Text-Book of Embryology 2-6 (1919)

Embryology - 20 Oct 2026        Expand to Translate
Google Translate - select your language from the list shown below (this will open a new external page)
العربية | català | 中文 | 中國傳統的 | français | Deutsche | עִברִית | हिंदी | bahasa Indonesia | italiano | 日本語 | 한국어 | မြန်မာ | Pilipino | Polskie | português | ਪੰਜਾਬੀ ਦੇ | Română | русский | Español | Swahili | Svensk | ไทย | Türkçe | اردو | ייִדיש | Tiếng Việt    These external translations are automated and may not be accurate. (More? About Translations)

Kerr JG. Text-Book of Embryology II (1919) MacMillan and Co., London.

- Currently only early Draft Version of Text -

Historic Disclaimer - information about historic embryology pages

Pages where the terms "Historic" (textbooks, papers, people, recommendations) appear on this site, and sections within pages where this disclaimer appears, indicate that the content and scientific understanding are specific to the time of publication. This means that while some scientific descriptions are still accurate, the terminology and interpretation of the developmental mechanisms reflect the understanding at the time of original publication and those of the preceding periods, these terms, interpretations and recommendations may not reflect our current scientific understanding.     (More? Embryology History | Historic Embryology Papers)

Chapter VI Vascular System

As has already been indicated the vascular system of the animal body consists of strands of highly specialized niesenchyme——-the cells (corpuscles) along the axes of the strands being detached from one another and Heating freely in a fluid intercellular substance (plasma), while the superficial cells are united together to form the walls of. tubular channels-—~—the vessels. The vessel walls are provided with a coating of muscle-fibres and this muscular coat becomes greatly thickened and specialized at one or more points to form hearts which serve as pumps to force the blood through the system of vessels. '

The fundamental plan of the Vertebrate vascular system appears to have been like that of an Annelid worm, with two main longitudinal blood-vessels, situated respectively one on the neural side of the alimentary canal and one on the side opposite to this, connected together by a series of half-hoop shaped vessels encircling the alimentary canal laterally. In the Vertebrate the longitudinal vessel on the neural side of the alimentary canal is the dorsal aorta and in it the blood runs in a tailward direction. The longitudinal vessel on the other (ventral) side of the alimentary canal develops the heart on its course: its precardiac portion is the ventral aorta, its postcardiac the subintestinal vein. In this ventral vessel the blood passes in a headward direction. Half-hoop shaped vessels lying in front of the heart and connecting ventral aorta and dorsal aorta are the aortic arches.

0 c YER'1‘EBRA'rEs.— mongst holob as 10 er e rates the first steps in the deve opment of the vessels have been investigated in the Newt (Triton) by Mollier (1906) and his account will here be followed.

In an embryo with six mesoderm segments the lateral sheetsof mesoderm have met ventrally except in the region of the liver where they terminate in a free edge. This free edge is thickened and the thickening extends back along the mid-ventral line towards the cloaca as the rudiment of the subintestinal vein—the entire thickening having thus a Y-shape (Fig. 173, A).

360 (:11. VI ORIGIN OF THE HEARTANI) VESSIELS 361

At a stzigu with ‘twelve segments this Y-shaped vascular rudiment is continued forwards as a couple of strands of cells, lying on each side on the inner surface of the splanchnic mesoderm and apparently derived from it. These are destined to give rise in their hinder portions to the two vitelline veins and in their anterior region to the first rudiments of the heart (_ Fig. 17 2, A, (me).

At a stage with fifteen segments the paired strands of cells have assumed a disposition like that shown in Fig. 173, B. They approach one another as the mesoderm extends downwards and presently fuse across the mesial plane (Fig. 172, B and O), the fused portion being the rudiment of the heart while the two :1.nteri.or limbs represent the first (mandibular) pair of aortic arches and the two posterior

FIG. 172.——Ventral portions of transverse sections of young Amphibians to illustrate the tlcvclopment of the he:u't:.. (R.-is-a_sal on ligln-o_-s by M olli¢_.:r. 1906.)

A, B, I), E, ’1'*r-Him (A twelve segments, B sixt.oa_-n ¢ln., l) 1.\u-nty «ln., I41 twenty-six do.); (.‘, R--nu. a.l.mc, dorsal mesocardium; eat, ectoclernn; rm-, «-mlm-.:n-«lium; mm‘. t'lI(l(Nlt‘l'lll; me, myouardium; mes, mesoderm; pa, pm-icardiac cavity; -v.mn, ventral lIlcs()(‘u.l'Ilillll1.

limbs the vitelline veins. The heart rudiment is at first extremely short in an antero—posterior direction being much broader than it is long. This is correlated with the shortness of the foregut. As the latter lengthens the heart rudiment keeps pace with it, and becomes elongated (Fig. 173, D). As it does so the tissue within the rudiment becomes loosened and takes the form of a syncytial network with wide meshes. .

In the meantime the mesoderni on each side, now containing a wide coelomic (pericardiac) space, has grown down to the mesial plane ventral to the heart, so as to give rise to a ventral mesocardium which however only persists for a short time (Fig. 172,

C and D, mnc). About this same period fluid begins to collect in.

the interstices between the cells of the subintestinal strand, with the result that some of the cells in its interior assume a spherical 362 EMBRYOLOGY or THE LOWER VERTEBRATES on.

form and are recognizable as embryonic blood eorpnscles. Mollier notes that about this stage the subintestinal strand comes into extremely close relation to the yolk—cells, there being in places apparently complete continuity between the two-——hence the conclusion on the part of observers who did not study earlier stages that the vascular strand was actually derived directly from the endoderm.

About the stage with sixteen to eighteen segments the rudiments of the Duct of Guvier and dorsal aorta become apparent, in the form of cells at iirst scattered and later joined into strands. The aorta cells anteriorly often show eonnexions with the sclerotomes and

Fm. 173.-——Rongh diagram to illllstl.':1tc the form of the early rudiments of heart, vitel1ine_

veins, and -‘x-ul'»intestinnl vein in Triton as seen in plan. A, six mesoderm segment stage; B, fifteen segments; (2, eighteen segments ; l), twenty segments.

u..e.l, mandibular aortic arch ; cl, position of clones; H, heart; L, position of liver; .~:.-a'..v, subintestzinal vein; -v.-r, vitelline vein.

Mollier admits that some of them may actually be derived from the sclerotomes (see p. 364) though he considers that the main source of origin is the upper angle of the lateral mesoderm.

At the stage with twenty segments a network of fine channels has appeared over the surface of the yolk, between it and the mesoderm, foreshadowing the vitelline network of blood-vessels. The subintestinal strand has become still looser in texture and prolongations may be found passing from it inwards amongst the

~yolk—cells.' ” The heart has now attained the form of a. straight tube

the protoplasmic strands in its interior disappearing while its superficial cells take on an endothelial character, and are recognizable as the endocardium. The splanchnic mesoderm has become closely moulded round it ventrally and laterally (Fig. 172, D) forming the rudiment of the myocardium and the latter begins to show contractions causing slight movement of the fluid contents of the heart. VI ORIGIN OF THE HEART AND VESSELS 363

By the twenty-seven segment stage the anterior limbs of the subintestinal strand have become detinite (vitelline) veins with well-defined lumen filled with fluid in which spherical young corpuscles float freely. The large flat cells forming the wall are probably simply the modified superficial cells of the strand though Mollier thinks these may be reinforced by additional mesoderm cells from without. The vitelline veins are continued in front into the posterior venous limbs of the hea.rt and the heart itself is seen in transverse sections (Fig. 172, E) to be now completely enclosed in myocardium, the inner wall of the pericardiac space having become moulded right over its dorsal side. Where the two sheets of mcsoderm, one from each side, have met dorsal to the heart there still persists a septum——the dorsal mesocardium (Fig. 17 2, E, cl.m.<:) which serves to sling up the heart to the ventral side of the foregut.

The dorsal aorta is at this stage particularly instructive. Posteriorly it is represented by scattered cells, lateral in position, thus betraying their lateral origin. .Further forwards these have approached the mesial plane and form a pair of cellular strands. Further forwards still in the region of the first eight segments—— these have become still more nearly mcsial in position and ever part of their extent have undergone actual fusion to form the unpaired aortic rudiment.

About this stage the dorsal aortic rudiment is connected up to the vitelline network by a series of segmentally arranged vessels (segments 5-17) which had made their appearance about the twentysegment stage as segmentally arranged strands of cells.

The rudiments of the ducts of Ouvier make their appearance even earlier than that of the dorsal aorta, in the form of cells derived according to Mollier from the somatic mesoderm at the cranial side of the pronephros. These rudiments develop extensions iii a headward and in a tailward direction to form the cardinal veins. ’l‘he vessels of the head region develop in situ from the mesenchyme and the same may probably be said of the smaller vessels generally.

The Crossopterygian fish 1’ol_2/pterus (Graham Kerr, 1907) is, apart from its generally archaic character, particularly suitable for the study of the first beginnings of the vascular system owing to the fact that the long axis of the embryonic body is straight, so that horizontal as well as sagittal sections may be made passing through practically the whole length of the dorsal aorta during its early stages, when in its hinder portion it has not yet taken definite form.

The first conspicuous stage in the development of the dorsal aorta consists in the collecting together of irregular multinucleate masses of yolky protoplasm in a row beneath the hypochord (Fig. 174, A). Vacuolar spaces develop in these masses and foreshadow the aortic cavity. The masses of protoplasm become more closely aggregated into a cylindrical shape while the vacuolar spaces increase enormously in size and eventually flow together to form the continuous aortic cavity. In the specimen figured in 364 EMBRYOLOGY or THE LOWER VERTEBRATES on.

Fig. 174, B, the cavity was perfectly continuous towards the head end, while posteriorly it was still in the form of isolated vacuoles. The cells which form the rudiment of the dorsal aorta are from their coarsely-yolked character clearly derived ultimately from the primitive endoderm, but the question remains whether they are derived from the definitive endoderm directly or through the inter FIG. 174.-—l’m‘fiunH of horizontal sections through I‘ul‘:_/[1/wrI1.s' larv:u+ 0l'.~t1i.ge.s' ‘Z4 + (A) zunl 25 (B) showing the rudiment of the aorta in longitudinal section.

A, aortic rndinn-nt; m..r..-, vacuole-.~;.

mediary of the mesoderm. Such a section as that shown in Fig. 175 indicates that the latter is the case. The definitive endoderm shows a perfectly sharply defined surface, clearly marked off from the aortic rudiment, while the mesoderm of the sclerotome on the other hand is continuous with the aortic rudiment. We may say therefore with high probability that the aortic cells are derived from the sclerotome.

A remarkable feature has been noticed'in the development of the VI ORIGIN OF DORSAL AORTA IN POLYPTERUS 365

dorsal aorta of Polypterus which requires further investigation both in that genus and in any other Vertebrates in which it may be found to occur. In Polypterus in the stage immediately preceding that in which the aortic cells collect together the position of the future aorta is distinctly marked out by the arrangement of the delicate reticulum that is visible connecting up the various organ-rudiments of the larva. This reticulum is usually regarded as an artifact caused by the action of the fixing and preserving solutions upon the albuminous substances contained in the fluids of the embryonic body but the fact that it becomes arranged in this peculiar fashion to._ foreshadow the future aorta at once raises the question whether it is not really a reticulum of living substance.

The aortic cavity in Polypterru-s has been seen to originate by the fusion of intracellular vacuoles. The cavity is filled with clear fluid and T1 this condition persists even after the main channels of the vascular system are ].aid down. The blood is at first simply fluid or plasma without corpuscles. This plasmatic condition may persist even after circulation has commenced and the heart propels through the vessels simply the clear cell—less fluid. Here we find repeated in ontogeny an extremely archaic condition of the circulation. The Plasma’ beconles peopled with Ifie. ]75.—Portion of transverse section corpuscles comparatively suddenly. through Polyptcms of stage 25

The portions of vessel wall lying “1“;?"i“gt“:f :°1‘;;i°““l°_ft‘1‘° “°'l'°i° external to the lining endothelium rm ""9" (_ ) “C m °me('9°)' appear to arise from mesenchyme W’ °"‘°"“l’,,‘,"“n::Zc‘1,:;;,”d’_""'°t°""" cells.

SOURCE on THE CORPUSCLES.—'-The blood corpuscles are to be looked on, broadly speaking, as mesenchyme cells which have lost their connexion with their neighbours and float free

_in the plasma. Their precise sources in ontogeny ‘appear to be

various :—

(1) They can frequently be seen in process of being budded off by the wall of the embryonic blood-vessel into its cavity.

(2) In other cases the vessel with its contents is seen to arise as a solid mass of cells, those at the periphery becoming the wall (endothelium) of the vessel rudiment while those more deeply placed round themselves off, becoming separated by chinks containing fluid, and develop into corpuscles. This may be regarded as a 366 EMBRYOLOGY OF THE LOWER VERTEBRATES t CH.

modification of (1) brought about by a hurrying on of the develop ment of the corpuscles.

(3) In still other cases the cells of the mesenchyme reticulum in certain localities e._q. in the spaces between the tubules of the pronephros draw in their processes, round themselves ofi' and are carried away in the blood stream as primitive corpuscles.

This last mode of origin may account for the fact that the

Flu, 176.——-"Portions of transverse sections through Ela.~nnobrancl1 enibryos illustrating the origin of the heart.

A, ’1’m'pedo, stage with one gill cleft; B, 7'()1'pe(lo, stage with two gill clefts; (3, .1’r-istiu:-us, stag.-,-e with twenty-five st-.,«_»;ments; D, I-’ri..-I2'.u-rw.«:, stage with forty segments. (After figures by Riiekert ([388) and Mollier (l906).) cl.-nue, dorsal mesocardinm; mo, endoeardimn; and, endoderm of foregut: me, myoeardiunl; mes, mesodenn; N, notochord ; pt‘, pericardiuc cavity; 3.0, spinal cord; splc, splanelmoeoele.

corpuscles have been observed to make their appearance suddenly in large numbers in the circulating blood. The sudden setting free of large numbers of corpuscles is possibly due ‘to an epidemic of mitosis in the mesenchyme cells, as it is well. known that the onset of the mitotic process frequently induces a retraction of the processes of the cell.-body and its assumption of an approximately spherical shape. ’

It is not proposed to trace out in this volume the further development of the blood corpuscles—-—how the originally similar indifferent V1 ORIGIN OF THE HEART AND VESSELS 367

corpuscles become differentiated with further development into specialized strains—Erythrocytes or Red corpuscles and the various types of Leucoeytes; for an account of this in Lepidosiren the student may be referred to the beautiful memoir by Bryce (1905).

The mode of origin of the vascular system in the holoblastic vertebrates in general seems to resemble in its main lines that described above. The chief question that has given rise to discussion is one which in the opinion of the present writer resolves itself very much into a question of mere verbal expression—namely whether it is more correct to state that the first rudiments of the vascular system (or parts of them) are derived from mesoderm or from endoderm. When it is borne in mind that the mesoderm of the Vertebrate is essentially a derivative of the endoderm, it will be realized that the questioll is of minor importance whether or not special portions, such as rudiments of the vascular system in particular cases, lag behind the main 1nesoderm_ in their separation from the endoderm, so as to originate fron1 the latter directly instead of from the already differentiated mesoderm.

On1c1N or THE lI1«:An'r IN MEROBLA IC VEIm::s1iA'1‘Es.——The heart 111 its earliest stages shows in mere » astic vertebrates generally a set of conditions quite similar to those met with in holoblastic forms. In Elasmobranchs (Fig. 176, A) the first obvious rudiment of the heart is in the form of a number of cells of irregular shape which make their appearance on each side of the foregut, between it and the splanchnic mesoderm, from which latter they are apparently derived. As the foregut separates off from the endoderm of the yolk the irregular row of these cells shifts downwards and towards the mesial plane, so as to form a single elongated group underlying the foregut 1 (Fig. 17 6, B). The individual cells unite together and form a syncytial mass containing vacuolar spaces (Fig. 176, C, ens). Finally (Fig. 176, 1)) this elongated mass assumes a tubular form, its vacuoles coalescing and increasing in volume to form a wide cavity, while the protoplasm becomes thinner and forms the endothelial wall. The myocardium (mo) comes to surround the endothelial tube in exactly the way described for holoblastic forms.

In the Sauropsida again the phenomena are similar. To take as an example the Fowl: about the stage When two or three segments areformed isolated cardiac cells begin to appear on each side between the endoderm and splanchnie mesoderm. These cells increase in number and form a longitudinal tract on each side, the two converging and meeting anteriorly. As the foregut becomes constricted off the two cardiac strands come together from before backwards in front of the yolk-stalk though even at the eight-segment stage they are not completely fused. The endothelial rudiment now forms a loose syncytial spongework with large meshes containing

1 Ruckert (1888) believes them to be reinforced by cells derived from the ventral endoderm of the foregut. 368 EMBRYOLOGY OF THE LOWER VERTEBRATES C11.

fluid. The myocardium develops as in the other forms already mentioned.

ORIGIN or 'rm+: l’ERIPHERAI. BLOOD-VESSELS IN THE MERo13LAs'r1o VEl{'1‘l¢3BRA'l‘ES.-——ln those Vertebrates which have meroblastic eggs the concentration of yolk in the highly modified ventral endoderm accentuates the need of an efiicient transport system by which this food material may be taken up and conveyed to the yolkless and actively developing parts of the embryo. ln accordance with this we find that such vertebrates show a precocious development of a rich network of blood-vessels over the surface of the endoderm or yolk and this vitelline network allbrds admirable material for the study of the earliest stages in the development of peripheral bleed-vessels.

it will be convenient to consider in some little detail the early development of the vitelline network in the Fowl, material for the study of which is easily obtainable and which further has been worked out in detail by numerous investigators. Riickert has furnished an excellent modern account (1906) a11d upon it the following description is based: for fuller detail reference must be made to the original.

The first signs of bloed—vessel formation make their appearance extremely early, at a stage when the primitive streak is present but hardly any trace, or only a small stump, of the so-called head-process at its front end. The mesoderm has at this stage spread slightly beyond the edge of the pellucid area, especially posteriorly. Round its posterior edge the mesoderm assumes a mottled appearance owing to the development in it of small cell condensations the first trace of the blood-islands as they were called by Pander. These bloodislands are sometimes arranged in two separate series one on each side but more usually they form a U-shaped arrangement parallel and in close proximity to the posterior limit of the mesoderm.

In a slightly later stage the mottled area containing bloodislands -— the vascular area — is of a U-shape, extending through about the posterior half of the extent of the mesoderm. The bloodislands are less conspicuous in front, gradually fading away, as they do also on the side next the primitive streak. They are most strongly marked towards the external margin of the mesoderm and here they, as well as the whole sheet of mesoderm, are being added to by delamination from the endoderm of the germ-wall. The bloodislands, the first rudiments of blood-vessels, are simply thickenings and condensations of the mesoderm. They at first have the appearance in sections of occupying its whole thickness but later it is seen that each blood-island is roofed over by a layer of unmodified mesoderm -—demareated apparently by a simple process of splitting-olf of the superficial layer of cells.

As development goes on, the area of vascular rudiments spreads inwards into regions of the mesoderm which have been for some time completely separated from the endodérm by a well-marked split. VI ORIGIN OF THE VESSELS 369

Blood-islands developing in such regions are therefore clearly derivatives of the mesoderm and there is no possibility of the endoderm playing a direct part in their formation as might be the case peripherally in the region of the germ-wall.

The vascular rudiments become joined up by strands of cells to form a network and this network gradually spreads inwards, its extension being brought about by a progressive differentiation in situ from the mesoderm : there is no actual sprouting inwards of the already formed strands of the network as is suggested sometimes by the study of whole blastoderms and as was once supposed to take place.

Of the network of cell strands which traverses the rudiment of the vascular area the bulkier portions give rise to masses of‘ blood corpuseles surrounded by an endothelial wall, the more attenuated portions to endothelial tubes without any corpuscles in their interior. In the former case the superficial layer of the cells forming the bloodisland becomes raised up from the main mass of cells, fluid accumulating beneath it in spaces which are at first isolated but later become continuous. The flattened cells which are raised up represent the endothelial wall while the main mass of cells left behind represent developing corpuscles. It is to be noted that the endothelial wall separates from the mass of corpuscles first below ('i.e. on the side towards the yolk) and laterally, so that after fluid has accumulated in the rudimentary vessel the mass of corpuscles still remains attached to its, as yet undifferentiated, roof. The narrower strands between the main blood-islands and also all those in the pellucid area, except sometimes a few near its posterior end, give rise simply to endothelial tubes containing fluid plasma. As the circulation begins the masses of embryonic corpuscles gradually break up, first in the region of the sinus terminalis} the individual corpuscles being whirled away by the current and carried to the heart and thence through the circulation.

The origin of the vitelline network has also been investigated in Elasmobranchs (especially Torpedo) by numerous workers. It agrees in its main features with what occurs in the Fowl.

As regards the peripheral vessels in general, of the Vertebrata, we may say that they take their origin as chinks within the n1esenchyme filled with a clear fluid secretion (plasma). These chinks are at their first appearance in some cases clearly intercellular while in others they at first have the appearance of intracellular vacuoles. As has already been pointed out in dealing with connective tissue (p. 292) this difference though at first sight impressive loses most of its apparent importance when regarded critically. In this particular case the protoplasmic masses in which the vacuoles appear are as a rule multinucleate and it is clearly impossible to draw a sharp line of morphological distinction between spaces in such masses with

1 The topography of the vascular area will be found illustrated later in the special chapter on the development of the Fowl.

partially broken down walls, and the ordinary intercellular spaces of syncytial embryonic connective tissue.

GENERAL CONSIDERATIONS REGARDING THE MORP F rue

RATE EART.-—- may )6’. regarded as apr1m1 ivec aracteristic o oo -vessc s t rat their walls are contractile, peristaltic waves of contraction serving to propel the blood in their cavities. It is usually the case however in the more complex animals that this contractility becomes concentrated in one or more localized portions of the vessels, known as hearts, in which the vessel becomes much enlarged and its muscular coating thickened and rhythmically contractile.

In the craniate Vertebrates there is one heart present and it represents an enlarged portion of the Von tral vessel in the region immediately behind the gills. During ontogeny the heart still repeats the archaic evolutionary phase in which it was tubular in character. As development goes on the primitive heart, or cardiac tube, shows rapid increase in size within the pericardiac chamber of the coelome in which it lies. This chamber is relatively small in size and in the lower, fish-like, Vertebrates is bounded by rigid unyielding walls. The confined nature of this space in which the heart has been evolved has, by imposing restrictions upon it during its increase in size, exercised a profound influence upon the modelling of the vertebrate heart. It is therefore desirable to have a. clear idea of the general relations of the heart to the perieardiac cavity, during its increase in size, before attempting to study its development in detail in the various groups of Vertebrates.

The portion of vessel originally included between the anterior and posterior limits of the pericardiac cavity will be referred to here as

the primitive heart or cardiac tube. As development proceeds the increase in size of the primitive heart reveals itself in (1) increase

in length and (2) increase in diameter.

(1) As regards the former, the cardiac tube is at its posterior and anterior ends-——where it enters and leaves the pericardiac cavity respectively—firmly embedded in the tissues of the pericardiac wall. These ends being consequently in the lower, fish-like, vertebrates rigidly fixed in position, it has of necessity come about that the cardiac tube, while in the course of evolution it has increased in length, has lost its original straight form and has been thrown into a system of bends or kinks which have had an important influence upon the structure of the fully evolved heart. This bending process is repeated, though with obscuring of some of its detail, during ontogeny and it is an interesting morphological problem to endeavour to unravel the details of the process from the data of comparative anatomy and embryology.

Apparently the primary flexure of the cardiac tube is represented by a simple loop or bulging towards the right side of the body, which is visible in the embryos of most Vertebrates during early stages of heart development. VI MORPHOLOGY OF THE HEART 371

With increasing growth in length of the cardiac tube this simple curvature becomes converted into a double flexure the heart taking on a S-shape. Of the two curves which 1nake up the 5 one which has its concave side towards the head represents the original loop, while the other which is convex towards the head has developed in the portion of cardiac tube lying posterior to the primary loop. Of these two curves the one last mentioned, that which is morphologically posterior, is in an approximately vertical plane. The anterior or primary curve on the other hand shows much variation in position in different Vertebrates. While on the whole it still bulges towards the right side, as did the primary loop, the portion of it formed by the originally headward section of the tube comes in many cases to lie ventral to the other limb of the curve. In other cases this, originally anterior, portion of the tube lies for a time dorsal to the other, as is the case in balamamtlm. The difference will be appre ciated by comparing the

relative Positions of 0 and V Flo. 177.-—-A, diagram to illustrate the llexure of

- - the cardiac tube in the adult I.epz'u’o.s-[rem as seen

In F138’ 184’ A’ and 178' from the ventral side. The portion of thelliagram 01 3:11 the IOWCT Vicrtf?‘ above the horizontal line represents the conus:

}_)1-a,t',es in which the per}- the portion below the horizontal line would

- _ - - represent the rest of the heart on the assumption cardiac space 1S Sun bounded that this portion of the cardiac tube possesses a

by rigid illextensillle Walls similar curvature to that of the conus. Longill} iS the gI'011p Of Lung-fishes tmlinal lines drawn along the tube mark the

that Shows the heart.’ at originally dorsal (1)), ventral (V), right (R), and ‘ left (L).

the highest level of evolu- B I I I L t d I b L It - - , - s rows t ne spira wis mg pro ueec y s rang] ening non‘ Anll In Lorrelatlon out a tube possessing the same llexure as the ‘L-onus

 VVO filld that portion in diagram A.

in these fishes the kinking

of the cardiac tube attains its maximum. In a fully developed Lepidosiren (see below, pp. 376-378) the anterior portion of the cardiac tube (the “conus arteriosus ”) has developed a further 372 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

flexure in addition to those already described. The nature of this flexure is shown in the upper portion of Fig. 177, A, which represents diagrammatically the eonus of Lepiclosiren as seen from the ventral side.

The extreme anterior end of the tube, being fixed firmly at its exit from the pericardiac cavity, retains its primitive morphological position: its originally dorsal side is actually dorsal.

Traced backwards the tube is seen to become sharply bent upon itself in a headward direction, in such a way that the side of the tube which was originally on the left side comes to be ventral, as is indicated by the finely dotted line L in the figure. Tracing the tube onwards a second sharp flexure is found and the tube resumes its antero-posterior direction. This second flexure involves a complete reversal of the tube. Its originally right-hand edge (indicated by the coarsely dotted line 13), which had come to be dorsal as a result of the first flexure, is now ventral.

The changes in the position of the tube caused by the two fiexures may be summed up by saying that the half of the tube which was originally dorsal, and which remains dorsal at its anterior or headward end, has come to be situated on the right side at the posterior or ventricular end of the portion of the tube now under consideration (eonus). Similarly the half of the tube which at the headward end is ventral, has come to be at the ventricular end on the left side.

The lower half of the diagram represents the portion of cardiac tube which gives rise to the main part of the heart and it is to a certain- extent hypothetical, inasmuch as it does not rest on a complete series of observations, but it is clear that the morphologically right side of the cardiac tube, which is topographically ventral in the middle part of the figure, has to get back to its original right-hand position at the hinder end of the cardiac tube (which like the front end is firmly fixed in position), and it is reasonable to infer that the flexure of this portion, which gives rise to the atrium and ventricle, would be found, were its unravelling possible, to be symmetrical with the anterior flexure already dealt with.

It is interesting to take such a model as that represented in Fig. 177 and subject the eonus portion to a process of straightening out——such as would happen in nature if the eonus were to shrink in length, its anterior and posterior ends remaining fixed. The result is shown in Fig. 177, B. The eonus assumes a twisting in a right-handed spiral through three right angles. In the Amniota it will be found that the representatives of the eonus of the Lung-fish~ the roots of the great arteries, pulmonary and systen1ic—as they pass headwards from the ventricular part of the heart, twist round one another in just such a spiral.

(2) As regards the increase in diameter of the cardiac tube, it is characteristic that this does not take place equally throughout. At VI MORPHOLOGY OF THE HEART 373

certain levels the increase in diameter is much less pronounced than it is elsewhere, with the result that the tube appears to be constricted at these points wl1ile it bulges out between them. This development of a series of dilated portions of the heart-tube is the first step in its segmentation into a series of chambers. Of these chambers there are typically in the lower vertebrates four——sinus venosus, atrium, ventricle and conus arteriosus.

Allusion must be made in passing to an unfortunate confusion of nomenclature which is apt to prove a stumbling-block in the way of the student who is trying to ‘get his ideas clear regarding the morphology of. the heart.. The name conus arteriosus was first used, so far as the comparative anatomy of the lower Vertebrates is concerned, by Gegenbaur (1.866) who used it to designate the structure lying between ventricle and ventral aorta in Elasmobranchs and Ganoids, and characterized by its possessing a muscular, rhythmically contractile wall and by its containing longitudinal rows of pocket valves. 'l‘he name’ was introduced in order to accentuate the supposedly fundamental difference, already suggested by Johannes Muller (1845), between the structure in question and the bulbus arteriosus of Teleostean fishes. This latter is not provided with striped muscle in its wall, it is not rhythmically contractile: in other words it does not form physiologically a. part of the heart. Objections, and quite valid ones, have been raised against the use of Gegenbaur’s name from the side 01' Human anatomy, it being pointed out that the “conus” of the lower fishes corresponds rather with the “ bulbus ” of the human heart. Human anatomists working at the embryology of the vertebrate heart in consequence commonly use the_,na1ne bulbus cordis for the part oi the heart under discussion. Gegenbaur’s name however has come to be so universally used by comparative anatomists in reference to the heart of the lower vertebrates as to indicate the desirability 01 using it in a work on comparative morphology such as this. It will be understood then that the name conus arteriosus is used in this volume as equivalent to what is by many writers termed bulbus. cordis,‘ without prejudice however to the question whether or not Gegenbaur was justified in his belief that conus arteriosus and bulbus arteriosus are fundamentally distinct structures.

ELASMOBRANCHII.-—Tl1e Elasmobranch heart passes through the typical early stages, first as a straight tube (see Chap.,XI.), then as a tube which bulges towards the right side, and then as a tube with the characteristic S-shaped double flexure already alluded to (Fig 17 8). The three limbs of the 5 during further development become converted into (1) atrium with sinus venosus, (2) ventricle, and (3: conus arteriosus. The well-marked constriction which demarcates atrium from ventricle forms the auricular canal. The progress

1 I avoid in this book using the term trwncus arteriosus as it is unnecessary and

is liable to cause confusion owing to the want of precision with which it is commonly used. 374 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

towards the condition of the fully developed heart is marked mainly by the increase in relative size of the atrium and ventricle. Whereas however the increase in the size of the atrium takes the form mainly of a mere process ..of dilatation, that of the ventricle is accompanied by a much more marked thickening of its wall. This is brought about by the inner surface of the niyoeardium forming numerous projections into the lumen which, becoming more and more pronounced and interlacing and fusing with one another, form eventually a spongework and encroach considerably on the ventricular cavity round its periphery. The eiidocardiuin fits closely over the surface of each of these inyoeardiac trabeculae.

The physiological iiieaning of the formation of the trabcculac during the evolution of the ventricle probably lies in the fact that a bundle of muscle which has for its function the pulling together of the ventricular wall can carry out this function more efficientl if it runs straight between its two ends, in ot)lIier words if it is in the position of a chord to the curve of the ventricular wall rather than simply a portion of that curve.

Attention must now be directed to a very characteristic and important proliferation of the endocardiac cells which makes itself "apparent in particular regions. In the eonus such proliferation takes place along the course of four longitudinal lines,- giving rise to cells 1,.“178___,l.“,0Smg0mW which lie in the space b(?:l)W€eI1.0Ild0S3iLl‘dl11II1

.1m10p..,e,,t of ti... ;...;.,.., and myocardium. As this proliferation goes 0fAcrmt/_w7a8 see1_1i'r<>mtlur on the endocardium is eventually made to

S1‘;;3'6‘()A”°"1i“"°}“ bulge into the lumen as four . prominent M m_mm_ F “mm endocardiac ridges. In 1icant7zw,s (GegenW; .M, Sim, .v,e“(')_s_l'l_\,; V, baur, 1894), one of the four ridges——-that which ventricle, is ventral in p,osition——is reduced in size.

_ . _ In the auricular canal sinii.lar endocardiac prohferatious take place, one upon the headward and one upon the tailward wall respectively of the canal. Herc also each causes a prominent bulging of the endocardium into the cavity——-the atrioventricular cushion (anterior and posterior).

Both the ridges of the eonus and the atrioventricular cushions constitute a valvular apparatus in that, by the contraction of the niyocardium lying outside them, they can be jammed together so as to occlude the lumen into which they project. In both cases, as development goes on, they undergo metamorphosis into a purely mechanical and automatic valvular apparatus. In the eonus each ridge becomes excavated into a number of pocket valves (“semilunar ” valves), the cavities of which open in a headward direction. Greil and others explain these cavities as being produced simply by the backward pressure of the blood but it is advisable to regard VI HEART OF ELASMOBRANCHS 375

such simple mechanical explanations of developmental phenomena with S1].Spl(:1UI1. Similarly the atrloventricular cushions become excavated on their ventricular side and form the two atrioventricular

valves of the adult. A pair of laterally placed valves also develop

FIG. 179.-—Vie\\'s of l.l|4,- hc:u‘t. of 1.4‘/n'«I...gu'ren as seu1'1 from the nlorphologically n.-ntral side. (ll and (,9 afta-.1‘ J. Hol_n-rtson, 1913.)

A, stage :52; I3, .s‘l...-1;,-er :51 ; ('.‘, stage 3-3. «I, atrium ; (‘, «onus :u-t-uriosils; 1.4-, In-ft a-uricle; 1211., right auriclc; I", \'«-lltl‘irln.

at the opening of sinus into atrium but according to extant descriptions (Rose, 1890) arise merely as infoldings of the heart wall.

DIPNo1.—-The development of the dipnoan heart has formed the subject of a careful and exhaustive study by Robertson (1913) and her account of the (lm-“cl.0p1ne1'1t of the heart in Lepidoshen forms the basis of the following description. 376 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

The first peculiarity to be noticed in comparison with the heart of the Elasmobranch is correlated with the fact that the head and anterior trunk region of the embryo are bent downwards and closely applied to the surface of the yolk. As a consequence, the pericardiac space is reduced to a flat chink, the point of entrance and the point of exit of the cardiac tube being in close proximity to one another at its upper end. The result is that the cardiac tube as it grows in length assumes the form‘ of a flattened loop, first V-shaped (Fig. 179, A) and later Z5-shaped, the apex of the loop being directed in a ventral direction, and the originally posterior (tailward) limb of the loop (at) coming to lie to the left of the anterior limb (C). The cardiac tube becomes demarcated into the same set of chambers as in the Elasmobranch-——sinus venosus, atrium formed from the posterior limb of the cardiac loop, ventricle formed from the apical portion of the loop, and conus formed from the greater part of the anterior limb. The dilatation of the walls of the several chambers is not uniform. In the ease of the atrial wall the dilatation is most marked dorsally and more especially laterally: the posterior wall on the other hand lags behind in its growth and the result is that the sinu-atrial and atrioventrieular openings remain comparatively close to one another (compare adult condition as shown in Fig. 180). Similarly in the case of the ventricular portion of the heart the increase in size is mainly on the ventral and lateral sides, the dorsal wall lagging behind so that the communication between atrium and ventricle and that between ventricle and conus also remain in close proximity (cf. again Fig. 180).

After the demarcation of the chambers there come about two important changes in the general form of the heart-—the first is the assumption of bilateral symmetry on the part of the ventricle, correlated with a rotation of the ventral side of the ventricle towards the animal's left side. The other consists of a very marked increase in the length of the conus which, owing to the fixation of its two ends, is made to assume the characteristic double flexure already described and illustrated (see also Fig. 179, 0).

DEVELOPMENT or SEPTA IN THE HEART. —- By far the most important feature of the Dipnoan heart, as compared with that of the Elasmobranch, is that now, for the first time, there comes about that separation of the heart into an arterial and a venous half, which is so characteristic of the higher Vertebrates. In Lepidosiren this separation is inaugurated at a relatively early stage of development (stage 27)—-at a time when the cardiac tube, as yet, shows no indication of a division into chambers—-by proliferation of the cells of the endocardium on its outer surface. This takes place along a. line which passes along the posterior wall of the U-shaped heart from the left side of the sinu-atrial opening, through the auricular canal down towards the apex of the cardiac loop. As this proliferation goes on it causes the endocardium to bulge into the lumen so as to form a prominent ridge traversing the hinder wall of atrium and v1 HEART OF LEPIDOSIREN 377

ventricle. It is of morphological interest to notice that this atriaventricular ridge extends from its dorsal end not directly ventralwards but towards the animal’s right side, so that, if the ridge in question be taken as marking an originally longitudinal line along the wall of the cardiac tube, it indicates that this part of the cardiac tube has undergone a process of twisting like that of a left-handed screw, in other words a twisting of the kind which might be expected on the hypothesis that the flexure of the atrioventricular portion of the heart was originally that suggested in the discussion on p. 372.

A second endothelial proliferationtakes place along the atrial wall facing that on which the atrioventricular ridge has developed. The projection formed in this way grows towards and fuses with the atrioventricular ridge to form the atrial septum which divides the atrium into a larger right and a smaller left auricle. Owing to the left-handed position of the atrioventricular ridge at its dorsal end the sinus opens into the right auricle. The pulmonary vein as it develops comes to open on the left face of the septum, z'.e. into the left auricle.

The ventricle becomes similarly divided into a right and a left chamber1 the foundation of the septum consisting of the atrioventricular ridge already mentioned. I11 this case however Robertson does not describe any endocardiac proliferation mls-ct-mls to the ridge but says the septum is completed by muscular trabeculae growing towards and eventually fusing with the edge of the ridge.

The conus arteriosus is characterized by the development of a series of longitudinal endocardiac ridges similar in nature to those of Elasmobranchs. A conspicuous difference in detail is that each ridge is markedly discontinuous, the portions situated in the anterior and in the posterior section of the conus developing independently. We may take it that the ridges were primitively in the Vertebrata longitudinal and continuous and the secondary discontinuity visible in Lung-fishes and also in the higher Vertebrates may be associated with two probable causes: (1) interference with the development of the middle region of the conus by the flexure into which it is thrown, and (2) the tendency, as seen in Elasmobranchs and Ganoids, for the terminal members of the longitudinal rows of valves to become enlarged relatively to the rest. There can be no doubt that the longitudinal ridges as we see them in the Lung-fishes and the higher Vertebrates are revertive rather than persistent primitive features. In other words the ancestors of these Vertebrates passed through the phase of evolution in which each ridge had become converted into a row of pocket valves. This seems clearly indicated by the fact that the latter condition holds in modern Elasmobranchs and primitive Ganoids. But if so then the tendency for the terminal valves of the row to be specially developed in that earlier phase of evolution may show itself

1 The separation does not normally become quite complete in the adult either in the case of atrium or ventricle. 378 EMBRYOLOGY OF THE LOWER VERTEBRATES OH.

on reversion to the ridge condition in more or less great suppression or diminution of the middle portion of the ridge.

In the anterior section of the conus four longitudinal ridges develop, situated respectively on the right-hand side (1), dorsally (2), on the left-hand side (3), and ventrally (4). This anterior end of the conus retains as already explained its primitive position and we shall therefore always refer to the four ridges according to the position they have in this undisturbed portion of the conus as Right, Dorsal, Left and Ventral respectively, the adverb morphologically being understood before the adjective in each case. The right and left ridges make their appearance first and they alone become prominent, forming thin shelf-like structures which project right in to the centre of the cavity so that their edges overlap. For a short distance at the extreme anterior end they become fused together so as to form a continuous septum. The left ridge is comparatively short, tapering off posteriorly, but the right extends back through the anterior and middle section of the conus. At the point of ilexurc between middle and posterior sections there is a break during early stages but later on the ridge becomes continuous witl1.a portion of ridge which projects from the ventral wall of the posterior section of the conus. There is no reason to doubt that this is really part of the same morphological structure as that with which it is in line in the anterior section of the conus and we shall therefore term it the posterior portion of the right ridge. The whole of this right ridge forms what is often called the spiral valve of the conus.

The dorsal and ventral ridges of the anterior section of the conus are later than the lateral ridges in making their appearance and soon disappear again. The ventral ridge is especially feebly developed.

In the posterior section of the conus the right ridge——now ventral in position—is alone well developed. The other three appear as rudiments, they are at no time prominent and they become resolved into vestigial pocket valves which may still be detected in the adults. This latter fact justifies the conclusion already reached that during the ancestral history of the Lung-fishes a stage was passed through during which the conus was provided with longitudinal rows of functional pocket valves, in other words that the primitive ridges seen in the conus of the modern Lung-fish are revertive rather than persistent. VALVES OF THE HEART.——The sinu-auricular opening is guarded on the right side by a valve. This develops out of the inpushed fold of the cardiac Wall in the constriction between sinus and atrium. The atrioventricular opening is guarded by a highly characteristic bevelled plug (Fig. 180, AV.p) which when the ventricle contracts is pulled downwards so as completely to occlude the opening. Developmentally this plug arises as a thickening of the atrioventricular ridge. This ridge, which, as already indicated, VI HEART OF LEPIDOSIREN 379

forms the common foundation of auricular and ventricular septum, traverses the auricular canal, projecting into it from behind so as to give the atrioventricular opening a horse-shoe shape. It is the part which lies above (dorsal to) the opening which becomes thickened and eventually assumes a cartilaginous character to form the plug. The plug is to be regarded as the homologue of the posterior atrioventricular cushion of Elasmobranchs but Robertson failed to find any trace of an anterior cushion.

The conus in the completely developed state is characterized by the absence of the functional valvular apparatus found in its homologue in other Vertebrates. On the one hand the endocardiac ridges, functional in the young Elasmobranch or Ganoid, are no

I

Fm. 180.-—-Heart of an adult Lepi«.losiren with the right side removeal. (After J. Robertson, 1913.)

AV.p, atrinventricuhw plug; ¢i'.'v, corona:-_v vein (cul'.); d.(j'. ducts of (Juviur; pm, pulnum:u'_\' vein; -p.V.4:, pu.~;t.orio1- \'t‘l'l{1('.‘l\'{l. at its opening into the sinus \'¢-.n0.\'uS: .~'..-L atrial sq-pt.um ; .~'.l', \'eut1'icul:11' h'l!])l-lllll; .~-. 1', sinus \'a-.nosus (its opening into the rigI1t:uu'ix-lo- illulit-:|lI'!l lay an :n-1-nu‘); .-p, .~:piI':Il \':l1\'4!: Ill, V1, aortic arches out near their ventral ends.

longer in a condition to fulfil their original function and on the other the pocket valves are here vestigial.

MUSCULARIZATION or rm: HEAnr.——Bud-like projections from the myocardium grow into the cavity of the heart, meet together and in the ventricle form a muscular spongework of the same type as that seen in Elasmobranchs. In the case of the ventricle numerous trabeculae arising in this way converge upon the free edge of the atrioventricular ridge and become continuous with it. As development goes on this spongy mass of trabeculae undergoes condensation arid acquires the solid character of the fully developed septum. The septum is continuous dorsally with the atri'ove11tricular plug and it forms a muscular apparatus by which the plug is pulled down so as to fit into and close the opening. The myocardium of the auricular 380 EMBRYOLOGY OF THE LOWER VERTEBRATES CII.

canal and of the conus never forms a spongework but remains as a compact layer of muscle. In the case of the conus this muscular coat is very feebly developed in the middle and cephalic portionsin which fact we may probably recognize a degenerative feature seeing that in the Elasmobranch the conus musculature is well developed right to its front end. The meshes of the ventricular spongework, as development goes on, come to spread somewhat round the auricular canal and round the ventricular end of the conus, so that each of these structures has the appearance of being drawn into the ventricular cavity.

SAUROI*sI1)A.~——'l‘hc most exhaustive account of the development of a Sauropsidan heart is that dealing with Lucerta by Grreil (1903) and upon it the following description is based. ln its early stages the heart passes through the familiar tubular form and becomes bent, first bulging in a simple curve towards the right and then assuming a double S-like curvature just as in the Elasmobranch. About stage 17-18 the constriction of the heart into sinus venosus, atrium, ventricle and conus becomes apparent—the three last mentioned chambers bulging outwards between the grooves which limit them. The atrial portion does not in these early stages take up the purely dorsal position seen in the Elasmobranch or Lung-fish but remains for a time well to the left.

The conus, in its early stages, much reduced in relative size as compared with that of the Elasmobranch, undergoes a marked increase in length, which causes it to assume a bayonet-shaped curvature in which we may see a reminiscence of the sharp double flexure seen in the conus of the Lung-fish. In the Lizard however this curvature of the conus is merely temporary. As development goes on the increase in length of the conus instead of being more pronounced than that of the heart as a whole becomes less so with the result that between stages 21 and 26 the anterior flexure of the conus becomes pulled out and replaced by a right-handed spiral twist.

DEVELOPMENT or SEPTA.-—Tl1e septation of the heart is inaugurated by the appearance of localized proliferation of the endocardiac lining. In the auricular canal, which runs in an antero-posterior direction rather than dorsi-ventrally, owing to the atrium lying anterior to the ventricle instead of dorsal to it as was the case in the Lung-fish, there develop two endocardiac cushions, one dorsal (posterior), the other ventral (anterior). Of these the ventral or anterior one which was not apparent in Lepidosiren is well developed and is continued as an endocardiac ridge round the anterior (headward) wall of the atrium on to its roof (compare Fig. 183, C, was). As development goes on this projects more and more prominently into the cavity of the atrium and forms the main part of the septum between the two auricles. By about stage 26 it has grown halt’-way across the atrial cavity, and by about stage 29 it reaches the auricular canal. While in a sense the atrial septum is now complete it is not so physiologically VI HEART OF LACERTA 381

as secondary perforations have made their appearance in the septum so as to keep the two auricular cavities in free communication. The two endocardiac cushions of the auricular canal become joined together by a bridge of endocardiac tissue which forms the free edge of the auricular septum. This is followed by a complete fusion taking place between the middle parts of the two cushions, so that the atrioventricular opening becomes completely divided into a larger right and a smaller left portion. The, at first, thick mass of tissue which separates these two openings becomes gradually converted into a thin plate, situated in the plane separating atrium from ventricle, and therefore perpendicular to the plane of the atrial septum. This plate is divided sagittally into a right and a left half by its line of attachment to the septum. The free edge of each half is concave and projects freely into the corresponding auriculoventricular opening—forming the mesial or septal valve of that opening.

In the meantime a new endocardiac cushion develops on what were the right and left sides of the auricular canal. These also become thin flaps and form the lateral auriculoventricular valves. It will be noticed that there have developed round the original atrioveutricular opening few proliferations of endocardium—--the same number as was found in the conus of the Lung -fish and as will be found in the conus of the Amniota, thus supporting the idea that there are four longitudinal endocardiac ridges potentially present throughout the cardiac tube of the higher Vertebrates though they may become actually apparent only in the conus region.

In the Lizard the atrioventricular ridge, which was so conspicuous in the Lung-fish, has practically become reduced to the portion lying within the auricular canal—the dorsal (posterior) endocardiac cushion. The ventricle is undivided.

The conus on the other hand undergoes a complete and somewhat complicated process of septation. This is inaugurated by localized proliferation of the endocardium to form longitudinal ridges. As in the Lung-fish these arise discontinuously there being distinct anterior (headward) and posterior rudiments. Anteriorly the normal four ridges develop, the dorsal and ventral appearing in this case at an earlier stage (17 or 18) than the lateral ones. Towards the ventricular end two ridges first make their appearance in a dorsal (ridge B, Greil) and ventral situation (A, Greil) respectively. Of these the ventral one becomes eventually continuous with the right-hand anterior rudiment. It clearly corresponds with the similarly situated ridge in the hinder portion of the conus of Lepirlrisiren and like it is to be interpreted as the hinder portion of the morphologically righthand ridge. The ridge mls-ft-vie to that just mentioned, here dorsal in position, would similarly represent the hinder portion of the morphologically left-hand ridge. Later on a small and transient ridge (0, Grreil) makes its appearance on the right-hand wall of this hinder portion of the conus and this would represent the hinder portion of the morphologically dorsal ridge. 382 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

As development goes on the wall of the conus becomes changed in histological character, its striped muscles become replaced by smooth and in general it takes on the ordinary features of arterial wall so that it resembles a portion of the ventral aorta rather than of the heart. As may be seen in a living embryo this histological change is accompanied by a physiological one, for the rhythmic contractions of the heart are seen now to extend forwards as far as the anterior limit of the striped Inuscle but no farther. Altogether the superficial ‘appearance is just as if the ventral aorta (“ truncus arteriosus ”) were extending backwards at the expense of the conus, and the word truncus is frequently used to include the whole as far back as the limit for the time being of the smooth non-striated muscular wall. It must however not be forgotten that in the strict morphological sense all that part of the heart is conus which corresponds to the conus of Lepvldosiren. 'l‘he special criterion which identifies it is

Fm. 18l.—~—l)iagraminatic transverse sections through conus of I.a.ce~rta ( A) and Gallus B to show the endocardiac rid res and the iocket valves. la

1, morphologically right ridge; 2, dorsal; 3, left; and 4, ventral. a and I2, problematieal ridges discussed in text; 12, pulmomuy cavity ; 8, main systemic; (.5, left systemic cavity.

the appearance of double flexure or the resultant spiral coiling during its development. The muscular coating, so characteristic a feature in the vertebrates below the Amniota, is associated with a definite type of functional activity: in the Amniota that type of functional activity has disappeared and with it the characteristic type of wall. The ventral aorta is in its hinder portion, where it becomes continuous with the front end of the conus, divided1 into a dorsal (pulmonary) and a ventral (aortic) cavity by a horizontal septum and this is prolonged backwards along the wall of the conus by the right and left ridges. Of these the right is very large, it projects across the lumen and gradually fuses with the left ridge (Fig. 181, A). This ridge (3) is low and double and it is with its dorsal portion,

i.e. the portion next the dorsal ridge, that the fusion takes place. By the spreading backwards of this process of fusion of the right and left ridges the horizontal septum of the ventral aorta becomes prolonged back as a septum in the conus——no longer horizontal

1 See below, p. 393. VI’ HEART OF LACERTA 383

however but spirally twisted owing to the twisting of the conus already mentioned. Parallel with and preparatory to this process of fusion the distal ridge rudiments spread backwards, pursuing a spiral course and thus making evident the spiral twisting of the conus as a whole. It is interesting to notice that the line of insertion of the ridges, and therefore of the septum formed by their fusion, becomes marked on the outer surface of the conus by a distinct incision--a preliminary step towards the complete splitting of the conus in the plane of the septum which takes place in Birds and in Mammals.

In addition to the dividing of the cavity of the conus into a pulmonary and an aortic portion in the manner just described there takes place also, in the Lizard, a splitting of the aortic portion into two parts, corresponding to the right and left halves into which the systemic portion of the ventral aorta is divided} The septum separating these becomes prolonged backwards at its hinder end, on the one hand, into the ventral ridge of the conus (Fig. 181, A,‘4) and on the other into a quite similar ridge developed upon the surface of the septum which separates the aortic from the pulmonary (Fig. 181, A, 1). These two ridges facing one another across the aortic cavity gradually extend backwards and undergo fusion just as in the other case so as to form a complete septum dividing the aortic cavity into two (S and l.s).

In this waythen the original conus becomes replaced by a set of three tubes twisted spirally round one another, forming the roots of the two systemic aortae and of the pulmonary artery, still however enclosed in a common wall.

VALVES on THE HEART.———-The right and left valves which guard the opening from sinus into atrium are formed simply by the exaggeration of the fold of the cardiac wall which delimits these two chambers from one another. The origin of the auriculoventricular valves has already been described. The pocket valves of the systemic aortae and pulmonary artery are derived from the endoeardiac ridges of the conus as in the Elasmobranch. According to Langer (1894) the outer valve in each of the three vessels (pulmonary artery, left systemic aorta, right systemic aorta) are derived from the Dorsal, Left and Ventral endoeardiac ridges (2, 3 and 4) respectively, while the inner valve in all three is derived from the hypertrophied Right ridge (1), which with its outgrowth takes part in the formation of all three vessels (Fig. 181, A).

The question as to whether or not the pocket valve is formed from the extreme ventricular end of the conus ridge, or whether on the other hand a considerable portion of this end of the conus with its contained ridges becomes incorporated in the ventricle as maintained by Langer and Greil does not appear to the present writer to be satisfactorily settled. It is advisable that the point should be re-investigated upon abundant material.

GALLUS.-—-(Figs. 182 and 183.) The most detailed investigations

1 See below, p. 393. 384 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

of the development of the Bird's heart are those of Greil, of which unfortunately there is available so far only the abstract given by Hochstettcr (1906). As we should expect from the close genetic relationship between Birds and Reptiles there is a close correspondence between the general features of the development of the heart in the two cases. It will suflice then to draw attention to the more i1npu1‘l):'mt points in which the development of the Fowl’s heart has

FIG. 182.—Il1ustrating the development of the heart in the fowl. (After original drawings by (_i1'cil.)

at, atrium; h.u..f, bulho-auricular fold :'«-_, l-onus; Lu, ll-ft zuu-iule; Li, left im|mnin:u«- urlo-ry; I.-,2, left pulmmnat-_v; I.l-', left. \‘l'ntl'i('l4-; r.vI, ri;_'l1l Ill1l‘l(,‘ll'f, r.'i, 1-iglnt. lllll()llllll{l.tl'. :u-Lory; r.p, right. pulmonary; r.l-’, l'i;.',‘l1t vent:-ii.-la-; .-..l, s_ysl.<-urn-:u.1-t:.a.

been found to differ from that of Lace-rta. So far as external form is concerned the most striking difference is that the sinus venosus loses its identity as a distinct chamber of the heart. It becomes as it were incorporated in the right auricle, all except its left portion which persists as the cardiac end of the left duct of Cnvier or anterior Vena Cava.

An important advance upon the condition in Lacerta is found in the division of the ventricular part of the heart into a right and a VI HEART OF BIRDS . 385

left ventricle. This (li_visi<.m. comes about in a somewhat complicated fashion the chief points in which, judging from Greil and Hoolistetter’s descriptions and figures, appear to be as follows. The ventricular portion of the cardiac tube is at an early stage encroached upon by the deep (“ bulbo-auricular ”) fold which separates the conus

Fm. lRfi.--~--St.:1._r_ges in the development of the heart of the Fowl, viewed from the right Shite. 'l‘l1crigllt wall of the heart has been removed in each case. (After original ilI'it\\'lll;:_'S lay (‘nu-.il.)

.1, 1.’, proxini.-:1. «_-nals oi‘ l'l(l".."'I‘S of eonus; uI.._-. atrial .-.«-.pl.un'1; h.n,.r, bulbo-auricular ri«l_:_-,-e ; a._-, r-onus arteriosus; i.-v./; llll.(‘i'\‘l‘lIl..l‘iC|ll:ll' 1'<.n':I.m«-u; r'.r..~', tralu__-<-.ul:u- portion of ventricular septum; La, left auriele; p, pu|mcnar_\- ;u-1..-,-_y; -r.u, ri_v_,--ht. auriclo-; r.r.c, rut. surface of endocardiae tissue of atrioventricular opening; 2-.1’, cavil _y oh-i;_,rl'it ventricle : .-, .~:ystennic aorta ; V, ventricle ; 1, ‘2, 3, conus ridges.

from the atrial part of the heart (Fig. 182, A, b.at.f). The encroachment of this fold gives the ventricular part of the tube a squat

U-shape. As the ventricle dilates the extent of the encroachment becomes reduced and the fold may now be called the bulbo-auricular

ridge of the heart-lining (Fig. 183, A, b.a,.'r). During the fourth

day this becomes extended tailwards along the ventral wall of the VOL. II . 2 0 386 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

ventricle (Fig. 181?», B, C, b.a..'r) the ridge and its extension forming the rudiment of the anterior portion of the ventricular septum. The remaining and larger portion of the septum arises otherwise, from a local exaggeration of the muscular trabeculae which in the Bird, as in lower forms, sprout into the ventricular cavity so as to convert the peripheral portion of that cavity into a sponge—work. This spo11gc—work becomes exaggerated in the prominence and thickness of its trabeculae along a plane marked on the external surface of the heart by a distinct groove-——thc interventricular groove (Fig. 182, (J). This trabecular part of the septum (Fig. 183, O, i.'v.s) is at first loose and spongy but it gradually becomes condensed, at first along its thickened free edge, and loses its spongy character. It gradually extends forwards and becomes continuous on the one hand with the bulbo-auricular portion and on the other with the septum of the auricular canal. The ventricular cavity is now divided into a right and a left chamber except at its anterior end where there remains an interventricular foramen (Fig. 18;}, 1), lt will be realized that but for the presence of this foramen the blood could not circulate, as the only means of exit from the ventricle the opening leading into the conus——lies completely on one side (right) of the original bulbo-auricular fold and, therefore, of the ventricular septum of uhich the fold in question forms a part. As a matter of fact this interventricular foramen never disappears, though it loses its right to that name, for it becomes continued as a groove over the surface of the mass of endocardiac tissue lying between it and the conus. Eventually this groove becomes overgrown by its edges and converted into a tubular channel, continuous on the one hand—— through the original interventricular foramen——-with the cavity of the left ventricle and on the other with the systemic or aortic cavity of the conus. rfhis tubular channel persists in the adult condition——as the communication between left ventricle and systemic aorta. _

Finally, before leaving the intcrventricular septuln, it has to be mentioned that dorsally (Fig. 183, 1)) it becomes continuous with the bridge of endocardiac tissue which divides the atrioventricular opening into a right and left half, the ventricular side of this bridge growing out to meet the trabeeular part of the septum. The atrial side of the bridge is continuous with the atrial septum, which develops here as in Lacerta, and the result is that the main part of the heart is now divided into two halves, the left auricle opening into the left ventricle and the right auricle opening into the right ventricle (Fig. 183, E). It is to be noted however that secondary perforations appear in the atrial septum (Fig. 183, E, at.s)—so as to allow the systemic blood which enters the right auricle from the sinus venosus to reach the left auricle, and through it the_left ventricle, without having to traverse the pulmonary circulation during the period before the lungs are functional. VI HEART OF BIRDS 387

of its development the ventricular portion of the heart undergoes a certain amount of rotation, the right ventricle becoming displaced somewhat towards the left side, ventrally to the left ventricle. The result of this is to undo to a small extent the spiral twist of the oonus.

'l‘he conus of the Fowl develops typical endocardiac ridges, here however only three in number, and the individual ridges retain a more nearly primitive condition in that in early stages they are not so completely divided into two distinct rudiments, while in later stages two of the three are obviously continuous. One of these ridges is clearly the morphologically Right. Here again it is Inuch enlarged (Fig. 181, B, 1) and grows right across the cavity to form a complete septum’ between the pulmonary (7)) and systemic portions of the cavity.

Regarding the identity of the two other ridges there is some doubt. They are identified by Grreil as the Dorsal (2) and Left (3) while the Ventral (4) is supposed to have disappeared. It appears to the present writer however that the possibility should be considered whether they do not together represent the Left ridge, with which in Lacerta the free edge of the enlarged Right ridge comes in contact and which in the latter animal shows an incipient division into two parts by a longitudinal groove.

’l‘he septum formed by the enlarged Right ridge follows a spiral course, its line of insertion being indicated by a spiral groove on the outer surface of the conus. In the Bird this groove gradually deepens into a slit which splits the septum into two halves and as a consequence divides the oonus into two separate vessels which course spirally round one another———the roots of the pulmonary arteries and the systemic aorta respectively. N o vestige of a septum subdividing the aortic cavity has so far been described.

V ALvES.——l’ul1nonary artery and systemic aorta are each provided -with three pocket—valves at their ventricular end. These arise in the manner indicated in Fig. 181, B. Each vessel receives a valve split off from the enlarged Right ridge. The pulmonary and the systemic cavities receive further a valve split off from the endocardiac thickenings marked a and b respectively. Following Grreil these would be attributed to the Dorsal and Left ridges, while accepting the alternative interpretation suggested above they would both be referred to the morphologically Left ridge. There remains a third pocket-valve in each cavity. That in the systemic cavity no doubt represents the otherwise missing Ventral ridge, while if a and 12 together represent the Left ridge then the third poeket~ valve of the pulmonary cavity would represent the Dorsal ridge. As there is no reason to doubt the reliability of a pocket-valve as evidence of a once existing endocardiac ridge we should be driven—— if we reject the explanation here suggcsted—-to assume the former existence of an additional ridge between the Right and the Dorsal and there seems no justification otherwise for doing this. 388 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

The pocket-valves are stated not to develop at the extreme hinder limit of the ridges, the septum stretching back beyond them to become continuous with the interventricular septum.

In the right auriculoventricular opening the inner or septal valve is not developed, the ventricular septum fusing with what in Lacerta becomes converted into the valve in question.

The main features of heart development having been illustrated from these three dilfercnt groups, Elasmobranchii, Dipnoi and Sauropsida, it will be convenient now to indicate the more important peculiarities which have been detected in other groups of the lower Vertebrates. It should be understood however that in the case of several of these, such as Cyclostonies, Ganoids and even Amphibians, apart from Urodeles, our knowledge is still fragmentary.

In I’ol_2/plerus (_ Graham Kerr, 1907) the cardiac tube when in the form of a loop shows a similar displacement to that which occurs in fM])li([0-5'?:'}‘(’-7'!»-—-lille lower end of the loop being pushed forwards in front of the yolk. In this case however the displacement has gone farther than in the Lung-fish so that the cardiac loop is completely inverted-its apex being directed forwards, while the ventral aorta passes OH’ in a tailward direction. A similar displacement occurs in Teleosts.

'l‘he conus of Ganoid fishes shows the usual endocardiac ridges which become converted into longitudinal rows of pocket-valves as in Elasmobranchs. In Poly/pm-us these ridges are six in number, alternate ones being much reduced in size, with the result that in the adult three rows of large pocket-valves alternate with three rows of small ones. In all these fishes the endocardiac ridges and their resultant rows of pocket-valves run straight along the conus and there is no reason to doubt that this is the primitive condition. To determine the primitive number of the ridges in Fishes more research'is needed although there is little doubt that four was the number present in the primitive Tetrapocls.

In the Teloostcan fishes we find in place of the conus arteriosus the structure known as the aortic bulb. As already indicated (p. 37 this is distinguished from the typical conus by well-marked histological and physiological difi'erenees. And it is frequently regarded as being morphologically a part not of the heart but of the ventral aorta.

If however we take, as we are probably justified in doing, the point of exit from the pericardiac cavity as being relatively fixed and as marking the headward limit of the cardiac tube or primitive heart, then it becomes clear that the aortic bulb, lying as it does within the pericardiac cavity, is really a portion of the primitive cardiac tube and of that part of it which lay between the ventricle and the ventral aorta-——in other words the conus arteriosus. What has happened in the evolution of the. Teleostean heart is in all probability entirely analogous with what has taken place in the ‘ curve lies 111 a nearly vertical plane

VI HEART 389

Amniota, namely that the conus arteriosus has gradually lost its power of rhythmic contraction while pawl passu its myoeardiac coating of striped muscle has degenerated and its primitive histological characteristics have been replaced by others resembling more closely those of the ventral aorta.

.During ontogeny it would appear from Hoyei-’s work on Salmo (1900) that the conus in the embryo possesses the characteristic features——a layer of striated muscle in its wall, and longitudinal ridges (two in number) projecting into its 1uu1en——and differs from that of an Elasmobranch merely in ' the fact that these features do not extend throughout the whole of the distance between the ventricle and the anterior limit of the pericardiac space, but only through about the posterior half of that distance. In the adult the two ridges are represented by the two pocket-valves.

In Urodele amphibians (Salamandm—-Hoehstetter, 1906) the heart during the period when it is in the form of a tube with an S-like curvature is conspicuously different in appearance from that of the Vertebrates already described, owing to the fact that the two curves of s the 8 lie in different planes from those which they occupy elsewhere. The morphologically posterior or tailward curve lies here nearly iii the

horizontal plane While the anterior FIG.184.—-Views of developing heart of ' Salamamlm as seen from the ventral side. (After Hochstetter, 1906.)

——the limb of the curve which will become conus dorsal to the at,‘atriu111;c,eonu's‘u.rte;.ios11s;d.(.',.duc.t ventricular portion, so that it is ;’if,,,(,’,,“,V.,',‘,i,’,,;,,:f,";‘ mm cm’ 8")’ hidden in a View of the heart from

the ventral side (Fig. 184, A). Special interest is lent to this curvature of the atrioventricular portion of the cardiac tube by the fact that it reproduces accurately the type of curvature which we inferred as being present in this portion of the vertebrate heart in our general discussion of its morphology (compare Fig. 184, A, with lower portion of Fig. 177).

As development proceeds the left-hand end of the ventricular portion, z'.e. its actually headward end, swings ventrally and tailwards so that the long axis of this portion of the heart comes to be perpendicular to the sagittal plane of the body (Fig. 184, B). As the ventricular part of the heart shifts backwards the conus becomes

visible in a view from the ventral side. The backward shifting 390% EMBRYOLOGX or THETLOWER VERTEBRATES on.

continues‘ imtil the ventricle comes to lie ventral -to the sinus (Fig. 184, 0) instead of well in front of it as it did originally. The ventricle is now on the tailward side of the atrium which bulges out on each side, its ventral wall fitting closely to the dorsal side of the conus. The sinus becomes marked oft‘ from the atrium by a constriction, which deepens most markedly on the right side so that tl1e sinuatrial opening becomes displaced towards the left. It is to be noted that during these stages in development the portion of the ventricular wall lying on its concave (anterior) side, undergoes relatively very slight increase in size. The result is that the ventricle bulges in a tailward direction and the openings by which it comniunicates with atrium and conus respectively remain relatively closely approximated together. ‘ i

The atrial septum arises as a ridge or fold of endocardium which,

as in the Sauropsida, projects into the lumen from the anterior (head— i

ward) wall. The concave free edge of this grows towards the atrioventricular opening while its base of attachment spreads, on the one hand ventrally until it becomes continuous with the anterior atrioventricular ciisliioii, and on the other dorsally and backwards, to the left of the sinus opening and to the right of the pulmonary vein opening, till it becomes continuous with the posterior (tailward) atrioventricular cushion. The conspicuous openings in the atrial septum known to exist in adult Urodeles though not irr‘Anura are secondary perforations. e t

The conus develops discontinuous ridges. Of these there are four anterior riidiments of which the Dorsal and Ventral develop first and the Right and Left later. Of posterior rudiments only three have been described but that the missing one is at least sometimes present is shown by the adult arrangements of the resulting pocket-valves in rliiiereiit Amphibians. Thus four pocket-valves have been observed in the posterior circle..i;»n specimens of Strait, Nectwrus, the Axolotl, and Salmnandm (Boas). In the anterior circle cases are known of one of the four valves being reiiuced in size (Sz"ren, Axolotl, Triton, Sa.lama./ndm), vestigial (Pqlpa), or_:§i.g6ne entirely (Proteus, Rana). Other cases occur in which an additional valve makes its appearance through one of the original ones becoining split (Nectwrus). All these variations in the pocket-valves of the adult are of importance in relation to the embryonic ridges which the valves represent. Details will be found in Boas (1882).

Of, the anterior rudiments the right-hand ridge (1) is prolonged backwards as thevspiral fold which projects across the lumen and disides it iinpepfectly into an aortic and a pulmonary cavity.

In some cases the s iral fold apparently makes an abortive attempt’

to pursue the course of development which it went through in the ancestral fish-like form, as it segnients up into a row of little knobs each of which, we may take it, represents a pocket-valve (Tm'ton pwnctatus, '1‘. cmlstatus). In other cases (Necturus, Ooecilia‘) the spiral fold has in the adult completely disappeared (Boas- J-R823. ’ s -VI HEART OF ' REPTILES 391

The hearts of Reptiles in general agree closely in their developmental features with that of Lcwerta. The most important variations are seen in the Crocodiles (Hochstetter, 1906*‘). The conus here is- of interest in that it still repeats with particular clearness the sharp double flexure seen in the .l‘)ip1'ioar1 (F 185).

The ventricular portion of the heart becomes completely divided" into a right and left ventricle by a septum which is formed for the most part from trabceular projections of’ the m'yoeardi'um but in part also from the endoeardiac bridge which divides the atri.oventricular opening as in the Bird. _ This septum becomes quite complete, the interventricular foramen which exists for a time closing up and the interventricular septum becoming continuous. with the aortic septum of the conus. This is rendered a possible physiological arrangement in the crocodile by the fact that here the opening from the ventricle into that cavity of the conus which is continuous with the right systemic aorta, is farther to the left than in Lizards, While on the other _hand the right atrioventricular opeuin g is farther to the right. The result is that the opening of the right systemic aorta... and the leftauriculoventricular opening lie to the left of the s_ep'tun1;while the openings of the left aorta and pulmonary artery together with . _ . the right auricuhiventricular opening lie to F“"' 185';“’“H°“"t"l“{"br"°“'°

-_ 7. _ __ w o , Crocouhle. (Alter H0('-llthe right of the septum. buch diiierences Stetter’19O6*.) in the position of the ventricular openings of the great arteries are regarded by Hoolistetter as due to varying degrees of i11c01'po1‘ati01.1 of the obliquely placed conus into the ventricle. i

The foramen of Panizza, that re1narkable_ communication which exists in the crocodile between the two systemic aortae close to their point of exit from the ventricles, arises as a secondary perforation of the aortic septum comparatively late in development just before the closing of the last remains of the interventrieular foramen. _

The splitting up of the coffiis into independent vessels remains as a rule in the Reptiles, as in Lacerla, in an incipient condition, indicated merely by a slight grooving of the surface. In Ophidia however the superficial groove becomes -so deepened as to split the conus completely‘ into an independent aortic and pulmonary root as in Birds. _ , .

Bassing in review the broad features of lieartdevelopliient in the lower Vertebrates the following general principles seem to emerge :(1) The primitive heart or primitive cardiac tube is that portion of the ventral or subintestinal vessel included within the limits of the pericardiac coelome.

(2) Annular segments in the wall of th.i§,i.§,,t11sbe lag behind in increase of diameter so that the tube beg‘oi;1e§f;»§§pnstricted into a

Ir, conus arteriosus; I-", ventricle. 392 EMBRYOLUGY OF THE LOWER V ERTEBRATES cu.

series of chambers~——si1ms venosus, atrium, ventricle and conus arteriosus——-the originally peristaltic waves of contraction tending to become reduced at each constriction so that the chambers come to contract in series.

(3) The primitive valvular apparatus consists of a series of longitudinal ridges. These are best marked in the conus where they were originally at least four in number. The presence of six in 1’ol;z/pterus and the occurrence of more than four rows of valves in various other relatively primitive Ganoids and Elasmobranchs suggests that the number may have been greater. Individual ridges may in various forms become reduced or disappear, while in other cases apparently an increase in number may take place. The ridges show very generally a tendency to develop from two distinct rudiments, an anterior one and a posterior one, but the continuity of the rows of valves in the lower fishes indicates the probability that this discontinuity is secondary. In the portions of the heart behind the conus the1'c is no complete series of ridges like those in the conus though it is possible that vestiges of such ridges are represented by the endocardiac cushions and septal rudiments.

(4) Using the ridges of the conus as marking morphologically longitudinal lines it is seen that the conus has in the Lung—1ishes undergone a double llexure of such a kind as to produce ‘wlfen straightened out a right-handed twist of the conus through approximately three right angles. 3

(5) In tctrapodous Vertebrates the conus shows a similar righthanded twist and this is adequately explained by the relative reduction in length which the conus has undergone if it be assumed that there was an ancestral condition resembling that of the existing Dipnoan. ...

(6) From the Lung-fishes upwards the originally right-hand ridge of the conus becomes hypertrophied and, either alone or by fusion with its mls-d-m's, forms a longitudinal septum dividing the cavity of the conus into two parts——pulmonary and aortic——twisted spirally round one another. "

(7) Correlated with this process is a tendency for the wall of the conus to lose its striped muscle and its rhythmic contractility. This process takes place from before backwards and the portion which has suffered this change, and the wall of which has assumed the ordinary arterial character, is in common usage included under the name truncus arteriosus.

(8) The atrioventricular part of the cardiac tube undergoes a double flexure similar in nature to that seen in the conus of the Lung-fish but of such a kind as would if straightened out give a left-handed twist.

(9) The valves and septa of this part of the heart originate in the endocardiac cushions and in Leptdosiren atrial septum and ventricular septum are at first continuous with. one another.

(10) In the evolution of the valves of the heart there has come VI HE ART AND ARTERIES 393

about a substitution of automatically acting pocket— or flap-valves for the original endocardiac ridge or cushions which for their functioning were dependent upon a complex and fallible neuromuscular mechanism.

ARTERIAL SYs'rEM.—-— -The conus arteriosus is primitively prolonged forwards into the ventral aorta which gives off on each side the series of aortic arches. The most important e.volutionary change wl1iel1 the originally simple tubular ventral aorta undergoes is a process of splitting whereby the stream of blood to the lungs is separated from that to the tissues generally. This splitting is seen for the first time in the Lung-fishes. Here a dorsal portion of the cavity is separated oil’, continuous behind with the pulmonary cavity of the conus and ending blindly in front by its floor meeting its roof. The anterior termination of this cavity is ust in lront of the point of origin of aortic arch V, and this arch along with arch VI, owing to the fact that they branch oil’ from the ventral aorta somewhat dorsally, open from the cavity in question. The horizontal partition which forms the floor of this dorsal pulmonary cavity begins in ontogeny as a rudimentary ingrowth on each side between the origins of arches IV and V. The two rudiments grow back, fuse, and form a horizontal partition continuous at its hinder end with the Right and Left ridges of the conus which, as already explained, form the floor of its. pulmonary cavity.

A’ similar splitting off of a dorsal, pulmonary, part of the ventral aorta occurs in air-breathing Vertebrates in general.

As regards the remaining, or systemic, portion of the ventral aorta, the main point to notice is its tendency to split into two separate halves in its anterior portion, a process correlated probably with" economy of material, allowing as it does the. origins of the aortic

- arch'e.s to be displaced outwards so as to shorten these arches. This

splitting or bifurcation of the ventral aorta spreads backwards for a variable distance, commonly to about the level of aortic arch III or IV. ' .

In the Reptiles there comes about an independent splitting into two lateral halves of the aortic cavity posterior to the region of the bifurcation just alluded to. A vertical septum grows backwards from the anterior lip of the opening into aortic arch IV of the left side and becomes continuous with the septum separating the two systemic cavities of the conus.

When this aortic septum is formed the ventral aorta in its hinder portion contains three cavitics—a dorsal, pulmonary, leading to aortic arches V and VI of both sides, a left ventral leading to aortic arch IV of the left side, and a right ventral leading to aortic arch IV of the right side together with the paired anterior part of the ventral aorta and the aortic arches springing from them. Each of the three cavities is continuous behind with the corresponding cavity of the conus. In the majority of Reptiles (not in Lacerta) the separation of these cavities is followed by splitting of the septa between them 394 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

so that the ventral aorta is resolved into three distinct vessels forming portions of the common pulmonary artery and of the right and left systemic aortae respectively. THE AORTIO Anenns AND THEIR DERIVATIVES.-——F1‘o111 the. ventral aorta there are given elf on each side a series of half-hoop-shaped aortic arches which pass in a dorsal direction, between successive visceral clefts, to open eventually into the dorsal aorta. In the region where it receives the aortic arches the dorsal aorta is frequeiitly paired (forming the aortic roots) either temporarily or throughout life. This assumption of the paired condition may not improbably be of similar significance to that of the ventral aorta 'o'.e. have to do merely with the eeonomizing of material. In any case the precise extent of the paired condition does not appear to be of any great morphological importance.

As regards the aortic arches themselves the following general features are to be noted: (1) that they develop in order of position from before backwards in agreement with the general principle of development of the vertebrate body and (2) that individual arches tend to become reduced in size in correlation with diminution of functional activity. Thus the mandibular and hyoid arches having lost or at least greatly diminished their respiratory activity even in the lower fishes we find a corresponding disappearance or reduction of their aortic arches. ’

An important point to notice is that where the particular visceral arch carries a true external gill (Crossopterygians, Lept'd0s'£ren and Protopterus, Urodele and some other Amphibians) the aortic arch passes out as a loop into the external gill. The aortic arch is in fact in these forms during early stages, before the gillclefts are perforated, the vessel of the external gill. Such relations on the part of vessels of the fundamental morphological importance of the aortic arches are not to be dismissed lightly as modern adaptive modifications. They appear to indicate that the archaic function of the aortic arch was to supply the external gill with blood. _

As the external gill ceases to function a short circuit is formed at its base, through which the blood passes directly to the dorsal part of the arch without traversing the external gill. In Urodeles, according to l\Iaurer, the short-eircuiting vessel sprouts downwards from the dorsal limb of the arch but in Lepidosrlren Robertson finds it arising simply by the enlargement of pre-existing chinks. Thus the definitive aortic arch, in those cases in which an external gill is for a time present, includes a portion secondarily intercalated in its

course and derived from the short-circuiting vessel.

In the typical fishes, where respiration is carried on by the wall of the gill-cleft, there becomes intercalated in the course of the aortic arch a respiratory network of capillaries, so that the arch is divided into a distinct ventral (afferent) and dorsal (efferent) portion. In such a fish as Lepidosiren, where the respiratory activity of the gills ‘VI ARTERIAL SYSTEM 395

is comparatively small, the aorticareh can be traced throughout as a distinct channel ; while in the Amniota, where the walls of the gill clefts have completely lost their respiratory function, the respiratory network never develops at all.

FIG. l86.—~—Scheme of aortic arches and their «lerivatives_aS seen from the Velltml Sid}?The parts of the original scheme which disappear dui-mg development aie shown in

pale tone. A, complete unmodified series of aortic arches; B, arrangement in a Urodele; 0, 111 ‘I Reptile: I), in a Mammal.‘ A, dorsal aorta; a.r, aortic root; mi, aiiastomotic vessel; ac, common carotid;

(LB, duct oi'Bota1lus; d.c, dorsal (internal) carotid; 'i, innoniinate artery; 512'’: left l’”l'“0"‘"'y artery 3 L3, left systemic; yr, pulmonary; 'r.p, right pulmonary; 5'. systemic 30ml; 3. Sllbcla-Via“ 31‘t9l‘)’ § v.n, ventral aorta; um, ventral (external) carotid ; 1, 11, etc., aortic arches. ‘

The longitudinal vessels with which the aortic arches are con nected are prolonged forwards as the carotid arteries which supply

the head with blood. The Ventral aorta is prolonged forwards as the ventral, or external, carotid while the prolongation forwards oi the aortic root forms the dorsal, or -internal, carotid. Of the alternative 396 EMBRYOLOGY OF THE LOWER VERTEBRATES CH’.

names dorsal and ventral (Mackay, 1889) are to be preferred to internal and external, for the latter though in common use are less precise. During the development of the young individual there is laid down a general scheme of aortic arches and associated vessels agreeing with that just described, and the processes of modification whereby there becomes evolved out of this the complicated and very different arrangement of the great arteries of the adult afford material for one of the most fascinating chapters in vertebrate

Fm. 187.-—Illustrating modification of the carotid arteries, correlated with elongation of the neck region.

A, Varanid Lizard ; B, Grass-snake (’l"mpi«Ionolus); }).f'., primary carotid. (Other letters as in Fig. 186.)

embryology. The general lines of these processes are best illustrated by an outline of what happens in the group Reptilia.

The arrangement which the main arteries assume in adult Reptiles shows much variety. The relation which the adult arrangement in the more important Reptilian types bears to the primitive scheme may he gathered from an inspection of Figs. 186, C, 187, A, B, and 187A, C3 Through the conspicuous differences in

1 It must be remembered that in the actual animal the various bends and turns of the vessels tend to become strai htened out. For example arch III becomes simply a portion of a straight interna carotid artery. In the diagrams the original curvature of the arches is retained for the sake of clearness. VI ARTERIES OF iREP'.I‘ILESi 397

detail there can be seen general agreement in the fate of various aortic arches and of other parts of the primitive arterial scheme. Thus the ventral aorta is continued forwards to form the paired ventral (“external”) carotid arteries (12.0) while the aortic roots similarly extend forwards as the dorsal (“internal”) carotids (al.c). Aortic arches I and II disappear. Arch III persists as the root of the dorsal carotid while the portion of ventral aorta behind it, when

Fm. 187A.—-Illustrating modification of the carotid arteries, correlated with elongation of the neck region.

C, Crocodile; 1); Bird; c, coeliac artery; .s-2, secondary subclavian. (Other letters as in Fig. 186.)

paired, is the common carotid (0.0). Arch IV is the Systemic arch which sends the blood to the hinder portions‘ of the aortic roots (am) and thence to the dorsal aorta (A). Arch V is reduced, appearing only as inconspicuous and transient vestiges during development. Of arch VI the proximal portion becomes the root of the pulmonary artery (rap and l.p) while its dorsal portion disappears.

The chief diiferences in detail are as follows: those affecting the

carotids will be given more fully later on. In Lizards and Chelonians the dorsal part of arch VI persists as

I 398 l EMBRYOLOGY or THE‘ LOWER VERTEBRATES

a duct of Botallus forming a connexio

A.

FIG. 188. —-Illustrating the modification of aortic arches III-VI" during ontogeny in Scyllium, according to Dohrn.

A, Dorsal aorta; of, afferent branchiul; cm, unastornotic dues frol-H

vessel; cf, efferent branchial; -v.A, ventral aorta; 'v.c, visceral clefts; I1 I, IV, V, V’ I, aortic arches.

C11.

11 between the pulmonary

artery and the aortic root just as is shown in Fig. 186, B (LLB) for the Urodele amphibian. In other cases it may persist as a ligainentous vestige, as is the case on the left side in Tropidovzotus.

As a rule the portion of aortic root lying between arches III and IV disappears during , development but in most Lizards (not in Chaineleons and Monitors) i.t persists in the adult, so that in a dissection arches III and IV appear to run into one another peripherally.

In Monitors (Varanidae), in correlation with the elongation of the neck, arches Ill and IV hecorne widely separated from one another and the intervening portion of ventral aorta shows a corresponding lengthening both in its paired (common carotid) and its unpaired (primary carotid) portions.

In Birds (Fig. 187A, 1)) arch IV completely disappears on the left side and with it the portion of the left aortic root lying posterior to it. Consequently in the adult Bird there is only a single systemic aortic arch and it passes down the right side of the body.

ELASMOBRANCIIII.-— In the Ichthyopsida, as we should expect, the departthe primitive scheme are less pronounced; VI ARTERIAL SYSTEM 399

they are mainly in details. In the Elasmobranchs perhaps the most conspicuous of these is to be seen in the relations of the efferent vessels, each of which emerges, not from an ordinary aortic arch traversing a gill septum, but from a vascular loop surrounding a gill-cleft, the general arrangement being that shown in Fig. 188, 0. According to Dohrn (1886) this arrangement comes about in the following way. As the Walls of the elefts form lamellae and develop respiratory activity, two branches grow downwards, one anterior and one posterior, from the dorsal end of each aortic arch (Fig. 188, A, V1). These branches become connected together by cross bridges as shown in the figure (cm). The aortic arch now undergoes reduction and eventually becomes obliterated, just ventral to the point where the two branches are given off (Fig. 188, A, arch lll), so that the arch is now divided into two distinct parts-— a ventral afferent and a dorsal elferent—the latter prolonged ventralwards into the two branches. Of these the posterior branch of each pair becomes somewhat reduced in size, it develops a secondary connexion at its upper end with the efferent vessel next behind and loses its connexion with its original efferent vessel. The result is that, after this has happened, each efferent vessel again possesses two branches at its ventral end but these, instead of passing into the same gill septum, pass into two adjoining septa one in front of and one behind the intervening cleft (Fig. 188, B). Eventually the. ventral ends of each pair of branches become joined so that the cleft is now surrounded by a complete efferent loop-— the loops of successive c1eI'ts being connected together by a single persisting anastomotic vessel (Fig. 188, C).

In C’/olamg/doselmtlms the modification of the aortic arches just described does not take place.

The number of aortic arches corresponds with that of the visceral arches and is normally six.

In correlation with the presence of the pseudobranch on the posterior face of the mandibular arch in Elasmobranchs the first aortic arch in these fishes is well developed but its primitive relations with the arterial scheme become much obscured owing primarily to the development of large new afferent and efferent channels connected with the pseudobraneh, which carries in its train the reduction of both the ventral and the dorsal portions of the original aortic arch.

1n the case of the second aortic arch, correlated with the fact that the anterior face of this visceral arch has lost its respiratory function, there is developed only a single, posterior, efferent downgrowth instead of two as is the case with the arches farther back. A Wide anastomosis between this and the first aortic arch

just below the spiracle provides the secondary afferent vessel to

the pseudobranch which as already mentioned supplants the primitive afferent vessel formed by the ventral portion of the first arch.

Q-If 400 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

CYeLos'1‘oMA'1'A. -——In the Lamprey it should be noted that according to Dohrn (1888) an aortic arch corresponding to aortic arch l of Gnathostomata makes its appearance and then disappears again. In Myxinoids the most important feature is that in them the number of aortic arches reaches its maximum for Craniata——-up to 14 in Jidellostovna.

CROSSO1’TERYGII.———Our knowledge is in this case very incomplete. The chief peculiarity (Graham Kerr, 1907) is that, correlated with the large size of the external gill belonging to arch II, which forms the sole respiratory organ during early stages of larval life, aortic arch II makes its appearance relatively early and the development of the other aortic arches is postponed. Distinct vestiges of aortic arch I make their appearance. The succeeding aortic arches rema.in small for a prolonged period. Aortic arch V I becomes much enlarged in its ventral part in correlation with the fact that it supplies the pulmonary artery.

AC'I‘IN01"l‘ERYGII.——-I11 the Teleostean fishes and in the Ganoids that approach must nearly to them (Lepido.steu,.s'—-—I*‘. W. Muller, 1897; Am'ia—-—-Allis, 1000) complicated changes, which need not be detailed, take place in arches I and II in relation with the blood supply of the pseudobranch which in these fishes (p. 159) comes to lie on the inner surface of the operculum.

DIPNOI.-—-I11 Lepidosiren (Robertson, 1913) aortic arches I and II never become complete well-developed vessels: they are vestigial and their ventral portions do not appear to develop at all. The remaining four aortic arches are well developed, each passes out into an external gill and in each an intercalary piece becomes developed to short-circuit the blood-stream at the time the external gills atrophy. In 0’emtodus and P9-otupterus efferent downgrowths make their appearance as in Elasmobranchs but they remain connected with the dorsal portion of their own aortic arch and do not undergo fusion at their ventral ends so that the condition in the adult departs less from the primitive than it does in the Elasmobranch.

AMPHIBIA»-—In U rodela the arrangement of aortic arches (Fig. 186, B) closely resembles that in Lung-fishes. Arches I and II are reduced: the latter in fact is according to Maurer (1888) no longer to be detected at all in the case of ’['7~zJt0n. Arches III, IV and V are prolonged outwards into external gills and in each case a short-circuiting piece becomes intercalated as the external gills lose their functional activity. According to Maurer the intercalary portion makes its appearance as a downgrowth from the dorsal or efferent limb of the aortic arch but this may perhaps be doubted in view of the fact that in Lung-fishes the corresponding piece of vessel develops by the widening out of pre-existing chinks (Robertson, 1913).

AMNIO'1‘A.—-Arch IV along with the aortic root into which it passes forms the main systemic aorta on each side. That of the left side is connected, in correlation with the spiral twist of the VI ARTERIAL SYSTEM 401

conus and its derivatives, with the 'r'£_qht side of the ventricular cavity in proximity to the opening of the pulmonary artery. With the. separation of the two ventricles the arch in question remains connected with the right ventricle. With increasing efliciency of the pulmonary circulation the venous blood of the right ventricle would he drawn off more and more to the pulmonary artery and, correlated with this, we find in those Sauropsida in which metabolism is most active and respiration most efiicient that this fourth arch on the left side with its aortic root disappears completely (luring development, leaving only the single right-hand arch and

VNIII

aczz 0?.

4/I ’ V)

Fm. l89.——lll()0(l-Vessels of (Jroeodile of stage 55-56. (After Hochstetter, 1906*.) Arteries are shown in outline, veins black.

u.I'.r, :mte1mr cardinal vein; .l.;, :lm't:(' root; /mt, b:1.~.Il:u artery; ¢l.a-, (l()l:~.'ll em-ot,ul ; oi, otocyst; p,«-.4‘, 1u)sfL'I‘I0l'('ardInal win; 1.)). right pulmonary an Lt-I) : ear. venttal e.'uot.d ; HI-\ I. aoilic an-hes.

root to form the proximal part of the systemic aorta. (Birds, Fig. 187 A, 1)). '

Arch V, in the Anmiota, appears only transiently and so greatly reduced in size as to have completely escaped the notice of the earlier investigators. Hence in Rathke’s classical scheme of the aortic arches which is given in the older text-books only five arches are shown, the posterior one being called the fifth. With our present-day knowledge of the homology of the lungs of Amniota with those of Orossopterygians and Lung—fishes, such a scheme is clearly erroneous, as it would involve the pulmonary artery, which is certainly the same vessel throughout, taking its origin in the Ainniotes from the fifth and in the Ichthyopsida from the sixth aortic arch. So without any special embryological data we should

VOL. 11 2 D 402 EMl%RY()L0(l‘rY OF THE LOWER VERTEBRATES CH.

be justified in believing that in the Ainniota an aortic arch has disappeared in front of the last one. The persisting vestiges of this fifth arch, which are now known to occur commonly in the embryos of the Amniota, we"e detected first by van llemmelen (1886) in Reptiles and Birds. A good example of such a vestigial fifth arch is seen in the embryo of the Crocodile (Fig. 189). ‘

The cause of the reduction of this fifth aortic arch is probably to be recognized in the fact that it receives its blood from the pulmonary cavity of the conus and of the ventral aorta (Graham Kerr, 1907*‘). As a consequence of this, during the evolution of the lungs as the main organs of respiration a larger and larger proportion of the blood in the cavity mentioned has become drawn off to the lungs, leaving less and less for arch V, with the natural result that the latter has become reduced to the verge ol'disappea1'ance.

Before leaving the subject of the aortic arches it is necessary to point out that their diagrammatic arrangement as shown in Figs. 186 and 187 is commonly much obscured in the adult. For during development there occur not merely the disappearance of large. portions of the original scheme of arches and the straightening out of the unnecessary curves, but also other complications. The chief of these are due to the longitudinal vessels———ventral aorta or aortic roots-——lagging behind in their growth in length. This leads, according to the position in which it takes place, to the crowding together of the ventral or the dorsal ends of consecutive aortic arches, and their mere approximation may he succeeded by actual fusion so that two or more arches may come to have a common root emerging from the. ventral aorta, or a common terminal portion opening into the aortic root.

PULMONARY AR'1‘En\'.——The pulmonary artery makes its first appearance in Crossopterygian fishes (1’0l_7/ptelrus) as a branch from arch VI towards its ventral end which passes to the lung and adjoining parts of the pharyngeal wall. Throughout the series of lung—breathing Vertebrates it develops similarly as an outgrowth of the sixth aortic arch. A result of the main blq rod-St1'(‘a.1n of this arch passing olfinto the pulmonary branch is that the dorsal part of the arch lying beyond the point of origin of the pulmonary artery becomes as a rule reduced in size, forming the duct of Botallus. Except in the case of certain Reptiles (p. 397) the duct cl‘ Botallus becomes in normal individuals of the Amniota completely obliterated soon after birth.

In the Lung-fishes the point of origin of the pulmonary artery is displaced up to the dorsal end of arch VI where it is fused with arch V. Thus arch V is able to carry blood directly to the pulmonary artery and correlated with this it does not undergo the reduction in size which has taken place in the Amniota.

In Lepiclosiren and Protopterus an important development of the pulmonary artery takes place inasmuch as its area of distribution extends on to the lung belonging morphologically to the other side of VI ARTERIAL SYSTEM 403

the body. Thus the right pulmonary artery comes to supply the dorsal side of both lungs, and the left artery the ventral side of both lungs. Each lung in other words receives a supply of blood from. both pulmonary arteries and this illustrates an initial step towards the condition in Am/in, where the right and left sides of the air-bladder— the homologue of the right lung—-are supplied with blood directly by a typical right and left pulmonary artery. ln the Actinopterygian fishes apart from Amid. the pulmonary artery has disappeared entirely from development and the air-bladder receives its blood-supply by se.condary connexions with the dorsal aorta and its branches.

CAROTII) ARTERIES. As has already been indicated the great longitudinal arteries ventral and dorsal aortae-—-are prolonged forwards into the region of the head as the carotid arteries——ventral and dorsal. Of these the latter, receiving as it does, in the ease of the more primitive Vertebrates, blood which has been oxygenated by passing through the gills, becomes the more important and is responsible for supplying blood to the brain. .

The ventral carotids are found from the Lung-fishes onwards—— perhaps in correlation with the reduction of the first two aortic arches. They are known eonnn only under the name external earotid( = lingual artery of Amphibia) and supply blood to the ventral side of the head, though cases are known amongst animals no longer having functional gills (certain Mammals) in which they take over the bloodsupply of the brain also.

lt will be convenient to consider first the carotids as they occur in the development of the Amniota. The simplest condition is found in an ordinary Lizard (Lacerm) where they are seen as apparent prolongations forwards of the aortic root and of the ventral aorta respectively. This simple arrangement becomes in otherAmniota modified during the course of development i11 different ways, of which the following are the chief. In various Lizards ag. Chameleons and Monitors (and the same holds for the great majority of Amniota ——Fig. 187, A) the portion of aortic root between aortic arches III and IV disappears during development, the consequence being that arch III comes to form the posterior portion of the internal carotid artery, becoming drawn out in the process of growth so as to be in line with the front part of that artery derived from the aortic root. The paired part of the ventral aorta from which the third arch was given off now becomes the common carotid artery (ac). The unpaired portion of the ventral aorta, from which the common carotids spring, is known as the primary carotid and in the long-necked monitors this becomes much elongated, the growth in length of the neck taking place in the region between aortic arches Ill and IV (Fig. 187, A).

In the European Grass-snake (Tropidonotus, Fig. 187,- B) an anastomosis is formed between the two internal carotids just behind the head (Fig. 187, B, an) and, correlated with this, the right common carotid as a rule disappears except for a small branch at its hinder 404 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

end which supplies the Thyroid gland (O'Donoghue, 1912). The blood-supply of the head-region therefore passes to it entirely by the persisting left common carotid (c.c). In Snakes other than ’1'rop7ld0notus considerable variety exists in the condition of the common carotids. Thus amongst the Boidae the two arteries may remain of approximately equal size (Bria) or on the other hand the left may be reduced (Pg/tIa.0'n.).

In Chelonians and Crocodiles (Fig. 187A, C) the growth in length of the neck takes place in the region in front of arch 111 so that here it is the portions of the carotid arteries in front of this level which undergo elongation. In both of these groups an anastomosis forms in the head-region between dorsal and ventral carotids and, correlated with this, the main blood-stream tends to pass to the head by the dorsal carotids, the ventral vessels becoming to a less (Crocodiles) or greater extent (Chclonians) reduced in size (van Bemmelen, 1887; Mackay, 1889). In the Crocodiles a still further modification takes place, inasmuch as the two dorsal carotids become for a considerable, part of their length fused together into a single vessel, and following upon this aortic arch lll of the right side atrophies, so that here as in '/'7"0})id07L0tus, though for a different reason, the main blood-supply of the head comes from the left side.

In the Birds the condition closely resembles that of the Crocodile. Here also the ventral carotids become reduced —in this case to the point of complete disappearance——in the neck region as a (‘.O11b'e(_111OIlC0, of an anastomosis with the dorsal carotids in the head. Here also the. enlarged dorsal carotids approach one another on the ventral side of the vertebral column. In those birds which depart least from the primitive condition in this respect (Ostrich, Emu, Casuari, Tinamus, Penguins, Divers, Gulls, Plovers, Snipe, Rails and their allies, Fowls, Pigeons, Ducks, Ihises, Storks, Herons, Cormorants and some Gannets, Birds of Prey, Parrots, Hornbills, Motmots, Goatsuckers) the two definitive carotids merely lie in proximity to one another. I11 many birds however they become fused together into a single vessel over a great part of their length and in such a case there may be no further modification (certain Herons such as the common Bittern, some Cockatoos, some Uannets), or, as is the general rule, this fusion is followed by the disappearance of the third aortic arch of the right side just as was the case in the Crocodiles (Rhea, Apterg/ac, Grrebes, Quails, some Cockatoos, Capitonidae, 'l‘oucans, Hoopoe, Meropidae, Trogons, VVoodpeckers, most Swifts, Humming birds a11d Passerine birds). I11 a few cases on the other hand it is the third aortic arch of the left side which becomes reduced to a small vestige (l3‘l-amingo) or disappears entirely (.E'upod0t'is—-— the African Bastard).

Amongst the anamniotic Vertebrates typical dorsal and ventral carotids are present in Amphibians a11d Lung-fishes. The arrangement in Urodeles is illustrated by Fig..186, B: the chief peculiarity to be noted is that the posterior portion of the external carotid (12.0) VI - ARTERIAL SYSTEM 405

has disappeared in the adult, the persisting anterior portion receiving its blood through a new anastomotie channel (cm) from the third aortic arch. The posterior portion of the “external carotid” or “lingual artery ” of the adult Amphibian is really constituted by this new development.

The connexion of this newly developed portion of vessel with arch III is just at the point where the short circuit is formed between the afferent and efferent parts of the arch, and in Lungfishes (.Lep'id0si7*em——Robertson, 1913) the blood-supply for the external carotid comes to it, for a time during early stages, from the dorsal end of arch III or from the aortic root, through what seems to be a precocious development of this same short-circuiting channel. In this case the vessel in question is at first simply continued from its dorsal origin forwards into the external carotid: it is only later that it communicates with the ventral or afferent end. of arch III so as on the one hand to form the short circuit, and on the other to permit the blood to pass to the external carotid from the ventral aorta.

In the more typical fishes the ventral carotid is not yet present. The dorsal carotids of the two sides develop an anastomotic connexion beneath the base of the skull so to form with the aortic roots a complete “ cephalic circle” which shows characteristic differences in different Teleostean fishes (Ridewood, 1899).

INTE'RSE(_T}MENTAL Anrsmes or TIIE BoI)Y-WALL.-—-'1‘he dorsal aorta gives off on its dorsal side paired arteries which run out into the body wall between the Inyotomes. One of the most important features of this series of intersegmental vessels is that for a time during early stages of development they provide the blood-supply to the limb rudiments.

The main artery of the fore-limbw-the subclavian artery-— appears during the stages in question to be simply a prolongation into the limb rudiment from one of these intersegmental arteries»not necessarily the same artery of the series in different types of Vertebrate, or even in different developmental stages of the same Vertebrate. Thus in Lacerta it is said to be the seventh intersegmental artery (van Bemmelen, Hochstetter, 1906) which becomes the subclavian artery and in the Fowl the fifteenth (eighteenth or nineteenth if the three or four intersegmental vessels in the headregion are ineluded—-—Hochstetter, 1890). In the Duck, Rabl (1907) found that during the fifth day of incubation the subclavian varies from the eighteenth to the twenty-first intersegmental artery and that in some cases two or even three such Vessels may pass out into the limb rudiment at one time.

Probably we may take it that the‘ general principle at work is this——-that the limb, as it became shifted along the side of the body in the course of evolution, received its blood-supply from successive intersegmental arteries as it came to be opposite to them, and that during ontogeny there takes place an imperfect repetition (bf this process.

The fact that the pectoral limb is supplied with blood. by an 406 EMBRYOLOGY OF THE LOWER VERTEBRATES CII.

intersegmental artery raises the question whether or not this is to be regarded as the primitive mode of blood-supply. For if this question he answered in the affirmative We should be confronted with a11 important point which would have to be borne in mind in all speculations as to the evolutionary origin of the limbs of Vertebrates. As a matter of fact, however, recent investigations tend to answer this question in the negative.

In the Chick (Evans, 1909) the limb rudiment in its earliest stages is traversed by an irregular network of blood spaces and this receives its blood-supply directly from the dorsal aoi ta by a number of slender ehannels——it may be as many as ten or eleven on the right side where they are commonly most numerous. These vessels are scattered irregularly over an antero—posterior extent of from three to five mesoderm segments and they take their origin from the dorsal aorta quite independently of and considerably ventral to the intersegmental arteries. As developmeiit proceeds a few of these supply channels those. which happen to be most nearly intersegmental in position———beco1ne relatively larger and finally a single one, at about the level of the eighteenth intersegmental artery, becomes especially enlarged and carries the main stream of blood to the limb rudiment while the others gradually diminish in size and eventually disappear. The persisting enlarged Vessel becomes the subclavian artery and secondarily its origin from the aorta becomes displaced in a dorsal direction until eventually it arises by a common root along with the intersegmental artery of which it now appears to form a branch.

In the Duck similar observations have been made so we are probably justified in stating that in the earliest stages of ontogenetie development the blood-supply of the pectoral limb is not metameric, and that the relation with the intersegmental artery observed in slightly later stages is a secondary acquii-ement.

In most vertebrates the subclavian artery which arises in the manner above. described (primary subclavian) persists throughout life. In Birds however a cross connexion develops between the primary subclavian just at the base of the limb, and the ventral end of the third aortic arch. This cross connexion gradually increases in size while the proximal part of the primary subclavian, arising from the aortic root, becomes correspondingly reduced and eventually disappears entirely. The result is that the permanent artery of the fore-limb in the adult branches off’, not from the dorsal aortic root but from the definitive ('z3.e. morphologically “intcrnal”) carotid, close to its hinder end. . A similar substitution takes place in Chelonians and Crocodiles (Fig. l87A, O) and this is the explanation, first given by Mackay (1889), of the otherwise puzzling fact that in certain Vertebrates the subclavian artery passes out ventrally to the vagus nerve (“secondary subclavian ”) instead of dorsally as it does normally.

The iliac artery to thehind limb has also been traced back in at least Lizards and Birds to the series of intersegmental arteries but vr ARTERIAL SYSTEM 407

here again it would appear from later investigations that the definitive artery is merely a surviving and enlarged representative of a number of original supply vessels (Evans, 1909).

’tepresentatives of the series of intersegmental arteries are recognizable in various Vertebrates in the hinder part of the headregion. Thus in Lacerta (van Bemmelen, Hochstetter, 1906) three have been detected in the head-region. Of these the first two disappear while the third becomes prolonged forwards immediately beneath the brain to become continuous with the internal carotid at the level of the mid-brain. These prolongations fuse together in the mid-line and form the basilar artery.

A similar basilar artery continuous with the internal carotids is of common occurrence in Vertebrates though the relations of its paired fore-runner to the series of intersegmental arteries differ in different forms.

The vertebral artery of Sauropsida is in its origin intimately related to the interserrmental arteries. Thus in Lacertrr. the intersegmental vessels posterior to the subclavian become connected together by a longitudinal anastomotic vessel, which persists and forms the cervical portion of the vertebral artery. In Snakes the two similarly-arising vessels apparently usually undergo fusion together in their anterior portion while farther back the unpaired condition is reached by the disappearance of the rudime.nt on the right side (Hochstetter, 1906).

The arterial blood-supply of the urino-genital organs is also generally provided by branches from the intersegmental arteries of the embryo.

M1«JsEN'1‘E1tIC Au'1‘E1uEs. The digestive tract receives a varying number of branches from the dorsal aorta which may be at first paired and undergo fusion secondarily or may be impaired from the beginning. In those vertebrates which have a bulky yolk—sae a pair of these are precociously developed as vitelline arteries. In some cases there is a remarkable relation between the chief mesenterie artery (coeliac) and the prone] )ll1'OS. Thus in L6})’id08'I:I'(3'Il. a connexion becomes established between the blood-spaces of the. right pronephros and those of the gut wall, and the branch of the dorsal aorta which supplies the right pronephros persists as the root of the definitive eoeliac artery. In such Teleostean fishes as the. Trout the connexion with the pronephric arterial supply is only temporary, a new anastomotic channel arising farther back between dorsal aorta

and mesenteric artery which remains as the definitive root of the latter vessel.

VENOUS SYsTEM.——-As an example of the development of the

venous system in a holoblastic Vertebrate we will take that of Lep1Idos'£1'e7t (Robertson, 1913)}

‘ For comparison with Lcpidosirmz, an account of the development of the vascular system of Uemtodus will be found in Kellicott (1905). 408 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.vI

The first part of the venous system (Fig. 190, A)—indeed of the, vascular systen1——-—to take definite form consists of the two vitelline veins, which pass tailwards ove.r the anterior surface of the yolk (stage 24). Anteriorly they become conjoined to form the heart while ,‘posteriorly they are continued into the rudiments of the vitelliue network (stage 24-25). The ap]»earanee ol' the vitelline. veins is followed almost immediately by the development of a longitudinal venous channel on each side anteriorly, superlicial to the aortic arches ——the anterior cardinal vein. At its hinder bud the anterior cardinal is continued into a set of venous spaces in the region of the pronephros (pronephric sinus) and onwards behind this as the posterior cardinal vein.

The pronephrie sinus is continued along its oute.r edge, by a number of channels, into the vitelline network of venous spaces, lying in the, splanchnic niesoderni over the surface of the yolk. In this network a conspicuous channel becomes apparent, leading from the anterior end of the ]_)l'()ll£’1)lll'lC sinus outwards to the vitelline vein, and so, by way of the anterior part of the vitelline Vein, to the heart. This vessel so constituted, which at iirst makes a wide sweep over the lateral surface of the yolk, is the Duct of Cuvier (Fig. 190, b, (L0). As development goes on the Ducts of (luvier becoine greatly shortened and at the same time widened until e\'e.ntually they form in the adult very short wide channels for the conveyance of the blood from the cardinal veins into the sinus venosus (see. Fig. 190, b and e and d).

The posterior cardinal vein on each side ap])ea1‘s about stage 24 in the form of spaces along the course of the archinephrie duct which become joined up so as to form two longitudinal Vessels running parallel to the duct, one on its mediodorsal the other on its ventrolateral side, the two vessels being joined round the duct by numerous anastomoses (Fig. 190, h and e, ;u.c.v).

These. posterior cardinal veins accompany the archinepliric duets throughout their length and just in front of the cloaca are joined by the hind ends of the bifurcated subintestinal vein (see below) and of the dorsal aorta. The vessel formed on each side by the union of these three elements is continued back last the cloaca and unites with its fellow of the opposite side to form a vessel lying iinmediately beneath the post-anal gut. This vessel is to be interpreted 1norpho_ logically as a post-anal ])o1‘tion of the sul )lllbC‘Sl}l]l{1]. vein and as will be shown later it is destined to become the caudal vein of the adult.

It will now be convenient to trace. out the subsequent fate of what may be called the dorsal venous system, consisting primarily of the anterior and posterior cardinal veins.

POSTERIOR CARDINAL V1«11Ns.——’l‘he posterior cardinal Vein Was left in the form of a pair of vessels, an inner and an outer, lying close to the archinephric duct and connected together by numerous anastomoses. As the opisthonephros develops between these channels they consequently co1ne,over a considerable part of their length, to be VI . VENOUS SYSTEM 417

blood-stream passes through these communications into the posterior cardinals and the latter take on the appearance of a direct forward prolongation of the caudal vein.

Dorsally there develops on each_ side a cardinal trunk which swells out into a great irregular sinus (pa) in the region of the pronephros, On its outer side branches pass from the pronephric sinus into the general vitelline network. In this network a specially wide channel develops, starting from the pronephric sinus and sweeping outwards over the yolk to join the lateral’ vitelline vein and so reach the heart. This channel, which becomes gradually more and more sharply defined, is the duct of Cuvier (Fig; 193, A, (L0). The pronephric sinus is continued backwards into the posterior cardinal vein. This is at first distinct from its fellow but at an early stage fuses with it to form a median inter—renal vein (Fig. 193, B, 7I7')—--the fusion being foreshadowed by the development of anastomotic conncxions between the two veins while still separated from one another by a distinct space (Fig. 193, A, an). Towards the cloaca the posterior cardinals taper off and are connected by irregular anastomotic ehannels_ with the subintestinal vein, as already mentioned, and also with the dorsal aorta. Eventually, as -already indicated, the inter-renal vein and the caudal vein form a continuous vessel.

During the later stages a striking asymmetry becomes apparent in the anterior, unfused, portions of the posterior cardinals——the left becoming greatly reduced as compared with the right (Fig. 193, D). The main blood-stream thus passesforwards on the right side, and upon this side a special direct channel develops on the ventral side of the pronephros, through which the blood-stream is able to reach the duct of Cuvier without passing through the tangle of pronephric tubules. The asymmetry affects also the ducts of Cuvier-— showing itself first in that of the left side becoming relatively shorter than its fellow, which retains for a time its wide sweep over the surface of the yolk (Fig. 193, B and C, d.(/"9. Eventually it too becomes shortened and its calibre becomes considerably greater than that of the left side (Fig. 193, D).

From the pronephric sinus a branch (Z/2:) develops about stage 30 which passes dorsalwards and then backwards beneath the lateral line nerve-—the lateral cutaneous vein. This, a large vessel about stage 33, becomes reduced to an insignificant vestige later on.

The anterior cardinal vein runs along the side of the head region, passing through the angle on the ventromesial side of the otocyst, between the latter and the brain-wall. At its front end the vein dilates into a large sinus, which gives off irregular branches to the mesoderm of the head. At an early stage an anastornotic channel makes its appearance on the outer side of the otocyst continuous anteriorly and posteriorly with the anterior cardinal. When this channel has been established (Fig. 193, A, La) the blood-stream from the head divides in front of the otocyst and passes backwards, part

VOL. 11 2 E 410 EMBRYOLOGY OF THE LOWER VERTEBRATES CII.

separated from one another by a considerable space in which the renal organ lies. As this happens the inner components of the two posterior cardinals become approximated and eventually undergo fusion with one another to form an inter-renal vein (Fig. 190, d, 'i9'.'v). In _Lep1Idosi'ren this fusion is only temporary and the two components again recede from one another—remaining, however, connected by a small number of anastomotic vessels (Fig. 190, e). There now takes place a severance of the continuity of the inner component at its hinder end (Fig. 190, e,"*), and a little later a similar severance of the outer component at its front end. The physiological result of these interruptions of continuity is that the blood from the caudal region now reaches the opisthonephros entirely by way of the outer component, it then passes through the substance of the kidney and is draine.d away entirely by the inner component. In other words the outer component and its backward continuation has now become the renal portal Vein.

While these changes are going on in the opisthonephric region of the posterior cardinal the portion of that vein in front of the opisthonephros takes on the form of a single channel, the original inner component becoming enlarged while-the outer component becomes reduced and eventually disappears. The rightand left posterior cardinal veins in this region in front of the opisthonephros become connected by numerous transverse vessels (Fig. 190, d). A short circuit now becomes established by which the, blood from the right posterior cardinal can pass direct to the sinus venosus through the substance of the liver (Fig. 190, d, 12/0.0).

This short-circuiting channel is of great morphological importance. It constitutes the intrahepatic or headward section of the posterior (“inferior”) vena cava, which in the higher vertebrates becomes the largest vein in the body. lts appearance here is followed by two important results. (1) The main blood-stream from the kidney region tends to pass to the heart, more and more by this direct channel, which in correlation with this becomes larger and larger. (2) The portion of right posterior cardinal vein lying behind its junction with the intrahepatic channel becomes correspondingly e11la.I'ge( l.

These two components together constitute the definitive posterior vena cava, in which vessel therefore. we recognize two fundamentally distinct portions, an anterior or intrahcpatic, and a posterior or cardinal. A secondary result of the diversion of the blood-stream from the right posterior cardinal vein through the hepatic component is that the anterior portion of the former vein, lying in front of the junction of the two components, becomes relatively reduced in size. It loses its continuity with the rest of the. right posterior cardinal and persists as the small vein shown at 12 iii Fig. 190, e.

An important clue to the origin of the. posterior vena cava in phylogeny is given by the condition seen in the adults of existing L1.1ng-fishes. The opisthonephros in these vertebrates retains its VI VEINS OF LEPIDOSIREN 411

primitive elongated form, extending far forwards in the splanchnocoele, and the iront end of the right opisthoncphros is in immediate apposition to the tip of the liver which is also situated dorsally and on the right side. It is no doubt the approximation of the tips of these two organs, still persisting iii the adult Dipnoan, which paved the way for the establislnnent of direct continuity between their vascular networks and the consequent short-circniting of the renal blood through the hepatic vci11 into the heart (Graham Kerr, 1910)

The ladder-like connexions between right and left posterior aardinals in the region anterior to the inter-renal vein gradually disappear in turn from before backwards (Fig. 190, d and e).

ANTERIOR CARDIN AL V EIN s.——Apart from relatively less important details, the chief change which comes about in regard to the anterior cardinal is the diversion of the blood-stream about the level of the otocyst into a more laterally placed channel called the lateral cephalic vein, which eventually becomes intercalated in the course of the anterior cardinal, and to all appearance forms simply a portion of that vein (Fig. 190, (1, L0). The anterior cardinal vein at first passes back along the side of the head region (following the course shown by the dotted outline i11 Fig. 190, d), ventral to the otocyst and internal to the posterior cranial nerves, to join the front end of the pronephric sinus. Presently (stage 30) a branch of the anterior cardinal vein makes its appearance and extends backwards external to the ganglion of the seventh cranial nerve, bending inwards and n-joininlg the anterior cardinal between the eighth and ninth cranial nerves. A little later (stage 31) the vessel forming the outer side of this loop becomes prolonged back and forms a second loop external to the ninth cranial nerve and rejoining the main Vessel between nerves IX and X. Finally (stage 31 + ) a similar extension backwards occurs external to the vagus, rejoining the anterior cardinal vein just behind it. ’l‘he lateral cephalic vein develops from the outer portions of these three vascular loops, the development of each of the three segments being followed by the atrophy of the corresponding section of the original anterior cardinal, except in the case of the most posterior section which persists as a short wide vein opening along with the posterior cerebral vein (ce”) into the definitive anterior cardinal.

VENTRAI. VENOUS SYSTEM.-—In addition to the dorsally placed cardinal veins there. exist certain important veins situated more ventrally and developed in relation with the vitelline veins. The vitelline veins spread backwards on each side as a wide vessel consisting of an enlarged channel of the vitelline network which covers the whole surface of the yolk. Posteriorly they unite in the. midventral line to form a very short subintestinal vein in front of the an11s (Fig. 190, O, S.’t.’I)). Later on the space bounded by the two vitelline veins becomes bisected by a median ventral vein which looks like a prolongation forwards of the subintestinal vein and 412 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

is known by the same name. (Fig. 190, D, s.z';v”). It will be realized that this anterior section of the subintestinal vein is developmentally of a different nature from the posterior portion for it is formed by a short-eircuiting of the blood-stream through the vitelline network, while the posterior portion represents rather the conjoined hinder ends of the paired vitelline veins (Fig. 190, O and D). This difference in development is no doubt purely secondary and we may take it that the later condition, where the subintestinal vein is continuous right forwards to the heart, represents the really primitive condition of this vein in evolution. The continuity of the subintestinal vein at its front end with the heart is brought about in ontogeny through blood-sinuses which make their appearance in the liver (stage 3]).

The right vitelline vein and the anterior, secondarily formed, portion of the subintestinal vein now gradually disappear (about stages :52 - ‘%5). The left vitelline vein ceases to form a continuous channel over the surface of the liver to the heart, so that the blood in it is forced to traverse the system of blood sinuses within the liver. 111 other words the whole of the blood which streams forwards in the subintestinal vein is diverted along the persisting left vitelline vein into the network of blood spaces in the liver. Subintestinal vein and left vitelline vein have thus come to constitute the hepatic portal vein. 'l‘he latter becomes complicated by a branch sprouting out from the front end of its subintestinal portion. This branch spreads round the alimentary canal along the line of the spiral valve, fusing with the subintestinal vein at each point of intersection (Fig. 190, F). As the liver increases in length a special supply channel from the portal vein lengthens out along its left side, giving off numerous branches into the liver substance (Fig. 190, F). From this the blood drains by numerous efferent vessels into the intrahepatic portion of the posterior vena cava.

CAUDAI. VI«:IN.——’.l‘hc post-anal portion of the subintestinal vein was left (p. 408) at a stage when its anterior bifurcated portion was continuous on each side not only with the pre-anal portion of the same vein but also with the dorsal aorta and the posterior cardinal. As development goes’ on the first two of these connexions disappear so that the post-anal subintestinal vein is new continuous anteriorly only with the posterior cardinal. VVith the atrophy of the post-anal gut, it comes to lie immediately beneath the caudal portion of the dorsal aorta and is now known as the caudal vein, its anterior forked portion forming the binder ends of the renal portals.

The veins which have been described constitute the main trunks of the venous system; in the later stages of development a number of other important vessels appear which are indicated in the diagrams. The subclavian vein (Fig. 190, d, .9) appears about stage 31 + leading from the pectoral limb into the pronephric sinus. As this sinus atrophies the point of opening of the subclavian vein comes to be situated on the anterior cardinal vein (Fig. 190, e, s). From the subclavian vein a small lateral cutaneous vein (Lo) passes back VI VENOUS SYSTEM 413

in tl1e body-wall. An inferior jugular vein (7}.j) passes back from the head-region, lateral to the pericardiae cavity, and opens into the anterior cardinal close to its hind end and on its ventral side. In later stages the po_int of junction of the inferior ji1g11la1' with the anterior cardinal comes to be shifted relatively forwards, as was the case with the suhclavian. Farther forwards the anterior cardinal is joined by an anterior and a posterior cerebral vein (ce’ and ca”) from

d.C.

A B C

D

Flu. 191.—Diagrams illustrating early stages in the development of the venous system of Elasmobranchs according to Rah] (1892) and Hochstetter (1906).

«.0. v, anterior cardinal vein; om, fused portion of \-'it'..ellim- veins behind liver; rl, posii.ion of cloaca; 4l._('_', duct oi‘ Uuvier; h..:', hepatic vein; p.(_'.'1‘, posterior cardinal vein; s.i.'v, subintestinal vein; ruv, vitelline vein ; :0/..s'.v, main vein from yolk sue.

the inside of the head. At the hind end of the system rectal (7') and pelvic (pl) veins open into the renal portal. The former appear to be the persistent remains of the anastomotic branches which in early stages connected the hind end of the subintestinal vein with the posterior cardinal.

ELASMOB_RANCIIII.——It is instructive to compare with the development of the venous system in a holohlastic vertebrate the corresponding phenomena as they occur in the Elasmobranchs, the lowest of the nieroblastic gnathostomes. An inspection of Fig. 191 brings out the most conspicuous difference, one that could be foretold a priori, 414 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

namely that, in agreement with the general principle «that animals with a large supply of yolk tend to show a precocious development of the vessels on the surface of the yolk, the vitelline veins and their derivatives make their appearance relatively earlier as compared with the cardinals than they do in Lepirlomlren.

Here again the first of the main trunks to make their appearance are the vite.lline veins, which by their fusion anteriorly form the heart (Fig. 191, A, 12.12). The most conspicuous difference in the next subsequent stages is that the two veins are for a considerable period strongly asymmetrical (Fig. 191, B), the right becoming greatly “reduced, and only the left being commonly traceable back into the subintestinal vein (1’r7Istiurus——Rab1, 1892 ; Accmth'ias——Hoffmann, 1893). There can be no reasonable doubt that this is to be looked upon as a secondary modification of a symmetrical condition of the two veins as seen in Le}7'id0s’iTe7b, and probably the impelling factor which has brought about the modification is to be recognized in the fact that the main channel for draining away the blood from the surface of the yolk-sac is situated on the left side and opens into the left vitelline vein (Fig. 191, 3/.s.v).

The unpaired mesial prolongation forwards of the subintestinal vein between the two vitelline veins which was so conspicuous in Lepidoselren has not been noticed in Elasmobranchs.

Both vitelline veins break up into a network as they traverse the liver, the parts of the veins in front of this network persisting as the hepatic veins of the adult (Fig. 191, l), /my). The portion of right vitelline vein behind this network is for a time much reduced, the left alone serving to supply the network with blood. Immediately behind the live.r, and in front of the yolk-sac vein, the two vitelline veins are fused together for a short distance into a single vessel (Fig. 191, C, an), and the same is the case again from about the level of the pancreas backwards where they form the subintestinal vei11, though it should be noticed that there still persist during early stages of the development of the subintestinal vein in Elasmobranchs distinct traces of the paired condition, the vein consisting of two parallel components situated on each side of the mid-ventral line of the intestine. These components soon become connected together by numerous anastomoses and eventually fuse completely to form a median vessel except where this is prevented by the presence of the cloaca.

The main vein from the yolk-sac (3/.s.-v) joins the left vitelline vein in the region in front of the pancreas and presently the portion of left vitelline vein behind the opening of the yolk-sac vein disappears. It thus comes about that the blood from the subintestinal vein passes forwards to the liver entirely through the right vitelline vein---in contrast with Lepidosiren where it did so by the left vitelline vein.

Cardinal veins and duct of Cuvier make their appearance, the chief difference from Lepidosdren being the comparative shortness of VI VENOUS SYSTEM 415

the duct of Cuvier which does not take the wide sweep over the surface of the yolk that it does in Lepitlosrkren during early stages. This difference is related to the fact that here the first rudiment ‘of the duct of Uuvier opens into a portion of the vitelline vein which has fused witl1 its fellow to form the hind end of the cardiac tube, while in Lep'izl0s'i7'e7t it opens much farther back, the result being that in Le_p'i(l0.‘i'r6'n a considerable stretch of free vitelline vein becomes incorporated in the definitive duct of Cuvier (compare Figs. 191, C and 190, b).

As in Lepwlclosirevi. the caudal vein becomes continuous with the posterior cardinals and loses its continuity with the pre-anal portion of the subintestinal vein. The inter-renal vein however develops here simply as a forward extension of the caudal vein according to Rabl. A number of anastomotic vessels connect up the inter-renal vein with the “posterior cardinal ”—-—the equivalent of the external component of this vein in Lep'idos'£7'en. The posterior cardinal now becomes obliterated behind the anterior one of’ theseanas tomotic vessels while the inter-renal becomes separated off from the caudal vein so that the whole blood-stream from the latter has to pass through the kidneys to reach the inter-renal. The latter vein splits into a pair of vessels eventually, thereby revealing that the inter-renal vein here is homologous with that of L6p'icl0si}'c7l in spite of’ its difl'erent———no doubt secondarily modified-—mode of development.

The anterior cardinal vein here again becomes in part replaced by a lateral cephalic vein.

In the lower vertebrates in general we may recognize the same main trunks as occur in Dipnoi and Elasmobranehs, with differences in detail. The following account gives an outline sketch of the development of the venous system in the various groups, the outline being filled in more fully in the case of 1’ol3/pte7"u.s on account of the very archaic character of this fish.

CYeI.0s'1‘0MA'1‘A.———ln the Lamprey, according to Goette (1890), the pair of vitelline veins appear first, spreading backwards on either side of the liver rudiment and meeting behind in the unpaired and much dilated subintestinal vein (Fig. 192, A). The vitelline veins break up into a network in the liver b11t on the left side the posthepatic section of vitelline vein disappears, so that the hepatic portal vein is formed by the subintestinal and right vitelline vei11 (Fig. 192, B and ())—somewhat as in the Elasmobranch, and unlike Legnldosiren where it is the right vein which disappears. 'l‘he vein of the “spiral valve.” of the intestine arises comparatively late, at the time of metamorphosis according to Goette, and on the opposite side of the gut from that on which the subintestinal vein lies. This latter is no longer ventral but high up on the right side, owing to a rotation which the gut has undergone.

The anterior and posterior cardinal veins present the peculiarity that they open at first separately into the vitelline veins. Later they become fused together to form the duct of Cuvier. Eventually the 416 EMBRYOLOGY OF THE LOWER VERTEBRATES on.

left duct of Cuvier disappears completely (Fig. 192, C) the blood from the left cardinals passing to the right side by a new anastomotic vessel which develops ventral to the dorsal aorta (Fig. 192, C, an)—-an arrangement presenting a remarkable analogy with what happens in certain Mammals.

CROSSO1’TERY(‘}1'I.——-II1 these archaic Teleostonies the main features of the development of the venous system have been investigated in P0lypterus—the less specialized of the two surviving genera (Graham Kerr, 1907).

In the earliest stage described there is a well-developed subintestinal vein which in front breaks up into a vitelline network.

Fm. 19‘3.-—-l)evelopment of veins in P(’tI'()uJ_7/2-;'n-N. as seen from the ventral side. (Al'1'er Goette, 1890.)

c_z.r.r, anterior c-ardinal vein; rm, anastomotic vein ; (l.l', duct. of (juvia-r; hm, lmpatic veins; Ii, outline oI.']i\'e1'; 1.11:9, left vitelline vein; p.rt.:', posterior ca1'tlina.l win; r. 121', right vitelline VI-in; s.i..a_-, suhintr-stinal win. Portions of the venous trunks which disappear during,‘ ontogeny are shown by dotted 0llhlllll'h' ; the outline of the liver is shown in IS by an into-rrnptetl line.

This drains into a pair of lateral vitelline veins which unite in front to form the heart. Postcriorly the subintestinal vein bifurcates to pass on each side of the cloaca and then joins again to form the post-anal. subintestinal vein. In its double cloacal portion, and also in front of this, wide communications pass between the subintestinal vein and the posterior cardinals. The ultimate fate of the subintestinal vein differs in its pre-cloacal and post-cloaéal portions. The former loses its identity in that it becomes entirely resolved into, portions of the vitelline network. The post-cloaeal portion becomes converted into the caudal vein in the normal fashion as already described for Lepid0s'i'r(m. It has already been mentioned that wide communications were established between the paired cloacal portion of the subintestinal vein and the posterior cardinals. With the breaking up of the pre-cloacal part of the vein the main VI VENOUS SYSTEM 417

blood-stream passes through these communications into the posterior cardinals and the latter take on the appearance of a direct forward prolongation of the caudal vein.

Dorsally there develops on each side a cardinal trunk which swells o11t into a great irregular sinus (gm) in the region of the pronephros. On its outer side branches pass from the pronephric sinus into the general vitelline network. In this network a specially wide channel develops, starting from the pronephric sinus and sWee.ping outwards over the yolk to join the lateral vitelline vei11 and so reach the heart. This channel, which becomes gradually more and more sharply defined, is the duct of Cuvier (Fig. 193, A, (1.0). The pronephrie sinus is continued backwards into the posterior cardinal vein. This is at first distinct from its fellow but at an early stage fuses with it to form a median inter-renal vein (Fig. 193, B, 7}/-)——the fusion being foreshadowed by the developme.nt of anastomotic eonnexions between the two veins while still separated from one another by a distinct space (Fig. 193, A, cm). Towards the eloaea the poste1'io1' cardinals taper off and are connected by irregular anastometic chan11els_ with the subintestinal vein, as already mentioned, and also with the dorsal aorta. Eventually, as -already indicated, the inter-renal vein and the caudal vein form a continuous vessel.

During the later stages a striking asymmetry becomes apparent in the anterior, unfused, portions of the posterior cardinals——the left becoming‘ greatly reduced as compared with the right (Fig. 193, D). The main blood-stream thus passes forwards on the right side, and upon this side a special direct channel develops en the ventral side of the pronephres, through which the blood-stream is able to reach the duct of Uuvier without passing through the tangle of pronephric tubules. The asymmetry affects also the ducts of (}uvier—— showing itself first in that of the left side becoming relatively shorter than its fellow, which retains for a time its wide sweep over the surface of the yolk (Fig. 193, B and (J, d.(/'). Eventually it too becomes shortened and its calibre becomes considerably greater than that of the left side (Fig. 193, D).

From the prenephric sinus a branch (lxz-) develops about stage 30 which passes dorsalwards and then backwards beneath the lateral line nerve——tl1e lateral cutaneous vein. This, a large vessel about stage 33, becomes reduced to an insignificant vestige later on.

The anterior zardinal vein runs along the side of the head region, passing through the angle on the ventromesial side of the otocyst, between the latter and the brai11-wall. At its front end the vein dilates into a large sinus, which gives off irregular branches to the mesoderm of the head. At an early stage an anastomotic channel makes its appearance on the outer side of the otocyst continuous anteriorly and posteriorly with the anterior cardinal. When this channel has been established (Fig. 193, A, Le) the blood-stream from the head divides in front of the otocyst and passes backwards, part Fm. 193.——Devc1opmeut of dorsal venous systeiii of 1"(')ly}Jte‘l"lt8_, as seen from the dorsal side.

Stages 27-29 (A); 31 (B): 33 (C); 30 mm. larm (D and E). a.c.w, anterior cardinal vein; an, aliastomotic vein; (1.! ‘, duct of Cuvier; hm, hepatic vein; h.('p.*v.('), pmzteriol‘ vena cavn; i.j, inferior jugular ; 1'7‘, inter-renal vein; l.c, lateral coplialic; I.«v,1ato,ru1 cutaneous; 31, pulmonary vein; p.c.v, posterior cardinal; pn, pronepin-ic sinus; s, .~:ubc1m'ian Vein ; ’l'h, thyro d.

418 CH. v1 VEIN S OF POLYPTERUS 419

on its outer and part on its inner side. Eventually the inner channel becomes constricted across and finally completely severed at its hinder end so that the whole blood-stream passes back by what was the outer channel. The inner channel persists as a small vein which opens at its front end into the definitive anterior cardinal. We see then that here, as in the Lung-fish or the Elasmobranch, the “ anterior cardinal” vein of later stages has intercalated in its length a segment of lateral cephalic vein.

Au inferior jugular vein on each side ('23./') drains the. blood from the ventral side of the head into the. duct of (juvier, and development goes on these become asymmetrical, the left becoming greatly reduced while the right forms a large vessel, trifid at its front end where it receives the blood from the thyroid (Fig. 193, D, T/1,).

'J‘he liver rudiment, at first a portion of the general yolk-mass, is at this stage supplied with blood by the portion of the general vitelline network which extends over it. The. portal vein develops simply as an enlarged channel of the network on the left side of the hepatic rudiment. There is no information as to whether this is really the persisting left lateral vitelline vein as we might expect: nor is it known whether the hepatic vein is derived from the prehepatic portion of this same vein.

An important feature is that there becomes established an anastomosis between the, blood-vessels of the liver at the hinder end of that organ and the inter-renal vein, so that a portion of the interrenal blood - stream becomes short - circuited direct to the heart through the hepatic vein, which in correlation with this forms a wide channel throughout the length of the liver (Fig. 193, E, h.(p.'r.c)). This enlarge.d hepatic vein is of morphological importance as it is clearly the equivalent of a posterior (or ‘‘inferior’’) vena cava. In the. Crossopterygian this vessel takes a step in evolution beyond the condition in Lung-fishes, inasmuch as its posterior portion becomes denuded of liver substance, so that it runs for a considerable distance through the splanchnocoele as a naked vessel.

Another important vein which makes its first appearance in the (lrossopterygian is the pulmonary vein. The hepatic vein, which towards its headward end lies on the dorsal side of the liver, comes into close contact with the pharyngeal floor in the region of the glottis, and venous spaces developing in the mesoderm sheath of the pharynx come to open into it. These are apparently the forerunners of the pulmonary vein. In the 30-min. larva, as in the adult, there is a main pulmonary vein which still opens into the hepatic vein on its dorsal side (Fig. 193, D, 1)). In addition to this main pulmonary vein, formed by tl1e. fusion of two branches coming from the ventral side of the two lungs, there is a small accessory vein coming from the dorsal side of the root of each lung and from the adjoining parts of the pharyngeal wall. These also open into the hepatic vein just in front, and on each side, of the opening of the main vein.

TEl.EOS'1‘El.——A1I10ngSt the Teleostean fishes We find the same 4'20 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

venous trunks laid down as in the Lung-fish or Elasmobranch. There is within the group considerable variability but the variations are as a rule derivable from a primitive type like that of Lepidosiren. Thus in Salmo the subintestinal vein bifurcates in front into the two paired vitelline veins while in numerous other Teleosts (Esom, Bclone, S3/ngnathas, Iflppocampas, Gobius) it passes forwards into a median unpaired vitelline vein. Each of these conditions is obviously derivable from that illustrated by Jjepidosiren, by the disappearance, on the one hand, of the median, and, on the other, of the paired vitelline veins.

AMPnmIA.—-In S'a.l(mtu.n¢l7'a, (Uhoronshitzky, 1900) two lateral vitelline veins are described, the right comparatively small in size. Behind the liver rudiment they lie close together near the midventral line and passing forwards they diverge, passing one on each side of the liver rudiment to unite in front of it and form the hind end of the heart. The two veins undergo fusion behind the liver to form the subintestinal vein and in front of the point of fusion the right vein disappears so that, as in the Lung-fish, all the blood passes to the heart round the left side of the liver. The mesenteric vein develops as a branch of the right vitelline vein close to its front end and after the disappearance of the greater part of the right vitelline vein the mesenteric is seen replacing it as the right limb of a horseshoe-shaped arrangement of veins which embraces the «liver rudiment from in front. The portion of the vitelline veins in front of the mesenteric breaks up into the hepatic network. The vitelline vein shrinks to an inconspicuous vestige while the mesenteric becomes relatively large and forms the hepatic portal of the adult.

The posterior cardinal veins (Hochstetter, 1888) run alongside the archinephric ducts, which they more or less surround, to the region of the pronephros where each dilates to form a large pronephric sinus. I11 the opisthonephric region the vein forms two main channels an inner and an outer (Fig. 194, B, op), the former eventually undergoing fusion with its fellow to form an inter-renal vein which becomes later the renal portion of the posterior vena cava. The outer channel, as in the Lung-fish, becomes continuous with the caudal vein to form the renal portal, while at the front end of the opisthonephros it loses its connexion with the part of the vein lying farther forwards. A venous connexion is established between the front end of the inter-renal vein and the tip of the liver, and the channel which so arises, commencing behind in the inter—renal vein, traversing the substance of the liver and ending in the hepatic vein, forms the posterior vena cava (Fig. 194, 0, 10.0.0). In later stages the liver tissue disappears over the greater part of the posterior vena cava so that it passes naked through the splanchnocoele.

The anterior cardinal vein with its intercalated section of lateral cephalic persists as the internal jugular vein of the adult. The Duct of (juvier also persists in the adult and is now termed the

anterior vena. cava. VI VENOUS SYSTEM 421

The anterior abdominal vein arises as a pair of small veins in the ventral body-wall. These unite in front in the region of the liver and open into the left duct of Cuvier according to Hochstetter. Later the two veins fuse into a single unpaired vessel except at their hinder ends where they become connected with the renal portal vein on each side. Anteriorly the opening into the duct of Cuvier becomes replaced by an opening into the hepatic portal vein.

d.C. Jill. act». /2 a. pn. si 11.. par op ’ (‘U A B C 0

Flu. 19-1.—~Development of main venous trunks in Salanuuzdra according to Hoclistctter (1888).

u.r.u, anterior ('.'1r(linal vein; c.v, vandal win ; ¢I.(', duct of Cuvwr; hm, hepatic \cm ; op, opisthonephros; p.r‘.v, p0St~l‘I‘l()l cardinal vein; p.'v.c, posterior wna vava; pn, prenephric sinus; r.;p, renal portal; s, subclavian ; s.’i.I', subinte-stmal wm ; 3.1‘, sinus venosus.

SAUROPSIDA.—-—Here we find the same main venous trunks as in the lower groups and, in correlation with the large quantity of yolk, the ventral or vitelline system of veins develops preccciously as compared with the dorsal or cardinal system. .

Lacerta, which has been investigated in detail by Hochstetter (1892), may be taken as an example of the Reptiles. The first veins to make their appearance are the lateral vitelline veins which pass forwards on the dorsal side of the yolk-sac, converging in front to form 422 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

the hind end of the heart. The two vitelline veins become connected by a transverse anastoniosis dorsal to the gut and just behind the dorsal pancreatic rudiment (Fig. 195, B). The parts of the veins in front of this anastornosis break up into a network in the substance of the li.ver, and the two networks become continuous with one another (Fig. 195, C). The two vitelline veins now form another anastomosis

Fm. 195.—-Diagrams to illustrate the development of the. ventral part of the venous system in Lacertu u._(/zll-is as seen from the ventral side. (After Iiochstetter, 1892.)

ab, abdominal vein; «LC, duct or (fuvler; :1.-v, ductus venosus; ant, alimentary canal: 1.0.11, left all:1.lll}(')l('. win; If, liver; 1.1.‘.-v, left vitelline vein; 7;, portal vein; 1». l‘.(.', posterior vona, cm-a; mull, right alluntoic vein ; -r.'v.-r, right vitelline vein.

with one another farther back and ventral to the alimentary canal (Fig. 195, C). The right vitelline vein becomes reduced and finally obliterated in the region in front of this ventral anastomosis so that the whole blood-stream passes forwards to the level of the dorsal anastomosis by the persistent left Vein (Fig. 195, D, l.v.v). In the region anterior to the dorsal anastomosis the left vein now diminishes in size and finally disappears, first in the region behind the hepatic VI VEIN S OF LACERTA 423

network (Fig. 195, D) and then in the region in front of the network (Fig. 195, E). Results of these changes are (1) that the hepatic network receives its blood supply by a single afferent vessel——the hepatic portal vein—-which curves round the gut and is derived in great part from the left vitelline vein——and (2) that its blood drains away to the heart by a single efferent vessel~—the hepatic vein——derived from the front end of the original right vitelline vein.

At a comparatively early stage in development a direct channel becomes established, by the widening out of the venous spaces along the middle of the hepatic network, so that a considerable proportion of the blood is able to pass forwards from the portal vein through the liver without actually traversing the network itself. This cha.nnel——the ductus venosus (Fig. 195, G, 11.1)) persists till nearly the period of hatching but then becomes obliterated so that all the portal blood has to traverse the hepatic network.

The posterior vena cava makes its appearance as a gradually widening channel through the hepatic network towards its right side (Fig. 195, D, E, [).'I7.6). This portion of the liver becomes prolonged backwards as a slender lobe ensheathing a prolongation of the blood channel mentioned. This prolongation fuses at its tip with the tip of the right opisthonephros, continuity becomes established between the venous spaces of the two organs and finally, as in the Amphibian, the liver tissue disappears ove.r a large stretch of the slender lobe already mentioned so that the vena cava is now for a considerable length free from either liver or kidney.

At an early stage a branch of the vitellinc vein develops close to its front end. This is the allantoic or umbilical vein (Fig. 195, roll and loll). These veins soon become asymmetrical, the left for a time being smaller than the right (Fig. 195, I), E). A little later however the left vein establishes a connexion with the hepatic network (Fig. 195, F); the portion of the vein posterior to this connexion becomes much widened, and the blood-stream from it courses by an enlarged direct channel of the network into the posterior vena cava (Fig. 195, F, G). The blood-stream being diverted through this channel, the portion of the left allantoic vein in front of it shrinks in size and disappears, as does the Whole of the right allantoic vein (Fig. 195, F, G). The result is that there persists a single (left) allantoic vein which drains the blood from the allantois into the posterior vena cava near its front end. The allantoic vein increases in size with the allantois but becomes obliterated at the time of hatching when the allantois is cast off. The mesenteric vein develops as a branch of the portal vein (left vitelline vein) a short distance behind its entry into the liver: it increases in size as the vitellinc diminishes with the consumption of the yolk and eventually it alone persists as the peripheral portion of the definitive portal vein of the adult.

The subintestinal vein is apparently present only in its post-anal portion which persists as the caudal vein of the adult. In front 424 EMBRYOLOGY OF THE LOWER VERTEBRATES CH.

of the anus, where the ventral wall of the primitive alimentary canal has become intensely modified in connexion with the storage of yolk, the subintestinal vein has disappeared from the course of development.

As regards the dorsal venous system (Fig. 196), the two posterior cardinal veins converge posteriorly and become continuous with the caudal vein. The portions in the region of the opisthonephros become resolved into their external and internal components connected by numerous sinus~like spaces and channels amongst the kidney tubules (Fig. 196, A). With the development of a capillary network

A B C

as

C“-"

Flo. 196.——Diagr:un illustrating the development of the dorsal venous system in l.m.-arm according to Hochstetter, as seen from the ventral side.

a..c.'v, anterior cardinal vein ; «-, (-.:audnl vein; d.(J, duct of Cuvier (-—-ant. vena cava); ii, iliac vein ; 1r.c.v, posterior cardinal vein; 7-:.zr.r.', posterior vena cava; .-, subclavian vein.

in the substance of the opisthonephros the larger blood spaces become divided into an afferent set connected with the external component and an efferent set connected with the internal one. The external channel remains continuous with the caudal vein and forms the renal portal vein. The two internal components fuse together in their anterior portion (Fig. 196, B) and become continuous with the intrahepatic portion of the vena cava. Posteriorly they remain separate and lose their continuity with the caudal vein (Fig. 196, B, C). The blood from the kidneys being now able to pass to the heart by the direct route through the posterior vena cava, the portions of posterior cardinal lying in front of the kidneys are no longer required and soon disappear (Fig. 196, C). VI VEN OUS SYSTEM 425

In the head region the anterior cardinal becomes in great part replaced by a lateral cephalic vein in a manner similar to that already described for Lc_p'id0.s"i7'07t.

B1nDs.——In Birds the development of the venous system pursues, as we should expect, a similar course to that already described for Reptiles. Amongst the differences in detail the most striking is that the two vitelline veins become completely fused into a single vessel, the ductus venosus, through the hepatic region, before there are any signs of a hepatic network. This may be regarded as a backward extension of the fusion of the two vitelline veins which gives rise to the heart. The ductus venosus secondarily becomes surrounded by tile. liver rudiment and a network of channels spreads out from it in the liver substance. The allantoic veins behave as in Lmrerta except that a small vestige of the left is said to persist throughout life.

The reduction of the tail region in modern birds has brought with it a modification of the caudal vein which is here paired, taking the form of a simple backward prolongation of the posterior cardinal. The main channel of the posterior cardinal runs along the outer edge of the opisthoneppros but later on a slender channel appears along its inner edge—that on the right side being continuous with the posterior vena cava of which it forms simply a backward prolongation. The two inner channels undergo fusion so that the blood from the kidneys can drain away entirely into the posterior vena cava and this is followed as in other cases by the atrophy of the. portion of posterior cardinal lying in front of the opisthonephros. This atrophy extends as far forwards as the subclavian vein which in the Fowl opens into the posterior cardinal vein some distance from its front end. The portion of posterior cardinal lying in front of this point is consequently saved from disappearance and persists as a portion of the definitive subclavian vein of the adult. It will be. understood that the blood of the outer channel of the posterior cardinal, which reaches it from the caudal vein, from the posterior limb, and from the body-wall, passes entirely through the_ opisthonephric network towards the posterior vena cava, in other words that there is at this time a typical renal portal system.

As the metanephros develops, its tubules are also mixed up with the sinuses connecting external and internal channels of the posterior cardinals, so that it too has for a time a functional renal portal system. Later on however one of the channels through the metanephros becomes enlarged and the blood-stream passes directly through it to the posterior vcna cava without traversing the meshes of the network. A true renal portal system then no longer exists and the reason for its disappearance is no doubt to be found in the fact that the vascular network of the kidney has become connected with the arterial system. Obviously this will give a much more efficient circulation than the original one, owing to the. higher blood pressure in the dorsal aorta and renal arteries than in the renal portal veins which have the systemic network of capillaries interposed between them and the heart. Further the quality of the blood supplied in this way to the kidney is better——being arterial instead of venous—-and for both these reasons we can readily understand the tendency in the more highly developed vertebrates for the renal portal system to disappear.

It has already been remarked that the posterior cardinals do not pass back into an unpaired caudal vein as in the Lizard. A vestige however of the unpaired condition may perhaps be recognized in the development of an anastomosis between the two vessels just behind the metanephros. From the transverse bridge so formed a connexion (coccygeo-mesenteric vein) is established with the portal vein in the mesentery.

The anterior cardinal vein together with an intercalated portion of lateral cephalic persists as the jugular vein of the adult bird.

The posterior cardinal vein undergoes in the Bird a curious change of position in relation to the root of the iliac artery which it crosses behind the mesoncphros. At first it lies on the ventral side of this artery: then it develops an accessory channel round the dorsal side of the vessel and finally the whole blood-stream passes by this dorsal channel while the ventral one disappears. This affords a good example of the way in which a vein may in the course of evolu tion pass an apparent barrier formed by an artery, nerve or other organ.

INTER-SEGMENTAL VE1Ns

In the body wall there develops a series of veins corresponding with tl1c interscgmental arteries and opening into the cardinal veins.

VEINS OF LIMBS

The vascular network of the limb-bud drains into the posterior cardinal vein. In the Bird (Evans, 1909) the, drainage during its earliest stages is into the allantoic vein. Later numerous cl1an11els arise connecting the network with the posterior cardinal and presumably one or two of these become enlarged and persist as the definitive veins draining the hind limb.

LYMPHATIC SYsTEM

The venous system has its obvious roots peripherally in the capillary network of blood-vessels, but it is also provided with a much less conspicuous set of tributary channels which constitute the lymphatic system. This extension of the vascular system retains a lower grade of evolution than the remainder. Its channels are less sharply defined, the lining endothelium over most of its extent having a much feebler development of the backing of connective tissue and muscle which forms the thick wall of the vein or artery. In its peripheral portions the lymphatic spaces may have remained practically in the primitive condition of intercellular chinks of the mesenchyme, while in its central portions, as it approaches the points at which it opens into the ordinary veins, its walls may be Well developed and muscular. The lymphatic system serves to drain oil’ the plasma which has oozed out from the capillary blood-vessels and forms the internal medium bathing the surface of the living cells of the body, and to return it to the blood-stream.

The fluid is peopledfby amoeboid eorpuscles but is without the red corpuscles which have no power to escape through the walls of the blood-vessels.

Our present knowledge of the ontogeny of the lymphatic system is in great part due to the labours of Huntington and McClure whose papers should be consulted as regards details. It seems clear that as a general rule lymphatic channels develop, later than the blood - vessels, as intercellular chinks in the mesenchyme which become continuous and form definite channels, the bounding cells becoming converted into thin endothelium.

Spleen

The spleen arises in Lepidoszveot and 1’r0topte7'us (Bryce, 1905 ; Purser, 1917), which may be taken as typical examples, in the form of a condensation of the mesenchyme of the gut-wall. Blood spaces soon make their appearance in the rudiment which later becomes intercalated in the course of the main venous channel leading from intestine to liver. Later on the main blood-stream passes to the liver by a direct channel, the spleen now lying on a lateral loop : later still the afferent part of this loop becomes replaced functionally by a new arterial connexion.

The spleen rudiment frequently arises in close proximity to that of the pancreas and this has led to statements that the spleen is actually derived from the pancreas but the probability seems to be that such statements are based upon erroneous observation.

Literature

Allis. Zool. Jahrb. (Anat.), xiv, 1900.

Bemmelen, van. Zool. Anzeiger, ix, 1886.

Bemmelen, van. Meded. tot de dierkunde. Amsterdam, 1887.

Boas. Morph. Jahrb., vii, 1882.

Bryce. Trans. Roy. Soc. Edin., xli, 1905.

Choronshitzky. Anat. Hefte (A1-b.), xiii, 1900.

Dohrn. Mitt. Zool. Stat. Neapel, vii, 1886.

Dohrn. Mitt. Zool. Stat. Neapel, viii, 1888.

Evans. Amer. Journ. Anat., ix, 1909.

Gegenbaur. J enaiseher Zeitsehrift, ii, 1866.

Goette. Abhandlungen zur Entwickelungsgeschiclate der Tiere, v. Hamburg and

Leipzig, 1890.

Greil. Morph. .laln'b., xxxi, 1903.

Hochstetter. Morph. Jahi-b., xiii, 1888.

Eochstetter. Morph. Jahrb., xvi, 1890.

Hochstettor. Morph. .Tahrb., xix, 1892.

Hochstetter. Hertwigs Handbuch der Entwicklnngslehre, iii, 1906. Hochstetter. Voeltzkows Wissensehaftliehc Ergebnisse, iv. Stuttgart, 1906*. Hofimann. Morph. Ja.hrb., xx, 1893.

Hoyer. Bull. intern. Acad. Sci. Cracovie. (Jomptes rendus des Séances, 1900. Huntington and others. Anat. Record, ii, 1908.

Huntington. Anat. Record, iv, 1910.

Kellicott. Memoirs N. Y. Acad. Sci., ii, 1905.

Kerr, Graham. The work of J. S. Budgctt. Cambridge, 1907.

Kerr, Graham. Proc. Roy. Phys. Soc. Edin., xvii, 1907*.

Kerr, Graham. Proe. Roy. Phys. Soc. Edin., xviii, 1910.

Langer. Morph. Jahrb., xxi, 1894.

Mackay. Phil. Trans. Roy. Soc., B, clxxix, 1889. 428 EMBRYOLOGY OF THE LOWER VERTEBRATES CH. VI

McClure. Aunt. Record, ix, 1915.

Maurer. Morph. .Iuhrl)., xiv, 1888.

Mollier. IIc1't\\ig'.-4 I-[mulbuch «lor Entwicklungslelm-, i, 1906. Muller. Abhand. Akad. Wiss. Berlin, 1845.

Miiller, F. W. Arch. mikr. Anat., xlix, 1897.

O’Donog'hue. I’:-nu. '/.00]. Soc.'Lond., 1912.

Purser. (311zn't. .lo1u'n. Mior. Sci., lxii, 1917.

Rabl, C. Lt-1u-kawts l"vstscln'ift. Leipzig, 1892.

Rabl, H. Arch. mikr. Anat., lxix, 1907.

Ridewood. l’l'0c. '/.001. S00. Lund., 1899.

Robertson. Quart. .lo1u'11. Micr. S(:i., li\', 1913.

R586. Morph. J:Lh1'b., .\'Vi, 1890.

Ruckert. Biol. ()9IItI':1l1)laLt, viil, 1888.

Rlickert. Hortwigs Hamlbuch dvr Entuicklungslohro, i, 1906.

Historic Disclaimer - information about historic embryology pages

Pages where the terms "Historic" (textbooks, papers, people, recommendations) appear on this site, and sections within pages where this disclaimer appears, indicate that the content and scientific understanding are specific to the time of publication. This means that while some scientific descriptions are still accurate, the terminology and interpretation of the developmental mechanisms reflect the understanding at the time of original publication and those of the preceding periods, these terms, interpretations and recommendations may not reflect our current scientific understanding.     (More? Embryology History | Historic Embryology Papers)

- Currently only Draft Version of Text -

Textbook Chapters: 1 Formation of the Germ Layers | 2 Skin and Derivatives | 3 Alimentary Canal | 4 Coelomic Organs | 5 Skeleton | 6 Vascular | 7 Internal Body Features | 8 Adaptation to Environmental Conditions | 9 General Considerations | 10 Common Fowl | 11 Lower Vertebrates | Appendix

Reference

Kerr JG. Text-Book of Embryology II (1919) MacMillan and Co., London.

---

Cite this page: Hill, M.A. (2026, October 20) Embryology Text-Book of Embryology 2-6 (1919). Retrieved from https://embryology.med.unsw.edu.au/embryology/index.php/Text-Book_of_Embryology_2-6_(1919)

What Links Here?

© Dr Mark Hill 2026, UNSW Embryology ISBN: 978 0 7334 2609 4 - UNSW CRICOS Provider Code No. 00098G