Paper - The development of the vascular system in the human embryo prior to the establishment of the heart

From Embryology
Embryology - 20 Aug 2019    Facebook link Pinterest link Twitter link  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)

M'Intyre D. The development of the vascular system in the human embryo prior to the establishment of the heart. (1926) Trans. Roy Soc. 40(1): 12-20.

Online Editor 
Mark Hill.jpg
This historic 1926 paper by M'Intyre describes early cardiovascular development. This describes two historic human embryos in detail (Teacher-Bryce Embryo 2) and compares these two with the other known (named) human embryos of that time. This represents an excellent summary of these historic embryos from various sources and researchers.

Modern Notes:

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 Textbook" and "Historic Embryology" 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 and interpretations may not reflect our current scientific understanding.     (More? Embryology History | Historic Embryology Papers)


IV. The Development of the Vascular System in the Human Embryo prior to the Establishment of the Heart

By Donald M‘Intyre, M.B.E., M.D., F.R.C.S.(Ed.) F.R.F.P.&S.(Glas.),

Assistant to the Muirhead Professor of Obstetrics and Gynaecology, University of Glasgow; Assistant Surgeon and Pathologist, Royal Samaritan Hospital for VVomen, Glasgow; Assistant Physician, Royal Maternity and Won1en’s Hospital, Glasgow.

Thomas Hastie Bryce
Thomas Hastie Bryce 1862 - 1946)

Communicated by Professor T. H. Bryce, F.R.S.

With Three Plates and Eight Text-figures.

(MS. received May 24, 1926. Read June 21, 1926. Issued separately November 11, 1926.)


Our knowledge of the earliest stages of blood-vascular development in the human embryo suffers from a dearth of suitable material. Early human specimens are not frequently available for examination; many are pathological; some, although of value for other purposes, are not sufficiently" well preserved to furnish observations on angiogenesis; direct microscopic observation of the tissues while undergoing development cannot be carried out as is possible, say, in the chick embryo. Our knowledge of the process must be based on descriptions of separate specimens representing different stages. Individual specimens, therefore, are worthy of careful record.

I have the opportunity of recording the blood-vascular development in two specimens of the presomite stage, viz. the TEACHER-BRYCE No. 2 and the MTNTYRE. These two specimens have recently been described by Professors BRYCE and TEACHER, and the reader is referred to their papers for the anatomical detail. The development of the system of which this paper treats has been considered in broad principle by BRYCE, but only in so far as it completes the anatomical description. Herein it is proposed to set down in some detail the vascular picture as it appears in these specimens.

The data accumulated will then be considered alongside that supplied in descriptions of other early embryos. Many controversial points are encountered in the literature. Take, as an example, the origin of the earliest angioblastic tissue. Does it arise from mesoderm or from endoderm? Isolated fixed specimens do not lend themselves readily to a solution of this problem, and recourse has been had to experimental work on living material. Needlessto say, no human material is available. Of this type of work the most recent and most important is that of Professor SABIN, who finds by direct observation of the blastoderm of the chick that the angioblast is developed from mesoderm. WANG’s work on ferret embryos, wherein specimens of different ages are followed through successive stages, is another recent addition to our knowledge of vascular development.

While material, other than human material, may be employed to settle general questions, the detail of early vascular development must come from human specimens. STREETER, in describing his specimen (Mateer), pays particular attention to the vascular system. INGALLS (1920) describes, with special reference to the vascular system, an embryo representing a very important stage in the development of thisisystem, viz. the stage at which the vascular elements in the different regions (chorion, yolk-sac, blastoderm) have commenced to establish connection with one another, ‘although the process is not yet completed. His description is most beautifully and clearly illustrated. This specimen is of special value when considered alongside the older of the two specimens to be dealt with in this paper.

Only one paper embodies the views of an observer after personal examination of a series of early human ova. I refer to BREMEB.’s article of 1914. His views must be regarded as holding a most important place in the literature. While reference in detail will be made later in this paper to the work of SABIN, WANG, STBEETER, and others, to appreciate properly the bearing of this new material on BREMER’s work it will be necessary at once to state some of the conclusions at which he arrives.

After examining many of the known specimens, BREMER concludes that “the earliest bloodivessels arise separately in the yolk-sac and in the body-stalk, by multiple anlages. The anlages in the body-stalk (and perhaps also in the yolk-sac, of. JUNG’s fig. 17) are funnel-shaped ingrowths of the surface mesothelium. . . . By partial fusion of the walls of an ingrowth a portion of the coelom, still bordered by mesothelium, may be cut off as a separate cavity, lying deep within the substance of the body-stalk. The endothelium seems to arise either (a) by delamination from the walls of such a detached portion of the coelom, or (b) by direct extension, in the form of an angioblast cord, from the mesothelial ingrowth. . 1. . Extension within the limit of the areas covered by the mesothelium is achieved by confluence of the detached portions of the coelom, or union of the cords; the result is a net comprising the various vascular units. Extension into the chorion, where the mesothelial layer is absent in the early stages, appears to be by direct centrifugal growth of the angioblast cords, without the addition of new elements from the surrounding mesenchyma. . . .”

After delamination of the endothelium we have, therefore, spaces regarded as isolated portions of the coelom, referred to by BREMER as “ unlined spaces,” containing strands which will go to form endothelium.

Teacher-Bryce Ovum No. 2

The ovum was found by Professor TEACHER at autopsy. The fixation and histological detail are excellent. This is worthy of emphasis, as it renders the specimen particularly valuable in supplying data regarding the many unsettled questions which arise in connection with vascular development. The following summary of measurements and anatomical notes indicate the stage of development reached :-

Measurements - Teacher-Bryce Ovum No. 2
Feature Dimensions (mm)
External dimensions of chorionic vesicle (roughly) 4 x 4.5 x 3.5 mm
Cavity of chorion 2.8 x 2.6 x 2.25 mm
Blastoderm 0.2 x 0.1 to 0.15 mm
Yolk~sac 0.05 to 0.2 x 0.396 mm
Amniotic cavity 0.09 to 0.1 X 0.16 mm
Ref: M'Intyre D. The development of the vascular system in the human embryo prior to the establishment of the heart. (1926) Trans. Roy Soc. 40(1): 12-20.

The yolk-sac, greater in size than the amnion, is conical in shape and is prolonged in a tapering process to form a second attachment to the chorion.[1] An allantoic diverticulum is not recognised. The villi are well developed and show simple but not extensive branching. The sections are 6'25 microns thick.

BRYCE in his memoir refers to the part of the chorionic vesicle at the point of entrance as the “vegetative pole,” and the area opposite on the decidua basalis and iii the vicinity of the body-stalk as the “ embryonic pole.” These terms will be employed similarly here.[2]


Before considering the question of vascular development, reference to the general arrangement and structure of the chorionic mesoderm is necessary. This layer has on its inner surface, and in direct contact with it, the granular reticular coagulated contents of the blastocyst which have been described by BRYCE. This material which had settled towards the embryonic pole lies on the inner surface of a very loosely arranged tissue, and makes it difficult to define the inner limit of the mesoderm. The nuclei of the chorionic mesoderm are of varied form, spherical, ovoid, or rod-shaped, and the size varies within limited range. This tissue cannot be resolved into individual cells, although usually there is a condensation of protoplasm around the nuclei. Most often this takes an elongated form with tapering extremities and with the long axis corresponding to that of the contained nucleus. A distinct cell-envelope is absent. The rod-shaped nuclei are most numerous at the inner aspect of the chorionic mesoderm, the rounded variety at the outer margin.

While the connecting stalk has a practically complete layer of mesothelium, the mesoderm at the embryonic pole and frequently at some little distance from the stalk shows small patches of what resembles a mesothelial covering (Pl. I, fig. 1). These small patches are not continuous with the mesothelium of the connecting stalk. Where they exist there is a zone of fine reticulated structure devoid of nuclei separating them from the underlying mesoderm.

M'Intyre1926 text-fig01.jpg

Text-Fig. 1. A flat reconstruction (20 sections) of the angioblastic strands and spaces in an area of the chorionic mesoderm at the embryonic pole directly opposite the operculum of the Teacher-Bryce ovum No. 2. The reconstruction is equivalent to a View at right angles to the microscopic sections, or to a silhouette view through the chorionic membrane.

Round the wall of the chorionic vesicle spaces occur in the mesoderm. In many of these there are present strands of nucleated protoplasm staining deeply with eosin. The significance of these strands and their relationship to the spaces in which they lie requires consideration.

The spaces form a complicated branching network disposed parallel to the trophoblast layer (text-fig. 1). The system is not continuous throughout. Some channels disappear when traced through several sections, while a few appear to open into the lumen of the chorionic vesicle. The walls of the spaces cannot be said to have an endothelial lining. Now and then a rod-shaped nucleus is seen in the wall, but this is no more frequent than in the mesoderm elsewhere. More often the nuclei in the wall of a channel are ovoid and without any more definite condensation of protoplasm around them than exists in relation to the mesoderm nuclei in general. Often no nuclei are seen in one wall of the space for a considerable distance, and this occurs mostly on the side towards the lumen of the vesicle. Again,at times a space appears in part of its course without any nuclei in the immediate vicinity of its walls, and the walls themselves are ragged and irregular. One is left with the definite impression that these are spaces which have opened up in the mesoderm and that there is very little, if any, special condensation of the mesoderm cells around them. The spaces are present principally in the inner region of the mesoderm, although occasionally they are seen near the trophoblastic covering but are never in direct relationship to it. They are not continued into the mesoderm of the villi. They form a striking feature at the embryonic pole, where they are of greatest calibre. Passing round the blastocyst towards the vegetative pole they are seen to diminish in size, until near the operculum where the mesoderm layer is thin they are infrequent and inconspicuous. At the area on the equator of the blastocyst, where the prolongation of the yolk-sac is attached, they are specially large, interrupting the gradual transition from pole to pole. "Where patches of mesothelium-like cells exist, no connection of these with the spaces could be made out.

The nucleated strands in the chorion can be recognised only in relation to the spaces or channels described above (Pl. I, fig. 2). In its typical form a strand consists of an elongated syncytial mass running in the lumen of a space. The protoplasm is small in amount and drawn out into a thin thread between nuclei where these are far apart. The protoplasm stains more deeply with eosin than that around the ordinary mesoderm nucleus. In the neighbourhood of the base of the stalk the protoplasm in places may swell out to contain several nuclei which are oval, rounded, or kidney-shaped, and are a little more regular in size than those of the mesoderm. The protoplasm has usually a clean-cut edge, but now and then the surface appears to have very minute thread-like processes passing from it. The nuclei where the strand is a more slender one are elongated and sometimes show a curved axis. In individual sections a strand may appear to be entirely free in the lumen, but often a narrow attachment to the wall is present (Pl. I, fig. 2). This attachment may be broader, and at times the strand forms one side of the wall of the space. Where a strand of some length appears, it may run obliquely from one side of the space to the other with attachment at either end.

Now and then a strand is seen to terminate free in a space, and this takes the form of a tapering, sometimes curved, tail-like ending. Plate I, fig. 3 represents a high-power drawing" with the aid of the stereoscopic eyepiece, by Mr A. K. MAXWELL, of a medium-sized strand. Like the spaces in which they lie, these syncytial strands do not form a continuous network. The larger ones branch and form isolated network systems throughout the mesoderm. No continuity of these structures with the mesothelium-like patches on the inner surface of themesoderm could be established. Excepting one strand described and figured by BRYCE in his memoir (Pl. iv, fig. 18), to which reference will be made later in this paper, these structures do not show the presence of a lumen. The arrangement of the nuclei and the general contour of many of the strands strongly suggest the possible, early appearance of a lumen. SABIN has observed the actual differentiation of mesoderm to angioblast in the blastoderm of the chick, and the angioblastic tissue unites to form cords of cells. STREETER describes, in the chorion of a human embryo, multinucleated, protoplasmic strands and intervening stages up to complete endothelial tubes.

In the next specimen to be described the earliest representation of vessel-forming tissue is found in solid, deeply-staining, nucleated strands, and the intervening stages up to corpuscle-containing channels are present. One has little hesitation, therefore, in concluding that these nucleated protoplasmic strands described are angioblastic strands.

The spaces or channels in the chorionic mesoderm possess no lining which might suggest that they are, or, in themselves are likely to form, vessels. Although it cannot be denied that some of these spaces communicate with the extra-embryonic ccelom, such communications are not numerous in comparison with the many spaces present. The chorionic mesoderm has a loose open arrangement of its tissues, and its inner limit is indistinct. No matter what type of spaces or cavities were distributed generally throughout this zone, it would be surprising if a few of these did not communicate, as mere accidental occurrences, with the extra-embryonic coelom. On the other hand, the spaces, whether they communicate with the coelom or not, may have been produced by the identical process which brought about the formation of the coelom. The extra-embryonic coelom in this ovum is established, and the arrangement of the spaces makes it unlikely that they will go to produce its further development, so that they are not a part of the process of coelom formation. The fact that most of the spaces do not communicate with it merely indicates that they are not formed by tubular extensions of the coelomic cavity into the chorionic mesoderm. It should be noted also that they never contain any of the granular coagulum present in the coelomic cavity.

A point which is worthy of consideration is the possibility that these spaces have been merely potential during life, and that they have become actual only in the course of preparation of the specimen. Against this we have the absence of distortion of the tissues in their vicinity and the presence of many with a Wide and sharply defined lumen. In the preparation of the specimen slight exaggeration of the spaces, present as actual channels in the living state, may have occurred.

The solid protoplasmic strands in these channels are regarded as angioblastic tissue, they are destined to form endothelium and blood corpuscles, as will be seen in the earlier forms in the chorion of the M‘Intyre ovum. There is no evidence to support the possible view that the spaces represent vascular channels and that the strands will go to form blood-cells only.

I venture to suggest in explanation of these channels that they are present to facilitate the rapid diffusion of material for the nutrition of the tissues of the ovum. As might be expected, they are found in direct relation to the angioblastic tissue which at this stage requires special provision for its rapid, even precocious, development. They persist only for a very short period because when the extra-embryonic coelom is fully formed, and the vesicle has become larger, the chorionic mesoderm is relatively a much thinner layer and has a greatly increased area. This improvement in facilities for diffusion having been established, the spaces disappear, the vascular elements lie in direct contact with their surroundings, and the chorionic mesoderm as a whole becomes a more condensed layer. The M‘lntyre ovum will show this stage completed.


Angiogenesis in the base of the stalk takes the same form as in the chorion. Here the mesoderm forms a considerable mass, but is broken up by the spaces which contain the angioblastic strands. The spaces are irregular in shape and taper off in several directions in the plane of the chorion into the more regular channels seen in that membrane. The contained strands are here less slender. Sometimes quite a mass of protoplasm with numerous nuclei is present, and this, at intervals, shows a broad attachment to the wall of the space. Again, the mass may appear broken up into small portions containing one, two, or three nuclei only. The spaces in the stalk come into close relationship with the amnion duct, but, close to the amnion itself---that is nearer the embryonic end of the stalk-another mode, or perhaps phase, of blood-vascular development is found. Collections of cells situated in the mesoderm of the stalk have all the appearances of early blood-cells. The two largest of these can be traced through four and five sections respectively. The remainder are smaller and consist of four to six cells. The cells stand out from the surrounding mesoderm by virtue of their nuclei being more rounded and more regular in size than those of the mesoderm in general, also, each nucleus has a narrow rim of deeply-stained protoplasm making the individual cells spherical. Although the protoplasm is not frankly limmoglobimcoloured, the deep staining is highly suggestive. The spaces in which these cell masses lie differ from those already described in that they have the lumen almost completely filled by the contained cells. Further, there are cells arranged around the wall, in appearance similar to those in the lumen of the space except that here and there the nucleus is elongated and is disposed along the wall. In one section (Pl. I, fig. 4) the collection of cells appears as if it had been formed by a sinking in from the surface in the direction indicated by the arrow. The space in which the cells lie would thus be regarded as continuous with the ccelom. Traced from this section, the space with its contained cells is found to run obliquely to the middle of the stalk in a direction away from the chorion. This arrangement and the appearance of the cells gives quite a different picture to that of the spaces at the base of the stalk and in the chorion.

If the interpretation that the space mentioned above communicates with the coelom is correct, it lends support to BREMER’s View that fu11nel-shaped ingrowths of the mesothelium of the body-stalk occur in association with the earliest stages of vessel-formation.


The yolk-sac is cut very obliquely to its long axis. For the most part both endoderm and mesoderm are composed of a single layer, andin both of them localised thickenings are seen sparsely scattered throughout. In the mesoderm this takes the form of a duplication of the single layer with a suspicion of radial arrangement of nuclei. Otherwise they differ no way in appearance from the rest of the layer in which they lie. There is nothing which enables one to label these collections “ angioblastic.”

In one instance situated in the mid-lateral wall a small mass lies between the mesoderm and endoderm (Pl. II, fig. 5). This mass, which is present through seven sections, at one end in two sections shows quite definite continuity with mesoderm (Pl. II, fig. 6). Elsewhere it lies free between the two layers. It is clearly a mass of mesoderm which has projected in towards the endoderm, and has then made its way along between the two layers. There is no hwmoglobin coloration, but the protoplasm stains fairly deeply, and the nuclei are more spherical than the adjacent nuclei of the mesoderm and endoderm. The nuclei are not arranged in any definite manner, and the mass cannot be resolved into individual cells. There is no actual proof that this is the precursor of a blood-island, but in all probability such an interpretation is correct. In the narrow prolongation of the yolk-sac to the chorion no angioblastic tissue is recognisable. As already stated, spaces and strands are larger and more numerous in the mesoderm of the chorion where this attachment of the yolk-sac is placed. The endoderm cells of this yolk-sac stalk terminate in the wall of a specially large space containing angioblastic strands. This space with a strand is continued into the slender mesodermic process which replaces the yolk-sac prolongation and terminates in the cavity of ‘ the blastocyst. The appearances in this particular area have been described by BRYCE. A further brief description will be necessary but is postponed until similar appearances in other ova have been indicated in the next section of this paper.


In the mesoderm of the villi there is no evidence of commencing vessel formation.

Minute spaces or clefts are present without any special condensation of protoplasm or nuclei in their vicinity, and without protoplasmic strands in their lumen. Many of the spaces were traced in the direction of the chorion, and in no case was communication with a space containing angioblastic tissue in the chorionic mesoderm established; nor do the spaces of themselves form a continuous network in a villus. Probably these spaces in the villi _ represent the fluid parts of the mesodermic tissue. They may have become slightly exaggerated in the course of preparation of the specimen, as shrinkage of the mesoderm from its trophoblastic covering is not at all common in the villi, while it is seen in places in the wall of the chorion. p Although vascular development is present in the mesoderm of the body~stalk, no evidence of its commencement is found in the prolongation of this on to the amnion. There is no evidence of the commencement of the process in the tissues of the embryonic shield.

M‘Intyre Embryo

This specimen was found by the writer in the course of his work as Pathologist at the Royal Samaritan Hospital for Women, Glasgow, in a uterus removed by supravaginal hysterectomy. The uterus was partially opened until the blastocyst projected, was washed in running water and then placed in Kaiserling’s formalin and salt solution. The specimen was immersed in the latter fixative less than an hour after operation. The shrinkage of the uterine wall ruptured the decidua capsularis and caused complete separation of the ovum, which came away free when the specimen was next handled.

Measurements - M‘Intyre Embryo
Feature Dimensions (mm)
External dimensions of chorionic vesicle including villi (after fixation) 14 X 13 X 8 mm
Anterior extremity of yolk-sac to posterior extremity of amnion 1.75 mm
Blastoderm (including primitive streak) length 1.37 mm
Blastoderm (including primitive streak) greatest breadth 0.5 mm
Primitive streak 0.32 mm
Yolk—sac (approximately) 1.44 X 1 mm X 1 mm
Amnion, length 1.28 mm
Ref: M'Intyre D. The development of the vascular system in the human embryo prior to the establishment of the heart. (1926) Trans. Roy Soc. 40(1): 12-20.

Sections 10 microns thick are numbered from the head end of the embryo to the base of the stalk and pass through the embryonic axis exactly transversely. In this embryo the primitive streak is fully developed. There is a notochordal plate and a neurenteric canal, and the,fore-gut has just commenced to close in. The yolk-sac is relatively large and is greater in size than the amnion. The villi show plentiful branching.


The yolk-sac as yet cannot be said to possess any vessels. Blood-islands are numerous and consist of multi-nucleated protoplasmic masses projecting from the surface of the vesicle. Nowhere is an endothelial-lined lumen with free individual blood-cells found.

As seen in transverse section the endoderm, in the middle line on the ventral aspect of the sac, consists of a single layer of cells of a low columnar type with centrally-situated ovoid nuclei. Traced dorsally to the embryonic area the cells gradually become cuboidal or spheroidal with spherical nuclei. For a short distance before reaching the embryonic shield the cells are flattened out in the line of the yolk-sac wall. This area is the thinnest part, and here there is a longitudinal infolding of the wall. The mesoderm cells have a less regular arrangement and are less constant in size, form, and in shape of nuclei. The intensity of nuclear staining also varies. In the middle line ventrally the mesoderm forms a thicker layer than elsewhere. Towards the embryonic area it is represented by a single layer. There is a tendency for the mesoderm cells toclump together, and the nuclei in these groups of cells have a radial arrangement relative to the surface. The blood-islands are situated principally in the lateral wall midway between the embryonic area and the ventral aspect (text-figs. 2 and 3). This corresponds to an area just ventral to the thinnest part of the wall. The cephalic end of the yolk-sac for a short distance is totally devoid of blood-islands, but the'se are seen to appear before the anterior end of the blastoderm is reached. On the left side of the yolk-sac the blood-islands are larger and more numerous than on the right. Again, on both sides they are more plentiful opposite the head end of the embryo and in the area of yolk-sac wall bridging across the tail-fold. Between these two areas they are scanty in number on the left side and still less numerous on the right. Away from the mid-lateral Wall towards the blastoderm margin, or towards the ventral aspect of the yolk-sac, bloodislands are rare and when present are small and in an early stage of formation.

M'Intyre1926 text-fig02.jpg M'Intyre1926 text-fig03.jpg

Text-Figs. 2 and 3. M‘Intyre embryo. The drawing representsa sagittal section of the embryo, amnion, and body-stalk, with superimposed alateral elevation of the yolk-sac. The blood-islands of the yolk-sac have been plotted in to scale from individual drawings of tlie sections. The numbers of the sections are indicated. It was not- considered necessary to correct the reversal of the View as taken from the sections.

The blood-islands show a variety of forms, and these may be divided for purposes of description into four types which are taken to represent different stages in development.

Stage 1. This form, which is regarded as the earliest evidence of blood-island formation, consists of a hemispherical mass with the central nuclei becoming spherical (Pl. II, fig. 7). The mass projects slightly from the surface of the yolk-sac. The protoplasm possesses no haemoglobin colouring. This form is taken to represent a stage following on the grouping of mesoderm cells already described. There is no proof that it is the precursor of a blood-island except that the next stage is similar to it, with the addition of haemoglobin colouring of protoplasm in the centre of the mass.

Stage 2. In this stage (Pl. II, fig. 8), the central nuclei are mostly spherical, have a bold outline, and stain less deeply with haemalum than the adjacent mesoderm nuclei. In the periphery of the mass showing haemoglobin colouring one or more deeply stained crescentic nuclei may be seen. These are not constantly present, but when present appear to belong to the central mass rather than to the surrounding protoplasm.

Stage 3. Increase in size of the blood-islands results in definite projection from the surface, and the mass originally hemispherical becomes almost spherical in section. The bulk of it is composed of protoplasm showing haemoglobin colouring and numerous faintly stained spherical nuclei of regular size (Pl. II, fig. 9). This mass cannot be resolved into individual cells. Surrounding it there is a narrow zone of .protoplasm uncoloured by haemoglobin, with nuclei in a single layer. These nuclei are well spaced and irregular in shape. The protoplasmlof this surrounding zone cannot be sharply demarcated from the central mass.

Stage 4. This still more advanced picture (Pl. II, fig. 10) is seen only in a few of the largest blood-islands in the posterior extremity of the yolk-sac at the tail-fold and, therefore, near the commencement of the body-stalk. The central mass now shows the presence of clefts in the protoplasm, dividing. it up into irregular masses closely packed together. Some of these masses have attachment to the surface zone. In one or two instances what appeared to be individual cells were seen, but examined alongside neighbouring sections these were decided to be merely narrow terminations of a mass cut transversely so as to show only one nucleus. A greater number of the nuclei of the outer covering now assume a flattened shape.

The blood-islands vary considerably in size, and the size is not related to the intensity of haemoglobin colouring of the contained protoplasm. Some of the smallest, having only four or five central nuclei, show the protoplasm as intensely coloured as the largest. The blood~islands vary in shape as viewed on the flat, and have an irregular arrangement. So irregular is the distribution that attempts to make a transparency reconstruction of them from the sections were only partially successful. It was possible, however, to appreciate that they do not form a continuously connected network although connection between neighbouring blood-islands is established. At the head end of the vascular zone the bloodislands appear to have their greatest measurement in the axis of the embryo, whereas at the tail end there is a definite tendency for them to be disposed at right angles to the embryonic axis, and here also they are more elongated. In addition, they are not so sharply limited by their outer zone with its elongated nuclei.

Although there is often difficulty in resolving mesoderm and even endoderm into individual cells, there is no diflicultyin differentiating mesoderm and endoderm. This enables one to state that the vascular tissue present is more closely connected to mesoderm than to endoderm.


In section No. 115 the yolk-sac merges into the funnel-shaped diverticulum which becomes the allantoic duct. This is taken to represent the upper limit of the body-stalk. In section No. 190 the stalk merges with the chorion. Its length, therefore, is 0'75 mm. The tail-fold of the embryo reaches as low as section No. 150; the amnion with its lower limit in section No. 175 passes off from the dorsal aspect of the stalk (text-fig. 2). The allantoic duct definitely established in section No. 122 ceases in section No. 166. The stalk in transverse section may be takenas roughly circular in outline, with the allantoic duct situated almost in the centre (text-fig. 4). The stalk increases gradually in diameter towards the chorion until it breaks up unevenly into strands which turn outwards to join the chorion very obliquely. This gives a very complicated picture in the sections through the lower end of the stalk.

The mesoderm of the stalk consists of a finely reticulated protoplasm which stains faintly with the basic stain employed. The nuclei are vesicular, are lightly stained with heemalum, are of fairly regular size, and are slightly ovoid or short rod-shaped. This tissue cannot be resolved into individual cells, nor is there any sharp condensation of protoplasm around the nuclei. A mesothelial layer is present (Pl. III, fig. 11). It is most complete at the level of the middle of the stalk, where it consists of a thin film of protoplasm which has taken up the eosin stain and a single layer of round nuclei smaller than and more deeply stained than the ordinary mesoderm nuclei. The nuclei are irregularly spaced; sometimes wide intervals exist between two nuclei. This covering is incomplete at both extremities of the stalk.

M'Intyre1926 text-fig04.jpg

Text-Fig. 4. M‘Intyre embryo. Photomicrograph by Professor J. H. TEACHER of a section of the body-stalk. The prolongation of the amnion also appears, and is seen passing off from the dorsal aspect. The allantoic duct can be recognised about the centre of the stalk. On either side of it an umbilical artery is present; behind it on one side a venous space is seen and on the other side an angioblastic mass (the dark elongated area). In the mesoderm of the amnion wall the “blisters” referred to in the text are shown. x 120. Section No. 154. See text-fig. 2.

The presence of two large vessels (the umbilical arteries) containing free cells and of approximately equal size forms the most striking feature of the stalk (text~fig. 5). These are situated one on either side of the allantoic duct at the embryonic extremity. Both have a wide lumen. They commence in section No. 121, and as they pass towards the chorion they increase in size and gradually come to lie anterior to the allantoic duct. In section No. 156 they connect across in front of this duct by a narrow open channel; in section No. 159 they connect by a solid strand; in section No. 168 they unite to form a single large vessel which in section No. 180 again splits up into two main branches running to the chorion.

The vessel walls consist of a condensation of nuclei and protoplasm. The nuclei are of shape similar to the ordinary mesoderm nuclei, and lie one, two, or three deep in the wall of the vessel. This contrasts with the surrounding mesoderm, which is specially scanty in nuclei in the ventral half of the stalk through which the vessels run. The condensation of protoplasm corresponds to the nuclear zone and takes up the eosin stain.

Text-Fig. 5. M‘Intyre embryo. Front elevation of the body-stalk, showing the two arteries present, and a lateral view of the left vessel assembled from drawings of individual sections.

Elongated nuclei which might be regarded as belonging to endothelium are present only at wide intervals. The lumen contains free nucleated blood-cells. These are rounded or polygonal and have a relatively large amount of protoplasm which shows unmistakable haemoglobin colouring. The nuclei are very regularly spherical and are situated centrally. They vary greatly in intensity of staining. Some stain so faintly that the presence of a nucleus is made out only with difficulty. In others the nucleus stains very deeply. Mitotic figures are seen, and the presence of two nuclei, one sometimes larger than the other, in a single cell is not uncommon. No syncytial masses are seen in the vessels, but now and then blood-cells lie closely applied to the wall. In the latter, however, direct protoplasmic continuity does not exist. Occasionally a V-shaped depression or cleft in the wall of the vessel passes outwards - towards a small mass of angioblastic tissue. Such masses are not specially common in association with the vessels at present being described. They resemble the cells within the vessels, and are sharply demarcated from the surrounding unaltered mesoderm. Sometimes the vessels appear to communicate with the mesothelium by a cleft or depression in the vessel wall passing towards another cleft on the surface of the stalk, the interval being bridged across by a solid strand. This is not frequently noted nor is it a conspicuous feature. Complete open communication is not found.

Traced towards the embryo the two vessels end blindly, and there is no connection established with the angioblastic tissue of the yolk-sac, neither by open channels nor by solid strands. In the stalk above the termination of the vessels there is angioblastic tissue, but this is less in amount than at any other level of the stalk.

Although in the mid-line the yolk-sac is replaced by its posterior diverticulum in section No. 115, the lateral walls are continued downwards to bridge across the tail-fold of the embryo to section No. 131. The amnion is now interposed between the yolk-sac and the body-stalk (text-fig. 2). The possibility of angioblastic connection between the stalk and this part of the yolk-sac wall rich in blood~islands by way of the amnion wall was excluded. Several small isolated masses of angioblastic tissue were encountered in the amnion mesoderm here, but only close to the stalk, and they might be regarded as belonging to the latter.

Another picture of vascular development is found in the dorsal part of the stalk and would appear to be unconnected with that already described. At the junction of the embryonic and middle thirds of the stalk, behind the allantoic duct and frequently situated near the angle between the body-stalk and the amnion on either side, angioblastic masses are seen. Some of these have an appearance rather like the blood-islands of the yolk-sac. They lie just under the mesothelium and have a few crescentic nuclei disposed around them - the central mass showing haemoglobin-coloured protoplasm. They form, however, no projection on the surface. Passing towards the chorion this gives place to larger irregular masses of the same structure, only differing in shape, running. directly dorsal to the allantoic duct. About the middle of the body-stalk these become smaller in size, are diffusely distributed, and sometimes show connection by clefts with a network of spaces which has appeared at this level. The walls of the spaces have the same appearance as that of the vessels already described, except that condensation of nuclei is less marked, endothelium-like nuclei are less frequently seen, and the protoplasm in the wall stains less deeply with eosin. The spaces connect up with one another behind the allantoic duct in an irregular manner. In transverse section the lumen is often stellate. Numerous V-shaped depressions pass outwards to end in solid processes which sometimes run to isolated angioblastic masses. Two clefts passing outwards from a space may converge and almost isolate a mass, as it were, in the lumen. This mass has the same structure as the wall of the space elsewhere. There is no haemoglobin colouring of the protoplasm nor other indication that it will be any more closely concerned in the production of endothelium or blood-cells than the other parts of the wall.

The main channels of this network are represented in the diagram as they exist in the middle third of the stalk (text-fig. 6).

Text-Fig. 6. M‘Intyre embryo. The plexus of spaces in the dorsal region of the stalk in the same view as text-fig, 5. Nearer to the chorion they beconle Slnauer in size, more numerous, and more complicated in arrangement. They can be traced to the mesoderm of the chorion. At this end of the body-stalk they cannot be traced as continuously connected. Some of them establish connection by solid processes with the walls of vascular channels in the chorion which contain undoubted blood-cells. Although this network of spaces contains no blood-cells, it is certainly the commencement of a capillary network, the future umbilical veins. Throughout the examination of this area particular attention was directed to the relationship of the angioblastic tissue to the mesothelium. Mention of this has already been made in connection with the umbilical arteries. In one instance what is almost certainly a connection between a well-formed angioblastic mass and a funnel-shaped depression of the mesothelium was found (Pl. III, fig. 11). This mass, which can be traced through twelve sections, in one part of its course may be regarded as a vessel. At two points it appears to have connection with the mesothelium, while at another level it connects by a solid strand with the network of spaces behind the allantoic duct.


It will be necessary to refer to the structure of the mesoderm before describing the vascular picture. In the vicinity of the base of the body-stalk the mesoderm has a loose arrangement. It cannot be definitely resolved into individual cells, although here, unlike the mesoderm of the stalk, there is evidence of some concentration of protoplasm often in the form of a spindle around the nuclei. The nuclei vary in shape from short ovoid to rod-form with blunt extremities. They lie principally with their long axes parallel to the chorion. Around the base of the stalk one frequently finds the mesoderm marked off into two layers by a very loose arrangement of the tissues in its middle. Under such circumstances, the mesoderm lining consists of one sheet clothing the trophoblast and another lining the blastocyst cavity. The latter contains the largest chorionic vessels seen, and would appear to be formed as a result of these large vessels leaving the chorion to pass into the stalk. Away from the attachment of the stalk the mesoderm of the chorion is a thinner, more compact layer, and gives indication of the presence of wavy fibrillee. The nuclei are relatively less numerous and approach nearer in appearance to adult fibrous tissue nuclei. The inner surface of the mesoderm near the body-stalk has, in areas, fine irregular protoplasmic strands streaming of into the cavity of the blastocyst. A mesothelium as such is not recognisable.

Vessels and solid angioblastic strands are present in the chorion, but only in an area limited to the vicinity of the base of the body-stalk. In this area they are quite numerous.

The earliest stage of vessel formation recognisable consists of a thin, solid strand of protoplasm staining rather deeply with eosin and partially, but not completely, marked off from its surroundings. It is never situated in a space as described for the angioblastic strands in the chorion of the T. B. No. 2 ovum. Three or four rod-shaped nuclei are arranged in a single row in the protoplasm. The next stage is the appearance of a nucleated haemoglobin-coloured cell in the strand. This cell is sharply demarcated and is distinctly a free cell. It lies in a space provided for it in the protoplasm of the strand, sometimes at its middle, sometimes at one extremity. This cell may be as great in thickness as the strand which contains it, and the protoplasm of the strand where it passes on either side to enclose it is thinned out and may be readily overlooked by the observer.

More commonly, the strand, before the appearance of frank haemoglobin colouring in the cell, contains a double or -treble row of elongated nuclei (Pl. III, fig. 12). Next, several h2emoglobin-coloured cells appear in a row and are situated in the middle of the strand (Pl. III, figs. 13 and 14). These are sharply marked" off from their surroundings and from one another, and the ends of the cells where in contact are flattened so that the cells are frequently square in shape, the other two sides being flattened against the walls of the space in which they lie. The appearances are such as to suggest almost that the cells are under compression.

These strands may be isolated or may establish protoplasmic connection with others of any stage in development. Intermediate stages are seen right up to the largest vessels present. Some of the latter have a lumen almost equal in size to that of the vessels in the stalk. Their walls consist of a condensation of protoplasm with numerous rod-shaped nuclei disposed parallel to the lumen. The structure of the vessel wall, apart from its richness in eosin staining and the regular arrangement of the nuclei, has no special feature to distinguish it from the surrounding mesoderm. There is not as yet a lining to be compared with adult endothelium. In these larger vessels the lumen contains free nucleated blood corpuscles, sometimes of irregular shape but the majority spherical. They differ in no way in appearance from those in the vessels of the body-stalk. The angioblastic tissue present tends to run parallel to the chorionic membrane, and takes part inithe formation of a network. The more mature vessels in the vicinity of the attachment of the stalk communicate together in a complicated manner. The earliest representations of vessels may connect up over limited areas either by solid strands or in part by open channels. Angioblastic strands are seen isolated, and these sometimes already possess cells in which haemoglobin colouring of the protoplasm is beyond doubt.

No angioblastic tissue is found directly in contact with the chorionic epithelium; only a few of the earliest strands are seen near the epithelial layer. As already stated, the vascular development is specially prominent on the inner (cavity) aspect of the chorionic mesoderm. This, along with the short distance in which a vessel with a wide lumen may be replaced by a solid strand, forms two striking features of vascular development in this area.

Special attention was directed to the vascular connection between the body-stalk and the chorion. Tracing the rudiments of the umbilical arteries from their common trunk at the lower extremity of text-figure 5, section No. 172, this single vessel divides into two. In section No. 180 the two branches passing away from one another become continuous with the network of vessels in the chorion. The arterio-vascular system in the body-stalk and chorion thus communicate by open. channel. The venous plexus of spaces, at the junction of the body--stalk and the chorion, establishes connection with solid angioblastic strands which are specially numerous in the vicinity of its termination, and by way of some of these with a few of the smallest vessels in the chorion. A connection by open channel was not made out. As far as could be gathered by tracing vessels and angioblastic strands, no connection exists either directly in the body-stalk or indirectly in the chorion between the two sets of vessels in the stalk, viz. the two arteries on the one hand and the venous plexus in the dorsal part of the stalk on the other.


The villi are well formed, are of considerable length, and show intricate branching. The mesoderm of the villi differs from that of the chorionic membrane in that there is no concentration of protoplasm around the nuclei. The protoplasm, which is of finely granular structure with a very fine network of fibrillee, stains faintly with the basic stain. It, therefore, resembles the mesoderm of the body-stalk in its staining reaction. The nuclei, on the whole, are situated wide apart and are not equidistant from one another, but are sometimes grouped together in collections of three or four. The blood-vascular development in the villi has reached a critical stage.

There are, in the first instance, in the villi in the vicinity of the base of the stalk, angioblastic strands consisting of a single row of nuclei usually, but sometimes of a double row, surrounded by protoplasm which stains deeply with eosin. These, therefore, stand out sharply in contrast to the general mesoderm of the villus which is faintly stained with hmmalum. A detailed description is unnecessary, as they have the same appearance as strands of similar dimensions in the chorion. In several the origin of haemoglobin cell from angioblastic strand is seen. In the vast majority of instances these strands run in the middle of the mesoderm in the axis of the villus. Exceptionally a short strand may run obliquely until it reaches near to the epithelial covering but never comes in contact with it. There is no space separating the strands from the mesoderm.

Such angioblastic strands as have now been described exist as already stated in a very‘restricted area at the base of the body-stalk. Elsewhere the villi show no i "evidence of commencing blood-vessel formation. Even in the area. where these strands "*are present, the great majority of villi as yet show no indication of the commencement “of the process.

The strands may lie in the base of the villus or half way between the two extremities. In the few instances where a strand is seen near the distal extremity, it stops a considerable way short of the tip of the villus. One angioblastic strand in the base of a villus was traced back into the chorion and eventually into what was sufficiently well developed to be termed a vessel.

Text-Fig. M‘Intyre embryo. The peI‘icaI.‘dia.l (..‘.C‘3l0I'I'1 Vie'Wed &nt9I'0traced backwards are found to terminate posteriorly. Graph-reconstruction from individual sections.

short of the chorion, and, therefore, have no connection with the vascular system therein. In fact, in examining a villus, one may encounter two angioblastic strands at different levels unconnected with one another or with the chorion.

In only two cases was a villus foundto contain free nucleated cells showing a haemoglobin reaction, or, in other words, a “vessel.” In one case the vessel established connection with a vessel of the chorion. In the other its proximal extremity terminated at the base of the villus in a solid strand which did not establish connection with any of the vascular elements in the chorion.

Embryonic Rudiment

The pericardial coelom is present in the form of a U-shaped space (text-fig. 7). The limbs of the “U” extend backwards in the mesoderm near the lateral borders of the embryo. Here the mesoderm consists of a double layer of cells, and the cavity is formed by separation of this double layer. When formed, therefore, the walls of the cavity dorsally and ventrally consist of mesoderm with a single row of nuclei (text-fig. 8).

Here and there the roof and floor come into apposition with one another for a short distance dividing the lumen into two. On the right side the canal extends back to section No. 107, on the left to section No. 102.

At the anterior extremity of the embryonic area they unite with one another to complete the formation of the “U.” This portion of the “U” is bent slightly downwards conforming to the shape of the embryo. If the sections are examined from the head end backwards this union in front appears first in section No. 42, which is also the first section in which the cavity of the amnion is apparent.

Text-Fig. 8. M‘Intyre embryo. Section No. 65. See text-figs. 2 and 7. The pericardial coelom is seen on either side. (By permission of Professor T. H. BRYCE.)

The entire embryo was carefully examined, especially the floor of the pericardial cavity, and nowhere were any changes found which could be regarded as evidence of early blood or vascular development. Nothing which could represent the commencing formation of the endothelial heart-tubes was recognised.

Notes on Vascular Development in Early Embryos

Having completed the description of these two specimens, and before developing any general conclusions, it may not be amiss to supply brief notes of the vascular development in other early embryos. It will not be necessary to analyse the points in every specimen which has been described. It will suflice if I restrict my analysis to Br-yce’s Selected List with the addition of three others, Viz. MEYER’S, TRIEI-"EL’S, and INGALLS’ (1920) specimens. BRYCE’S list is an approximation to a sequence in respect of differentiation and the order is maintained. MEYER’S ovum I have placed between T. B. No. 2 and STRAHL-BENEKE’s. TRIEPEL’s and INGALLS’ (1920) ova are included after GRAF V. SPEE’S “ Gle.”

The source of each specimen is indicated and, where available, a measurement of the embryo and of the blastocyst are given. Very short notes of the two specimens just described are inserted to make the list complete.

J. W. Miller (1913)

Curettage. Embryonic rudiment solid. Blastocyst 0.44 mm. There is no extra-embryonic coelom. There are no villi. There is no indication of the commencement of vascular development.

Teacher-Bryce No. 1 (1908)

Abortion. Embryo about '15 mm. Blastocyst '77 X '63 X '52 mm. There is no eXtra-embryonic coelom. There are no villi. There is no indication of the commencement of vascular development. I have had the

opportunity of examiningthe specimen and I agree with this finding. BREMER, also, came to the same conclusion after examination of the sections.

v. Mollendorff (Soh.) (1921)

Abortion. Embryonic knot. '17 mm.‘ Blastocyst '26 mm. There is no extra-embryonic coelom. There are no villi. There is no indication of the commencement of vascular development.

Linzenmeier (1914)

Vaginal hysterectomy. Embryonic anlage '21 mm. Blastocyst '75 x 615 x 525 mm. The extra-embryonic coelom is present. V-illi have formed and in these as in the mesoderm in general there is no trace of the commencement of vessel formation. The body-stalk has not yet assumed its characteristic appearance, the embryo lying in a collection of mesoderm on the Wall of the blastocyst. The anlage of the allantois is described as lined with epithelium coloured like red-blood corpuscles. It is of interest to note that the cells of the allantoic duct in the M‘INTYRE embryo have a somewhat similar appearance; so much so that this structure might readily be mistaken for angioblastic tissue. No further note of vascular development is made.

Peters (1899)

Suicide autopsy. Embryo '19 mm. Blastocyst1'6>< '8><'9 mm. The extra-embryonic coelom is present. The villi have a mesodermic core. A body-stalk can hardly be said to exist. In the model made by

KEIBEL from drawings by SELENKA, the external surface of the yolk-sac appeared uneven, but it could not be decided if these represented the anlagen of vessels.

Jung (1908)

Curettage. Embryonic anlage about '25 mm. Blastocyst 2'5 X 2'2 mm. All trace of vessels is absent from the chorion and villi. In the body-stalk there are present collections of cells with a lumen in their middle. The lumina have no contents resembling blood corpuscles. He hesitates to decide Whether or not these are the first vessel anlagen. Vessels are not recognisable in the embryonic mass.

Schlagenhaufer and Verooav (1916)

Suicide autopsy. Embryonic shield 0.24 mm. Blastocyst 2 x 1.6 x 1 mm. A well-defined body-stalk is present. In the mesoderm of the yolk-sac are thickenings which are regarded as the anlagen of bloodislands. The commencement of vascular development is not noted in the body-stalk, chorion, or villi.

Fetzer (1910)

Curettage. Embryonic shield '23 mm. Blastocyst 1'6 X '9 mm. The villi are vessel free. There is no mention of vascular development in the chorion. No vessel anlagen are present in the body-stalk or in the amnion wall. The mesoderm of the yolk-sac shows numerous projections but there is nothing found which could be called a vessel anlage. A process passes out from the yolk-sac to end free and without attachment to the chorionic wall.

v. Mollendorff (Or) (1921)

Hysterectomy. Chorionic vesicle 2'25><2><2'5 mm. In one instance in the embryonic shield a collection of mesoderm cells is found near the entoderm. Similar cell groups are found in the mesoderm of the yolk-sac. He does not decide definitely that these are concerned in the formation of blood or of vessels. In the chorion he finds here and there channels lined with flattened cells. These are pronounced to be vessels. Their cavities do not yet connect up to form a continuous system. There is apparently no vascular development in the body-stalk or in the villi.

Herzog (1909)

Autopsy. Embryonic shield '154 mm. Blastocyst 2'3>< '8>< 12 mm. The villi consist of projections without dichotomous branching. At the junction of the yoll:-sac and bod y-stalk mesoderm there are found some solid and some open circular masses of mesoderm cells. These formations are taken to represent the earliest “ anlage” of the yolk-sac blood-vessels. BREMEB regards HERZOG’s interpretations of these cellular rings as incorrect. The mesoderm of the chorion and of the villi do not yet show any trace of blood-vessels.

There is no mention of blood-vessel development elsewhere on the yolk-sac or in body-stalk or in the embryo itself.

Teacher-Bryce No. 2 (1924-26)

Autopsy (acute rheumatism). Blastoderm '2 mm. Blastocyst cavity 2'8 x 2'6 x 2'25 mm. Angioblastic strands are found in spaces in the chorionic mesoderm. These are most plentiful around the base of the body-stalk and are not continuously connected. The body-stalk contains similar but larger strands and spaces at its base. Nearer the embryo several spaces contain collections of cells having the appearances of primitive blood~cells. What is probably the commencement of blood-island formation is found in the wall of the yolk-sac. No angioblastic formation is recognised in the embryonic shield or in the villi.

Meyer (P. M. 1923) (1924)

Curettage. Embryonic shield '41 mm. Chorionic cavity 2.6 x 2.1 x 2.72 mm. No indication of vascular development is found in the mesoderm of the chorion or in that of the villi. The yolk-sac possesses blood-islands which do not communicate with one another and are furthest developed over the ventral aspect. A blood-isl.and consists of a collection of cells between the yolk-sac entoderm and the mesoderm coat with a narrow connection to the latter. He finds in the body-stalk dorsal to the allantoic duct a collection of cells which he regards as the first anlage of the umbilical artery. There is no mention of vascular development in the embryonic shield. The yolk-sac possesses a long process which passes across the blastocyst cavity to terminate in a vesicle in the chorionic mesoderm at the implantation pole. The yolk-sac prolongation contains no blood-islands. The vesicle he regards as lined on the outer (trophoblast) side by mesoderm, on the inner (towards the chorionic cavity) by entoderm. This is covered by mesoderm which separates it from the lumen of the chorionic cavity and between the two layers he figures a blood-vessel anlage (fig. 12, p. 59). One is unable to recognise the entoderm lining of the vesicle from the figure, otherwise, apart from the absence of strands in the lumen, the vesicle has the appearance of the spaces in the mesoderm at the attachment of the yolk-sac duct in the Teacher-Bryce ovum No. 2. Blood-islands, he points out, are found 1‘1owhere away from entoderm, and he takes this as indicative that the blood-islands have some relation to entoderm.

Strahl-Beneke (1910)

Curettage. Embryonic shield '75 mm. Blastocyst 3.8 x 22 x 12 mm. In this ovum on the under surface of the yolk-sac there are thickenings of mesoderm containing cell masses which are regarded as the forerunners of vessels, but these have not yet acquired the definite characters of vessels. Sharply defined spaces in the mesodermic envelope of the embryonic body and specially marked on the yolk-sac side are referred to as “like capillary vessels.” These spaces do not form a closed tube system and are empty except in one place where free cells are present. These cells, however, show no hmmoglobin colouring of their protoplasm. BENEKE thinks he can recognise similar spaces in the mesoderm of the Teacher-Bryce ovum No. 1 in the photograph (Pl. ii) reproduced by the authors. In the mesoderm of the chorion of the .STRAHL-BENEKE ovum spaces similar to these described in the embryonic sections are present and are referred to as having the appearance of empty endothelial tubes. The authors’ fig. 1, however, shows mesoderm of the chorion which in the description of the figure is stated to be vessel-free. The villi have a mesoderrnic core in which there are no vessels. This ovum has a prolongation of the yolk-sac which runs to the chorionic wall. There is no note of vascular development in this structure, nor does there appear to be any particular development of spaces in the mesoderm of the chorion where this structure reaches it.

Graf von Spee (V. H.) (1896)

Abortion. Embryonic shield '37 mm. Blastocyst 4 mm. Spaces in the chorionic mesoderm and in the body-stalk have no visible content; neither blood nor endothelial cells are found. The yolk~sac wall has irregular protuberances of the mesoderm. Corresponding to these are blood-islands situated between mesoderm and entoderm. The youngest stage of blood-island lies nearest to the embryonic disc. The formation of blood-islands is noted as extending nearer to the embryonic disc than in his older embryo “Gle.” EVANS, in KEIBEL and MALL’S Ma/n/anal of Human Embryology, states that some of the vascular anlagen of the yolk-sac of this embryo show evident differentiation into endothelium and blood-«cells. Also, that in the body-stalk and chorion there are, as GRAF SPEE has described, “highly characteristic strands of spindle cells.” This he thinks suggestive of endothelium, although there is no blood-cell formation.

Lewis (Minot) (1912)

In the yolk-sac wall among the mesodermic cells are vessels with a true endothelial lining and containing nucleated blood corpuscles. Sometimes a corpuscle is closely applied to the endothelium as if arising from it. The vessels of the yolk-sac do not pass into the body-stalk, which contains numerous vessels. There are spaces in the chorion which are regarded as vessels. “Frequently these contain strands of darkly staining cells suggesting collapsed endothelium.” There are no vessels in the embryo proper. It is not definitely stated that the vessels in the body-stalk are not in continuity with the spaces in the chorion, but this may be inferred.

Streeter (Mateer) (1920)

Hysterectomy. Embryonic plate 1 mm. Blastocyst 6.1 x 5.6 x 2.5 mm. Evidence of blood-vessel formation is present in all parts of the chorion. All stages from simple multinucleated protoplasmic strands to completed endothelial tubes are seen. In the villi the stage of development is similar to that in the chorion except that the vessels appear to be more numerous, although a great many villi, as a rule the smaller ones, show no sign of vessel formation. Embryonic blood-cells are recognised. In the body-stalk vascularisation has occurred to the same degree as in the chorion, and as in the latter vessels are mostly empty. Blood-vessel formation is also recognised over the greater part of the parietal mesoderm covering the amnion. Angiogenesis in the yolk-sac is limited to the caudo-ventral half, and is represented by clumps of cells up to completely formed endothelial tubes. In this process of vessel formation in the yolk-sac relatively few complete cells (blood corpuscles) result, and none of these as yet show evidence of the presence of haemoglobin. The process does not appear to have commenced in the embryo proper. The continuity or otherwise of the vessel~forming tissues in the diff'ere.nt areas is not definitely stated.

Debeyre (1912)

Hysterectomy. Embryonic shield '85 mm. Blastocyst 5.6 x 2.1 mm. There is no intra-embryonic coelom. The yolk-sac possesses blood-islands over the distal. pole. These vary from a solid group of cells arranged concentrically to irregular festoons containing cells compactly arranged. Apparently the cells show no haemoglobin colouring. The islands do not form a network. In the body-stalk bloodislands are present. Two of these are specially large and one of them describes a third of a circle spirally round the allantois. In the embryo there is a cellular collection which DEBEYRE thinks may be the first cardiac formation. In the villi occasionally a lumen is seen but without endothelial lining. He states that if these are vessels they represent a very early stage of development. Vessels are present in the chorion. The question of continuity of vessel-forming tissue in the different areas is not discussed.

Thompson and Brash (1923)

Curettage. Embryonic shield including caudal fold '9 mm. Chorionic vesicle 10 x 7.5 X 4 mm. There is no intra-embryonic coelom. Blood-islands are found on the ventral aspect of the yolk-sac and appear to be more advanced in development in the cranial half. All stages from small clumps of cells to completely formed vessels containing developing blood-cells are encountered. The description and figure would suggest that the blood-cells arise from entoderm, while the endothelium arises in the vicinity of mesoderm, although the writers do not arrive at any definite conclusion with regard to this. The doubtful presence of angioblastic tissue in the body-stalk is noted, otherwise there is no further evidence of vascular development in this specimen.

Rossenbeck (Peh. I) (1923)

Hysterectomy. Embryonic shield 1.4 mm. In this specimen the presence of the anlage of the aorta is suggested but not definitely afiirmed. The mesoderm of the amnion contains endothelial-lined spaces not continuously connected. Two strands run from the amnion to the chorionic mesoderm, one traversing the body~stalk. Both end in connection with vessel anlagen in the chorion. The picture of angiogenesis in the yolk-sac is stated to be the same as in STREETER’s (Mateer ovum). In the body-stalk at its caudal ‘end, two endothelial-lined lumina vessel anlagen appear. These traced caudally end in an unpaired vessel anlage which lies near the ventral surface of the stalk, and which finally divides and passes into numerous but not continuously connected vessel anlagen in the chorionic mesoderm. Vascular development in the villi is not mentioned.

Strahl (1916)

No history. Embryonic anlage '7 mm. Chorionic vesicle with villi about 10 mm. diameter. There is no mention of vascular development in the embryonic shield or chorion. Vessel anlagen are not recognised in the body-stalk or villi. Vessel anlagen are present as thickenings in the mesoderm of the yolk-sac wall. Cells, free in the lumen of the yolk-sac, are regarded as nucleated red-blood corpuscles.

Grosser (1913)

Abortion (after operation.) Embryonic shield including primitive streak '67 mm. Chorionic cavity 6 to 8 mm. across. Vessels are absent from the embryonic area. The yolk-sac shows b1ood-islands on the distal pole. Blood corpuscles are recognised by their intensive staining, and the lumina of the bloodislands possess an endothelial lining. The blood-islands are already connected together in some parts. In the body-stalk and chorion there are empty cleft-like spaces resembling vessels and lined by fairly irregular endothelium. At a single place in the body-stalk there is a true blood-island. Vascularisation of the villi is not mentioned. A mesoderm strand passing from the yolk-sac right across the cavity of the blastocyst to the chorion shows the presence of entodermal cysts. In the wall of one of the larger cysts bloodislands are present.

Ingalls (1918)

Abortion. Blastoderm 2 mm. Ovum external measurements 9'1 X 82x 6 mm. In the embryo there are no vessels or blood-cells. Spaces present “might stand in some relation to the future pericardial coelom.” Over the fundus of the yolk-sac there is extensive formation of blood-cells and blood-vessels. The body-stalk contains numerous vessels filled with nucleated red cells. There are a few “funnel” ingrowths of mesothelium, but connection of these with “unlined” spaces and angioblast cords, which are also present, is not specially evident. Vessels and solid strands are present in the chorion and villi, and are most frequent near the attachment of the body-stalk. The question of continuity of vascular tissue in the body-stalk, chorion, and villi is not entered into, but it is definitely stated that the vessels in the stalk are not in connection with the vascular tissue of the yolk-sac.

Frassi (1907-08)

Vaginal hysterectomy. Embryonic shield 1.17 mm. Blastocyst cavity 9'4>< 3'2 mm. There is apparently no vascular development in the embryonic area. The yolk-sac possesses early blood and bloodvessel anlagen which are situated between mesoderm and endoderm (fig. 16, 1908). As far as I can make out from fig. 16 (1908), the cells in the centres of the blood-islands are individual cells, and the limiting cells approach more ‘nearly to endothelium than is the case in the M‘IN'.I'YRE ovum. Blood-vessels are present in the body-stalk and, according to GROSSER, one of these contains blood-cells, the remainder being empty. Vessel anlagen in the chorionic mesoderm are only recognised with certainty near the insertion of the body-stalk, and two of these at the base of the body-stalk (fig. 17, 1908) show the presence of free cells. Vessel anlagen are not found in the villi. Fig. 15 (1908) shows a small cyst on the wall of the yolk-sac in association with “the anlage of a blood-vessel. with blood.” The epithelium of this cyst is similar to the coelomic epithelium where it passes over the blood-vessel anlagen. J UNG, referring to this ovum in the description of his own ovum, implies that all trace of vessels is absent from the chorion.

M‘Intyre (1924-26)

Hysterectomy. Blastoderm (including primitive streak) 1.37 mm. Chorionic vesicle external dimensions including villi (after fixation) 14X 13_><8 mm. Blood-is'lands are seen in the lateral walls of the yolk-sac. They are more numerous on the left than on the right side. They do not form a continuously connected network. None of these structures has reached a stage in development which would permit one to call it a “ vessel.” The body-stalk possesses two large vessels regarded as the umbilical arteries, and dorsal to these a venous plexus of spaces. Neither of theseisystems is in connection with the blood-islands of the yolk-sac, nor do they appear to communicate with one another, but both can be traced into the chorion. In addition to the above elements, small solid angioblastic masses, some not unlike the blood- islands of the yolk-sac, are found scattered throughout. _All stages of vessel formation are seen in the chorionic mesoderm in an area around the base of the body-stalk, and are restricted to this area. The commencement of vascularisation of the villi is noted only in the villi belonging to the same area. In only two villi could a “vessel” be said-to be present. Angioblastic tissue is found isolated in the mesoderm of the villi. In the embryo no evidence of the commencement of formation of the heart or of vessels was found. The pericardial coelom has the form of a “ U ”-shaped tube.

Eternod (Vulliet) (1894)

Abortion. Embryo 1'3 mm. Blastocyst cavity 6><4r8><3'6 mm. There is in this embryo a horseshoe shaped heart of symmetrical outline giving rise to two primitive aortas (with three or four aortic arches) which are continued into two umbilical arteries in the body-stalk, and presumably through these into vessels in the chorion. The villi are commencing to become vascularised. On the venous side, veins run from the chorion to unite in the body-sta1k to form the future umbilical vein; this divides into two, the branches running forward in the mesoderm of the yolk-sac on either side to the primitive heart. Circulation is thus established. Posteriorly a venous loop is formed around the allantoic canal. There is evidence of vessel formation in the amnion wall, and some of these elements are already canalised. In addition to the veins mentioned above, the yolk-sac possesses blood-islands and vessels in which the lumen is already established.

Graf von Spee (Gle.) (1889)

Abortion. Embryonic shield 1'54 mm. Internal dimensions of chorion 7.5 x 8 mm. No vessel formation nor commencement of development of the heart is found in this embryo. A space is present in the embryonic mesoderm. Blood anlagen are present exclusively in the wall of the yolk-sac. The mesoderm of the body-stalk and yolk-sac is rich in spaces having a smooth lining of low cells like embryonic endothelium. In the yolk-sac, cell strands lying between mesoderm and entoderm are more closely connected with the mesoderm.

EVANS, in Keibel and Matt, reproduces a drawing of a section in which the above-mentioned space in the embryonic mesoderm is shown, and he regards it as the pericardial cavity. In the same section the anlage of the cardiac endothelium is seen. Vascular anlagen are recognised in the embryonic area, and these can be traced into the vessels in the body-stalk - the anlagen of the umbilical arteries. He also recognises the presence of vessels in the chorion, but does not state if these are in continuity with the vascular elements in the body-stalk.

Triepel (1917)

Abortion. Embryonic shield 1'6 mm. In the embryonic shield projections between the mesoderm and entoderm are taken to represent fine vessels. In these no trace of blood-cells is found. The endothelial heart-tube is not recognisable. The mesoderm of the yolk-sac possesses blood-cells and vessel anlagen in great numbers. Near the anterior end of the yolk-sac extra large blood-islands are found. He finds that both blood cells and vessels arise from the mesoderm, but considers that in a few places cell processes of endoderm take part in the vessel formation. He concludes, however, that the vessel-forming potentiality of the mesoderm is greater than that of the endoderm. In the middle of the yolk-sac a few quite separate “Erythrocytes” are found. These are nucleated cells. The body-stalk has at its middle short representations of vessels which can be followed only through one or two sections. The villi show the presence of channels regarded as the anlagen of vessels, but these contain no blood-cells. In the mesoderm of the amnion near the body-stalk are a few isolated blood-cells.

Ingalls (1920)

Curettage. Embryo, greatest length 1'38 mm. Ovum 7.5 x 10.5 x 12 mm. In this embryo the pericardial coelom is present, a heart plexus has formed, and the dorsal aortas and the rudiments of two aortic arches are evident. In the yolk-sac vessels are abundant, especially laterally and posteriorly, but all are not canalised. In the body-stalk are two umbilical arteries and a venous plexus. The chorion contains numerous vessels, many with a wide lumen. Formed elements are present, however, only at the base of the body-stalk. The villi have abundant slender anastomosing channels or cords and many detached strands. One umbilical artery establishes connection with the vascular elements in the yolk-sac wall. The venous plexus of the body-stalk is not in connection with the yolk-sac vessels. “Slender and circuitous” connection between the plexus in the body-stalk and the chorionic vessels is established by small, thick-walled vessels regarded as ingrowths from the chorion to meet the independently formed vessels of the stalk. Vascular structures in the villi are in continuity with the deeper vessels of the chorion. Circulation in the sense of a pulsating heart is not yet established.



Reviewing the specimens considered, we find the following steps in vascular development in the yolk-sac.

The earliest embryo in which vascular development is mentioned is that of SCHLAGENHAUFER and VEROCAY, who find the cmlagen of blo0d-z'slands present. The lirst definite statement that blood-'islomds are present is found in MEYER’s description of his specimen. Vessels (with endothelium and blood corpuscles) are first encountered in the LEWIS (Minot) embryo. Vessels having or definite course over the yolk-sac are found only in ETERNoD’S specimen.

Of the embryos earlier in the list than T. B. No. 2, only those of SCHLAGENHAUEER and VEROOAY and HERZOG show indication of vascular development, and in the latter this is seen not in the yolk-sac proper but at its junction with the body-Stalk. Following on T. B. No. 2, blood-islands are present in all the specimens. The thickenings or grouping of cells in the mesoderm of T. B. No. 2, present also in FETZEE and v. MoLLENDoREE’S Op., are almost certainly blood-islands in the process of formation. With reference to the collection of mesoderm cells in T. B. No. 2, which lies between mesoderm and endoderm, it is interesting to note that V. MOLLENDORFF in his ovum “ Op.” describes a similar solitary collection which he also regards as of mesodermal origin.

The M‘INTYRE specimen appears after several (LEWIS, GROSSER, INGALLS (1918), and FRASSI) in which vessels are clearly formed. From the description given it is obvious that the term “vessel” cannot be correctly applied to the most mature form of blood-island which I have described. This doubtless represents a variation within normal limits. Another example of this variation is found in STRAHL’S specimen, in which vessel anlagen are described merely as thickenings in the mesoderm of the yolk-sac wall.

Regowding the Distribution of the Blood-Islands. In all specimens where definite mention of the distribution of the blood-islands is made, only one (INGALLS (1920)), corresponds at all closely to the arrangement in the IWINTYRE embryo. In the remainder, with two exceptions, the blood-islands are present, or furthest developed on the ventral or distal pole. In the THOMPSON and BRASH case, although the blood-islands are situated on the ventral pole, they are more advanced in development at the cranial extremity. TRIEPEL, in his embryo, finds the largest blood-islands at the anterior extremity. The restriction of the blood-islands to the lateral walls in the M‘INTYRE specimen has permitted me to figure them graphically (text-figs. 2 and 3) as they appear in side view of the yolk-sac. I have no explanation to offer for this departure from the arrangement usually described.

The Source of Celts forming Blood-Islands

In the majority of cases the blood-islands are described as being situated in the mesoderm layer, or as producing thickenings or protuberances of that layer (SCHLAGENHAUFER and VEROCAY, STRAHL-BENEKE, LEWIS, STRAHL, and TRIEPEL). In a few cases they are said to lie between mesoderm and endoderm. (MEYER, GRAF SPEE v. H., and GRAF SPEE “ Gle.”). In MEYER’s and in GRAB‘ SPEE’s “ Gle.,” although situated between the two layers, connection to mesoderm is noted. THOMPSON and BRASH, in describing their specimen, hint at the possibility of blood-cell origin from endoderm and endothelium from mesoderm, although they come to no definite conclusion on the matter. TBIEPEL finds in a few places in his ovum that processes from endoderm take part in vessel formation. In the original description of all the human ova considered, with the exceptions mentioned, no statement that blood-islands are connected to or arise from endoderm is encountered.

These facts at first sight might appear to decide the question under consideration, but one has to contend with the possibility of the early migration of endoderm cells into the mesoderm layer, there to give rise to the formation of blood-islands. If we accept the thickenings of the mesoderm present in FETZER, v. MOLLENDORFF (Up) and T. B. No. 2 as blood-islands in process of formation, then, as proof of endodermal origin, we might expect to find in those cell masses or in parts of them, in addition to mere alteration in arrangement, some slight departure from the appearance of the mesoderm layer in which they lie. If, for instance, we imagine a yolk-cell as the nucleus of one of these cell groups, in well-preserved specimens one would expect to recognise, at least, a different staining reaction of the protoplasm. I have not encountered any statement that such an appearance is present in any specimen in this series, nor is it found in the T. B. No. 2 ovum, which represents a stage at which it would most likely be found if such were the case.

MANN in Quain’s Anatomy accepts the mesodermic origin of the yolk-sac vessels. In Keibel and dial! EVANS also practically accepts a mesodermic origin for the vascular anlagen, but with the reservation that they may actually have arisen from the entoderm. MINOT, in the Same work, holds that the angioblast is formed from cells which separate from the yolk, or from the layer of yolk-cells. The most recent work bearing on this problem is that of FLORENCE SABIN, who has been able to observe in the living chick the actual differentiation of mesoderm to angioblast. This work would appear to furnish conclusive proof that in the embryo of the chick, at least, the blood-vessels arise not only in, but from, mesoderm.

WANG believes that in ferret embryos the blood-cells are formed before the endothelium and in connection with the endoderm. Endothelium arises from the mesoderm layer, and growing round and engulfing the blood-cells, takes them into the circulation. The THOMPSON and BRASH ovum, as has already been noted, is the only one in the human series in which a somewhat similar appearance is described. The vascular elements in the yolk-sac wall of the ferret embryos produce projections on the endoderm side, another appearance not encountered in the human embryo.

The evidence regarding the source of origin of the blood-islands in the human embryo may still be regarded as inconclusive, but the weight of evidence is decidedly in favour of origin from mesoderm.

The Differentiation to Blood-Vascular Tissue

Accepting that the blood-islands in the wall of the yolk-sac take their origin from mesoderm, at what stage does differentiation of mesoderm to blood-island cease? From the descriptions of many of the older specimens, even when vessels are present, it‘ is apparent that early blood-islands also are found. Definite statements regarding the continuity or otherwise of these are not always encountered, but it would appear that the differentiation is progressive, and that these early representations are not the result of growth from the more mature elements. The arrangement in the wall of the yolk-sac in the M‘INTYRE ovum at any rate supports this view. The stage at which this differentiation ceases is, I think, a stage beyond that embraced by the material examined or reviewed here.

Origin of Corpuscles, Plasma and Endothelium

Turning now to the separate elements of the yolk-sac vessels, are we to regard the blood corpuscles and endothelium as both arising from the blood-islands? If we regard the blood-island as consisting of the whole thickness of the mesoderm layer where a blood-island exists, then both elements arise from the bloodisland. If, on the other hand, we regard only the central part of the total mass (different in appearance to the adjacent mesoderm) as the blood-island, then it is possible that the bloodisland is responsible for the production of blood corpuscles, while from the surrounding mesoderm endothelium takes its origin. The former is the View generally held (BREMER, SABIN), but here again in the descriptions of early human embryos there is little evidence one way or the other. JORDAN, in the yolk-sac of a 13-mm. human embryo, finds the latter method of differentiation present. In the M‘INTYRE embryo the appearances generally suggest, first, a differentiation into angioblastic tissue which is responsible for production of both blood corpuscles and endothelium.

The most mature form of blood-island in the M‘INTYRE ovum represents the stage at which plasma first appears. This arises as a result of splitting-up of the central mass into smaller nucleated masses which no doubt by further division will produce individual cells---the blood corpuscles. STREETER finds “considerable conversion into clear plasma.” SABIN in the living chick finds that whole masses are destroyed in the process of liquefaction to form plasma. A comparison of direct observation of the process with the picture representing an isolated stage is impossible, but it may be said that in the material I have examined there is no evidence of destruction of nuclei in the process of plasma formation. The formation of plasma would appear to occur pom} passu with the splitting-up of the protoplasmic mass into individual cells. lt is not suggested that there is an immediate production of single free cells by this process of cleft formation, but rather that for a time masses of nucleated protoplasm remain attached to the outer boundary of the blood-island (the future endothelium). They are already haemoglobin coloured and will produce blood corpuscles. They, therefore, represent blood-islands in the sense in which the term is employed by SABIN, and their appearance corresponds to her description of such._

The Blood-Islands are not continuously connected. lt will be remembered that in the M‘INTYRE ovum the blood-islands in the yolk--sac are not continuously connected. MEYER states that in his ovum the yolk-sac blood-islands do not communicate with one another. DEBEYRE states that they do not form a network. GROSSER finds the blood-isla11ds already connected together in some parts. In the other ova where blood-islands are present, if not definitely stated, in the majority of cases the description implies that they are not continuously connected. If, as MINOT suggests, the angioblast appears as a reticulate grouping of cells between the mesoderm and endoderm, and maintains its complete independence throughout life, then the yolk-sac blood-islands would be connected up together at all stages and progressive differentiation of mesoderm would not occur. The evidence in human material is entirely against this view. The blood-islands are at first isolated from one another, inter-communication being established later. There is indication that this linking-up of the blood-islands has commenced in the M‘INTYRE embryo.

I have not encountered any funnel-shaped arrangement of the mesoderm of the yolksac as described by BREMER in JUNG’s ovum. The projection of mesoderm cells noted in T. B. No. 2. is solid.

In two cases blood-cells are found free in the cavity of the yolk-sac. STRAHL in his "embryo finds nucleated red-blood corpuscles free in the yolk-sac lumen. TRIEPEL describes erythrocytes (nucleated cells) as present in the middle of the yolk-sac. In this connection it is interesting to note that a few cells, in appearance very similar to the nucleated red-blood corpuscles in the stalk, were found in the yolk-sac cavity of the M‘INTYRE embryo. They were situated at the angle of junction of the yolk-sac wall and blastoderm on the left side. After careful consideration it was decided that these were endoderm cells heavily laden with yolk which produced a staining reaction of the protoplasm simulating haemoglobin colouring.

MINOT states that “in man, if red blood-islands occur at all, they must break up very early.” That haemoglobin deposit occurs in the blood-islands of the yolk-sac before these break up is clear from the appearances in the lVI‘INTYRE embryo. This, however, may be a very transitory stage, as MINOT implies.

No endodermal “blisters,” as described by SABIN, were encountered in either specimen examined. What were taken to be “blisters” in the mesoderm of the amnion receive mention shortly.


The earliest ovum in which vessel formation is noted is that of STREETER, where the process is recognised over the greater part of the amniotic mesoderm. ROSSENBECK finds endothelium-lined spaces present. ETERNOD describes early vessel formation, some of the elements of which are canalised. TRIEPEL notes a few groups of blood-cells near the body-stalk. The early vascular tissue present in the M‘INTYRE embryo is so near to the body-stalk that I prefer to regard it as part of the vascular development in that area. In this embryo ring-like structures or “blisters” in the mesoderm of the amnion are seen projecting outwards in some of the sections (teXt-fig. 4), but these were not regarded as concerned in vascular development. I am unable to offer any opinion with regard to their significance.

It is rather remarkable that the STREETEB. embryo---the earliest in the series in which vascular development is recognised in the amnion--should show comparatively extensive development. More so is this the case when we meet with no mention of the commencement of the process in many older embryos.


Mention of vascular development is encountered at the earliest stage at which a body-stalk exists, viz. in JUNG’s ovum. The cellular collections with lumen“ described, he does not. however, definitely decide to be Vessel anlagen. The next reference is found in HERZOG’S specimen, where the anlagen of the yolk-sac vessels are located at the junction of that structure and the body-stalk. The next in order is the earlier of the two specimens under consideration in this paper, and in it two separate regions of the stalk are involved in the process. The angioblastic strands and spaces situated at the base of the stalk indicate essentially the same process as is going on in the chorionic mesoderm. Consideration of this area is deferred until the vessel formation in the chorion is discussed. The question as to whether this process extends upwards into the stalk to represent the ingrowths from the chorion described by INGALLS (1920) will also require consideration. In the body-stalk proper the early commencement of vessel formation in a different manner has been indicated. The drawing reproduced (Pl. I, fig. 4) is not so convincing as the actual specimen. It is reproduced mainly to show, as indicated by the arrow, the possible interpretation that the mass has sunk in from the surface, a point of interest in connection with BREMER’s theory of origin of the vessels. The figure, however, demonstrates the difference between the mass and the surrounding mesoderm. There is no doubt that this represents the early formation of blood and vessel, and from its position it probably represents one or other umbilical artery.

In DEBEYRE’s embryo the next step in development is found. In the body-stalk blood-islands are described. Two of these are particularly large and elongated, and might well be taken to represent the first recognisable differentiation into umbilical arteries. In R0ssENBECK’s specimen we meet with the first open channels. Two endothelium-lined spaces present are described as having a course strikingly like that of the two vessels of the M‘INTYRE embryo. The combination of two regular vessel channels with nucleated red cells free in their lumina is first encountered in the latter specimen.

With minor variations vascular development in this region advances steadily in the specimens considered. Two exceptions may be noted, viz. S'rRAHL’s and TRIEPEL’s. In the former no blood-vessel anlagen are present, in the latter a somewhat more advanced stage of development than is described might be expected.

Returning now to the M‘INTYRE embryo, in the body-stalk we find two umbilical arteries, a venous plexus and isolated angioblastic masses. These last named in some cases resemble the blood-islands of the yolk-sac. Blood-islands in the body-stalk are described by DEBEYRE and INGALLS (1920). From their descriptions it appears that in the former the blood-islands cannot stand in relation to vessel lumina, as lumina are not yet established, while in the latter the blood-islands are situated in the vessels. In my specimen the angioblastic tissue is not located in the vessel channels. No blood-islands, in the sense in which the term is employed by SABIN, or as described by INGALLS, are found. Nucleated blood corpuscles are numerous in the umbilical arteries, but no syncytial or cell masses are attached to the vessel walls. My interpretation of this is that the angioblastic masses, as they mature and produce blood-cells, are incorporated in the vessel lumen. This inclusion of these masses of cells must coincide with or follow the freeing of the cells, otherwise their attachment to the vessel wall would have been encountered. In the venous plexus of spaces no free cells are present, but communication partly open and partly solid between the spaces and the angioblastic masses is very common. Such connection withythe umbilical arteries exists, but is not so frequent. Speculating on this, I should offer as an explanation that in the umbilical arteries bloodcells are now being produced partly by multiplication of the existing free blood~cells, whereas, the venous plexus at this stage having none, is entitled to a more generous supply of blood-cell forming tissues. 2

In describing the walls of the vessels in the body-stalk it was stated that their structure does not entitle one to apply the term “endothelium-lined” to these channels. The inner lining has to undergo further differentiation before it can be so described. It is quite obvious, however, that the lining of these channels will form endothelium. If this is so, then in the plexus of spaces we have tissue which is undergoing differentiation into endothelium in the absence, as yet, of blood-cell contents. Here it may be noted that STOCKARD finds in teleost embryos that the “ endothelium is in all cases utterly incapable of giving rise to any type of blood-cell.” When spaces become lined by endothelium, blood-cell reproduction stops. “ Thered-blood corpuscles are always produced so as to be delivered into the vessels. . . This description might quite well be applied to the body-stalk of the M‘INTYRE embyro. SABIN, however, maintains that it is ‘proved for the chick that endothelium can form erythroblasts.

Finally, with regard to the body-stalk in the M‘INTYR.E embyro, the condensation of protoplasm and nuclei which forms the walls of the umbilical arteries is a relatively thick layer. This layer is of a greater thickness, one might suppose, than is necessary for the production of a simple endothelial lining. Is it possible that we have here, already, evidence of the commencement of formation of the extra-endothelial structures of the walls? I think one is almost justified in concluding that such is the case.

The commencement of vascular development can be recognised in the body-stalk and yolk-sac at about the same stage. It is usually stated that vascular development can first be recognised in the yolk-sac, but in the human embryo the evidence points rather to a simultaneous commencement of the process in these two regions.


Among the ova earlier than the T. B. No. 2, in one only (v. MoLLENDoRFF’s Op.) is there mention of vascular development in the chorion. v. MOLLENDORFF describes channels which are accepted as vessels, and these are lined with flattened cells. In the T. B. No. 2ovum we meet with the angioblastic strands and spaces which have been described. Vessels are present in the chorion of DEBEvRE’s and INGALLS’ (1918) specimens, and numerous large vessels with plentiful free nucleated red cells have appeared in the M‘INTYRE ovum. Recent work on angiogenesis has been directed principally to solving the problem of development of intra-embryonic vessels. In the literature, therefore, one finds the descriptions of vessel development in the chorion less satisfactory than for the

'regions already considered. A noticeable feature is that a gradual sequence in development is not so readily made out.

The appearances in the T. B. No. 2. ovum are of particular interest and are reproduced somewhat closely in only one specimen, viz. LEWIS (Minot). In the v. MOLLENDORFF (0p.), STRAHL-BENEKE, GRAF v. SPEE (v. H.), Gnossna, and INGALLS (1918) ova, spaces only are found in the chorionic mesoderm. In the LEWIS (Minot) embryo, however, spaces with contained strands, the latter being regarded as collapsed endothelium, are described. BREMER] describes spaces and strands in several ova he examined, but these are found more particularly in the body-stalk. The V. MOLLENDORFF (Op), STRAHL~BENEKE, GRAB‘ V. SPEE, and LEWIS specimens all represent a stage in development very near to that of T. B. No. 2, while GRossER’s and INGALLS’ ova are slightly older. It would seem almost as if the appearances in the T. B. No. 2 ovum, which have been detailed in the early part of this paper, were present only for a very short period. There is nothing comparable to these spaces found in the M‘lNTYRE ovum or, as far as one may judge from the descriptions, in other specimens representing a similar or a later stage in development.

In the descriptive part of this paper I have suggested a reason for the presence of the spaces in the chorionic mesoderm, and have recognised that they may have become exaggerated by adventitious influence. I have also advanced some proof that the strands which they contain represent early phases of vessel--formation, that, in short, they are angioblastic strands.

In T. B. No. 2, although an occasional space communicates with thecavity of the vesicle, no connection of angioblastic strands with the inner surface of the mesoderm layer was encountered. The strands and spaces, already mentioned, at the base of the body-stalk differ in no respect but size from those elsewhere in the chorion. The area where they lie it is impossible to allocate with any certainty to either body-stalk or chorion, but as the appearances are essentially the same as in the chorion, I prefer to regard it for descriptive purposes as chorionic. Away from the chorion there is no indication of downgrowth in the body-stalk of angioblastic tissue into the chorionic mesoderm. The possibility of migration of endoderm cells from the yolk-sac to the chorionic mesoderm, there to provide the origin for angioblastic formations, has never been suggested. We must conclude, therefore, that in the chorionic membrane angioblastic tissue arises by differentiation of the mesodermic elements.

With reference to the M‘INTYRE embryo, blood-islands as described in the yolk-sac are not encountered in the chorion. In the chorion the vessel and blood-forming tissue is present in more elongated form. Although wide vascular channels containing nucleated red cells are found, these can always be traced a considerable distance in the wall of the chorion. The earlier types in their narrow elongated form give the impression that there is considerable effort to produce rapidly channelstwith a lining destined to form endothelium. In the yolk-sac, on the other hand, the effort would appear to be directed more to the forming of blood-cells. Even in the chorion, nevertheless, as soon as a channel exists, the presence of primitive blood-cells can be made out.

The collected evidence points to the commencement of vascularisation of the chorion as occurring at about the same time as it commences in the yolk-sac and body-stalk. This would correspond to a stage in development represented by an ovum slightly younger than the T. B. No. 2. In practically all the specimens olderithan T. B. No. 2 vascularisation of the chorion is noted in some form or other.


Descriptions of the process of vascularisation of the villi are even less satisfactory than those of the chorion. In the list of specimens surveyed, the process receives notice first in STREETER’s ovum, in which solid strands and endothelial tubes are present. DEBEYRE refers to doubtful early vessel formation. In INGALLS’ (1918) ovum, and in the majority of those older than his, vessels can be recognised. It is apparent that vascularisation of the villi commences at a later stage in development than is the case in the yolk-sac, bodystalk, and chorion.

In the M‘lNTYRE embryo there is no difference in the process of vascularisation as seen in the chorionic membrane and in the villi except in degree. In the T. B. No. 2 there is no proof that the (empty) spaces present in the mesoderm of the villi are concerned in vascular development. It is possible that these spaces might later take on the characters of those in the chorionic mesoderm and contain similar angioblastic strands. If such a stage exists for the villi, it has not been described. The presence of angioblastic tissue isolated in villi, as seen in the M‘INTYB.E embryo and as described by TNGALLS (1920), entitles one to assume that here again such masses result from differentiation of the mesodermic elements in site.

Pericardium and Heart

Of the ova earlier than ETERNOD’s, in three only does mention of the presence of vascular development in the embryo appear. v. MOLLENDORFF (Op.) finds one group of cells of doubtful significance; DEBEYRE thinks a cellular collection present may represent the first cardiac formation; ROSSENBECK suggests the possible presence of the anlage of the aorta.

The failure to find, after a careful search, any appearances in the MTNTYRE embryo indicative of the commencementof the aortic or heart rudiments makes it more than probable that the findings in these three specimens cannot be interpreted as suggested by those who have described them. The description of the FRASSI embryo, also definitely more advanced in development than the three mentioned specimens, would bear out this contention. The sudden transition to the presence of a heart and an established circulation in ETERNoD’S ovum seems almost too sudden to be accepted as normal. In GRAF SPEE’s “Gle.” EVANS finds the cardiac endothelium commencing, and INGALLS’ (1920) embryo has a heart plexus, dorsal aortee and aortic arches. From the latter at least it is a very short step to an established circulation. The size of the ovum or of the embryo we now also know cannot be taken by itself as an indication of the stage of development. It must be conceded, however, that, as described, the cardio-vascular development of ETEB.NoD’s ovum has probably been a little precocious. For comparison, it is interesting here to note that SABIN finds in the chick that in the embryo the angioblast can be seen to differentiate from mesoderm at the stage of five somites, and that the heartbeat may be established at the stage of ten somites.

With regard to the formation of the pericardium, INGALLS in his 1918 ovum regards tubular ingrowths (two on either side) from the extra-embryonic coelom as standing in some relation to the future pericardial cavity. If this is the method of origin there is certainly no trace of any communication between the “U”-shaped cavity in the M‘INTYP.E embryo (which conforms to the description given by ROBINSON for mammals in 1903) and the extra-embryonic coelom. The material and the literature I have examined throws no further light on the earliest phase of pericardial formation in the human.

Yolk-sac Prolongations and Mesoderm Strands in the Chorionic Vesicle.

The yolk-sac connection to the chorion in the T. B. No. 2 ovum is so well established and so accurately described by BRYCE that particular attention is directed to this type of structure in relation to vascular development. In FETzER’s ovum a process from the yolk-sac ends free in the chorionic vesicle, and has no particular relation to vessel formation in “the yo1_k-sac. It is r_ather curious thatthe following three ova showing yolk-sac prolongations extending to the chorion should appear in direct sequence in the list. These are the T. B. No. 2, MEYER’S and S'rRAHL-BENEI§E’s. In two of these, viz. T. B. No. 2 and MEYER’S, the strand from the yolk-sac terminates in relation to a space or spaces in the chorionie mesoderm which are in some way related to vascular development. In the former, angioblastic strands are more plentiful there than at any other part of the chorion, with the exception of the base of the body-stalk, while in the latter a blood-island is present, although these are absent elsewhere from the chorion. The similarity in appearance in some respects at this area, in these two specimens, has already been indicated. In the T. B. No. 2 ovum the yolk-sac prolongation, after reaching and running on the inner surface of the chorionie mesoderm for a short distance, is replaced by an elongated protoplasmic mass of mesoderm which ends free in the chorionie cavity. This projection is traversed by a space which contains angioblastic strands. In the wall of this space the yolk~sac prolongation finally terminates. The strands in the space attract particular attention. BRYCE, in his memoir, reproduces a section through this area (Pl. iv, fig. 18), showing a lumen in the angioblastic strand present. I have made drawings of the strand or strands in the individual sections, and after very careful consideration I have come to the conclusion that the lumen present cannot be definitely accepted as an early vessel channel. As this question is one of some importance, it will be necessary to give a brief description of the appearances in this region.

From the first section, after the endoderm elements have disappeared from the wall of the space, the space can be traced through forty-nine sections. In the first section, after the yolk-sac duct terminates, the space has just been formed by the running together of several spaces. From here onwards it does not branch. Throughout the sections the space, as far as one can judge, has been cut more or less at right angles to its long axis, except in the last three or four sections where it has been cut obliquely. In the first section there appear three strands, two of which both show one nucleus present and the third five nuclei. The next shows four strands having respectively one, three, five, and six nuclei, and these strands lie in the space equidistant from one another, so that if their margins came in contact a lumen would be established. The third section shows only one strand which as cut across has the form of the letter “ C,” while, in the fourth, in which fourteen nuclei can be counted in the strand, the gap in the “ C ” is almost but not quite closed. In the next two the “ C ” shape is maintained, but has opened up considerably. In the succeeding seven sections the strand appears in one, two, or three portions. There is no suggestion of a lumen, but where the strand is in one portion only it has a double curve in the form of the letter “ S,” one extremity showing attachment to the wall of the space. The next section, the fourteenth from the one first considered, shows the strand “ U ”-shaped in section. In the succeeding section it is “L”-shaped, and the next section in order is that figured by BRYCE where a closed circle is undoubtedly present. In the next section (seventeenth) the strand appears in three parts where again, if one imagined the adjacent margins in apposition, a lumen would be present. The subsequent section shows four strands; the nineteenth, one strand (attached to wall of space); the twentieth, five strands (two attached to wall of space); the twenty-first, four strands, while in the twenty-second there is but one strand “C”shaped in section. From here onwards in the twenty-seven subsequent sections the strands vary in size and number, but never again suggest the possibility of a lumen complete or incomplete.

The possibility that the break in the “ C ”-shaped form, of which there are five examples, is accidental, and that there was in reality a lumen present, must be taken into account. Even if the break in the circle in these sections has been produced in the course of preparing the material, Istill think these forms, including the particular one described by BRYCE, do not represent channels or tubular cavities in a strand. My interpretation is that the angioblastic strand in this particular space is constantly dividing into branches, and being joined by new branches which run parallel to one another, and that the lumen represents a space around which several branches have joined up to form one. The space, therefore, I regard as not situated in a strand but as part of the space in which the strand lies. From the M‘INTYRE ovum it is clear that the earliest vessel channels arise as closed spaces within the angioblastic strands. I have come to the conclusion, therefore, that in tl1eT. B. No. 2 ovum there is no canalisation of the angioblastic strands in the chorion, but that these are still solid.

In the STBAHL-BENEKE ovum the yolk-sac prolongation to the chorion possesses no early vascular elements, and although spaces are described in the chorionic mesoderm these do not appear to be specially prominent at the attachment ofthe structure. RossENBECK describes two mesodermic strands running from the amnion to the chorion (one through the bodystalk), and these end in connection with vessel anlagen in the chorion. GrRoSsEP.’S ovum possesses a prolongation of the yolk-sac to the chorion, and in the wall of one of the entodermal cysts which this structure contains blood-islands are present.

These connections between the embryonic mass and the chorionic mesoderm must have some relation to vascular development, but the information at present available does not furnish an explanation of this relationship. BRYCE mentions GrRossER’s suggestion that this “might be a reminiscence of a stage in the phylogeny of the primates in which there was a yolk-sac placenta.” I am prepared to state that, after careful examination of one specimen in which such a communication is present, I can find no evidence to indicate that the yolk-sac endoderm has any relationship to the angioblastic tissue present. This is contrary to MEYER’s findings.

Continuity of Vascular Tissues

The question of the continuity of the vascular elements has already been discussed with regard to the yolk-sac, and may now be considered for the other regions of the ovum, first individually and later collectively. In the body-stalk, while still at the stage of angioblastic masses or blood-islands, the umbilical arteries are being defined. This stage is represented by DEBEYBE’s ovum where two elongated blood-islands might be regarded in this light. Plexus formation in the case of these arteries must represent a very transitory phase, as in R0ssENBECK’S ovum while still referred to as “anlagen ” they have a course corresponding closely to that of the umbilical arteries in the M‘INTYRE. ovum. In the case of the veins, the plexiform phase would appear to be of much longer duration, as it is still present in INGALLS’ (1920) specimen and to a lesser extent in ETERNOD’s.

As regards the chorion, definite statements are rarely encountered. ROSSENBECK, however, finds the chorionic vessel anlagen not continuously connected. This is in agreement with the appearances in the T. B. No. 2 ovum, and also in the M‘INTYRE ovum. At this stage it is interesting to note that in a number of these early ova the area in the vicinity of the bodystalk shows a more advanced vascular development than elsewhere in the chorion. This is found both in the T. B. No. 2 and in the M‘INTYRE specimens. INGALLS (1918) finds vessels by far most frequent near the attachment of the body-stalk. FRASSI recognises vessel anlagen with certainty only near the body-stalk. In the INGALLS’ (1920) specimen, where vessels are abundant, formed elements are found only at the base of the body~stalk. As it is already clear that in the human ovum the information we have ‘is entirely against a progressive growth of vascular tissue outwards from the body-stalk, this can beregarded only as an early indication of differentiation of placenta from chorion. It is probable that at no stage of development is the chorion equally well vascularised throughout. Similarly, the villi at different points of attachment to the chorion, from the beginning, are vascularised in varying degree. Two ova indicate an independent origin for the vessels in the villi. These are the M‘INTYRE and INGALLS’ (1920) specimens. From human material, therefore, the evidence collected would indicate that in the different areas the vessels have their origin not in one angioblast but by progressive differentiation at many points into angioblastic tissue.

Turning now to a consideration of the continuity of the vascular tissues in the different regions, we again find in many of the specimens considered absence of definite statements on this question. Where vascular elements are recognisable in both the yolk-sac and the bodystalk one is frequently disappointed to find absence of a definite statement that these are, or are not, connected up with one another.

In the T. B. No. 2 ovum, if the interpretation of the appearances is correct, then one may assume that there is no extension of vascular tissue from the yolk-sac to the body-stalk. Similarly one may conclude that there is at this stage no extension of angioblastic tissue from the body-stalk to the chorion. This ovum, however, suggests the possibility of extension of angioblastic strands from the chorion into the body-stalk to communicate with vessels formed independently in the latter structure. This has already been suggested by INGALLS from observations on his 1920 specimen, and the appearances in the T. B. No. 2 ovum lend some support to his view. Possibly it is not so much an extension of vascular tissue into the body-stalk as a taking up of the chorionic mesoderm and its contained vascular elements to assist in the formation of this extremity of the stalk.

Here one might again refer to the distribution of the angioblastic strands and spaces in the T. B. No. 2 ovum. These are well distributed throughout the chorionic mesoderm, and although smaller and less numerous in the vicinity of the vegetative pole, are nevertheless present in that area. In the M‘INTYRE. specimen no vessel or vessel-forming tissue is found in the chorion except around the attachment of the body-stalk. What has become of the angioblastic strands near the vegetative pole of the T. B. No. 2 ovum “.3 These strands must either have reverted. to mesodermic tissue, or the mesoderm in which they were situated has, in the course of enlargement of the chorionic vesicle to the size it has attained in the M‘INTYB.E ovum, gradually moved round to the vicinity of the body-stalk.

Continuity of vascular tissue in the body-stalk and chorion is established in RossENBEOK’s ovum. Vascular communication between the yolk-sac and body-stalk is found in INGALLS’ (1920) ovum where open communication with the right umbilical artery exists. The M‘INTYRE ovum is especially interesting in this respect, that it presents a stage at which the umbilical arteries are well formed but do not communicate with the blood-islands of the yolksac ; nor does it appear that they communicate with the venous plexus in the body-stalk.

In INGALLS’ (1920) and the M‘INTYRE specimens continuity of angioblastic strands or vessels in the chorion and in the villi in some cases is established.

From consideration of human material we may conclude that the vessels arise independently in the different regions of the embryo, and that at an early stage the body-stalk vessels are linked up with those of the chorion before vascular communication is established between the yolk-sac and the body-stalk. The only apparent exception to this might be found in the body-stalk as already explained. There is absolutely no evidence that the vessels in the body-stalk result from extension of the yolk-sac vessels as has been suggested.

The question of establishment of vascular communication with the embryo proper does not come within the scope of this paper, but it is clear from the descriptions of the various specimens that the vascular tissues in the areas here considered are connected up before communication with the embryonic area is established.

Much has been writen about the origin of the intra-embryonic vessels because a proof of their independent origin in the embryo settles the question of the possibility of origin of vessels in loco. A very complete review of the literature and experimental work in this connection is given by M‘CLUB.E in his presidential address to the American Association of Anatomists, 1921. He concludes that the angioblast theory of HIS, in which it is maintained that the vascular tissue in the embryo is an ingrowth from the yolk-sac, no longer holds, and “that the general principle of a local origin of intraembryonic endothelium has been completely confirmed by experiment.” Some additional evidence from human material that the “general principle” holds also for the extraembryonic vascular development is supplied in this paper.

Returning now to the work of Bremer, the appearances in the T. B. No. 2 ovum in some respects conform to the descriptions and theories he offers. Spaces and strands are present abundantly in the chorion and at the base of the body-stalk, but do not fulfil BREMER’s description in that neither forms a continuous network, and therefore, cannot have arisen by direct extension from the body-stalk. N 0 connection of spaces or of strands with the mesothelium of the body-stalk could be made out. Spaces in the chorionic mesoderm may communicate with the extra-embryonic coelom, but no mesothelium is present. The angioblastic mass shown in P1. I, fig. 4, lying at a higher level in the body-stalk might be regarded in the one section reproduced as having arisen by a sinking in of the surface. Even if this mass were in continuity with the surface, it is unconnected with the spaces and strands at the base of the stalk. These spaces with their contained strands I have already suggested belong to the chorion rather than to the body-stalk, and a theory regarding their method of formation and their significance has been propounded. In the case of the M‘IN'1‘YRE embryo, the conditions present are against BR.EMER’s contention that “no surely isolated endothelial cords have been found in the chorion.” In the bodystalk of this specimen, however, the appearances are certainly in some cases very suggestive of a communication between the surface and the developing vessels through the medium of solid strands. . One of these is reproduced in Plate III, fig. 11. We may, therefore, conclude that the mesothelium of the body-stalk may play some role in vascular development in that part of the ovum, but that extension of this vascular tissue into the chorion and villi does not occur. The latter postulation is dependent on assuming that the spaces and strands in the chorionic mesoderm of T. B. No. 2 are the same as those described by BREMER. The appearances correspond sufficiently closely to warrant this.

In reading through the literature some difliculty is experienced on account of the wide range of terms employed to denote vascular elements and their different forms. For instance, in reference to the yolk-sac, does “the first anlagen of blood-vessels” represent a less mature or a more mature stage than “blood-islands”? I have been content to employ some of the existing terms, but I think a simplification of the terminology merits the consideration of embryologists. Throughout, I have restricted the term “vessel” to those instances where there is a lumen containing free blood-cells. In so1ne of the specimens it is quite obvious that the vascular development has only received passing mention, and no doubt, if all the specimens here considered were examined by one and the same individual, the record would show many differences. In some cases, notably STREETER’s, DEBEvRE’s and INeALLs’ (1920), the vascular development receives special notice, The personal factor may to some extent come into play in recording the picture of the vascular system in isolated specimens. The identification of the very earliest representation of angioblastic tissue can besubmitted only as an expression of opinion. The method of recording in this case adopted by one whose work did not previously necessitate a knowledge of the literature was as follows. In the first instance a complete examination of the material was carried out, the findings recorded, and illustrations prepared. This explains the small number of references to the literature in the descriptive part of the paper. An investigation into the literature was then made, and subsequently the material was again examined and a few additional drawings prepared. Some alterations in the original description were necessary, but where doubt arose the first interpretation was allowed to stand. The very favourable plane of section in the case of the M‘INTYRE specimen was of the greatest help in investigating the continuity of vascular tissues. I

The work in connection with this paper was carried out partly in the Anatomy Department of the University of Glasgow, and partly in the Pathological Department of the Royal Samaritan Hospital for Women, Glasgow. I am responsible for the drawings reproduced, with three exceptions, viz. Plate I, fig. 8, by Mr A. K. MAXWELL; text-fig. 8, which is reproduced from Professor BRYCE’S paper; and text-fig. 4, a photomicrograph which Professor TEACHER was good enough to prepare for me. For these illustrations and for many helpful suggestions from Professors BRYCE and TEACHER I wish to tender my sincere thanks. I wish also to acknowledge the privilege of being allowed to examine the sections of the ova, Teacher-Bryce Nos. 1 and 2.

The Carnegie Trust for the Universities of Scotland has made a grant to the Author to defray the expenses of reproduction of the illustrations. This grant is very gratefully acknowledged.


From a consideration of the material here recorded, and from a review of the literature, the following conclusions are set forth. Some of these may not appear to be fully justified by the evidence supplied, and are, therefore, to be regarded to some extent as opinions rather than conclusions. I

1. The blood-vascular system in the earliest stages of development in human ova arises in the extra-embryonic areas by progressive differentiation of mesoderm at multiple points in situ. Its commencement may be recognised while yet absent in the embryonic area proper.

2. The theory of vascularisation of the chorion and villi by centrifugal growth of angioblastic tissue from the yolk-sac or body-stalk does not hold for the human ovum.

3. Differentiation to angioblastic tissue occurs at about the same time in the yolksac, body-stalk and chorion, while it appears at a slightly later stage in the villi. The first differentiation to angioblastic tissue occurs at a stage just earlier in development than that represented by the T. B. No. 2 ovum.

4. The vascular elements in the separate areas become connected together and the different areas establish connection with one another in the following order: 1. Bodystalk, and chorion. 2. Chorion and villi. 3. Body-stalk and yolk-sac. 4. Yolk-sac, body-~ stalk and embryo.

5. From examination of the M‘INTYRE embryo the impression wasformed that in the wall of the yolk-sac the main effort is directed to the rapid production of blood-cells, whereas in the body~stalk, chorion, and villi channel formation is of equal importance to blood-cell production.

6. Prolongations of the yolk-sac across the blastocyst cavity to the chorionic wall have some bearing, not properly understood, on vascular development in the chorionic membrane. It does not appear, however, that the endoderm elements which these may contain have any part in the origin of the vessels in the chorion.

7. Vessel formation in the body-stalk outpaces that in the other regions, so that the first channels which can be identified as vessels having a recognisable course are those destined to be the umbilical arteries.

8. The arterial and venous systems in the body-stalk do not communicate with one another or, if they are ever in communication in that structure, they become isolated from one another at a very early stage.

9. Blood-cell forming tissues in the body-stalk communicate with the vessel channels but do not project into these channels.

10. The mesothelium of the body-stalk may play some part in vessel formation.

11. Extension of vessel growth does not take place from the body-stalk into the chorion, but there is some evidence that the reverse process may occur in connection with the venous channels in the body-stalk.

12. In the earliest stage at which angioblastic tissue is found in the chorionic mesoderm, it appears in the form of nucleated protoplasmic strands which run in spaces. These strands are destined to form vessels and blood-cells, while the spaces are regarded as a temporary provision for the ready diffusion of substances for the nutrition of the ovum and, more particularly, of the angioblastic tissues which the spaces contain.‘

13. At the stage at which angioblastic tissue appears in the chorionic mesoderm, the latter possesses no mesothelium, and, therefore, mesothelium in this region has no role in vessel formation.

14. At a very early stage of development the distribution of the vascular tissues in the chorion indicates a differentiation of the area around the body-stalk towards placenta.


  1. An excellent idea of the general anatomy of this ovum may be obtained by reference to BRYCE’s memoir, pl. i, fig. 4.
  2. TEACHER uses “ Closing Pole” and, following GRAF SPEE, “ Implantation Pole.”

Literature Cited in the Text

BREMER, J. L. (1914), “The Earliest Blood-Vessels in Man,” Amer. Jou/rn. Anat., vol. xvi, p. 447.

BRYCE, T. H., TEACHER, J. H., and MUNRO KERR, J. M. (1908), Contributions to the Study of the Early Development and Imbedoling of the Human Ovum. MacLehose, Glasgow.

BRYCE, T. H. (1924), “Observations on the Early Development of the Human Embryo,” Trans. Roy. Soc. .E'din., vol. liii, pp. 533-567.

BRYCE, T. H. (1925), “Measurements of very Early Human Embryos,” Zeitschr. f. cl. yes. Anat., 1 Abt., Bd.1xxvii, Hefte iii/iv, p. 493.

DEBEYRE, A. (1912), “Description d’un embryon humain de O""“' 9,” Journ. de l’Anat. et de la Physz'ol., N o. 48, p. 448.

ETERNOD, A. O. F. (1898), “Premiers stades de la circulation sanguine dans l’oeuf et l’embryon humain,” Anat. Anz., Bd. xv, p. 181.

Evans HM. The development of the vascular system. In Keibel F. and Mall FP. Manual of Human Embryology II. (1912) J. B. Lippincott Company, Philadelphia. pp570-708. (p. 590)

FETZER (1910), “Ueber ein durch Operation gewonnenes menschliches Ei ,' dass in seiner Entwickelung etwa dem Peterschen Ei entspricht,” Anat. Anz., Bd. xxxvii, Erganzhft., p. 116.

FRASSI, L. (1907), “ Ueber ein junges menschliches Ei in situ,” Arch. f. Mikr. Anat., Bd. lxx, p. 492.

FRASSI, L. (1908), “ Weitere Ergebnisse des Studiums eines jungen Menschlichen Eies in situ,” Arch. f. Jllilcr. Anat., Bd. lxxi, p. 667.

GROSSER, O. (1913), “ Ein menschlicher Embryo mit Chordakanal,” Anat. Hefte, Bd. Xlvii, p. 653.

HERZOG, M. (1909), “A Contribution to our Knowledge of the Earliest Known Stages of Placentation and Embryonic Development in Man,” Amer. Journ. Anat., vol. ix, p. 361.

Ingalls NW. A human embryo before the appearance of the myotomes. (1918) Contrib. Embryol., Carnegie Inst. Wash. Publ. 227, 7:111-134. (p. 113)

Ingalls NW. A human embryo at the beginning of segmentation, with special reference to the vascular system. (1920) Contrib. Embryol., Carnegie Inst. Wash. Publ. 274, 11: 61-90. (p. 63)

JORDAN, H. E. (1910), “A Microscopic Study of the Umbilical Vesicle of a 13-mm. Human Embryo, with special reference to the Entodermal Tubules and the Blood-Islands,” Anat. Anz., Bd. xxxvii, p. 56.

JUNG, P. (1908), Beitrdge zur Fruhesten Ei-einbettung beim menschlichen Weibe. Berlin.

Lewis FT. The form of the stomach in human embryos with notes upon the nomenclature of the stomach. (1912) Amer. J Anat. 13(4): 477-503. (p. 302)

LINZENMEIER, G. (1914), “ Ein junges menschliches Ei in situ,” Arch. f. Gyn., Bd. cii, p. 1.

M‘CLURE, G. F. W. (1922), “ The Endothelial Problem,” Anat. Rec, vol. xxii, p. 219.

MANN, G. (1912), “ The Vascular System,” Quain’s Anatomy, vol. ii, p. 352.

MEYER, P. (1924), “ Ein junges menschliches Ei mit 04 mm. langem Embryonalschild,” Arch. f. Gyn., Bd. cxxii, p. 38.

MILLER, J. W. (1913), “Corpus Luteum u. Schwangerschaft. Das jiingste Operativ erhaltene menschliche Ei.” Berlin Klin. l’V0chensch7‘., Bd. 1, p. 865.

MINOT, C. S. (1912), “ The Development of the Blood,” Manual of Human Embryology, Keibel and Mall, vol. ii, p. 498.

MoLLENDonr«‘F, W. V. (1921), “Ueber das jiingste bisher bekannte menschliche Abortivei (Ei Sch.),” p. 352, “Ueber einen jungen operativ gewonnenen menschlichen Keim (Ei Op.),” p. 406, Zeitschr. f. Anat. u. E'ntwicl:., Bd. lxii.

PETERS, H. (1899), Ueber die Einbettung des menschlichen Eics. Leipzig u. Wien. Franz Deuticke.

ROBINSON, A. (1903), “The Early Stages of the Development of the Pericardium,” Journ. Anat. and Phys., vol. xxxvii, p. 1. RossENBEcK, H. (1923), “Ein junges menschliches Ei (ovum humanum) Peh. 1, Hochstetter,” Zeitschr. f. Anat. u. Entwick., Bd. lxviii, p. 325.

SABIN, F. R. (1920), “Studies on the Origin of l5]ood-Vessels and of Red-Blood Corpuscles as seen in the living Blastoderm of Chicks during the second day of Incubation,” Contributions to Embryology, vol. ix, Carnegie Inst. Publications, No. 272, p. 215.

SCHLAGENHAUFER and VEROCAY (1916), “Ein junges menschliches Ei,” Arch. f. Gyn., Bd. cv, p. 151.

SPEE, F. GRAF V. (1889), “ Beobachtungen an einer menschlichen Keimscheibe mit offener Medullarrinne und Canalis neurentericus,” Arch. f. Anat. ancl'Phys., Anat. Abth. Jg., 1889, p. 159.

SPEE, F. GRAF V. (1896), “N eue Beobachtungen iiber sehr fruhe Entwickelungsstufen des menschlichen Eies,” Arch. Anal. n. P/2,ys., Anat. Abth. Jg., 1896, p. 1.

STOCKARD, C. R. (1915), “The Origin of Blood and Vascular Endothelium in Embryos without a Circulation of the Blood and in the normal Embryo,” Amer. Journ. Anat., vol. xviii, p. 227.

STRAHL, H., and BENEKE, R. (1910), Ein junger menschlicher Embryo. Bergmann, Wiesbaden.

STRAHL, H. (1916), “ Ueber einen jungen menschlichen Embryo nebst Bemerkungen zu C.’s Gastrulationstheorie,” Anal. Hcfte, Bd. liv, p. 115.

STREETER, G. L. (1920), “A Human Embryo (Mateer) of the Presomite Period,” Contributions to Embryology, vol. ix, N o. 43, Carnegie Inst. Publications, No. 272, p. 391.

TEACHER, J. H. (1924), “On the Implantation of the Human Ovum and the Early Development of the Trophoblast,” Journ. of Obst. and Gyn. of Brit Emp, vol. xxxi, N o. 2, p. 166.

THOMPSON, P., and BRASH, J. C. (1923), “A Human Embryo with Head Process and commencing Archenteric Canal,” Journ of Anat., vol. lviii, p. 1. TRIEPEL, H. (1917), “Ein menschlicher Embryo mit Canalis neurentericus; Chordulation,” Anat. Hefte, Bd. liv, p. 151.

WANG, C. C. (1918), “The Earliest Stages of Development of the Blood-Vessels and of the Heart in Ferret Embryos,” Journ. of Anat., vol. lii, p. 107.

Description of Plates

(A scale indicating the magnification is supplied with each drawing taken from the sections.)

Plate I

Fig. l. Teacher-Bryce ovum N 0. 2. Wall of chorionic vesicle showing an inner mesothelium-like layer (Mth.), a reticulate zone (Rz.), mesoderm (Mes), and the trophoblast ('l‘r.).

Fig. 2. Teacher-Bryce ovum No. 2 Wall of chorionic vesicle including the base of a villus. The mesoderm contains spaces, in one of which a large angioblastic strand appears in cross section (A). C., cavity of chorionic vesicle ; V., villus; Tr., trophoblast. Fig. 3. Teacher-Bryce ovum No. 2. Drawing by Mr A. K. MAXWELL of a medium-sized angioblastic strand and space in the chorion. Fig. 4. Teacher-Bryce ovum No. 2. Drawing from a section through the body-stalk to show a space containing early blood~cells (Ba), situated near the embryonic extremity of the stalk. The mesothelial covering on the surface (M.) is apparent, and its possible relationship to the space is indicated by the arrow. S., surface of body-stalk.

Plate II

Fig. 5. Teacher-Bryce ovum No. 2. Yolk-sac wall showing a small mass (blood-island '3) between endoderm and mesoderm. M., mesoderm ; E., endoderm. Fig. 6. The same mass as in fig. 5 two sections removed, showing protoplasmic connection with the mesoderm layer. M., mesoderm ; E., endoderm Figs. 7-10. M‘INTYnE embryo. Examples of the four stages of blood-island formation in the wall of the yolk-sac as described in the text. M., mesoderm 3 E., endoderm.

Plate III

Fig. 11. M‘INTYRE embryo. The field embraces the dorsal part of the body-stalk on one side. The amnion wall (Aw.), is seen passing off from the stalk. Near the angle between the amnion, wall and the body-stalk a depression of the mesothelium (Mth.) runs towards an angioblastic mass (Am.), which at this particular level may almost be regarded as a vessel. A., cavity of amnion.

Fig. 12. M‘INTYRE embryo. Wall of chorionic vesicle showing condensation of protoplasm and nuclei to form the earliest representation of an angioblastic strand (A3,). Tr., trophoblastg M., mesoderm of chorion.

Fig. 13. M‘lNTYnE embryo. Chorionic mesoderm showing the presence of an angioblastic strand in which haemo globin--coloured cells have appeared. Fig. 14. M‘IN'rYRE embryo. A further stage in development of the chorionic vessel is shown. Contained blood cells are seen and a lumen is present for some distance.

Cite this page: Hill, M.A. (2019, August 20) Embryology Paper - The development of the vascular system in the human embryo prior to the establishment of the heart. Retrieved from

What Links Here?
© Dr Mark Hill 2019, UNSW Embryology ISBN: 978 0 7334 2609 4 - UNSW CRICOS Provider Code No. 00098G