLER, who regarded it as the efferent vessel of the right
pronephros. The common posterior cardinal now bends to the left, and passes inperceptibly
into the sinus venosus. It is not clear, judging from adult conditions, where one ends
and the other begins, and it is therefore difficult to be quite certain whether the hepatic
veins open into the sinus venosus or the common posterior cardinal, or both. Several
vessels from the gut wall (avoiding the portal system),* and a few segmentals, open into
the common posterior cardinal + sinus venosus, and isolated kidney glomeruli may occur
between the pronephros and the definitive kidney which drain into the common posterior
cardinal. I have also found efferent vessels from the pronephros which pass backwards
to open into the common posterior cardinal on the left side and the right posterior
cardinal on the right. The posterior cardinals receive only segmental and renal veins, the
genitals opening into the portal vein. It is unusual for the same segment to have both
an artery and a vein. There is no regularity or symmetry in the arrangement, but the
segmental arteries are generally more numerous than the veins.
The posterior hepatic or subintestinal vein (fig. 2, s.i.v.) arises on the ventral surface
of the gut. It may extend over the whole length of the gut as it does in the Ammoccete,
but generally it is confined to a short region behind the posterior lobe of the liver. On
reaching the posterior lobe it may divide into two large vessels—one coursing on the
concave dorsal surface, and the other, the main trunk of the vessel, on the convex ventral
surface of the posterior liver. The former is dissolved in the liver, and in my Part V I
concluded that it was an afferent hepatic vein, and that therefore the subintestinal acted
both as an afferent and an efferent vessel of the posterior lobe of the liver. This, of course,
is suggested by a knowledge of its development in other types. There are, however, no
valves anywhere in the subintestinal vein, so that the action of the portal heart must
inevitably drive the blood into the subintestinal, in spite of the fact that in the case of
the dorsal factor the blood would have to pass backwards in order to reach the main trunk
of the subintestinal. We must therefore conclude that the subintestinal is an efferent
hepatic vessel only. Assuming this to be true, as it is in Bdellostoma according to
* Both A. A. RBTZIUS and J. MULLER note that a vein from the "stomach" opens into the left posterior cardinal (see fig. 2).
ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. 319
JACKSON, the subintestinal cannot be regarded as an extra portal vein, since it does not
distribute any of its blood to the liver. The main trunk of the vessel is always more or
less superficial in position and of a fairly uniform calibre throughout, and as it courses over
the posterior lobe it receives from it a few large efferent vessels. The subintestinal is
always the principal, and may be the sole, hepatic vein from the posterior lobe. An
accessory posterior hepatic vein opening separately into the sinus venosus is however
usually present, and forms an anastomosis with the hepatic vein from the anterior lobe of
the liver. The subintestinal differs from all the other hepatic veins in retaining its con-
nection with the gut, but the extent to which it does so varies considerably. It opens
apparently into the sinus venosus. The subintestinal was described in Myxine by
KLINCKOWSTROM in 1890, who mentions it as extending back to the cloaca in one specimen.
It was afterwards independently found by JACKSON in Bdellostoma.
The anterior hepatic vein (fig. 2, a.h.) receives a number of efferent vessels from the
anterior lobe of the liver, and may also include a factor from the posterior lobe and an anas-
tomosis from the accessory posterior hepatic vein when present. It opens into the sinus
venosus. In addition there may be as many as ten large and small accessory anterior
hepatic veins, each having a separate opening into the sinus venosus.
At the point where the common posterior cardinal passes into the sinus venosus it
receives two vessels, which are fused in front and arise from the junction of the left
anterior cardinal with the cardinal anastomosis (figs. 2 and 9, l.a.c.', La.c."). These vessels I
have named the dorsal and ventral divisions of the anterior section of the left posterior
cardinal. Whether they in fact belong to the posterior or anterior cardinal system, or to
both, can only be determined when we know their development. They may conceivably
belong to the anterior cardinal, in which case a ductus Cuvieri is formed on the left side
only by the union of anterior and posterior cardinals. The figures which A. A. RETZIUS
gives of this region, if somewhat crude, agree essentially with my fig. 2. There are no
valves associated with the dorsal and ventral divisions, but presumably their contents will
flow backwards into the common posterior cardinal owing to the action of the cardinal
heart. Both vessels, and the fused portion, project into the dorso-median section of the
pericardial coelom (fig. 9, shown in black), and are partly surrounded by it. The dorsal one
of the two (La.c.') is the larger, and is wrapped round the pronephros which lies apparently
inside it. This member of the pair, and also the fused anterior part, may receive veins
from the gut and the constrictor branchiarum et cardiae muscle, and also segmental veins.
It is curious that in none of the many detailed memoirs on the Myxinoid pronephros which
have appeared in recent years is the true relation of the pronephros to the venous system
recognised. MAAS, for example, simply states that the pronephros is enclosed by a venous
sinus. KIRKALDY, it is true, identifies the sinus as the " post-cardinal vein," but says
nothing of the connections of this vessel.
On the right side the conditions are less different than they appear at first sight.
There is a vessel corresponding exactly to the dorsal division of the left posterior cardinal
as regards its topographical position and relations to the pronephros (figs. 2 and 9, r.a.c.').
Variations of this vessel establish its identity beyond question, since in many, if not most
examples, it develops posteriorly a definite anastomosis with the right posterior cardinal at
the point where the latter fuses with the left. This anastomosis can only represent the
forward extension of the right posterior cardinal (fig. 2, r.p.c.'), the corresponding vessel
of the left side being the so-called common posterior cardinal (c.p.c). In three dissections
the anastomosis had no connection with the portal vein behind the portal heart, and was
320 PROFESSOR FRANK J. COLE
the direct continuation forwards of the right posterior cardinal.* The dorsal vessel there-
fore must be homologous with the dorsal division of the anterior section of the posterior
cardinal of the other side. It may, however, terminate posteriorly by passing into the
portal vein posterior to the portal heart, and this posterior section may be very small.
Anteriorly it fuses with the portal heart to form an unpaired section which opens into the
junction of the right anterior cardinal with the cardinal anastomosis, and thus behaves
exactly as the corresponding vessel of the left side (cf. fig. 2).
The portal or supra-intestinal vein of Myxinoids lies dorsal to the gut and to the
right of the median plane, and acts not only as the efferent vessel of the abdominal
section of the gut, but also of the gonad (cf. my Part II, PI. I, p.v.). These facts indicate
that it is not comparable with the portal vein of the true fishes, which develops from the
sub-intestinal vein of the embryo and has no association with the gonad. In Ceratodus,
however, according to KELLICOTT, the hepatic portal of the adult is derived from an
anterior new formation dorsal to the gut and the persisting posterior section of the sub-
intestinal vein. The portal vein of the Myxinoids therefore may represent the anterior
portion of Ceratodus. An examination of fig. 9 shows that the portal heart, as regards
its topographical relations, must represent the ventral division of the anterior section of
the posterior cardinal of its side (figs. 2 and 9, cf. p.h., l.a.c"). The course of the
afferent and efferent portal veins (p.v., cp.v.) suggests that the portal heart is an addition
to the system. The heart differs from the other accessory hearts in having an intrinsic
musculature, and is further larger and more powerful. The three openings from the right
anterior cardinal, portal vein, and common portal vein are all guarded by effective valves,
which ensure that blood passes into the heart from the first two vessels and away from it
by the third. A cystic vein (or veins) is constituted on the surface of the gall bladder
and opens into the portal vein (fig. 2, cys.v.). In one Bdellostoma stouti dissected I found
that the relations of the great veins to each other and to the heart, although differing in
certain details, were essentially the same as in Myxine. i t seems therefore that the accepted
description of the venous system of Bdellostoma might be revised with advantage.
The portal heart was first discovered by A. A. RETZIUS in Myxine in 1822. He then
described it as a spongy sac very similar in appearance to the auricle of the heart. He
noted that the right anterior cardinal vein passed into it, and also that this connection
was paralleled by the union of the anterior cardinal vein and pronephric vein of the left
side (cf. fig. 2). He discovered the pronephros, and suggested that it might be a "repre-
sentative of the kidney." He is extremely surprised that the pronephric vein should open
into the portal system on the right side and the posterior cardinal on the left. This
difference, however, as we have seen, is more apparent than real. When RETZIUS' paper was
translated into German in Isis in 1825, the translator notes that he had received a letter
from the author describing the presence of muscle fibres in the portal heart, and stating that
he had watched its pulsation in a living animal. J. MULLER appears to have overlooked this
note, for in 1841 he denied that RETZIUS' "long sac" was a portal heart, holding that it had
no valves or musculature and was capable only of elastic resistance. In 1845, however, he
corrects this statement, having observed the rhythmic pulsation of the portal heart in 1841,
but he still says nothing of the striated musculature which the heart undoubtedly possesses.
The interpretation of this unusual type of circulation can only be properly attempted
* J. MDLLER appears to have seen the vessels lettered in fig. 2 as r.p.c' and r.a.c.', but his description of them is diffi-
cult to follow. JACKSON found a vein passing from the right pronephros to the posterior cardinal in one specimen of
Bdellostoma.
ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. 321
pretation of the cardinal system be accepted both are absent. According to HATTA, the
anterior cardinal system of the Ammocoste is also anomalous, but not in the same way as
in Myxine. He concludes from topograpical relations and development that the anterior
cardinal vein terminates just behind the second gill arch, and that- the section of the
vein posterior to this arch is formed secondarily and does not correspond to a true
cardinal vein. This posterior part is an extension of the mandibular vein which becomes
secondarily connected with the anterior cardinal. In another respect Myxine differs from
the Ammoccete. Veins are constituted anteriorly on the tip of the snout in the usual
way, but (and in this they differ from the arteries) they are originally post-otic vessels
which afterwards extend headwards. The true head veins are thus reduced to vessels
which belong to the posterior region of the head. The Ammoccete therefore represents a
later stage in the evolution of the venous system than the adult Myxinoid.
The lacunar species of the Myxinoid, which are all lined by a tenuous flattened
endothelium, are, for reasons to be stated later, divisible into two series, which I propose
to call the red lymphatics and the white lymphatics. The former always contain red
blood in more or less greater quantity, and may be regarded provisonally as a part
of the blood vascular system interposed between arteries and veins. The red lymphatics
are grouped further into two cycles—the subcutaneous cycle and the peribranchial cycle.
The white lymphatics contain a white lymph only, which is discharged into the venous
system. They are, moreover, not a part of the blood vascular system, and therefore
do not contain red blood. They fall into a definite system by themselves. This is
schematised in fig. 3, which illustrates the relations of all the white lymphatics except
the pairs of segmental lymphatics. The white lymphatics may be considered, also
provisionally, as constituting a true lymphatic system.
with a paired coarse plexus confined to the caudal fin (fig. l l ) , into which the contents
of the sinus flow quite freely. This plexus, which is easily injected from the subcutaneous
sinus, and also the sinus itself (fig. 10), is drained by a marginal vessel (s.c.s.1), which,
starting on the dorsal side, passes backwards round the tip of the tail and then forwards
to open by a valved aperture into the antero-ventral angle of the hearts. Valves prevent
any regurgitation from the hearts to the afferent vessels. At the same place the lateral
chordal sinus also discharges into the heart (l.c.s/). The marginal vessel is situated
at the roots of the fin rays, and may be regarded either as a single vessel which bifur-
cates to surround the base of every fin ray, or as a double vessel which fuses between
the fin rays (cp. fig. 10). It does not completely split to open into the paired caudal
hearts, but sends an extension upwards for this purpose on each side of the median
ventral bar of the caudal fin skeleton, at the ventral edge of which it is situated during
the whole of its forward course. In front of the opening into the caudal heart the marginal
vessel may be continued forwards for a short distance. In addition to the above,
ALLEN describes a lateral lymphatic which passes dorsal to the series of slime sacs
anterior to the caudal hearts. I have frequently found this vessel in Myxine also.
KLINCKOWSTROM, in a brief note (1891), mentions the plexus on the caudal fin and the
marginal 'vessel, which he regards as paired, and states that it opens into the caudal
hearts. He missed, however, the connection with the lateral chordal sinus. FAVARO
describes the caudal vein (c.v.') as arising by the splitting of the ventral marginal vessel
(s.c.s.'). I doubt whether this ever occurs, especially as these vessels are related in such
a way that it is easy to understand how such an error could arise. FAVARO'S account of
the caudal circulation in Myxine is otherwise incomplete.
Connections—direct.—1. Hypophysio-velar sinus. 2. Rostral sinus. 3. Dorsal oral
sinus. 4. Ventral sinus. 5. Subcutaneous anastomosis. 6. Segmental lymphatics.
Connections—indirect.—7. Hypophysio-velar sinus, ventral sinus, and lateral chordal
sinus—all via the subcutaneous anastomosis. 8. Lateral chordal sinus via the segmental
lymphatics. 9. Caudal vein and posterior cardinals via the caudal hearts.
but not vice versa. (3) Behind the velo-spinalis muscle the remainder of the sinus is
constituted within the velum or pharyngeal valve by a coarse but extensive plexus, which
forms the velar lymphatic system and arises blindly at the posterior extremity of the
valve. A special section of this plexus lies alongside the lymphoid organ of the velum
(? thymus) which is associated with the anterior half of the external lateral velar bar.
Red blood reaches the velar section of the sinus from the capillaries of a median artery,
which arises from the point where the right and left factors of the vertebralis impar
capitis fuse to form the median portion (fig. 1, vel.a.).
Connections.—1. Subcutaneous sinus (direct connections anteriorly). 2. Rostral sinus.
3. Dorsal oral sinus. 4. Subhypophysial sinus. 5. Subcutaneous sinus, ventral sinus, and
lateral chordal sinus (indirectly via the subcutaneous anastomosis). 6. Deep anterior
cardinal veins.
the latter muscle a break appears first dorsally and then ventrally, thus dividing it
into two portions. These, however, at the extreme posterior end of the club muscle join
up again, after which the sinus rapidly diminishes in bulk, curves ventrally to the right
of the cardiac aorta but to the left of the inferior chondroidal bar, bends to the left of and
below the tip of the perpendicular muscle, and opens by a well-marked valved aperture into
the inferior jugular vein (fig. 2). The passage between the vein and the sinus in an
average adult forms a vessel about 5 mm. long. The structure of the valve and the
results of numerous injection experiments establish conclusively that the contents of
the sinus can pass into the vein but not vice versa.
Connections.—1. Dental sinus. 2. Subcesophageal sinus. 3. Inferior jugular vein.
anastomose over the surface of the ventricle. This plexus abuts on the sinus venosus
and auricle, and is itself surrounded on all sides except the median sagittal by the
pericardial ccelom (cf. fig. 9). In front the plexus fuses to form a conspicuous sinus
which completely surrounds the cardiac aorta throughout the whole of its length (fig. 8).
On each side the latter part of the sinus gives off wide channels which transmit the
afferent branchial arteries and place it in communication with all the peribranchial
sinuses. The first pair of these channels to the first pair of peribranchial sinuses is
formed by the bifurcation of the anterior extremity of the cardiac sinus. The portion
of the sinus surrounding the ventricle was known to MECKEL, who erroneously regarded
it as the pericardial cavity. J. MtiLLER. himself failed to observe that the ventricular part
of the sinus is really a coarse plexus, and was therefore unable to understand why
it could not fully be inflated. He considered whether the cardiac "and peribranchial
sinuses were lymph spaces, and decided that they were not, holding that they corresponded
rather with the pleural cavities of higher vertebrates.
Connections.—Peribranchial sinuses.
3. Anterior cardinal venous system via the sixth peribranchial sinus + the peribranchial
anastomosis.
3. Peribranchial Anastomosis (Fig. 8, p.b.a.).
This is a longitudinal somewhat small paired sinus situated dorsal to the peri-
branchial sinuses and attached to the internal surface of the dorsal anterior limb of the
constrictor branchiarum et cardise muscle (cf. Part II, PI. II, figs. 2 and 3). It commences
blindly in front and has no connection in most cases with the first peribranchial sinus, but it
may at this point be in communication with the sub-chordal sinus. The anastomosis
opens freely into all the peribranchial sinuses behind the first, and appears to fuse com-
pletely with the last of them. The point, however, where the sixth peribranchial sinus
opens into the subchordal sinus is the place where the anastomosis has fused with it. Also
the place where the sixth peribranchial sinus opens into the anterior cardinal anastomosis
(fig. 2, p.b.s.') is where the peribranchial anastomosis has fused with it, so that the latter
may be described as the vessel which connects up the sixth peribranchial and subchordal
sinuses and discharges into the vein. Thus the anastomosis opens at each end into the
subchordal sinus, and in the intervening region into the peribranchial sinuses.
Connections.—1. Subchordal sinus. 2. Peribranchial sinuses. 3. Anterior cardinal
venous system.
4. Subchordal Sinus (Figs. 7, 8, s.ch.s).
First described by J. MULLER in 1836, as "probably a lymph sac." He describes its
connection with the subcesophageal sinus, but appears to confuse it with the lateral chordal
sinus.
A large extensive median sinus constituted anteriorly and coursing backwards. It
lies for the most part immediately below the notochord, and stretches from the region just
behind the skull to the level of the external branchial opening. At each extremity it
consists of an irregular and variable coarse plexus, which has no connection with any part
of the blood vascular system. This is established by injection experiments. Posteriorly
the sinus dips ventrally under the anterior median edge of the constrictor branchiarum et
cardise muscle (fig. 8), and it then lies ventral to the median unpaired section of the lateral
chordal sinus on the dorsal surface of the gut, where it terminates in a plexus as above
mentioned. This posterior region is connected at the side of the gut with the posterior
section of the suboesophageal sinus (fig. 8). The peribranchial anastomosis opens posteriorly
into the subchordal sinus in the region of the sixth peribranchial sinus (fig. 8), and at
its anterior extremity may also open into the subchordal. This places the latter sinus in
indirect communication with all the peribranchials (except usually the first).
In a letter to me written in 1906 KLINCKOWSTROM mentioned that on one occasion he
was able to inject the portal heart as well as the veins entering and leaving it from the
subchordal lymph sac, but that he could never succeed in injecting the lymph sac from
the veins, from which .he concluded that the connection between the two series of vessels
must be by a valved aperture. In the experiment in question the injection would pass
from the subchordal sinus into the peribranchial anastomosis, and from thence through the
valved opening into the anterior cardinal venous anastomosis. From this point there is
nothing to prevent it extending into the veins in any direction (figs. 8 and 2).
Connections.—1. Suboesophageal sinus (anterior)—frequent lateral communications.
2. Suboesophageal sinus (posterior). 3. Peribranchial sinuses via the peribranchial anastomosis.
330 PROFESSOR FRANK J. COLE
G. W H I T E LYMPHATICS.
the subcutaneous anastomosis. 4. Intestinal lymphatic trunk via the series of abdominal
anastomoses and the direct passage anteriorly of the intestinal lymphatic trunk into the
intestinal sinus (fig. 3). 5. Caudal vein via the caudal hearts. 6. Segmental lymphatics
of the whole body.
2. Segmental Lymphatics (Fig. 7, seg.l., seg.l.').
The segmental lymphatic has the typical dorsal and ventral rami of a segmental vessel
with the striking exception that both are double, each pair opening into the lateral chordal
sinus of its side (l.c.s.). Between the double segmental lymphatic vessels of both rami
courses the segmental artery or vein—each segment, whilst possessing a segmental nerve,
having either a segmental artery or vein, and only rarely both. Lymphatics and blood-
vessels pass anterior to their corresponding segmental nerve. The occurrence of the lym-
phatics is not regular; in some cases they occur in almost every segment, in others in alternate
segments more or less. On the tail of Bdellostoma ALLEN finds that there is a lymphatic
in every segment accompanied by either a segmental artery or vein. Immediately above
the dorsal edge of the obliquus muscle in the case of the trunk vessels, the posterior
member of each lymphatic doublet gives off an anastomosis which passes straight through
the parietal muscle to open intersegmentally into the subcutaneous sinus {seg.l.'). An
injection thrown into the latter sinus reaches the lateral chordal sinus via these anastomoses
and the segmental lymphatics, and may subsequently pass also into the intestinal lymphatic
trunk. On the tail the connection between the segmental lymphatics and the sinus is
intersegmental and dorsal to the slime sacs, and may be of a more complex nature than
the one just described. In some injections the segmental lymphatics were connected up
by a longitudinal anastomosing vessel just dorsal to the slime sacs.
Connections.—1. Subcutaneous sinus via the segmental anastomoses. 2. Lateral chordal
sinus.
3. Intestinal Lymphatic Trunk (Fig. 3, i.l.t., i.l.t.').
This is a longitudinal vessel which accompanies the portal vein, and is situated
immediately to the left of the latter. Into this vessel an elaborate anastomosing plexus
of lymph capillaries discharges. These capillaries arise blindly in the wall of the gut in
the usual way, and are not the vessels described by MAWAS (i.l.t.'). The plexus is situated
on the outer surface of the gut immediately under the serosa, and lies superficial to the
blood capillary system. If the portal vein and the lymphatic trunk are injected with
different coloured media, the two systems of capillaries, one lying over the other and
quite distinct from each other, are strikingly shown. The lymphatic trunk discharges in
front into the posterior margin of the intestinal (lateral chordal) sinus, and behind into
the anterior margin of the rectal sinus, which embraces the gut and is situated just
anterior to the cloacal opening. It may be regarded as an expansion of the trunk itself
(fig. 3). The intestinal lymphatic trunk, as demonstrated by numerous injections, is con-
nected at intervals by means of anastomoses with the left lateral chordal sinus, these anas-
tomoses accompanying the branches from the aorta to the intestine. There is also a special
larger anastomosis connecting the rectal sinus with • the left lateral chordal which passes
downwards between the two posterior cardinal veins just in front of their formation by the
bifurcation of the caudal vein. A successful injection of the lymphatic plexus in the wall of
the gut can be made from the left lateral chordal sinus, and even from the subcutaneous
sinus via the segmental lymphatics (cf. fig. 7).
There can be no question that the intestinal lymphatic trunk is a true lymphatic
332 PROFESSOR FRANK J. COLE
vessel for the following reasons: it originates as blind capillaries in the wall of the
intestine, its contents never exhibit any admixture of red blood, and it discharges
ultimately into the blood-stream.
The lymph from the gut probably enters the blood-stream via the lateral chordal
sinus, which communicates directly with the caudal heart (fig. 3). It might, however,
pass from the lateral chordal sinus via the segmental lymphatics into the subcutaneous
sinus, and again reach the blood-stream via the caudal heart.
In some specimens I have found a ventral lymphatic on the gut corresponding with
the subintestinal vein. This also opened into the intestinal sinus.
Connections.—1. Left lateral chordal sinus by numerous anastomoses. 2. At its two
extremities it passes into sinuses—the intestinal and rectal sinuses—the former of which
may be regarded as an expansion of the fused lateral chordals at this region, the latter
being also connected with the same sinus on the left side.
depth (60 fathoms or more, or a little less), I could not exclude or refute the objection that
the blood was derived from burst small blood-vessels in the walls of the subcutaneous
spaces. We have often here discussed this question, but could never solve the- problem.
Now more and more I have been convinced that the blood naturally belongs to the
lacunar spaces already in the fresh living animals." The belief, however, that the blood
in the subcutaneous sinus of Myxinoids was due to extravasation was held until quite
recent times, and JACKSON, in his paper on the vascular system of Bdellostoma published
in 1901, still believed that there were no red corpuscles in the lymphatic spaces in life
and in uninjected specimens. That this statement is not correct is now well known,
although, according to ALLEN, the lymphatic spaces of Bdellostoma contain great quantities
of red corpuscles in the embryonic stages, but only a few in the adult.
That red blood occurs normally in the so-called lymphatic spaces of Myxinoids
is beyond question. At four widely separated points (cf. fig. 2) definite provision is made
for the return of the contents of the spaces to the blood-stream. This indicates clearly
that they constitute a part of the general vascular system, and that there must be
a constant and considerable flow from the so-called lymphatics to the veins, but it
does not necessarily involve the presence of red blood in the spaces. The fact, however,
that they contain red blood in the earliest developmental stages (ALLEN) is an important
piece of evidence. Against the extravasation hypothesis, which is dependent on a deep-
sea habitat, we have the fact that red blood is, nevertheless, found in the spaces of the
shallow-water Myxinoids and in the corresponding cavities of the fresh-water Lampreys.
Further, I have never seen a trace of red blood in the pericardial and abdominal
cavities, where we should certainly expect to find it if extravasation had occurred. We
must therefore conclude that those lymphatics which contain red blood form an integral
part of the blood vascular system of the animal, and that any scheme of the circulation
of the blood must not only include the red lymphatics, but must also explain how
the red contents are acquired and dismissed.
An examination of the contents of the lacunar spaces reveals the fact that most
of them invariably contain red blood in more or less considerable quantity, but that
others exhibit only a very few isolated corpuscles or (normally) none at all. We must
therefore separate these two groups, and, as explained above, they may be referred
to provisionally as the red and white lymphatics. Further, the latter fall into an
independent system which is complete in itself, and corresponds, at least in part and
physiologically, with the true lymphatics of higher Vertebrates. The intestinal lymphatic
trunk, for example (fig. 3, i.l.t.), arises as an elaborate blind plexus in the wall of
the gut, and discharges finally into the blood-stream, and therefore may be compared
physiologically with a true lymphatic vessel. The white and the red lymphatics are only
connected with each other by the subcutaneous anastomosis (fig. 3), the segmental
lymphatics (fig. 7), and at one point of the caudal circulation (figs. 10 and l l ) . The
occurrence of a few red corpuscles in the white lymphatics is paralleled by the occasional
presence of red blood in the true lymphatics of higher animals. The white lymphatics
are connected with, and discharged into, the red blood-stream, as elsewhere described,
their contents entering the circulation especially by the caudal hearts—directly from the
lateral chordal sinus (figs. 3 and l l ) , and indirectly via the subcutaneous blood sinus
(cf. text-fig, l). The latter sinus is fed with white lymph by the subcutaneous anasto-
mosis and the segmental lymphatics, and it is returned to the blood-stream via the
caudal hearts. In the case of the caudal hearts there can be no question that the direction
TRANS. ROY. SOC. EDIN., VOL. LIV, PART II (NO. 4). 44
334 PROFESSOR FRANK J. COLE
of the flow is from the white lymphatics into the true blood-vessels, and in other cases it
is a legitimate inference, since, for example, the blood-pressure in the red subcutaneous
sinus is a negative quantity, and therefore the contents of the small white segmental
lymphatics would naturally pass into it.
Inferior Jugular l/em
Cardinal Heart
Subcutaneous
Subhypophysial
Anastomosis
*L ateral
Chorda I
*lntestmal
Lymphatic
Lingual
Suboesophageal
(anterior)
Inferior Jugular Vein
An analysis of the anatomical relations of the red lymphatics shows that they fall
into two cycles—an anterior subcutaneous cycle and a posterior peribranchial cycle. The
two cycles are largely but not entirely independent of each other, and are connected up
via the anterior suboesophageal sinus. The dominating feature of the former cycle is the
extensive subcutaneous sinus, with which a number of the other members of the cycle are
connected. Red blood enters the cycle from the arteries of the snout.* What happens
• ALLEN was unable to find in Bdellostoma any direct connection between the arterial system and the "lymphatics."
ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. 335
then is almost impossible to determine, and the direction of the flow indicated in the
scheme is based on anatomical relationships only. The blood, however, is undoubtedly
returned to the veins of the definitive blood vascular stream by valved openings at the
following three points (cf. fig. 2): (l) via the hypophysio-velar sinus and cardinal heart
into the anterior section of the jugular system (h.v.s.', cl.h., r.d.c.); (2) via the lingual sinus
into the middle section of the jugular system (l.s.', i.j.v.); (3) via the subcutaneous sinus
of the tail region and caudal hearts into the caudal vein and posterior cardinals (fig. 11,
s.c.s., s.c.s.', cd.h., c.v.).
The central sinus of the peribranchial cycle is the subchordal sinus, but the largest
members of the cycle, in fact the largest sinuses in the body apart from the subcutaneous
sinus, are the peribranchials. Red blood enters the cycle via the afferent and efferent
branchial arteries and also _ the dental sinus of the subcutaneous cycle, and is returned to
the veins of the posterior section of the jugular system via the peribranchial anastomosis
(fig. 8, p.b.a.) and the anterior cardinal venous anastomosis (fig. 2, p.b.s.', c.a.).
Cardiac
Peribranchials
Peribranchial Anterior
-cardinal
anastomosis anastomosis
Subchordal
Suboesophageal Subbesophageal
(anterior) (posterior)
Dental
FIG. 2.—Scheme to illustrate the interrelationships of the sinuses constituting
the peribranchial cycle.
J. SUMMARY.
1. A degenerating lymphoid or thymus-like organ occurs in the pharyngeal velum.
2. The liver is an hepato-pancreas. Special glandular tubules of a pancreatic nature
are associated with the branches of the portal vein within the liver parenchyma, as in
some other fishes.
3. There are two cavities associated with the heart. The inner one is a blood sinus
and surrounds the ventricle only. The outer one is related to the ventricle, auricle, and
portal heart, and is a true pericardial cavity. It communicates with the abdominal coelom
on the right side by the large pericardio-peritoneal foramen.
4. The heart is asymmetrical. The auricle and sinus venosus are situated on the
left side, and the sinus venosus is ventral to the auricle. This differs from the Selachian
arrangement.
5. The coeliac artery is paired, and encircles the gut. The hepatic artery arises from
the left member of the pair.
6. Vascular papillae occur throughout the whole length of the common carotid artery,
but on no other vessel.
7. A vertebralis impar capitis artery is formed by the union of the "internal" carotids
8. The arteries differ from the veins in being continued forwards to the extremity of
the snout, where their contents are discharged principally into the blood sinuses of the
snout and tongue.
ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. 337
9. The general characters of the venous system are represented in fig. 2. Attention
may be directed to the asymmetry of the anterior cardinal system, the presence of a heart-
like structure posterior to the auditory capsules, the existence of an anterior cardinal
anastomosis, and the relations of the pronephros to the cardinal system and of the anterior
and posterior cardinals to each other. The right posterior cardinal does not terminate by
fusing with the left, but is usually continued forwards symmetrically with the left. The
portal heart is probably a modified part of the cardinal system, and if so its connection
with the anterior cardinal is explained.
10. Apart from the skin and brain there are no definite veins anterior to the
auditory capsule, arteries and veins being connected by irregular blood sinuses instead of
capillaries.
11. The caudal artery and vein are connected by a capillary system. The relations
of the caudal hearts to the blood vascular and lymphatic systems are shown in figs. 10
and 11.
12. The lacunar system is lined by endothelium, and is divisible into red lymphatics,
which constitute a part of the blood vascular system proper and always contain red blood,
and the white lymphatics, which represent a true lymphatic system. The red lymphatics
fall into two cycles, largely but not entirely separate from each other. The red lymphatics
communicate with the venous system by seven valved apertures at four widely separated
points (fig. 2). The white lymphatics are connected indirectly with the veins by the red
subcutaneous sinus, but have a special direct communication via the caudal hearts. Ked
blood enters the red lymphatics directly from the arteries. Therefore the blood circulates
through these spaces and a lacunar circulation is established. A true lymphatic duct with
an associated capillary plexus in the wall of the gut accompanies the portal vein.
K. LITERATURE.
[In this list only those works are included which contain original observations or generalisations
on the vascular system of Myxinoids.]
L. EXPLANATION OF PLATES.
REFERENCE LETTERS.
1, 1/ 2, 3. Anterior and posterior cusps of tooth 1 and cp.t.c' Posterior) divisions of the copulo-tentaculo-
teeth 2 and 3 of outer row of ventral teeth. cp.t.c." Anterior J coronarius muscle.
S. Third tentacle with skeleton. c.p.v. Common portal vein splitting into two branches
3.' Skeleton of third tentacle just prior to junction which pass to the anterior and posterior lobes
with the lateral labial cartilage. of the liver.
a.d.p. Anterior arch of the dental plate. c.q.p. Copulo-quadratus profundus muscle.
a.h. Anterior hepatic vein. e.q.s. Copulo-quadratus superficialis muscle.
a.s.ao. Anterior portion of the dorsal or systemic aorta cr. Membranous cranium.
( = median anterior aorta or vertebralis c.v. Median section of caudal vein.
impar). C.V.' Paired factor of caudal vein expanding to form
au.c. Auditory capsule. In fig. 6 with contained the caudal heart = afferent vein of caudal
labyrinth. heart.
aw. Auricle of systemic heart. c.v. Efferent vein of caudal heart which joins at
b.p.l~s Anterior, middle, and posterior segments of once with its fellow of the opposite side to
the basal plate, with guiding rails on dorsal form the median caudal vein.
surface of middle segment (fig. 6) directing cys.v. Cystic vein.
the antero-posterior motion of the dental d.o.s. Dorsal oral sinus.
apparatus. d.s. Dental sinus. In fig. 6 fused with lingual sinus
c.a. Anterior cardinal anastomosis. ventrally, and on right side communicating
c.ao. Ventral or cardiac aorta. dorsally with the subhypophysial sinus.
cb.a. Cerebral artery (= internal carotid). In fig. 5 note that the dental sinus is divided
c.h.c. Branchial portion of the constrictor branch- by the dental plate into external and in-
iarum et cardise muscle. ternal portions.
c.h.c' Ventral longitudinal tract of ditto. d.s.' Latero-dorsal extension of dental sinus com-
c.h.c. municating with the subhypophysial sinus.
) Loops 3 and 4 of ditto (cf. Part II, Plate II).
c.b.cM e.b.p.' External bar of the anterior segment of basal
cb.v. Cerebral veins constituting the right and left plate.
superficial anterior cardinal veins. e.c.a. External carotid artery.
c.c. Cornual cartilage. e.l.b. External lateral velar bar.
c.car. Common carotid artery( = paired anterior aorta). ex.c. External carotid artery.
c.car.' Anastomosis between the paired and median f-r. "Fin rays" of the caudal fin.
anterior aortse. h.a. Hepatic artery.
cd.c. Cordis caudalis muscle. h.c. Naso - pharyngeal or hypophysial canal just
cd.h. Caudal heart. anterior to its junction with the pharynx.
cd.s. Cardiac sinus surrounding the ventricle and On the right it has already fused with the
cardiac aorta. velar portion of the pharyngeal gut.
c.e. Copulo-ethmoidalis muscle. h.c.g. Hyo-copulo-glossus muscle.
e.g.p.' Lateral \ heads of the copulo-glossus profundus h.c.g.' Anterior end of " tendon" of the hyo-copulo-
c.g.p." Median J muscle. glossus muscle.
c.g.s. Copulo-glossus superficial muscle. h.c.p. Hyo-copulo-palatinus muscle.
cl.h. Cardinal heart. In fig. 6 connected by a h.p. Anterior process of hypophysial plate.
valved aperture with the hypophysis-velar h.v. Factors of hepatic veins in the liver.
sinus. h.v.s. Hypophysio-velar sinus. In fig. 5 note the
CO. Coronarius muscle. forward extension of this sinus into the
coB.a.
coe.a.' J coeliac or gastric arteries.
median olfactory lamina.
h.v.s.' Portion of hypophysio-velar sinus which
cod. Abdominal coelome. communicates by valved aperture with the
c.p.' First division of the constrictor pharyngis cardinal heart.
muscle. h.v.s." Section of hypophysio-velar sinus which arises
c.p.c. " Common " posterior cardinal vein (= left in association with the lymphoid organ of
posterior cardinal). the velum. The two halves have fused at
c.pl. Copulo-palatinus muscle. the middle line.
340 PROFESSOR FRANK J. COLE
r.s.c. Right superficial anterior cardinal vein. s.o.s.' Posterior suboesophageal sinus and communica-
rs.s. Kostral sinus. tion with the sub-chordal sinus.
s.a.a. Position of sinu-aurioular aperture of heart. sp.c. Membranous neural tube.
s.ao. Dorsal or systemic aorta. s.s. Slime sacks numbered from before backwards.
s.c.a. Subcutaneous anastomosis. s.v. Sinus venosus.
s.c.a.' Connection of same with hypophysio-velar sinus. sy. Sympathetic nerve or ft. intestinalis vagi.
s.c.a." Connection of same with ventral sinus. i.e. Tentaculo-ethmoidalis muscle.
s.c.a.'" Connection of same with lateral chordal sinus th. Lymphoid organ of the pharyngeal velum
forming the anterior extremity of the latter (1 thymus).
sinus. thy. Thyroid vesicle.
s.ch.s. Subchordal sinus. t.o. Transversus oris muscle.
ts.c.s. Subcutaneous sinus. t.p. Tentacularis posterior muscle.
s.c.s.' Marginal vessel which drains the subcutaneous vel. Pharyngeal velum.
sinus and discharges into the caudal hearts. vel.a. Velar artery.
seg.a. Segmental artery. vent. Ventricle of systemic heart.
seg.l. Segmental lymphatic. The posterior member v.i. Eight and left anterior branches of the "un-
of the pair, which communicates with the paired vertebral artery of the head" or
subcutaneous sinus, is in black. vertebralis impar capitis.
seg.l.' Connection of above with the subcutaneous v.i.' Unpaired median or fused portion of same.
sinus. v.i." Branch of same to skin.
seg.v. Segmental vein. v.i.'" Paired posterior factor of same or "internal
s.g.i. First spinal ganglion. carotid."
s.h.s. Subhypophysial sinus, on right side com- vn.s. Ventral sinus, fn fig. 6 it communicates on
municating with the dental sinus. the right side with the subcutaneous anas-
s.i.v. Subintestinal or posterior hepatic vein. tomosis.
sn.b. Subnasal bar. v.q." Dorsal division of the velo-quadratus muscle.
s.o.s. Anterior suboesophageal sinus. v.s. Velo-spinalis muscle.
PLATE I.
Fig. 1. Scheme of the arterial circulation from the region of the heart to the tip of the snout as seen from
above. The lateral series of dots represent the position of the spinal ganglia, counting as number one the first
spinal nerve which possesses both dorsal and ventral roots. The longitudinal lines in the neighbourhood of the
first spinal ganglion indicate the antero-posterior position and extent of the auditory capsule (not its position
relative to the arteries). The vascular papillae on the common carotids are shown as black dots, but only some of
them are figured. In the specimen in question there were 95 on the left common carotid and 85 on the right.
The position of the gills in the vicinity of their respective pairs of efferent branchial arteries is marked.
Fig. 2. Scheme of the entire venous system and its connections with the lymphatic spaces as seen from
below. The four breaks represent more or less long stretches in which the vessels course unchanged. The figures
on the left mark the position of the respective spinal ganglia (cf. fig. 1), and the lines anteriorly indicate the
transverse level of the auditory capsules, but not the relative position of the latter to the veins. The small circles
in l.a.c.' and r.a.c! mark the position and extent of the left and right pronephros. The position of the cloacal
aperture is behind the place where the caudal vein splits to form the two posterior cardinals. The points where
the deep anterior cardinal veins pass through the fourth fenestra of the skull, and where the superficial anterior
cardinals pierce the cranium, are indicated by circles.
Fig. 3. Scheme to illustrate the entire course of the paired axial portion of the true or white lymphatic
system, i.e. the lateral chordal sinus, and its relations with other sinuses, the intestinal lymphatic trunk and the
venous system. Ventral view. Long stretches in front of and behind the systemic heart, where the vessels course
unchanged, are represented by breaks. The segmental lymphatics which open into the lateral chordal sinus throughout
its whole length are not shown. Only a small portion of the true lymphatic plexus in the wall of the gut is
represented. Caudal hearts displayed laterally.
PLATE II.
Fig. 4. Section 150 of my large series through the anterior margin of the mouth to illustrate the relations of
the hypophysio-velar sinus with the rostral and dorsal oral sinuses. Nerves and ligaments in black. The relations
TRANS. ROY. SOC. EDIN., VOL. LIV, PART II (NO. 4). 45
342 PROF. FRANK J. COLE OK THE GENERAL MORPHOLOGY OF THE MYXlNOID FISHES.
of the anastomosis between the hypophysio-velar and dorsal oral sinuses are shown, but the actual connection is
posterior to this region.
Fig. 5. Section 390 through the nasal capsule and olfactory laminae. Illustrates the connection between the
subhypophysial sinus and the dental sinus via the latero-dorsal extension of the latter sinus, and the splitting up of
the dental sinus by the dental plate (a.d.p.) and the "tendon" of the hyo-copulo-glossus muscle (h.c.g.'). Olfactory
nerves, tendons, horny caps of the teeth, and a portion of the anterior arch of the dental plate shown in black.
PLATE III.
Fig. 6. Section 660 through the cardinal heart and the base or root of the pharyngeal velum. Somewhat
diagrammatic. Naso-pharyngeal tube (h.c.) is about to fuse with the pharynx, and has already fused on the right
side with that portion of the pharyngeal gut associated with the velum. Several sections have been telescoped
in order to illustrate the interrelationships of the various sinuses.
Fig. 7. Section 900. Segmental lymphatic (double) and its connection with the subcutaneous and lateral
chordal sinuses figured on left side. The posterior member of the pair is shown in black. Velar portion of
hypophysio-velar sinus and the left lateral chordal sinus also in black. The velum or pharyngeal valve with its
skeleton projects into the cavity of the gut from its dorsal wall.
PLATE IV.
Fig. 8. Section 2830. On the right side the complete course of an afferent branchial artery and on the left of
an efferent branchial artery is shown. On the right also the anastomosis between the cardiac and peribranchial
sinuses is illustrated. Efferent gill ducts one to four numbered from before backwards. On the left the connection
of the peribranchial anastomosis with the peribranchial and subchordal sinuses is figured, as also is the connection
of the latter sinus with the posterior section of the subcesophageal sinus.
Fig. 9. Section 3220 illustrating the relations of the pericardial cavity (in black) to the systemic heart, portal
heart, and abdominal ccelome. • The nature and position of the connection between the abdominal coelome and
pericardial ccelome are indicated, but the actual pericardio-peritoneal foramen occurs posterior to this region. The
branches of the portal vein in the liver are shown surrounded by pancreatic tubules.
PLATE V.
•
Fig. 10. Transverse section through the tail, passing through the narrowing anterior extremity of the caudal
hearts. Slightly diagrammatic. The dorsal portion of the marginal vessel of the subcutaneous sinus (s.c.s.') is here
double so as to allow the fin ray to fuse with the median dorsal bar of the caudal fin skeleton. The ventral
portion is shown as between two fin rays, and at these points the vessel is single. Note the asymmetry of the
caudal arteries at this point. They become symmetrical further back. The lateral chordal sinus and subcutaneous
sinus are in communication with the caudal heart on the right side. Behind this point the dorsal portion only of
the lateral chordal sinus is continued backwards as a paired vessel, the left one being already detached. Note the
mode of connection between the subcutaneous sinus and the dorsal and ventral marginal vessels.
Fig. 11. Dissection of right side of tip of tail to show the relations of the caudal heart with the subcutaneous
sinus. Injected from the latter sinus. Above on the right the skin only has been removed, which exposes the
cavity of the subcutaneous sinus and also a portion of the right coarse plexus on the caudal fin connected with
the sinus. On the left and below, the myotomes and fin plexus have been dissected away so as to uncover the
caudal heart and its vessels and the marginal channel which drains the subcutaneous sinus both directly and
indirectly through the plexus on the fin, and finally discharges into the caudal heart. The posterior extremity of
the lateral chordal sinus and its connection with the caudal heart are shown, as also is the cartilaginous knob
which projects forwards from the anterior margin of the median ventral bar of the caudal fin skeleton. Remainder
of caudal fin skeleton not shown.
Trans. Roy. Soc. Edin. VOL. LIV.
Professor FRAXK J. COLE: "A Monograph on the general Morphology of the Myxinoid Fishes."—PLATE I.
cb v.
h.vs'.
s c.s.
positl
a. s. ao.
c.car.
• IS
31
rfxc i.p.c
s.c.s.
right left
r.ght
FIG. 1.
F. J. Cole, del.
Trans. Roy. Soc. Edin. VOL. LIV.
Professor FRANK J. COLE : " A Monograph on the general Morphology of the Myxinoid Fishes."—PLATE II.
F. J. Cole, del.
Trans. Roy. Soc. Edin. VOL. LIV.
Professor FKAXK J. COLE : " A Monograph on the general Morphology of the Myxinoid Fishes."—PLATE III.
F. J. Cole, del.
•vv 'n°o
—,,'saqstj jo uo y „ : UTOQ '£
-
'All
Trans. Roy. Soc. Edin.
VOL. L1V.
Professor FRANK J. COLE: " A Monograph on the general Morphology of the Myxinoid Fishes."—PLATE V.
• ^ 1 c.i/.
F. J. Cole, del.
FIG. 11.