Paper - The origin of the heart and blood vessels in felis domestica (1924)

From Embryology

{Header}} Watson KM. The origin of the heart and blood vessels in felis domestica. (1924) J Anat. 58: 105-133. PMID 17104001

Online Editor  
Mark Hill.jpg
This 1924 paper by Watson describes the development of the heart and blood vessels in the cat.



Modern Notes: heart | blood vessel | cat



Cardiovascular Links: cardiovascular | Heart Tutorial | Lecture - Early Vascular | Lecture - Heart | Movies | 2016 Cardiac Review | heart | coronary circulation | heart valve | heart rate | Circulation | blood | blood vessel | blood vessel histology | heart histology | Lymphatic | ductus venosus | spleen | Stage 22 | cardiovascular abnormalities | OMIM | 2012 ECHO Meeting | Category:Cardiovascular
Historic Embryology - Cardiovascular 
1902 Vena cava inferior | 1905 Brain Blood Vessels | 1909 Cervical Veins | 1909 Dorsal aorta and umbilical veins | 1912 Heart | 1912 Human Heart | 1914 Earliest Blood-Vessels | 1915 Congenital Cardiac Disease | 1915 Dura Venous Sinuses | 1916 Blood cell origin | 1916 Pars Membranacea Septi | 1919 Lower Limb Arteries | 1921 Human Brain Vascular | 1921 Spleen | 1922 Aortic-Arch System | 1922 Pig Forelimb Arteries | 1922 Chicken Pulmonary | 1923 Head Subcutaneous Plexus | 1923 Ductus Venosus | 1925 Venous Development | 1927 Stage 11 Heart | 1928 Heart Blood Flow | 1935 Aorta | 1935 Venous valves | 1938 Pars Membranacea Septi | 1938 Foramen Ovale | 1939 Atrio-Ventricular Valves | 1940 Vena cava inferior | 1940 Early Hematopoiesis | 1941 Blood Formation | 1942 Truncus and Conus Partitioning | Ziegler Heart Models | 1951 Heart Movie | 1954 Week 9 Heart | 1957 Cranial venous system | 1959 Brain Arterial Anastomoses | Historic Embryology Papers | 2012 ECHO Meeting | 2016 Cardiac Review | Historic Disclaimer
Historic Disclaimer - information about historic embryology pages 
Mark Hill.jpg
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)

The Origin of the Heart and Blood Vessels in Felis Domestica

By Katharine M. Watson, D.Sc., Department of Embryology, University of London, University College


General Introduction and Methods

Tue following observations were made on a beautiful series of cat embryos in the collection of Prof. J. P. Hill, and I wish to express gratitude to him not only for permission to work on the material but also for his guidance and criticism, which have served as a stimulus throughout my work. I am indebted to the Scientific and Industrial Research Committee for a grant which has enabled me to devote part of my time to research during Sessions 1919-20, 20-21, and 21-22.


In this paper I shall deal only with very early stages (4-5 somites to 15 somites). The problems considered are therefore those of the actual origin and very early development of the heart and blood vessels. With regard to the heart, I was anxious to find out how far the processes described by Wang (15) for the ferret occur in other mammals. In this paper Wang describes a “primary union” of the heart primordia and states ‘‘It should be specially noted that the term ‘primary’ proposed signifies that the right and left parts separate from one another and are again fused by a secondary union” (p. 169). Unfortunately Wang has misunderstood and actually misquoted my paper on the early development of the heart in Marsupials. He gives the impression that I also found the same “primary” union in Dasyurus viverrinus, which is contrary to fact. The principal object of the present study is to give an accurate account, based on a closely graded series of stages, of the early development of the heart in a Mammal. The material also affords an opportunity for observing the earliest stages of development of the blood vessels. In recent years the problems of vasculogenesis have received a considerable amount of attention, and, as the result of the work of a number of American anatomists, the theory of the origin in situ of the vessels of the embryo has been definitely established. It is unnecessary to give here an account of the recent work on the subject as McClure(6) has given an admirable summary so recently as March, 1921. The relative merits of the method of injection and that of study of serial sections have been freely discussed by various authors. Obviously, for the study of the development of the vascular system as a whole, both methods are necessary. For the purpose of discovering the actual origin of the cells which line the blood vessels it is clear that the injection method is inapplicable.


The present paper deals with a series of ten embryos. The sections are all 10 thick. I have made one or more graphic reconstructions of each embryo which is cut transversely, and these are reproduced here as text-figures. The main object is to make clear the relations of the heart, pericardium and gut. I have, however, also graphed the aorta and aortic arch and the umbilical vein from their inception onwards. In so doing I have repeated some of the work published by Schulte in 1914 (12), and my results in general simply confirm his. I have not, however, followed his precedent and distinguished “‘hyperentodermal” mesenchyme cells from vasofactive cells. In the material I have studied, I cannot distinguish these two types of cells. Moreover, Schulte’s account of their characteristic features appears to me unconvincing. Referring to the distinction between hypectodermal mesenchyme and vasofactive cells, he says (pp. 42, 43): ‘‘ These (vasofactive cells) are readily distinguished from the other mesenchyme cells by their transverse position, in addition to their slightly more abundant protoplasm and slightly paler stain....The transverse position of the vasofactive cells requires explanation. They stand out clearly against the other mesenchyme cells which have their long axis in general radial, and are further distinguished by their rather deep position, near or in the almost cell-free interval between the bulk of the mesenchyme and the compact layer of parietal mesoderm.” Apparently he means this distinction to apply also to hyperentodermal mesenchyme and vasofactive cells, Unfortunately he does not explain to what plane the vasofactive cells lie “transverse” or the mesenchyme ones “radial” and his figures do not make this distinction clear. In his fig. 24, plates of ““mesenchyme cells” (green) form the continuation of “‘anlages of the dorsal aorta” and in reference to fig. 24 he says (p. 48): “A second endothelial sac in the aortic lines is situated just cephalad of the somites, ending near'the first somite in continuity with a flattened plate of mesenchyme cells.”’ It seems evident that such plates of mesenchyme cells must become absorbed in the formation of the aorta and if so they are, ipso facto, vasofactive cells. In any case no distinction in position or staining capacity can have greater weight in identifying a cell as vasofactive than the fact that such a cell does contribute to the formation of the wall of a blood vessel. Schulte appears to me to have based his distinction on the grounds of position and staining capacity and to have failed to justify this distinction when tracing the subsequent fate of the cells. Furthermore, in all probability the vasofactive cell is simply a modification of the mesenchyme cell so that the distinction is in any case of minor importance. In making my reconstructions I therefore included as nearly as possible every individual cell which lay between entoderm and mesoderm in every stage. A study of the series of graphs reproduced in this paper shows that these cells disappear gradually and their place is taken by blood vessels; it is therefore natural to conclude that these cells do form the walls of the blood vessels, and if so, they are vasofactive. The majority of the loose ‘‘hyperentodermal” cells (adopting Schulte’s term for the moment) thus undoubtedly contribute sooner or later to the vascular system and I shall therefore use the term vasofactive for them, though it is obvious that a few cells may be erroneously included in the graphs and descriptions under this name. This error will, I think, have no bearing on any point under discussion.


The accurate representation of so many individual cells requires the use of a fairly high power and the process is rather laborious even for a small number of embryos. I adopted a simple method of graphing which is perhaps worth a brief note. I selected a lens and eyepiece giving a good magnification (a Zeiss 8 mm. and No. 6 ocular were used throughout) and a convenient ocular micrometer. The value of the unit of this scale with the given lens and eyepiece may be taken as z mm. (in this particular case z = -008 mm.). If now we treat each unit of the micrometer as a millimetre, it is equivalent to multiplying each measurement by 1/a (in this case =.125). Using millimetre squared paper and taking the median groove on the brain plate as an axis we can plot a cell which is situated at a distance from the middle line of 30 units on the micrometer as 30 mm. from the axis. If we plot each section (10,) at a distance of 1-25 mm. from the preceding section, we obtain a fairly accurate graphic reconstruction at a magnification of 125, without multiplying each measurement. This is a great saving of time when each graph represents some hundreds of measurements. Obviously its ease depends on the value of the micrometer scale, but by the use of different lenses it is generally possible to get a condition in which the method is practicable. Moreover, a small error, if produced, is generally of no importance as it remains constant throughout a given series of graphs and in a case like the present one, will have no effect on conclusions. In the case of endothelial tubes, blood vessels and vasofactive cells, each section was carefully graphed, but for the outlines of the gut, pericardium and myocardium it was generally possible to get a fairly good outline by graphing every fourth section.

In the graphs each cell is represented by a dot; unless otherwise stated, all the cells lie between the mesoderm and entoderm.


Descriptive Portion

Stage I.

Felis domestica (81.8.09, C.), 4-5 somites.

Fig. 1. represents a graphic reconstruction of the pleuro-pericardial mesoderm and the vasofactive cells of the right side of this embryo. The pleuropericardial canal does not yet possess a continuous lumen, but its position is marked by a thickening of the mesoderm which shows a tendency to split into dorsal and ventral layers. The outline of this mesoderm is represented by a broken line in the figure (P.p.m.), the transverse lines indicate the absence of a lumen, the interruptions in these lines represent the occurrence of a cavity.

1 N.B. In all the graphs the following conventions have been observed: Outlines of the pericardium and myocardium are indicated in fine broken lines. Outlines of the gut are represented in heavy broken lines. Outlines of the brain and layers of ectoderm are represented by an unbroken line. Vasofactive cells are shown as dots.

Blood vessels and angiocysts are represented by solid black. All graphs are reproduced at a magnification of 62:5.


It is thus seen that the pleuro-pericardial mesoderm is recognisable as a continuous horseshoe around the anterior end of the embryo, reaching its maximum development at the level of the first somite. Small, isolated clefts are present in the we yo ye™ ee .C. mesoderm throughout the extent of the ;AMP’“"Sa. horseshoe, and in the region of the greatest ‘ ve extent of the pleuro-pericardial cavity a aa continuous lumen is present. Behind this region the pleuro-pericardial canal narrows abruptly and becomes unrecognisable at the level of the posterior margin of the third somite.


The vasofactive cells in the head end of the embryo may be seen to be grouped in two main lines, the one nearer the median line representing the future dorsal aorta, the lateral one being the primordium of the endothelial heart tubes and vitelline veins. Only in the anterior end of the aortic line is a definite, hollow vessel present (4.D.A.). It is easily seen from the graph that the vasofactive cells of the aorta fall into anterior and posterior positions, separated by a gap of about 100 mm. This interruption in the line of the aorta is recognisable in each of the early embryos figured and is also seen in some of Schulte’s (12) figures. The vasofactive cells of the anterior portion of the aorta are not so numerous as those of the posterior, which moreover become denser as we pass backwards. It is a matter of considerable difficulty to determine the exact site of origin of the cells of the aorta line, but the appearances in this embryo and in the 7 somites and 8 somites stages are fairly conclusive in regard to the anterior portion of the aorta. The vasofactive cells in front of the gap in the aorta have every appearance of arising directly from the head mesoderm which is here fairly sparse (fig. 2). In the region of the gap, the mesoderm is distinctly more closely packed and less thick dorsi-ventrally. It is noteworthy that the medullary plate in this region is flat and widely open, whereas both in front and behind the gap in the aortic line, medullary folds rise up to a greater extent. This fact probably accounts | for the compactness of the mesoderm and possibly also for the absence of vasofactive cells in the aorta line. Probably thecells in the aorta line behind its interruption and in front of the first somite also arise from the mesoderm lying immediately dorsal to them. In the region of the somites many of the cells of the aorta lie directly ventral to the somite and, as pointed out by Schulte (12), appearances do not suggest the origin of vasofactive cells from somites in such a stage as this. On the other hand, in this region both the somitic stalks and the splanchnic mesoderm appear to be active in producing vasofactive cells and it seems possible that the cells of the aorta have arisen laterally and migrated towards the middle line (see fig. 2).



Fig. 1. Felis domestica, 4-5 somites (31.3.09).


Stage I. Graphic reconstruction of the pleuro-pericardial mesoderm and vasofactive cells of theright side of the embryo. The solid portions of the pleuro-pericardial mesoderm is indicated by lines drawn parallel with the sectional plane. The region where the canal is developed is left unlined. The arrow marks the anterior border of the first somite.



Fig. 2. Felis domestica, 4-5 somites (31.3.09). Transverse section. In position indicated in fig. 1. x 130.


Scattered cells lie between the two main lines of vasofactive cells which at the level of the second somite are connected together by a network of cells. The cells destined to form the endothelial heart tubes are scattered throughout the antero-posterior extent of the pleuro-pericardial mesoderm, reaching actually to the middle line anteriorly. They become most dense in the region of the somites; where they also become continuous with those of the extraembryonal area and with the aortic line. Behind the portion represented in the graph, the vasofactive cells gradually become less concentrated.


Fig. 3. Felis domestica, 4-6 somites (31.3.09, C.). Transverse section in region of first somite (position shown in fig. 1). x 130.

At the anterior end of the head of this embryo there are a few vasofactive cells which are destined to contribute to the formation of the vena capitis medialis. Some of these are shown in fig. 2 (V.C.M.), where on one side of the embryo the cells are actually grouped so as to surround a small lumen. These cells occur in the region of maximum development of the dorsal aorta (i.e. 110 where it actually forms an angiocyst), but are separated from it by a continuous layer of head mesenchyme. Moreover, in some sections the appearances point very definitely to the direct origin of such cells from the dorsal surface of the head mesenchyme. In the following stage, the same appearances will be noted.

This embryo, and, indeed, all this series, show a number of cells between the mesoderm and the ectoderm such as are described by Schulte (12), In the head region these all appear to belong tothe venacapitis medialis; in the region of the somites many of them arise from the somatic layer of the mesoderm of the pleuro-pericardial canal, but others are situated close beside the neural tube in the position characteristic of the vena capitis medialis and its posterior continuation (fig. 3).

Stage II.

Felis domestica. 31.3.09. A,and B., both embryos, having 7 somites,

Fig. 41 represents a graphic reconstruction of the pleuro-pericardial canals, aorta, endothelial heart tube and vasofactive cells of the left side of embryo A. The outline of the myocardium which is now beginning to develop is also represented. The relations of the parts in median longitudinal section are shown diagrammatically in fig. 4a.

The pleuro-pericardial canal is now luminated throughout the greater part of its extent, but a small portion in the middle line isstillsolid. Itresembles that of the preceding stage, but the narrow anterior portion has become longer in antero-posterior extent, and

point, it is obvious that growth in length has occurred in the head region. 1 In figs. 4, 6, 7, 10, 12, 18, for U.V. read V.V. (vitelline vein).


Fig. 4. Felis domestica, 7 somites (31.3.09, A.). StageII. Graphic reconstruction of the pleuropericardial canal, myocardium, vasofactive cells, dorsal aorta and endothelial heart tubes of the left side of the embryo.

the posterior expansion is wider than

Fig. 4a. Diagrammatic longitudinal section

that of Stage I. If the anterior border of the first somite be taken as a fixed through the anterior end of the embryo 62-5.



The dorsal aorta has now become converted into a continuous vessel through a considerable portion of the embryo. In the head region, it is somewhat irregular with lateral offshoots—the descendants of the scattered vasofactive cells between the aorta line and the heart line in the previous stage. These in some cases anastomose with each other. In this embryo we see again a well-marked interruption in the aorta, with only a few vasofactive cells beginning to bridge this gap. The relations of the mesoderm and of the brain plate in this embryo are similar to those in Stage I, i.e. the interruption in the aorta corresponds with a flattened region of the medullary plate and a denser and dorsi-ventrally thinner portion of the mesoderm.


It may be noted here that the dorsal aorta in the region of the somites instead of occupying a mechanically favoured position, as described by Bremer in the rabbit, lies beneath the somite at its point of maximum depth (see fig. 5).



Fig. 5. Felis domestica, 7 somites (31.3.09, A.). Transverse section through the first somite on one side and the first intersomitic cleft on the other. x 130.


The vasofactive cells of the heart and vitelline veins are most numerous in the region of the posterior somites, where they extend across a wide zone bounded by the aorta on the median side. At the level of the first and second somites the cells become limited to a strip underlying the pleuro-pericardial canal and penetrate right to the middle line slightly behind the median portion of the canal. Immediately in front of the first somite a portion of endothelial heart tube is definitely established, and this lies in a typical myocardial gutter. The extent of the myocardium is also shown in fig. 4.


Much confusion occurs in the literature of the early development of the heart, owing to the impossibility of drawing a clear line between the portions of the endothelial tube destined to form heart and vitelline vein respectively. Wang (15) uses the term “‘vitelline vein” for even the anterior portion of the structures commonly called lateral heart tubes (pp. 120, 169, etc.) and also applies the term to structures which I described as lateral heart tubes in Dasyurus (pp. 128, 183). Moreover, in fig. 18, he shows a “‘vitelline vein seen through the pericardium” and obviously situated in a deep myocardial gutter. Observations on slightly later stages of Amniote embryos show that the ventricular region can be distinguished from the atrial region by the fact that in the former the myocardium is separated from the endothelium by a space which is bridged, as development proceeds, by strands of protoplasm which foreshadow the thick muscular walls of the ventricle (see (7), fig. 14), whilst in the auricular region the myocardium is closely applied to the endothelium (see (7), figs. 17 and 18). In Perameles ((7), fig. 9), Dasyurus ((7), fig. 24). and Felis, as well as in the ferret ((15), fig. 18), this distinctive feature of the ventricle is noticeable even before union of the lateral heart tubes. It therefore appears correct to refer to that region which is situated in a definite myocardial gutter as a lateral heart tube rather than as a vitelline vein. In my previous paper I described the endothelial tubes from the bulbus aortae anteriorly, to the opening of the Cuverian ducts posteriorly, as lateral heart tubes, and I think that the whole of this portion of the endothelial tube is destined to be converted directly into the median heart.

In the Felis embryo now under discussion, the only portion of heart tube present shows the features characteristic of the ventricular region and we may therefore assume that the bulbo-ventricular portion is the first part of the heart to develop in Felis.

The myocardial gutter becomes shallow and disappears where the endothelial tube passes over into scattered vasofactive cells. At the level of the third and fourth somites, a portion of the vitelline vein is completely established. It follows the antero-posterior line of the lateral heart tube and so lies lateral to the pleuro-pericardial canal which here curves in towards the middle line.

In both embryos of this stage, portions of the vena capitis medialis are present and in some places a tubular piece of it extends through nine or ten sections.

At the level of the cleft between the first and second somites on the left side of the embryo, there occur a few cells dorsal to the somatic mesoderm of the pleuro-pericardial canal (see fig. 5, V.C.U.V.). These represent the first trace of the dense mass of vasofactive cells which develop in this region and give rise to the capillaries from which are developed the Cuvierian duct and the umbilical vein (cf. Schulte (12)). On the left side of the section represented in fig. 5, a single cell between the somite and the neural tube represents the continuation of the vena capitis medialis of this region.

The vasofactive cells continue for some distance behind the region shown in the graph (actually through about -9 mm.), and then disappear. There is no aorta established behind the region depicted in the graph.

Stage III.

Felis domestica (31.8.09, D.), 8 somites.

Fig. 61 represents a graphic reconstruction of the pleuro-pericardial canal, endothelial heart tube, aorta and the vasofactive cells of the anterior end of this embryo, which resembles the last stage very closely. The pleuro-pericardial canal is still not luminated in the middle line; it has increased slightly in width throughout the whole extent of its lateral limbs. Two distinct portions of the dorsal aorta are again recognisable and a considerably greater length of the endothelial heart tube is now established. The myocardium has also increased in length and a comparison of figs. 4 \---.. and 6 indicates that differentiation of , myocardium hastaken place both anterior |— and posterior to the portion existing in Stage II. The endothelial heart tube is continuous posteriorly with a short portion of vitelline vein, but this is not yet connected with the capillaries of the vitelline network. Vasofactive cells lying below the pleuro-pericardial canal continue forwards but do not actually reach the middle line in this embryo.


1 See footnote, p. 110.


The relations of the brain plate to the ectoderm and entoderm of the proamniotic area and to the pleuro-pericardial canal are somewhat peculiar (see fig. 6 a). The brain plate in the middle line is much sunk relatively to the ectoderm lying in front of it; laterally the medullary folds rise up level with thenonneural ectoderm. This elevation of the proamniotic region may be interpreted | as a head fold of the amnion; if so, it is precociously developed in this embryo, | Ho ane for it is not present in any other stage Fig. 6. Felis domestica, 8 somites (31.3.09, D.). described in this paper. It may be noted Stage HII. Graphic reconstruction of the pleuro-pericardial canal, myocardium, vasohere that an amniotic tail fold is present ¢, ,tive cells, dorsal aorta and endothelial in this embryo.’ heart tube of the left side of the embryo.

In this stage, as in the embryos having Fig. 6a. Diagrammatic median longitudinal 7 somites, traces of the vena capitis section through the anterior end of the medialis are present, both in the region °™>"Y° of the fore brain and in that of the somites, but this vein shows no advance on the preceding stage.

At the level of the first somite, the lateral and ventro-lateral walls of the pleuro-pericardial canal, as well as the dorsal wall are actively producing cells. From their subsequent history it appears that these cells contribute to the formation of the group of capillaries from which the umbilical vein, Cuvierian duct and posterior cardinal vein are all derived. The early development of the umbilical vein has already been described by Schulte (2).

Stage IV.

Felis domestica, 9 somites (8.2.10, 4.), G.L. 3-6 mm. (neglecting posterior end which was bent under).

The most marked step which has occurred since the preceding stage is the development of a head fold, resulting in the presence of a small crescentic fore gut. The brain, whose anterior border is indicated in Fig. 71 (4.M.P.), has grown forward so that its anterior end now lies in front of the anterior wall of the pericardium, instead of behind its posterior border as in the preceding stage. The median portion of the pericardium has increased considerably in antero-posterior extent, and is luminated throughout. Comparison of figs. 6 and 7 shows that the shape of the pleuro-pericardial canal has changed considerably, for whereas in Stage III the widest part of the canal occurs some distance behind the antero-median limb, in Stage IV, itis situated directly behind this portion. The myocardium has not increased greatly in length; it lies as in Stage III in the region of greatest expansion of the pleuro-pericardial canal and therefore it also is situated much nearer to the antero-median limb than in the preceding stage. It should be noted that in both stages, III and IV, the first somite lies near the posterior end of the differentiated myocardium (cf. figs. 6 and 7).



Fig. 7. Felis domestica, 9 somites (8.2.10, A.). Stage IV. Graphic reconstruction of the fore gut, pleuro-pericardial canal, myocardium, dorsal aorta, endothelial heart tubes, etc. (left side of embryo on right in figure). Pericardium and myocardium are incomplete on the right side of the figure.


Fig. 7a. Diagrammatic median longitudinal section through the anterior end of the embryo. Fig. 76. Dorsal view of the anterior end of the dorsal aorta.


The endothelial heart tube is now continuous with a wide vitelline vein posteriorly, and runs forward to shortly behind the anterior intestinal portal. It stops short abruptly but anterior to this there are a number of vasofactive cells and on each side of the middle line, immediately in front of the anterior intestinal portal, there is a small angiocyst. These are very thin-walled and slender and lie isolated, posterior to the median limb of the pericardium.


1 See footnote, p. 110.


The relations of the pericardium to the gut and fore brain are best understood by reference to the diagrammatic longitudinal section (fig. 7 a). From this it will be seen that the fore brain projects freely in front of the gut which is short, whilst the pericardium lies entirely ventral to the ectoderm underlying the head; thus it is situated in front of the ectodermal wall of the headfold bay, instead of lying between that and the anterior intestinal portal. Moreover, it thus comes to lie between ectoderm and entoderm, instead of lying between two layers of entoderm as it usually does. The small angiocyst is situated entirely posterior to the median pericardium.


Fig. 8. Felis domestica, 9 somites (8.2.10, A.). Transverse section through the anterior extremity of the lateral heart tubes in the position indicated in fig. 7. x 100.


The myocardium at its anterior extremity is entirely detached from the pericardium for a few sections and here surrounds the blindly ending endothelial heart tube (see fig. 8). The vasofactive cells lying outside the pericardium in fig. 8 (V.C.V.A.) are continuous with the endothelial heart tube and represent the primordium of the ventral aorta in that region.


In the anterior portion of the lateral heart the endothelium and myocardium are widely separated, but the space between them becomes less as we pass backwards.


The dorsal aorta is greatly enlarged at its anterior end and extends to the anterior wall of the fore gut, whilst a slight prolongation of the vessel on to the ventral face of the fore gut is present. This represents the antero-ventral portion of the first aortic arch. The aorta continues back into the region of the somites where it attains a considerable size and lies under the somite itself not under the somitic stalk. The wide anterior and the trunk portions of the aorta are joined by a slender irregular vessel which has now bridged the gap between the two detached parts of the aorta seen in the preceding stage.


A slender vena capitis medialis is present lying alongside the fore brain and opening into the anterior extremity of the dorsal aorta. It is not extensive, but isolated portions of it are recognisable both in the anterior and the somitic regions.


The primordium of the umbilical vein has increased considerably and in a few sections the cells thereof are so arranged as to surround a lumen. A continuous vessel is not, however, established.


This stage then is one in which there are well-developed paired endothelial heart tubes, surrounded by myocardium. Portions of the ventral aorta are present as strings of vasofactive cells and as one isolated angiocyst on each side of the middle line, but these are not continuous either with each other or with the dorsal aorta so that the first aortic arch is incomplete.


A small head fold is developed in consequence of the forward growth of the brain plate, but the pericardium has not yet attained its normal position in the lip of the anterior intestinal portal, but lies entirely outside of the head fold.

Thelateral hearts are further forwards relatively both to the brain plate and to the median pericardium than in the preceding stage, but they retain approximately the same relation to the first somite. As the myocardia have not increased much in length it appears that this change in relative position is not merely due to growth of the myocardium.

Stage IV a.

Felis domestica (8.2.10, B.), 9 somites. Cut longitudinally and an situ.


This embryo, although it has the samenumber of somites as Stage IV, shows a distinct advance in the condition of its gut, pericardium and heart. Unfortunately, the sections are very oblique, so that no one is actually median, but that represented in fig. 9 a passes more nearly than any other through the middle line of the heart, whilst figs. 9 6 and c are lateral. The brain is partially closed in this embryo, the gut has increased considerably in length. The pericardium has grown both in antero-posterior extent and in dorsi-ventral depth and has altered its position so that in the middle line it now lies wholly posterior to the oral plate, in its normal position in the lip of the anterior intestinal portal. On each side of the middle line, however, there is a diverticulum of the pericardium running forwards and lying between the extraembryonal ectoderm and entoderm beneath the head fold (fig. 9 5).

The right and left myocardial tubes are not fused in the middle line. The endothelial tube shown in fig. 9 ¢ can be traced continuously from the auricular region through the ventricular limb which passes into the bulbus aortae. The bulbus aortae is separated from the ventricle by a constriction of the myocardium, the bulbo-ventricular sulcus of Schulte(13), but no definite limit between the ventricle and auricle is present. The endothelium of the bulbus aortae on both sides is continuous with that of the ventral aorta, the right and left ventral aortae being now continuous across the middle line (fig. 9 a). From the median ventral aorta thus formed there arises the first aortic arch which is now complete (fig. 9c, 4.4.1). From the anterior extremity of the first aortic arch arises a relatively large vessel, which runs up alongside the brain and is continued backwards for a short distance as the vena capitis medialis (fig. 9c, V.C.M.).


In the region of the somites the dorsal aorta gives rise to small intersegmental offshoots, closely resembling those described and figured in Perameles (7), but these do not yet appear to be connected by a longitudinal vessel.


The conditions in this embryo appear to resemble closely those described and figured by Schulte (13) in his stage with 11 somites. His figures both of the endothelial tubes and of the myo-epicardial mantles of an 11 somites cat embryo represent approximately: the conditions of the embryo here described.


Fig. 9a, bande. Felis domestica, 9 somites (8.2.10, B.). Stage IV a. Three longitudinal sections through the embryo. Fig. 9a is nearly median through the heart, fig. 9 b is lateral, fig. 9c is still more lateral. x 50.


Stage V.

Felis domestica (9.5.12, A.), 10 somites, G.L. 3-2 mm.

This embryo closely resembles Stage IVa in its general condition. It is unfortunate for purposes of comparison that it is a small embryo, its greatest length being less than that of Stage IV, whilst the distance of the first somite from the anterior end in Stage V is -98 mm. and in Stage IV it is 1-21mm. The embryo is normal in all other respects, and shows a marked advance on Stage IV though in some respects it is probably earlier than Stage IVa.

In Stage V practically the whole of the pericardium has receded from the 118 Katharine M. Watson

extra-embryonal position which it occupied in Stage IV and only two small solid diverticula of the walls of the pericardium now remain to represent the hollow lateral pockets of the preceding stage.

Fig. 101 represents a graphic reconstruction of the endothelial heart tubes, aorta and aortic arch of this embryo, as well as the outlines of the gut, pericardium and myocardium. Fig. 10 a represents a very diagrammatic longitudinal section through the embryo.

Comparing this stage with Stage IV, represented in the corresponding figs. 7 and 7 a, we see that the fore gut has grown in length considerably, as also has the pericardium. The shape of the pericardium has altered again so that the median portion is distinctly shorter and deeper than in Stage IV (cf. figs. 7 a and 10 a). The lateral limbs of the pleuro-pericardial canals reach their greatest expansion immediately behind the median pericardium. The shape of the anterior intestinal portal has changed from a broad crescent to a reversed U or V. The myocardia are situated relatively further forward than in Stage IV so that their anterior ends now project into the median pericardium and lie ventral to the closed gut (see fig. 11). The well-marked constriction shown in fig. 10 (B.V.C.) presumably corresponds with the bulbo-ventricular suleus of Schulte.


1 See footnote, p. 110.


Fig. 10. Felis domestica, 10 somites (9.5.12, A.). Stage V. Graphic reconstruction of the endothelial heart tubes, ventral aortae of both sides and the dorsal aorta of the right side and the outlines of the fore gut, pericardium and myocardium of the left side of the embryo.

Fig. 10a. Diagrammatic longitudinal section through the anterior end of the embryo. Fig. 10 6. Dorsal view of the anterior end of the dorsal aorta.


The ventral aortae do not communicate across the middle line, but each is a wide vessel, interrupted at intervals by non-vascular loculi and the two aortae approach each other very closely in the’ middle line. Numerous vasofactive cells are scattered between the right and left ventral aortae. The first aortic arch is complete and has the form of a very short, wide vessel running along the antero-lateral border of the fore gut to join the dorsal aorta. The anterior portion of the dorsal aorta is a very wide vessel, but it narrows abruptly in the region of the anterior intestinal portal and disappears on both sides of the embryo for a distance of about -04 mm. Behind this gap there is a slender and incompletely established portion of the aorta which, however, becomes larger and perfectly continuous in the region of the somites. Thus in this embryo the original interruption in the line of the aorta which was noticeable in Stages I, II and III has not yet been bridged.


Fig. 11. Felis domestica, 10 somites (9.5.12, A.). Stage V. Transverse section through the anterior extremity of the lateral heart tubes, in the position indicated in fig. 10. x 80.


A portion of the vena capitis medialis is present, communicating with the dorsal side of the apex of the first aortic arch. Further back in the head there are, as in preceding stages, a few cells close beside the neural tube; these cells are destined to contribute to the vena capitis medialis, but nowhere else is this vessel definitely established. In the region of the somites, there are minute intersegmental offshoots running medianwards and dorsalwards towards the neural tube, but they are not connected by a longitudinal vessel.


The primordium of the umbilical vein now lies entirely dorsal to the pleuropericardial canal and more of the cells are now grouped so as‘to surround a lumen.


Stage VI.

Felis domestica (B.6.12, D.), 12 somites. Cut longitudinally and in situ.

Only a brief reference to this embryo is necessary, as it differs in no important features from Stages IVa and V. The right and left endothelial tubes are fused in the region of the bulbus aortae and appear to communicate with each other at one or two points in the ventricular region. The myocardia are also fused in the bulbar region but not elsewhere. The ventral aortae are still separate and are smaller more definite channels than in the preceding stage. The first aortic arch is complete and there are traces of the ventral portion of a second. Portions of the vena capitis medialis are present in the fore brain region, but I have failed to trace any connection between this vessel and the dorsal aorta or first aortic arch on either side in this embryo.

This embryo appears to be very similar to that with 12 somites described by Schulte (13).


Fig. 12. Felis domestica, 15 somites (26.3.09, C.). Stage VII. Graphic reconstruction of the heart, pericardium, fore gut and first aortic arch. The outline of the ventral part of the second aortic arch is indicated represented by a dotted line to indicate that it lies dorsal to the ventricle.

Fig. 12a. Dorsal view of the dorsal aorta of the right side of the embryo and the umbilical vein of both sides.


Stage VII.

Felis domestica (26.3.9, C.), 15 somites. Felis domestica (11.38.09, E.), 14 somites (cut longitudinally).

Figs. 12 and 12 a! represent ventral and dorsal views of a graphic reconstruction of the former of these two embryos, whilst fig. 14a represents a median and fig. 14} a lateral section through the second embryo, cut longitudinally.

The condition of the heart, pericardium and blood vessels is simple and stages of this type are so well known that no detailed description need be given. Three main divisions of the heart are easily recognisable, viz. the bulbus aortae, the ventricle and an auricular limb which passes imperceptibly into the vitelline veins. The ventricle is separated from the auricular limb by a well-marked constriction which affects both myocardium and endothelium.

1 See footnote, p. 110.


The endothelial tubes are united throughout the bulbus aortae except for a short distance at its anterior end; they are united through most of the ventricular limbs, but separate at two points for a short distance. The myocardia are united also in the bulbus aortae and ventricle, but a groove on the ventral face of the myocardium indicates the line of fusion and in one portion of the ventricle, the twomyocardial walls diverge from each other, being hereseparated by a flat stretch of mesoderm, the middle cardiac plate of Schulte (13) (fig. 4).

The asymmetry described by Schulte in his embryo with 14 somites has made its appearance, but is only slight in this stage; the anterior end of the left side of the ventricle is slightly in advance of that of the right side, whilst in the auricular region the asymmetry is more marked.

The pericardium has the normal form and relations, except for the presence of a peculiar tubular diverticulum on the left side. This runs far forwards, between the extra-embryonal ectoderm and entoderm, lateral to the head of the embryo. It is probably a slight abnormality of no special interest for our present purposes.

The first aortic arch is complete and in both embryos the ventral portion of the second arch is present. From the apex of the first aortic arch a small vessel arises and runs dorsalwards close beside the brain. This vena capitis medialis can be traced forwards dorsal to the anterior end of the head and it also runs backwards for some distance. Just behind the optic vesicle there is present a small portion of the vena capitis lateralis, opening into the vena capitis medialis. Behind this again is a small angiocyst apparently also part of the vena capitis lateralis, but it seems to be completely isolated. Further portions of these veins are present as indicated in fig. 12 a. In the region of the somites inter-segmental offshoots are again present, but are not connected by a longitudinal vessel.

The umbilical vein shows a striking asymmetry, lying at the level of the auricular segment on the left side of the embryo, whilst on the right side it is situated in the region of the widely separated vitelline veins (see figs. 12 and 13)1.


Summary and Conclusions

1. Development of the pericardium and heart.

Since the original description by Hensen in 1875 of the bilateral origin of the heart in Mammals, it has been regarded as established that the median heart is derived from lateral heart tubes which, in Mammals, take their origin in the hind brain region by the proliferation of cells from the ventral wall of the pleuro-pericardial canal. These cells become grouped to surround a lumen and the vessel thus formed grows both forwards and backwards and constitutes the endothelial heart tube. At the same time, it causes a bulging of the ventral wall of the pleuro-pericardial canal into its cavity, forming the myocardial gutter or myo-epicardial mantle from which the muscular wall of the heart is developed. It is further recognised that on closure of the gut, the myocardia and endothelial tubes are brought into position ventral to the gut, and those of the right and left sides eventually fuse. Various authors (e.g. Rouviére (9), Robinson (8)) have discussed the method by which these changes are brought about, but the main outlines of the process have only been questioned by Wang (15) who states (p. 172): “It will be noticed then that there is evidence to show that, at a period before the formation of the fore gut the first rudiment of the heart in the ferret is single and is situated in the median plane of the embryo, caudoventral to the pleuro-pericardial cavity, in the form of a transverse endothelial tube which is destitute of any blood cells. Laterally it is in direct communication with the two vitelline veins one on each side of the embryo.” Again (p. 178): “‘The next phase of development of the heart is represented by the separation of the ‘primary’ heart rudiment into two distinet endothelial tubes lying closely together one on each side of the embryo, not far from the median plane.”

1 See footnote, p. 110.



Fig. 14. Felis domestica, 14 somites (11.3.09, H.). Stage VII. (a) Longitudinal section, approximately through the middle line of the heart. (6) Longitudinal section through the bulbus aortae, and first aortic arch of the right side of the embryo. x 50.


Attention has been drawn above to the confusion caused by the loose application of the term vitelline vein. Wang has applied it in his Stage III to a well-defined endothelial tube, surrounded by a myocardial wall (15) (fig. 18), which corresponds exactly with the structure usually described as a lateral heart tube. In his Stage II the pleuro-pericardial canal appears to be incompletely differentiated, whilst the endothelial tube is relatively advanced. There is, however, no evidence in his paper that his ‘“‘median heart” actually developed earlier than the so-called “‘vitelline vein.”” Both the “‘vitelline vein” and the “‘median heart” are absent in Wang’s Stage I, and are completely established in Stage II so that no direct information is available as to the order and mode of development of these two structures in the ferret. It may be noted that in Wang’s fig. 6a (Stage II) the “‘vitelline vein” is much larger than the “‘median heart,” a fact which suggests that the “vitelline vein” appeared first. Moreover, Wang’s median heart rudiment lies caudo-ventral to the pericardium (15) (pp. 170, ete.), and appears to correspond exactly with the two angiocysts seen in my Stage 4 (fig. 7) and these lie outside the limits of the heart proper, and form the ventral aorta. Unfortunately, in his fig. 18, Wang does not give the complete outline of the myocardium, but in his figs. 19, a, b, ¢ and d, he shows the “‘median heart rudiment” entirely without a myocardial wall, whereas in fig. 18 he shows a distinct myocardial gutter around the “vitelline vein.” The correct interpretation of Wang’s material therefore appears to me to be that his “‘median heart” represents a ventral aorta which has developed and fused in the middle line earlier than it does in most Mammals, whilst the heart proper has arisen in the typical way from lateral endothelial tubes (Wang’s “‘ vitelline vein” which makes its appearance between his Stages I and II) and is surrounded by a myocardium of the ordinary type. Moreover, the fact that in Wang’s Stage IV the right and left endothelial tubes are separate does not prove conclusively that the heart is first median and later separates into two tubes, for reconstructions of a number of embryonic hearts show that even within a species, there is a great variation in the relative development of different parts of the heart and in the mode of union of the right and left endothelial tubes. The ventral wall of the pleuro-pericardial canal appears to be capable of producing vasofactive cells throughout its entire length and it is evident that if such cells be produced at an early stage right across the middle line (as is the case in my Stages I and II) a median angiocyst is likely to be produced before the lateral heart tubes reach a position in which fusion is possible. On the other hand, if few endothelial cells arise near the middle line (as is the case in my Stage III) no median angiocyst will be formed prior to the fusion of the lateral heart tubes. The work of Schulte (13) and the graphs in the present paper taken together show that in the cat, fusion takes place piecemeal, rather than at one step or even in steady progress craniocaudally. It is therefore quite possible that in the particular embryo described by Wang as Stage IV, the early development has differed slightly from that occurring in the embryos described as Stages II and III. To this point I will return after discussing the corresponding processes in the cat.

Several other points raised by Wang (15) in his paper on the development of the heart in the ferret may be mentioned here. He refers repeatedly to my description of a Dasyurus embryo (Dasyurus viverrinus, Stage II, 8-5 mm. A. in my paper of 1915(7)), and on p. 188 he states: ‘In the Dasyurus the endothelial tubes have come actually in contact with each other at their extreme cranial ends and presumably have united across the median plane of the embryo. It should be noted that the significance of this connection between the two vitelline veins across the median plane was not dwelt upon, and was considered only as being remarkable by Miss Parker, who states also ‘that lateral and caudal to the median union, each endothelial tube gives rise to the first aortic arch, which follows the antero-lateral margin of the gut almost to the median plane, and there becomes continuous with the corresponding dorsal aorta, the two aortae being well developed at this stage ’.”’ The quotation from my paper though in inverted commas is inaccurate, my actual words being “Some distance behind its cephalic extremity, the lateral heart tube gives rise to the first aortic arch, which follows the antero-lateral margin of the gut almost to the middle line and there becomes continuous with the corresponding dorsal aorta, the two aortae being completely established in this stage.”

I have re-investigated this question of union of the heart tubes in this Dasyurus embryo and find that the right and left tubes are not fused. Professor J. P. Hill, F.R.S., to whom this specimen belongs, has also examined its heart and agrees with this statement. Even if fusion does occur at about this stage in Dasyurus (i.e. about Stage II of my paper(7)), the fact would not support Wang’s view that the first rudiment of the heart is a median structure, for typical lateral endothelial tubes, surrounded by a myocardium are present in an earlier stage as described in my paper (Dasyurus viverrinus, Stage I, 7-5 mm. (7)).

I greatly regret that in Plate I, fig. 4(7), a line which should have divided the right and left endothelial tubes throughout the greater part of their length, was lost in reproduction. This brought about the inconsistency between the text and that figure to which Wang (15) refers on p. 175 (see fig. 15).

With reference to Wang’s further criticism (p. 175) that I omitted to state the magnification of my figures of Stages IV and V and the lengths of these embryos, it may be remarked that I refrained from comparison of lengths between Stages IV and V for several reasons, In the first place the embryos are of diferent species and are therefore even more liable to variations in size than two embryos of the same species; in the second place, Stage IV is markedly in advance of Stage V in almost every respect except the development of the heart. Stage V is, in fact, very similar to Stage III as regards the development of its brain. For these reasons I did not attempt to make comparable figures of Stages IV and V. I regret very much that I did not make these points clear. .

The series of catembryos described Fig. 15. Perameles nasuta (2PA). Outline reproabove enables us to get a clear idea of duction of fig. 5, Pl. I, from the “Early developthe development of the heart and ment of the heart, etc., in Marsupials”? (7), pericardium from the stage of the appearance of the pleuro-pericardial canals and the origin of the endothelial cells of the heart and aorta up to the complete establishment of a median heart with a recognisable bulbus aortae, ventricle and auricle. The details of the process of fusion of the lateral heart tubes in the cat have been given clearly by Schulte (13), and will not be repeated here.

The first stage in the development of the heart is one in which the pleuropericardial mesoderm is recognisable and forms a horseshoe around the anterior end of the embryo, with its median portion shortly in front of the anterior margin of the medullary plate. The median limb of the horseshoe is narrow and the lateral limbs become narrower as we pass backwards, and in the widest part of the pleuro-pericardial mesoderm a cavity is already present.

Vasofactive cells are numerous and occur in greatest numbers in the posterior part of the pleuro-pericardial canal and caudal to this region. Two or three angiocysts in the anterior part of the aorta are the only ones present; the vasofactive cells ventral to the pleuro-pericardial mesoderm extend right to the middle line in front of the anterior margin of the medullary plate.

Development proceeds by the spreading of the cleft constituting the pleuropericardial cavity, by the increase in the number of vasofactive cells, and their re-arrangement to form angiocysts, which coalesce to form vessels. These develop along two almost parallel lines one representing the dorsal aorta, the other the heart and vitelline vein. The development of the blood vessels will be dealt with in another section. The endothelial heart tube appears first (Stage II, fig. 4) just in front of the level of the first somite, and there it causes a slight bulging of the ventral wall of the pleuro-pericardial canal, the myocardial gutter. In this first developed portion of the lateral heart tube, the endothelium and myocardium are widely separated and it seems probable that this region represents the bulbo-ventricular limb.

The pleuro-pericardial canal expands both in lateral and dorsi-ventral extent and the endothelial heart tubes and myocardia continue to grow both forwards and backwards. If we compare Stages II and III (figs. 4 and 6) we see that the main change that has taken place consists in the conversion of vasofactive cells into endothelial tubes. The curious position of the median limb of the pleuro-pericardial canal in Stage III may be regarded as of nosignificance in the present connection.

After this stage there follows one of very rapid development. The most important factor in the change of relations occurring between Stages III and IV is undoubtedly the forward growth of the brain. If we take, as a startingpoint, Stage II (which is not complicated by a precocious head fold of the amnion), and imagine that the brain plate grows so rapidly that it pushes its anterior border forwards and at the same time the pleuro-pericardial canal increases in the antero-posterior direction, the condition seen in Stage IV is brought about, i.e. a small fore gut is developed, the median limb of the pleuropericardial canal is left lying between extra-embryonal ectoderm and entoderm. The vasofactive cells lying caudo-ventral to the median limb of the pleuro-pericardial canal in Stage II are, in Stage IV, converted into two angiocysts which still lie caudal to the posterior wall of the pleuro-pericardial canal, showing that no reversal of the pericardium has taken place up to this stage. At the same time the pleuro-pericardial canal has undergone a considerable change in shape and relations. Its median limb has increased in anteroposterior extent and its lateral limbs attain their greatest width very shortly behind the median limb. Moreover, the myocardia, which have not grown much in length (see figs. 6 and 7), now lie further forwards so that their anterior extremities reach almost to the median pericardial limb of the horseshoe. Whilst part of this change in relative position of the myocardium and pericardium is due to forward growth of the myocardium, it is evident that other factors have been at work. If we note the relations of the myocardium and pericardium to the somites we see that the first somite in both stages lies level with the posterior part of the myocardium, whilst the distance between the first somite and both the anterior and posterior walls of the median pericardium is considerably less in Stage IV than in Stage II.

The explanation of these facts seems to be that the anterior (median) limb of the pericardium is increasing in size at the expense of the lateral limbs and that as it enlarges it also moves backwards. In Stage IV no other changes in the pericardium have taken place, i.e. it has not undergone reversal as described by Robinson (8) nor have its lateral limbs begun to close inwards as described in Perameles and Dasyurus(7), for the anterior intestinal portal still has a wide crescentic shape, and the vasofactive cells and angiocysts lie posterior to the median pericardium. The Origin of Heart and Blood Vessels in Felis domestica 127

Throughout this period, vasofactive cells, wherever they may be, continue to be converted into angiocysts and blood vessels. The presence in Stage IV of two angiocysts near the middle line indicates the manner in which the “median heart” of Wang(15) might be produced, but it may be noted here that in the embryo described as Stage ITI vasofactive cells do not extend across the middle line in relation to the pericardium. It seems probable that in such an embryo, the development of Wang’s ‘‘median heart”? would be omitted and the completed endothelial tubes might be found separated at a relatively late stage as described by Wang (15) in his Stage IV.

Following on Stage IV in Felis domestica there occurs a rapid expansion of the median limb of the pericardium and a change in its position, for in Stage IVa the pericardium lies almost entirely behind the ectodermal limit of the head fold, between that and the entoderm of the anterior intestinal portal. The two angiocysts seen in Stage IV have in Stage IVa united across the middle line and form a ventral aorta which is now continuous with the heart on the one hand and with the first aortic arch on the other. It must be noted that this median vessel has no myocardial gutter surrounding it. It lies at the anterior end of the median pericardium, i.e. in the reverse of its position in the preceding stage. On each side of the middle line a portion of the pericardium has retained its former position between the extra-embryonal ectoderm and entoderm and traces of these lateral diverticula are present in the next stage. This fact sheds light on the process by which the pericardium shifts from its position in Stage IV to that which it occupies in Stages V and VI, for it must take place by the backward growth of the postero-ventral wall of the median pericardium in Stage IV, in the manner indicated by an arrow in fig. 8 a coupled with the backward movement of the median pericardium as a whole. This must be imagined as taking place in such a way as to bring about the reversal in position of the ventral aorta between Stages IV and IVa whilst permitting of the “lagging behind” of the lateral diverticula of the pericardium in the extraembryonal region.

The anterior ends of the myocardial tubes surrounding the endothelial tubes in Stages IVa and V have reached a position in the median pericardium and lie in Stage V (fig. 11) ventral to the closed gut but still separated from each other. From this point the heart tubes in Stage V follow a spirally curved course round the anterior intestinal portal so that their posterior ends lie in the ventral wall of the lateral pleuro-pericardial canal. At the same time, between Stages IV and V the lateral limbs of the pleuro-pericardial canals have converged towards the middle line, so that in Stage V the anterior intestinal portal has the shape of a narrow U instead of the wide crescent of Stage IV.

The process by which the head fold is formed and the lateral heart tubes are brought into position ventral to the newly established fore gut thus consists in the enlargement and backward growth of the pericardium, accompanied by a twisting of the pericardium. This twisting must be due to 128 Katharine M. Watson

differential growth which brings about a complete reversal in the middle line first and thus causes the heart tubes at one period (e.g. in Stage V) to take a spiral course between their anterior ends which lie dorsal to the median pericardium and their posterior ends which lie ventral to the lateral limbs of the pleuro-pericardial canal.

It may be noted that actually in the middle line in Stage V the pericardium is smaller in the antero-posterior direction than in Stage IV. It is, however, deeper in the dorsi-ventral direction (N.B. the embryo described as Stage V is smaller than that described as Stage IV) and a shortening and deepening of the pericardium in the middle line seems to me a natural phase in such a process of back-growth and twisting as I have described.

By a continuation of the process of extension of the median pericardium, accompanied by its rotation, the lateral heart tubes are gradually shifted from the ventral to the dorsal side of the pericardium.

This account of the formation of the head fold merely amplifies that given by Rouviére (9). Robinson (8), on the other hand, attributes a large share in the process of gut closure to the forward growth of the brain. I have previously brought forward evidence against this view(7). The reversal which I have described above I regard as being due not to the forward growth of the head, but to active differential growth of the pericardium.

In the guinea-pig, according to Strahl and Carius (14) (figs. 21, 22, and 22 a) and Yoshinaga (16), the processes of gut closure and heart development do not appear to follow exactly the course described here in the cat. The fundamental point of difference between the two forms appears to be that the formation of the head fold takes place at an earlier stage relatively to the development of the heart and pericardium in the guinea-pig than in the cat, i.e. in the guineapig the brain plate grows forwards and produces the type of head fold seen in Stage IV of the cat series at a stage when the pleuro-pericardial canal is still very small.

In Stage V (fig. 10) the vasofactive cells ventral to the median pericardium have become converted into a maze of capillaries constituting the ventral aorta and the ventral portion of the first aortic arch which is here complete. These capillaries are very thin-walled and therefore difficult to trace, but they appear not to anastomose in the middle line. As already indicated, however, this appears to be a point of minor importance, as fusion would probably occur if the vasofactive cells develop early in the middle line and might fail to occur if circumstances were unfavourable.

The developmental processes following on Stage V are those concerned firstly with fusion of the right and left heart tubes and secondly with their curvature and both these phases have been adequately described in the cat by Schulte (3). As already remarked, considerable variation seems to occur in details of fusion. Schulte’s stages as designated by somites do not agree with mine, but wherever possible I have pointed out a correspondence between his stages and mine. In his 12-somite stage, which shows a stage of heart development very little earlier than my 15-somite stage, there are points of fusion in the bulbus aortae, whilst in his 14-somite stage there is fusion both in the bulb, the ventricle, and at the cardio-venous angle. In my 12-somite stage there is fusion in the bulb, the ventricular limb and the auricular limb whilst in my 15-somite stage conditions as regards fusion are almost identical with those of Schulte’s 14-somite stage, but the heart in my embryo is uncurved, whilst in Schulte’s embryo it is markedly curved. The general conclusion is that different embryos vary considerably in the details of the rate of growth of various organs and that fusion occurs differently according to the relative advancement or retardation of the development of the pericardium and heart tubes.

In my previous paper I referred to a suggestion made originally by Prof. J. P. Hill that fusion of the right and left heart tubes is rendered possible by growth in antero-posterior length of the pericardium exceeding its growth in width(7) (p. 492). This explanation was criticised by Wang (15) (p. 174), who slightly misunderstood my point. The process of growth in length of the median pericardium I regard as occurring between the stages of widely separated heart tubes and that of approximated (and possibly curved) tubes, i.e. between Wang’s Stages III and IV, between Stages III and IV in my Perameles series and between Stages V and VII in the Felis embryos described above. Moreover, I expressly stated that “approximation” (not fusion) of the heart tubes results from such a process of longitudinal stretching. The loops into which the heart tubes are thrown are doubtless due to a phase following on this period of growth in length (i.e. a phase following on Stage IV in my Perameles series and Stage VII in Felis) in which the rate of growth of the heart tubes exceeds that of the pericardium and the former are accordingly forced to curve. Both these stages have occurred between Wang’s Stages III and IV with the result, as I believe, that the heart tubes have become first approximated and then curved.

It has been pointed out by Yoshinaga (16) (pp. 802, 808) that in the guineapig the phase in which curvature of the heart tubes occurs is preceded by one in which “the myocardial tube grows excessively in the cranio-caudal direction and decreases in lateral width.” He further states that the endothelial tubes are brought together in the median plane, where they come to fusion by the extreme longitudinal stretching of that part of the myocardial tube in which the endothelial tubes are enclosed.” In my view, growth in length of the pericardium itself is the main factor but as Yoshinaga’s figures show clearly that a stretching of the myocardium in the antero-posterior direction also occurs.

2. Development of the blood vessels.

The most important recent work on the development of the blood vessels is that of the American anatomists to whom reference was made in the introduction. This work has led to the general acceptance of the theory that the blood vessels of the embryo develop in situ, in contrast to the angioblast doctrine of His. 130 Katharine M. Watson

The observations on cat embryos described above in general confirm those of Schulte(12). As his figures and mine show, the earliest stage of vasculogenesis is one in which vasofactive cells occur scattered in the embryo and not connected into an angioblast network as described by Bremer (1) for the rabbit. Moreover, the vasofactive cells in the embryo from a very early stage are grouped in two main lines, an aortic line and a cardiac line. At first there are many .cells scattered between these two lines, but as development proceeds, and the aorta and heart tube become definitely established, the cells lying between them disappear. These cells are doubtless used to build up the walls of the vessels and in order to do so some of them must migrate from their place of origin. In the case of the endothelium of the heart, the cells arise apparently directly from the splanchnic mesoderm. They become grouped into clumps and arranged to surround a lumen, retaining approximately their original relation to their place of origin. The aorta in the cat arises as two distinct parts, separated by a gap which is not bridged until a stage having 9 or 10 somites. In the anterior part of the head, in stages with 4 to 7 somites, the medullary plate has risen up slightJy into folds, the mesoderm is loosely packed and the cells of the aorta appear to arise directly from the ventral side of this head mesenchyme. Behind this is a region in which the medullary plate in a 4-5 somite stage is perfectly flat, the underlying mesoderm is dense and there appear to be no vasofactive cells in the aortic line. Posterior to this the medullary plate again rises up somewhat. In the embryos with 7 and 8 somites also, there is a widening of the medullary groove and the aorta is interrupted in this region. It therefore appears that the fact that the aorta arises in two portions is simply due to the presence of a region in which the shape of the brain plate causes a certain compression of the mesoderm, and this in its turn inhibits the formation of vasofactive cells.

In the embryo with 9 somites, the medullary folds have risen up more completely and the mesoderm has become less dense. The aorta is here complete, but the site of its former interruption is marked by a reduction in its size.

The vasofactive cells of the aorta as far back as the first somite, appear to arise directly from the overlying mesenchyme. In the region of the somites, however, as remarked by Schulte, it is difficult to determine the exact origin of the vasofactive cells. The somites themselves do not appear to be actively producing them at any stage. On the other hand, it may be noted that in Stage I (fig. 1) there are vasofactive cells in considerable numbers lateral to the somites, whereas in Stage ITI (fig. 6) there are few of these ¢ells in a similar position. This suggests that a number of vasofactive cells arise lateral to the aortic line and migrate towards that line. The sectional appearances also confirm this view, for in Stages I and II in the somitic region there are numerous vasofactive cells arising from the splanchnic mesoderm, as well as from the somitic stalks (fig. 3). At the same time it is quite possible that, even in these early stages, the somites produce some vasofactive cells in addition to those described by Schulte(12) as heing contributed later by the sclerotome. In many sections, including that represented in fig. 5, the outline of the somites in Stages I and II is not very regular. On one side, in fig. 5, the section passes almost through an intersomitic cleft and here as in other similar sections a cell is seen detached from the paraxial mesoderm and projecting towards the aorta, This suggests that vasofactive cells may arise from the segmented mesoderm. Schulte (12) has suggested that some of the cells of the aorta arise from the paraxial mesoderm before the formation of the somites, but I have found no evidence of this. In Stages I and II in the region posterior to the somites, there is abundant evidence of the origin of vasofactive cells from the lateral mesoderm, cells in all stages of separation being numerous, but the cells lying ventral to the paraxial mesoderm are completely detached and give no clue as to their place of origin. +

It seems probable then that the cells which form the walls of the aorta are derived partly from the splanchnic mesoderm, lateral to the final position of the aorta, and partly from the somites (see also Schulte (12)) and somitic stalks.

In the earliest stages described above, the formation of the main vein of the head, the vena capitis medialis, from the mesenchyme lying alongside the brain, can be observed clearly. The vasofactive cells simply become detached from the mesenchyme and a lumen is produced partly by a process of vacuolation of the cells and partly by re-arrangement of the cells to surround a cavity.

Vasofactive cells which are destined to contribute to the posterior continuation of this vessel are seen lying close alongside the spinal cord in the region of the somites. As in the case of the cells of the aorta it is difficult to locate their origin precisely but they probably arise from the somitic mesoderm. Until Stage IV these cells remain isolated, but in Stages IVa and VII a series of intersegmental offshoots from the dorsal aorta can be seen and these run medianwards and dorsalwards towards the spina] cord. Probably both the vasofactive cells lying alongside the nerve cord and the intersegmental offshoots from the dorsal aorta contribute to the formation of a longitudinal vessel lying close alongside the neural tube. This vein in the cat, as in other vertebrates (Salzer (11), Grosser(3), Sabin(10), Parker(7), etc.), contributes to the formation of both the anterior and posterior cardinal veins. In the cat its anterior portion is undoubtedly developed in situ and not derived as described by Sabin (10) for the chick from the first aortic arch. In the embryos of Perameles and Macropus described in an earlier paper (7) it was noted that the vena capitis medialis is present in the head before it becomes connected with the dorsal aorta, It is therefore probable that in these species also the anterior portion of this vein arises in situ.


The origin of the umbilical vein in situ from the somatic mesoderm has already been described by Schulte (12). From my observations it appears that some of the cells that are destined to contribute to the umbilical vein-Cuvierian duct complex arise from the splanchnic mesoderm and migrate laterally round the pleuro-pericardial coelom to reach their normal position in the splanchnopleure.


Final Summary

  1. The heart arises as two lateral endothelial tubes, surrounded by a myocardium and the first portion of the heart to appear is the bulbo-ventricular limb.
  2. The mode of union of the lateral heart tubes and their forward continuations varies in different species and in different individuals of the same species. In some cases the right and left ventral aortae are continuous in the middle line before the fusion of the heart tubes takes place (Wang (15), ferret). In other cases the bulbar region of the heart tubes is the first part to fuse. These differences depend mainly on variations in the rate of production of vasofactive cells and the rate of growth of the pleuro-pericardial canal.
  3. The formation of the head is initiated by the forward growth of the brain plate, but this by itself produces no reversal of the pericardium, which attains its normal position immediately ventral to the gut by a process of backward movement and enlargement of the pericardium as a whole. The median pericardium increases in antero-posterior and dorsi-ventral dimensions partly at the expense of the lateral limbs of the pleuro-pericardial canals and in this way the heart tubes which are at first lateral, come to be surrounded by the median pericardium. At the same time the pericardium undergoes a rotation in such a way that a complete reversal is accomplished first in the middle line and the lateral heart tubes for some time necessarily follow a spiral course between their anterior ends which lie dorsal to the median pericardium and their posterior ends which lie ventral to the lateral limbs of the pleuropericardial canal.
  4. The lengthening of the fore gut is brought about by growth in anteroposterior extent of the pericardium, accompanied by a backward movement of the anterior intestinal portal.
  5. The approximation of the lateral heart tubes is due to the fact that for a time the growth in length of the pericardium exceeds its growth in width to such an extent that there is an actual decrease in the width of the pericardium, so that the lateral heart tubes are brought closer together.
  6. After the heart tubes have reached their position in contact with each other, ventral to the gut, they grow rapidly in length and become curved. Fusion of the right and left endothelial tubes takes place at this period, beginning in the bulbar region. It does not take place regularly in an antero-posterior direction, but portions of the bulbus aortae and ventricle remain unfused for some time.
  7. Blood vessels arise in situ, the cells which constitute their walls being derived from the mesoderm in the region of the vessel. The exact place of origin of the cells of the aorta is not clear in this material, but they appear to arise chiefly from the lateral (splanchnic) mesoderm and migrate from their place of origin to the aortic line.


List of Papers Quoted

(1) Bremer, J. L. “The development of the aorta and aortic arches in rabbits.” Amer. Journ. of Anat. xm. 111-128. 1912.

(2) Evans, H. M. “On the development of the aortae, cardinal and umbilical veins and other blood vessels of Vertebrate embryos from capillaries.” Anat. Rec. m1. 498-518. 1909.

(3) GrosssR, O. “Die Elemente des Kopfvenensystems der Wirbeltiere.”” Verh. d. Anat. Ges., Erg.-Heft z. Anat. Anz. Bd. xxx. 179-192. 1907.

(4) Hensen. “Beobachtungen iiber die Befruchtung und Entwickelung des Kaninchen und Meerschweinchens.” Zeits. f. Anat. u. Hntwickel. Bd. 1. 213-353. 1875-76.

(5) Horrmann, C. K. “Zur Entwickelungsgeschichte des Venensystems bei den Selachiern.” Morph. Jahrb. Bd. xx. 289-304. 1893.

(6) McCiors, C. F. W. “The endothelial problem.” Anat. Rec. xx. No. 4, pp. 219-238. 1921.

(7) Parxer, K. M. “The early development of the heart and anterior vessels in Marsupials.’ Proc. Zoo. Soc. of Lond. xxxu. 459-499. 1915.

(8) Rozrinson, A. “The early stages of the development of the pericardium.” Journ. of Anat. and Phys. xxxvu. 1-15. 1902.

(9) Rovvitre, H. “Etudes sur le développement du péricarde chez le lapin.” Journ. del’ Anat. XL. 610-633. 1904.

(10) Sasry, F. R. “On the origin of the duct of Cuvier and the cardinal veins.” Proc. of the Amer. Assoc. of Anat., Anat. Rec. 1x. No. 1, pp. 115-117. 1915.

(11) Satzer, H. “Uber die Entwickelung der Kopfvenen des Meerschweinchens.” Morph. Jahrb. Bd. xxm. 1895.

(12) Scuutre, H. von W. “Early stages of vasculogenesis in the cat, with especial reference to the mesenchymal origin of endothelium.” Mems. of the Wistar Institute of Anat. and Biol. No. 3. 1914.

(13) “The fusion of the cardiac anlages and the formation of the cardiac loop in the cat.” Amer. Journ. of Anat. xx. 45-72. 1916.

(14) Srraut and Cantus. “Beitrage zur Entwickelungsgeschichte des Herzens und der Korperhéhle.” Arch. fir Anat. und Entwick. xv. 231-248. 1889.

(15) Wana, C. C. “The earliest stages of development of the blood vessels and of the heart in ferret embryos.” Journ. of Anat. LI. 107-185. 1917.

(16) Yosurvaaa, T. “A contribution to the early development of the heart in Mammalia, with special reference to the Guinea-pig.” Anat. Rec. xxi. No. 3, pp. 239-308. 1921.


(13)

ABBREVIATIONS

A. angiocyst; A.A. 1, 2, etc., first, second, etc., aortic arch; A.D.A. angiocyst of dorsal aorta; A.1.P. anterior intestinal portal; A.M.P. anterior margin of medullary plate; A.U.V. angiocyst of umbilical vein; A.V.A. angiocyst of ventral aorta; Au. auricle; B.A. bulbus aortae; B.V.C. bulbo ventricular constriction; B.V.S. bulbo-ventricular sulcus; D.A. dorsal aorta; D.O.D.A. dorsal offshoot of dorsal aorta; E.H.T. endothelial heart tube; Hc. ectoderm; Hn. entoderm; End. endocardium; F.B. fore brain; F.@. fore gut; G. gut; H.M. head mesoderm; L.U.V. left umbilical vein; M.B. mid brain; M.F. medullary fold; M.P. medullary plate; M.T. medullary tube; My. myocardium; P. pericardium; P. 1 lateral diverticulum of pericardium; P.p.c. pleuropericardial canal; P.p.m. pleuro-pericardial mesoderm; R.U.V. right umbilical vein; S. 1, S. 2, etc., somites 1, 2, etc.; U.V. umbilical vein; V. ventricle; V.A. ventral aorta; V.C. vasofactive cell; V.C.D.A. vasofactive cell of dorsal aorta; V.C.H. vasofactive cell of heart tube; V.C.L. vena capitis lateralis; V.C.M. vena capitis medialis; V.C.M. 1 vasofactive cells of vena capitis medialis; V.C.U.V. vasofactive cells of umbilical vein; V.C.V.A. vasofactive cells of ventral aorta; V.V. vitelline vein.


Cite this page: Hill, M.A. (2020, September 27) Embryology Paper - The origin of the heart and blood vessels in felis domestica (1924). Retrieved from https://embryology.med.unsw.edu.au/embryology/index.php/Paper_-_The_origin_of_the_heart_and_blood_vessels_in_felis_domestica_(1924)

What Links Here?
© Dr Mark Hill 2020, UNSW Embryology ISBN: 978 0 7334 2609 4 - UNSW CRICOS Provider Code No. 00098G