"The Carnegie Stages of Early Human Embryonic Development: Chart of all 23 Stages, Detailed Descriptions of Stages 1 - 6"


Stage 5

Approximately 0.1-0.2 mm
Approximately 7-12 postovulatory days
Characteristic features: implanted but previllous; solid trophoblast in 5a; trophoblastic lacunae, cytotrophoblastic clumps, and primary umbilical vesicle in 5b; lacunar vascular circle and some mesoblastic crests in cytotrophoblastic clumps in 5c

Stage 5 comprises embryos that are implanted to a varying degree but are previllous, i.e., that do not yet show definite chorionic villi. Such embryos are believed to be 7-12 days old. The chorion varies from about 0.3 to 1 mm, and the embryonic disc measures approximately 0.1-0.2 mm in diameter. The significant dimensions of Carnegie specimens of stage 5 are listed in Table 5-l. The external and internal diameters of the chorion are listed as "chorion" and "chorionic cavity," respectively. Additional features of stage 5 include the definite appearance of the amniotic cavity and the formation of extra-embryonic mesoblast. The appearances at stages 2 to 5 are shown in figure 5-1.

Implantation, which began in stage 4, is the characteristic feature of stage 5. It should be appreciated that both maternal and embryonic tissues are involved in the complex process of implantation: "in the normal process they are mutually supporting and neither can be regarded as chiefly responsible" (Boyd and Hamilton, 1970). An indication of a decidual reaction appears during stage 5 and, from this time onward, the term "decidua" (used by William Hunter) is commonly employed. The decidua, at least in the human, "is a tissue made up of endometrial connective tissue cells which have enlarged and become rounded or polyhedral due to the accumulation of glycogen or lipoids within their cytoplasm, and which occur either in pregnancy, pseudo-pregnancy or in artificially or pathologically stimulated deciduomata" (Mossman, 1937).

Successful implantation may depend on the ability of the embryo to produce an immunosuppressive factor (or factors) having a direct suppressive effect on the maternal immune response (Daya and Clark, 1986).

Failure of implantation may result from rejection of the antigenic embryo by the maternal immune system. Heuser's technique of opening the uterus laterally and searching for a young conceptus has been described on several occasions (e.g., by Heard, 1957, and by Hertig and Rock, 1973). No correlation has been found between the side of the uterus on which the conceptus becomes implanted and the ovary from which the oocyte originated. Normal specimens, however, are more commonly found implanted on the posterior wall of the uterus, abnormal ones on the anterior wall (Hertig and Rock, 1949).

Both walls are considered to be antimesometrial in comparison with a bicornuate uterus. Furthermore, "it is interesting to note that cases are known of a double discoid placenta in man very similar to that of the monkey. It seems entirely possible that in some cases the human blastocyst may attach both dorsally and ventrally and therefore fail to undergo complete interstitial implantation" (Mossman, 1937).

The trophoblast from stages 4 and 5 onward comprises two chief varieties, namely, cytotrophoblast and syncytiotrophoblast. That the latter is derived from the former had long been suspected and has been shown by organ culture and also indicated by electron microscopy (Enders, 1965). An amniotic cavity is found by stage 5. If duplication of the embryo occurs after the differentiation of the amnion, the resulting monozygotic twins should be monochorial and monoamniotic (fig. 5-2). It has been estimated that the frequency of monoamniotic twins among monozygotic twins is about 4 percent (Bulmer, 1970). About once in every 400 monozygotic twin pregnancies, the duplication is incomplete and conjoined ("Siamese") twins (e.g., the second specimen of Shaw, 1932) result.

The following description of stage 5 is based largely on the work of Hertig and Rock, in whose publications (1941, 1945a, 1949) much additional information (including descriptions of the ovaries, uterine tubes, and uterus) can be found. Based on the condition of the trophoblast and its vascular relationships, stage 5 is subdivided into three groups: 5a, 5b, and 5c (Hertig, Rock, and Adams, 1956). Although a brief description of the trophoblast at each stage is provided in this account, the main emphasis is devoted to the embryo itself. This is justifiable inasmuch as comprehensive books on the human tro- phoblast (Hertig, 1968) and the human placenta (Boyd and Hamilton, 1970; Ramsey, 1975; Ramsey and Donnet, 1980) have been published.

Stage 5a

The characteristic feature of subdivision 5a is that the trophoblast is still solid, in the sense that definitive lacunae are not yet evident. Specimens of this stage are believed to be 7-8 days old. The chorion is less than 0.5 mm in its greatest diameter, and the embryonic disc is approximately 0.1 mm in diameter. Because of the collapse of the conceptus during implantation, the blastocystic cavity is flattened. The rarity of specimens of stage 5a has been attributed to the circumstance that they are "impossible to discern in the fresh, and probably often unrecognizable even after fixation" (Hertig, 1968).

Endometrium. The endometrial stroma is edematous (fig. 5-4). Two specimens (Nos. 8020 and 8225) show early, superficial implantation. The conceptus has eroded the surface epithelium of the uterus and has barely penetrated the underlying stroma (fig. 5-5). Apparently an attempt has been made by the maternal epithelium to repair the defect, and occasional mitotic figures are found. A portion of the conceptus, however, is still exposed to the uterine cavity. A third specimen(No. 8155) shows later, interstitial implantation. The conceptus is almost embedded within the endometrium, so that its abembryonic pole, which is barely exposed, is nearly flush with the epithelial lining of the uterine cavity (fig. 5-6).

Trophoblast. The trophoblast may encroach on the surrounding endometrial glands. At the abembryonic pole, the wall of the conceptus is merely a thin layer of cells that resembles mesothelium. Because this region is not in contact with maternal tissue, it probably presents the structure of the wall of the blastocyst as it was at the time of implantation. Because of the collapse of the blastocyst during implantation, the mesothelioid layer is closely applied to the ventral surface of the embryonic disc (fig. 5-5).

As the mesothelioid layer is traced laterally, it becomes continuous first with indifferent trophoblastic cells, which, at the embryonic pole, become differentiated into cytotrophoblast and syncytiotrophoblast. Definitive trophoblast is found only in the area of endometrial contact, presumably under the influence of an endometrial factor. A ventrodorsal gradient of trophoblastic differentiation is noticeable. In other words, the most highly developed trophoblast tends to be found deeply (i.e., away from the uterine cavity). The cytotrophoblast is located nearer the embryonic disc. The cells are large and polyhedral, and show distinct cell boundaries. Mitotic figures are moderately frequent.

The syncytiotrophoblast is described by Hertig (1968) as "invasive, ingestive, and digestive." It presents adark, homogeneous cytoplasm, and large, densely stained nuclei. No mitotic figures are seen. Near the maternal tissue, the syncytiotrophoblastic mass displays numerous small nuclei, which appear to be formed amitotically, although some may perhaps be derived from the endometrial stroma. The syncytial masses project into, and frequently partly surround ("eat their way into"), the uterine stroma, giving the surface of the trophoblast a lobulated appearance. In only rare instances are vacuoles found in the syncytiotrophoblast, and they contain no maternal blood (No. 8020).

Macroscopically, no congestion or hemorrhage is visible in the endometrium. Microscopically, the capillary plexus and sinusoids are moderately dilated but contain very few blood cells. It seems that the endometrium itself is adequate for the nourishment of the conceptus at this stage.

It has been found (in Nos. 8020 and 8225) that "the course of several capillaries can be followed through the syncytiotrophoblast. The endothelial walls of the capillaries are intact to the point where each vessel enters and leaves the trophoblast, but between these points red blood cells can be seen to occupy a series of irregularly shaped spaces, which are in continuity with one another" (Harris and Ramsey, 1966). These spaces, however, "are unlike the well-rounded vacuoles occasionally observed in syncytiotrophoblast," and they are "much smaller than the definitive lacunae" present within a couple of days. It is assumed that "the syncytium advances by a flowing movement that engulfs the blood vessels" in the capillary plexus of the adjacent stratum compactum (ibid.). Isolated endothelial cells in the lacunae may lend support to the supposition that capillaries have been engulfed (Dr. Ramsey, personal communication, 1972). In other words, the future lacunae, which are usually in continuity with maternal vessels (No. 8020) "are interpreted to be derived from engulfed maternal vessels" (Böving, 1981).

Trophoblast and endometrium (in No. 8020) are intimately related, and no cellular boundaries can be seen by light microscopy. It is probable that uterine epithelial cells have been phagocytized prior to autolysis, although it is possible that they are fused with the trophoblast (Enders, 1976). Indeed, it has been claimed that the appearances seen in the rabbit (fusion of a uterine "symplasma" with the syncytiotrophoblast) may apply also to No. 8020 in the human (Larsen, 1970). Implantation in the rhesus monkey has been studied by electron microscopy, including the spreading of trophoblast along the basal lamina of the uterine epithelium, the breaching of the basal lamina, and cytotrophoblastic proliferation (Enders, Hendrickx, andSchlafke, 1983).

Extra-embryonic mesoblast. The term mesoblast is preferred by Hertig and Rock (1941) to mesenchyme ("rather nonspecific"), primitive mesoderm ("one might unintentionally imply some connection with the embryonic mesoderm"), or magma reticulare (which "refers to the more mature characteristics of this tissue"). The magma has also been considered as merely a "degenerative remnant of primary yolk sac endoderm" (Luckett, 1978).

The formation of mesoblast begins in stage 5a. The theory that the extra-embryonic mesoblast develops in situ by "delamination" (i.e., without cellular migration) from the cytotrophoblast (Hertig and Rock, 1945a and 1949) remained current for many years, although other possible sources (embryonic disc, amniotic ectoderm, and endoderm) were also considered.

In his study of the Miller (5c) specimen, Streeter (1926) concluded that the primary mesoblast "must have either separated off from the inner cell mass during the formation of the segmentation cavity or have been derived from the trophoblast. Since it is, in reality, so largely concerned in the differentiation of the trophoblastic structures, the latter is the more probable explanation." From his investigations of the same specimen and more particularly of macaque embryos, Hertig (1935) believed in the "simultaneous origin of angioblasts and primary mesoderm by a process of delamination of differentiation from the chorionic trophoblast....".

According to the alternative view, namely that extraembryonic mesoblast is not of trophoblastic origin, the trophoblast is considered to give rise to additional trophoblast only: i.e., cytotrophoblast, syncytiotrophoblast, and trophoblastic giant cells (Luckett, 1978). Hill's (1932) studies of the primate embryo led him to believe in "the existence in early embryos of the Pithecoid and man of a mesodermal proliferating area involving the postero-median margin of the shieldectoderm and the immediately adjoining portion of the amniotic ectoderm which contributes to, if it does not entirely form, the connecting stalk primordium.

This proliferating area, it may be suggested, functionally replaces, if it does not actually represent, the hinder end of the primitive streak of the Tarsioid and the Lemuroid...." In pursuance of this idea, Florian (1933) concluded that, in the Werner (5c) embryo, an "area of fusion of the ectoderm of the caudal part of the embryonic shield with the primary mesoderm" was a site where "at least a part of the primary mesoderm originates."

This theory of Hill and Florian has been supported by Luckett (1978) who believes that "the caudal margin of the epiblast is a precociously differentiated primitive streak, which gives rise to the extraembryonic mesoderm of the chorion, chorionic villi, and body stalk." The term caudal mesoblast-proliferating area will be retained here and, as in previous studies, identification of the primitive streak will await stage 6b, an interpretation that has long met with general agreement.

Amnion. The cells of the inner cell mass that are adjacent to the mural trophoblast at stage 4 may already be those of the amniotic ectoderm. In stage 5a, the small space that appears within the inner cell mass, or in some instances seemingly between the mass and the trophoblast, represents the beginning of the amniotic cavity. Either the amnion itself or the amniotic cavity may be noticeable first.

In one instance (No. 8225), a single layer of flattened cells is found attached to the trophoblast although the amniotic cavity is scarcely present. In another case (No. 8155), by contrast, a prominent amniotic cavity (formed by the curved epiblast) is present although amniogenesis has not yet begun (fig. 5-6). In a third specimen (No. 8020), a small cavity is visible and amniogenesis is under way (fig. 5-5). The amniogenic cells, which are believed by Hertig and Rock (1945a) to be delaminating from the trophoblast dorsal to the embryonic disc, appear to be in the process of enclosing the amniotic cavity by fusing with the margin of the germ disc.

In summary, the roof and lateral walls of the amniotic cavity are, in Hertig's (1968) view, derived from the cytotrophoblast, and the cells are mesoblastic. The floor, however, is constituted by the epiblast. In general terms, "two distinct methods of amnion formation are ordinarily considered: formation by folding and formation by cavitation, the latter beingconsidered the more specialized" (Mossman, 1937).

Both methods have been invoked by Luckett (1975) who, in an interpretation quite different from that of Hertig and Rock, has proposed that a primordial amniotic cavity appears within the embryonic mass (e.g., No. 8020), followed by opening of the roof to form a temporary tropho-epiblastic cavity (e.g., No. 8155). In stage 5b, it is maintained that the definitive amniotic cavity is formed by "upfolding of the margins of the epiblast" (e.g., No. 8215).

In a recent electron-microscopic study of the rhesus monkey (Enders, Schlafke, and Hendrickx, 1986) it was concluded that the amniotic cavity appears as a result of a rearrangement of epiblastic cells (a "change in cell association occurring within epiblast") whereby they are separated into amniotic ectoderm and epiblast proper.

The chief function of the amnion is not mechanical protection but rather the enclosing of "the embryonic body in a quantity of liquid sufficient to buoy it up and so allow it to develop symmetrically and freely in all directions" (Mossman, 1937).

Embryonic disc. The term "germ disc" was formerly employed for "the epithelial plate that is derived directly and exclusively from the blastomeric formative cells" (Heuser and Streeter, 1941). The plate may more conveniently be referred to as the epiblast. When the underlying primary endoderm is also included, the term "embryonic disc" (formerly "embryonic shield") is used.

The embryonic disc (figs. 5-5 and 5-6) is bilaminar, composed of the epiblast and the primary endoderm. It is concavoconvex from dorsal to ventral. The epiblast consists of variably sized, polyhedral cells which either show no precise pattern of arrangement (No. 8020) or are in the form of a pseudostratified columnar epithelium (No. 8155). One or more mitotic figures may be encountered. The primary endoderm consists of a cap of small, darkly staining, vesiculated cells without any specific arrangement. Mitotic figures are not noticeable.


Stage 6

Approximately 0.2 mm in size
Approximately 13 postovulatory days
Characteristic features: chorionic villi and secondary umbilical vesicle in 6a; primitive streak in 6b

The appearance of recognizable chorionic villi is used as the criterion for stage 6. The villi begin to branch almost immediately. The space bounded by the internal surface of the chorion begins to expand greatly toward the end of stage 5 and during stage 6 (fig. 6-1). The secondary umbilical vesicle develops. Axial features are not evident, or at least have not been described, in all embryos of stage 6. Moreover, in some instances, their presence is in dispute. It is possible, however, if the fixation and plane of section were always suitable, the series complete and free from distortion, and an adequate search made, that axial features would be found. Thus, in the well-known Peters specimen, which is frequently considered not yet to show axial features, the possible presence of an allantois was raised originally (Peters, 1899); apparently Grosser believed for a time that a primitive streak was present.

Hence it is convenient to distinguish (a) those embryos in which little or no axial differentiation has occurred or been noted, from (b) those embryos in which axial features, particularly a primitive streak, have definitely been recorded (fig. 6-10). This distinction corresponds more or less to Mazanec's (1959) groups V and VI, respectively. In the latter, according to Mazanec, the chorionic villi have already begun to branch.

Size and Age

The maximum diameter of the chorion varies from 1 to 4.5 mm, that of its enclosed cavity from 0.6 to 4.5 mm. The embryonic disc varies from 0.15 to 0.5 mm in maximum diameter; the age is believed to be about 13 days. It is of interest to note that those specimens (referred to here as 6a) in which definite axial features have not been described are characterized by a slightly smaller chorion (1-3 mm, as compared with 2-4.5 mm in the remainder, 6b), contained cavity (0.6-2.2 mm, as compared with 1.3-4.5 mm), and embryonic disc (0.15-0.22 mm, as compared with 0.15-0.5 mm).

Histological Features

A chart indicating the derivation of various tissues and structures is presented as figure 6-2.

Decidua. The decidua (fig. 6-8) varies considerably in thickness at the implantation site but is generally between 3 and 12 mm. A decidual reaction is present, a variable amount of edema and leucocytic infiltration is found, and actively secreting glands and prominent blood vessels are noticeable. Compact, spongy, and basal strata are distinguishable from superficial to deep (Krafka, 1941). At about this time the well-known subdivision of the decidua into three topographical components may be employed: (1) the decidua basalis is that situated at the deepest (embryonic) pole of the conceptus, (2) the decidua capsularis is reflected over the rest of the chorionic sac, and (3) the decidua parietalis lines the uterine cavity except at the site of implantation Chorion. The increasing structural complexity of the trophoblast from superficial to lateral and basal aspects has been attributed by Ramsey (1938) to the more advantageous nutritive conditions prevailing at the latter sites. The coalescence of the lacunae forms the intervillous space, which, placed as an offshoot on the uterine circulation, may be regarded as "a variety of arteriovenous aneurysm" (Hertig, 1968). The syncytiotrophoblast is more active enzymatically than the cytotrophoblast, and is believed, among other functions, to be responsible for the secretion of chorionic gonadotropin.

The cytotrophoblastic clumps and mesoblastic crests of stage 5 have now progressed to form processes that are commonly known as chorionic villi (fig. 6-3). Severa1 authors have pointed out that "the initial villi do not arise as free and separate outgrowths from the chorionic plate" into the lacunar spaces (Boyd and Hamilton, 1970). Trophoblastic trabeculae, which initially are syncytial in character, come to possess a central process of cytotrophoblast and are generally termed primary villi. None of the trabeculae, however, possesses a free distal end, because these "primary villous stems" do not arise as individual and separate sprouts from the chorion (Hamilton and Boyd, 1960) but rather by invagination of syncytiotrophoblast (fig. 6-3).

The mesoblastic crests form the cores of the villi, and the cytotrophoblastic clumps form caps from which cytotrophoblastic columns proceed externally. These columns, indications of which have been seen also at stage 5c, make contact with the stroma and form a border zone (penetration zone, Greenhill, 1927), the fetal-maternal junction, characterized by pleomorphic fetal cells (Krafka, 1941) and necrotic maternal cells. The columns also make contact with each other peripherally to form the cytotrophoblastic shell (fig. 6-3).

The cytotrophoblastic shell (of Siegenbeek van Heukelom) is the specialized, peripheral part of the cellular trophoblast in contact with maternal tissue. As the shell forms, syncytiotrophoblast is left both internally, where it lines the intervillous space, and externally, where it forms masses that blend with the decidua. The development and arrangement of the shell have been discussed by Boyd and Hamilton (1970). Defective development of the trophoblast, especially of the shell, results later in villous deficiencies (Grosser, 1926).

Moreover, it is important to appreciate that, even in instances where the embryonic disc either fails to form or remains rudimentary, a large chorion with luxuriant villi may still be found. Because the number of mitotic figures in the cytotrophoblastic covering of a villus is approximately equal to that in the mesoblastic core, Krafka (1941) suggested "that the mesoblast, once established, proliferates at the same rate as the Langhans layer, and hence furnishes its own growth zone." The layer featured by Theodor Langhans (in 1870 and subsequently) is the villous cytotrophoblast, "which constitutes the cellular (as opposed to syncytial) investment of the villi" (Boyd and Hamilton, 1970).

Park (1957) found a low incidence of sex chromatin in the trophoblast and chorionic mesoblast of an embryo (No. 7801) of stage 6, and in the chorionic mesoblast and umbilical vesicle of a second specimen (No. 7762).

Extra-embryonic mesoblast. The chorionic mesoblast is well formed and extends into the villi even in stage 6a, as seen clearly in the Linzenmeier and Peters specimens. From his acquaintance with Hill's observations of primates and from his own studies of the Fetzer embryo, Florian (1933) believed in "the existence of an area in the most caudal part of the embryonic disc where the primary mesoderm is fused with the ectoderm." The zone in question may be the site "where primary mesoderm originates (at least in part). This area is situated close behind the primordium of the cloaca1 membrane" (ibid.), and involves the disc epiblast and the adjoining amniotic ectoderm. This theory of the origin of the primary mesoblast from a localized proliferating area has already been discussed under stage 5a.

At stage 6a, angiogenesis (see Hertig, 1935) is occurring in the chorionic mesoblast (Nos. 6900 and 6734), blood vessels are found in the villi (No. 6900) and blood islands are seen on the umbilical vesicle (No. 6734). By the end of stage 6b, vascularization of the chorion is almost constant and blood vessels are generally present in the villi (ibid.). Thus, the blood vascular system first arises in extra-embryonic areas. From his studies of chorionic angiogenesis, Hertig (1968) became convinced of "the independent in situ origin of angioblasts and mesoblasts from trophoblast."

Amnion. The amnion is well formed. At its upturned margins, the epiblast of the embryonic disc changes abruptly to the squamous cells of the amnion (fig. 6-9). Although the amnion appears largely as a single (ectodermal) layer, an external coat of mesoblast can also be detected and, in some places, it "runs along as a layer of mesothelium" (Heuser, Rock, and Hertig, 1945). The amniotic cavity may be either smaller or larger than the umbilical vesicle at stage 6 (fig. 6-4).

In the Fetzer embryo, among other specimens, Florian (1930a) was able to confirm von Möllendorff's finding in Op of enlargement of the amniotic cavity by epithelial degeneration. An active extension of the amniotic cavity toward the tissue of the connecting stalk took place in a dorsal and caudal direction from the embryonic disc.

The vault of the amniotic cavity may give rise to a diverticulum known as the amniotic duct. The appearances vary from a localized amniotic thickening (in No. 7801) to a pointed process directed toward the trophoblast (in No. 7634, Krafka, 1941, fig. 2, and in No. 7762, Wilson, 1945, fig. 8). The amniotic duct is generally regarded as an inconstant and transitory developmental variation.

Embryonic disc. Terms such as "embryonic disc," "embryonic shield," Keimscheibe, Embryonalschild, or "blastodisc" are used in measuring embryos from the rostral amniotic reflexion to, frequently, the caudal end of the primitive streak (Odgers, 1941). Although certain authors (e.g., Grosser, 1931a) therefore do not include the cloacal membrane, others (e.g., Florian and Hill, 1935) do include it. When the connecting stalk and allantoic diverticulum are also included, terms such as "embryonic rudiment," Embryonalanlage, Keimanlage, or Keimling are employed. Many writers do not make clear their points of reference, nor do they always specify whether a measurement has been taken in a straight line (caliper length) or along the curvature of the disc (contour length). In the present work, the "embryonic disc" will be taken to include, where possible, the cloacal membrane, and, except where otherwise specified, the length will generally refer to that measured in a straight line.

As seen from the dorsal aspect, the embryonic disc generally appears elongated, and the long axis of the disc usually (in ten out of twelve specimens of stage 6b) coincides with that of the primitive streak rather than lying at a right angle to it. Although the dorsal surface of the disc may present some localized convexities and concavities, it is, as a whole, fairly flat. No. 7801, a particularly excellent specimen, is slightly convex (fig. 6-9) but the marked curvatures illustrated in some embryos (such as T.F.) may be assumed to be artifactual.

A detailed study of the "cytodesmata" in early human embryos was undertaken by Studnicka (1929). In two specimens (Bi I and T.F.) of stage 6, cytodesmata were described between the germ layers ("interdermal cytodesmata") and between the mesothelial and the adjacent germ layer in both the amnion and the wall of the umbilical vesicle (Studnicka and Florian, 1928).

With the appearance of the primitive streak during stage 6, a process is begun whereby certain cells of the epiblast enter the streak, and the remaining cells on the dorsal aspect of the embryo will become the embryonic ectoderm. The epiblast is continuous with the amniotic ectoderm at the margin of the embryonic disc. The basement membrane (membrana prima of Hensen) of the epiblast is clearly visible in a number of embryos of stage 6, such as Harvard No. 55 (where it has been demonstrated histochemically by Hertig et al., 1958) Peters, E.B., Bi I, etc. The embryonic mesoblast will be discussed under the primitive streak.

The endoderm shows a marked concentration of glycogen, and some of its cells may be primordial germ cells (Hertig et al., 1958). Rostral to the primitive streak, the endoderm consists of large vesicular cells (Heuser, Rock, and Hertig, 1945). It should be kept in mind that, in the chick embryo, it has been shown that the rostral part of the primitive streak (including the node) always contains a large population of presumptive endoblastic cells.

Primitive streak. The primitive streak is a proliferation of cells lying in the median plane in the caudal region of the embryonic disc (figs. 6-5 and 6-10). The streak on its first appearance and in narrow usage, is "the thick caudal end of the germ disk' (Heuser and Streeter, 1941). In a much broader and more functional sense, however, "the essential features of the primitive streak are the pluripotential nature of the cells that compose it and the continued segregation of more specialized cells which migrate or delaminate from the less specialized remainder" (ibid.).

The primitive streak enables cells from the outer layer of the embryo to pass inside and become mesodermand endoderm (Bellairs, 1986). The process by which cells leave the epiblast, become part of the primitive streak, and then migrate away from the streak is termed ingression. It seems to depend on loss of the basal lamina beneath the streak.

On the basis of radio-autographic studies of grafts in chick embryos, the primitive streak is believed to be not a blastema but rather an entrance in which occur movement of epiblast toward the streak, invagination at the streak, and subsequent migration to both homolateral and heterolateral mesoderm. At primitive streak stages of the chick embryo, zones have been established for future ectoderm, mesoderm, endoderm, and notochord. It is likely that the mammalian pattern is basically similar. In the rabbit, the formation of the embryonic disc and primitive streak "is primarily achieved by migrations of cells that are being rapidly proliferated over the entire surface of the embryonic area. Cell death occurs but is an insignificant factor in this phase of embryonic growth" (Daniel and Olson, 1966). In the human, the primitive streak appears first during stage 6. The possibility of a streak in some embryos (e.g., Peters) here classified as 6a has been raised. Conversely a streak in at least one 6b embryo (No. 8819) has been denied (Krafka, 1941).

Brewer's (1938) criteria for the presence of a primitive streak are (1) active proliferation of the cells (shown by a large number of cells in mitosis), (2) the loss of the basement membrane separating the epiblast and endoderm, (3) migration of the epiblastic cells, and (4) intermingling of the cells of the epiblast and endoderm of the disc.

The shape of the early streak is not entirely clear. Brewer (1938) described it as a crescentic formation at the caudal margin of the disc (similar to that seen in the pig), but, as already mentioned, the presence of a streak at all in his specimen (No. 8819) was denied by Krafka (1941).

In other young 6b embryos (Op, Fetzer, and Wo) the streak possesses the form of a node, and indeed initially "seems to correspond in its position with Hensen's knot" (Florian, 1930a). Thus, in the Fetzer specimen (Fetzer and Florian, 1930) the streak appeared almost circular in dorsal view, was situated not far from the middle of the embryonic disc, and did not reach the cloacal membrane. The question naturally arises whether the primordium is not in fact the primitive node rather than the primitive streak in these early specimens. This idea is supported by Dr. J. Jirásek (personal communication, 1970), who believes that the fusion of the epiblast with the endoderm found in this region indicates that the node rather than the streak is involved. According to this interpretation, the primitive node appears during stage 6 and the streak would not appear until the following stage. In the chick embryo, although the streak is said to appear before the node, there is reason to believe that the rostral end of the very young primitive streak already contains the material of the future primitive node.

When the primitive streak attains a rostrocaudal measurement of 0.1 mm or more, as in Beneke, Am. 10, Bi I, and T.F., its elongation fully justifies the name "streak." By the end of stage 6, both a node and a streak (separated by a "neck') have been recorded in one (somewhat abnormal) embryo (HEB-18, Mazanec, 1960).

In the first specimens of stage 6b (Liverpool I and II, No. 7801, No. 8819, No. 7762, Op, Fetzer), the length of the primitive streak is less than one-quarter that of the embryonic disc. In Wo and Beneke it is less than one-third, and in Am. 10, Bi I, and T.F. it is less than one-half. Finally, in the transitional and somewhat abnormal HEB-18 specimen, the streak attains one-half the length of the embryonic disc.

The primitive groove appears probably during stage 6b. At any rate its presence has been claimed in some specimens (Liverpool II, Op, and T.F.) of that stage (fig. 6-10). Although it may be possible, at least in some instances, to ascertain the rostrocaudal axis of the embryo at stages 5c and 6a, unequivocal manifestation awaits the initial appearance of the primitive streak during stage 6b (O'Rahilly, 1970). With the establishment of bilateral symmetry, the embryonic disc, in addition to its dorsal and ventral surfaces, now has rostral and caudal ends and right and left sides. The median plane may be defined,0 and the terms "medial" and "lateral" are applicable. Moreover, it is proper to speak of coronal (or frontal), sagittal, and transverse planes. The last-named, in anticipation of the erect posture, may be termed horizontal. Certain other terms, however, such as "anterior," "posterior," "superior," and "inferior," should be avoided at this period because of their special meaning in adult human anatomy.

Embryonic mesoblast. At first, the embryonic mesoblast is scarcely recognizable as such (Nos. 8819 and 7762) or is quite scanty in amount (No. 7801). Although the main bulk of the embryonic mesoblast is believed to come by way of the primitive streak, other sources are not excluded. Lateral to the streak, for example, it is possible that some epiblastic cells bypass the streak and migrate locally into the mesoblast (Heuser, Rock, and Hertig, 1945). In addition, the possibility of contributions from the gut endoderm has been raised in the case of the macaque (Heuser and Streeter, 1941).

Finally, the degree of incorporation of some of the primary mesoblast into definitive body mesoderm is unknown. Conversely, in the Beneke specimen, Florian (1933) "could trace the secondary mesoderm behind the caudal end of the primitive streak around the cloacal membrane into the connecting stalk." (The very closely arranged cells of the secondary mesoblast were distinguishable from the looser cells of the primary tissue.) Hence, the convenient distinction between primary and secondary mesoblast should not be interpreted too rigidly. A need exists for further detailed studies of the distribution and spread of mesoblast during these early stages.

In the chick embryo, at the stage of the definitive streak, it has been established that each topographical kind of mesoblast has a definite place along the rostrocaudal axis of the primitive streak, precise enough to be demonstrated experimentally.

Prechordal plate. The earliest human embryo in which a definite prechordal plate has been recorded seems to be Beneke at stage 6b. In that specimen, "in front of, and below the cranial extremity of the primitive streak... the endoderm is distinctly thickened and proliferative" in "a horseshoe shaped area"; that area "must be regarded as a mesoderm producing zone" (Hill and Florian, 1963). Study of later specimens, such as Manchester 1285 and Dobbin, led Hill and Florian to "regard the thickened area in question as prechorda1 plate." Dr. W. P. Luckett has called the writer's attention to several stage 5c specimens (Nos. 7950,8558, and 8330) in which the endoderm appears to be thickened at one end of the embryo, as shown in the photomicrographs published by Hertig, Rock, and Adams (1956, figs. 35, 37, and 38).

It is of interest to note that, in Tarsius, the prechordal plate, which later adopts the form of an annular zone, is at first represented by a continuous sheet of thickened endoderm underlying most of the embryonic epiblast (Hill and Florian, 1963).

Umbilical vesicle (fig. 6-9). The umbilical vesicle functions "as a specialized nutritional membrane" (Streeter, 1937) and, in addition, serves as the site of origin of primordial germ cells as well as an important temporary locus of hematopoietic activity (Hoyes, 1969).

From stage 5c to stage 6, the primary is transformed into the secondary umbilical vesicle (fig. 6-4), although the mode of the transformation has long been disputed. (See Stieve, 1931; Heuser and Streeter, 1941; Strauss, 1945; Hamilton and Boyd, 1950; Starck, 1956.) According to Hertig (1968), the primary vesicle "blows up" or "pops," and the torn edges that remain attached to the endoderm coalesce to form the secondary vesicle. The secondary sac "soon takes on a second or inner layer of epithelial nature" which is, in Hertig's view, derived from the wall of the sac itself. In other words, "endodermal" cells differentiate in situ from the wall of the umbilical vesicle (Streeter, 1937).

According to a number of other authors (such as Stieve, 1931), however, "it seems likely that endoderm from the edge of the embryonic disc proliferates and migrates round the interior" of the wall of the sac (Hamilton and Boyd, 1950), using that wall "as a guiding surface" (Mazanec, 1953). A further possibility is the dehiscence of cells from the disc endoderm so that a new cavity is formed between the two endodermal layers or possibly between the disc endoderm and the dorsal part of the wall of the sac.

In any case, it appears likely that, at least in some embryos, the distal part of the primary umbilical vesicle becomes detached from the proximal part, thereby forming one or more isolated vesicles or cysts (see below). As a result of these processes, the secondary vesicle is at first smaller than the primary sac (fig. 6- 4). Thus the secondary vesicle may develop "as a result of the collapse of the abembryonic and lateral walls" of the primary sac, "with subsequent pinching-off and vesiculation of the distal collapsed portion" (Luckett, 1978). The part of the primary vesicle immediately under cover of the embryonic disc persists as the secondary umbilical vesicle.

It is maintained that the secondary vesicle is unilaminar in stage 6 and that adherent epithelial remnants of the primary sac merely give an impression that the secondary vesicle has already acquired an external mesodermal layer (Luckett, 1978). In a number of embryos of stage 6 (such as No. 7634) and some subsequent stages, a diverticulum of the umbilical vesicle has been recorded. These outgrowths arise generally at the abembryonic pole of the main sac. They vary from slight evaginations to long processes (0.45 mm in Liverpool II, for example), and are frequently associated with cysts. It has been suggested that the diverticula and cysts are remnants of the primary umbilical vesicle (Heuser, Rock, and Hertig, 1945).

Cloacal membrane. The cloacal membrane appears during stage 6b. Although at least its site may be detected in the first specimens of that stage (such as No. 7801), the membrane is probably present in all specimens that possess a primitive streak 0.05 mm or more in length. The membrane is at first a cell cord that connects the epiblast with the endoderm (Florian, 1933) and is ofvariable length (about 0.015-0.025 mm). Later (stage 7) it increases in size and begins to assume the form of an actual membrane.

Allantoic diverticulum. The existence of an allantois in the human remained controversial until the end of the nineteenth century (Meyer, 1953). In early embryos, the recognition of an allantoic primordium presents considerable difficulty. A recess of the umbilical vesicle, such as appears in embryo Op, should not be assumed to be necessarily the allantois. According to Florian (1930a), "the yolk-sac penetrates into the connecting stalk in the form of a narrow diverticulum which enlarges and eventually opens out again into the cavity of the yolk sac. This process may probably be repeated several times."

Hence some reservations must be made concerning the "allanto-enteric diverticulum" of Liverpool I. In No. 7801, all that is found is merely "a tiny recess in the wall of the yolk sac at the spot where the allantoic duct presumably originates" (Heuser, Rock, and Hertig, 1945). In Wo, a solid Allantoisanlage has been claimed. In Beneke, the diverticulum of the umbilical vesicle is not the allantois (Florian and Beneke, 1931). In Bi I, the appearance of the diverticulum has been attributed to "a ventral downbulging of the underlying wall of the yolk-sac" produced by the end node of the primitive streak (Florian, 1930a). The condition in T.F. may well be similar.

In conclusion, it is difficult to find a convincing example of an allanto-enteric diverticulum at stage 6.

Connecting stalk. Florian (1930a) pointed out that the connecting stalk (as exemplified in embryo Op) comprises two portions (fig. 6-5): (1) the amnio-embryonic stalk, an attachment of the entire amnio-embryonic vesicle to the chorionic mesoderm, and (2) the umbilical stalk, by which the caudal end of the embryo is anchored to the chorion. The umbilical stalk, which is in all stages covered on its cranial surface by amniotic ectoderm, later becomes transformed into the umbilical cord.

The Peters embryo is situated in a thickening of the chorionic mesoderm but the connecting stalk is "not yet present" (Florian, 1930a). Although "a body stalk proper has not yet fully formed" in No. 7634 (Krafka, 1941), its primordium is present and comprises the amnio-embryonic and umbilical stalks of Florian. The condition is similar to that in Op, in which Florian has pointed out that the axis of the connecting stalk has already begun to form an acute angle (open caudally) with the embryonic plate. By the end of stage 6b, blood vessel primordia are present in the developing stalk and indicate the future umbilical vessels (Hertig, 1935).

In the opinion of Hill and Florian (1963), the vessels of the connecting stalk in Tarsius "can arise directly as invaginations of the mesothelium" covering the stalk.

FAIR USE NOTICE: This may contain copyrighted (© ) material the use of which has not always been specifically authorized by the copyright owner. Such material is made available for educational purposes, to advance understanding of human rights, democracy, scientific, moral, ethical, and social justice issues, etc. It is believed that this constitutes a 'fair use' of any such copyrighted material as provided for in Title 17 U.S.C. section 107 of the US Copyright Law. This material is distributed without profit.

1, 2, 3,