Previous   Next


Early Embryology

George D. Snell and Leroy C. Stevens

This description of the early embryology of the mouse is a revision of Snell's ( 1941) description in the first edition of the Biology of the Laboratory Mouse. Most of the new information was obtained from a search of the literature published since then, and some of this was based on investigations on the embryology of the rat. In general there is much similarity in the development of Muridae, but we will make it clear when we discuss information obtained from the rat. Since 1941 there have been many investigations devoted to certain aspects of mouse embryology, but only a few have attempted a comprehensive study. For other descriptions the reader is referred to Boyd and Hamilton ( 1952), Bonnevie ( 1950), and Rugh ( 1964). Graves ( 1945) described early development of the golden hamster. Grüneberg ( 1943) has provided a very useful description of mouse development from 9 1/3 to 18 1/3 days, and Otis and Brent ( 1954) have compared human and mouse developmental stages. Grüneberg ( 1963) surveyed the field of developmental pathology of the skeleton, including many inherited malformations of the mouse.

The material used in this study is the same as that used in 1941 and consists of sections of embryos spaced at 6-hour intervals from 4 to 9 days. In some cases 10 or more embryos of a single stage have been sectioned. The sections were prepared by Olive Bartholomew, Elizabeth Fekete, and G.D. Snell, using a technique described by Fekete et al. ( 1940). In most cases the females used as mothers were hybrids between two strains and fathers were from a third strain, thus giving both embryos and mothers a maximum of hybrid vigor. Because of this the stages described here are usually earlier, often by as much as a day, than comparable stages described by other authors. Although this procedure gave embryos which developed rapidly and were usually normal, it did not eliminate variability. We have not attempted to describe the variations noted in the rate of development of embryos or in the rate of development or form of separate parts. It should be emphasized, however, that the range of variation in these respects is considerable.



In several mammalian species, including the mouse, mitotic activity of oogonia ceases and the prophase of meiosis is known to be initiated about 5 days before birth ( Mandl, 1963), and primary oocytes may progress to the dictyate condition by birth ( Mintz, 1960). It is these cells which transform the definitive germ cells, and there is no somatic contribution. In adult mice the interval between the dictyate (resting) stage and metaphase of the second meiotic division which occurs just before ovulation is about 12 hours ( Edwards and Gates, 1959).

According to Edwards and Gates ( 1959), oocytes apparently remain in the dictyate stage until stimulated by luteinizing hormone to complete their maturation. They found that 2 hours after injection of luteinizing hormone oocytes were in late prophase, and from 5 to 8 hours all were in metaphase of the first meiotic division. At 14 hours all eggs were ovulated and were in the metaphase of the second maturation division.

The ovarian oocyte of the rabbit, mouse, and mole has been shown to be bilaterally symmetrical, and this symmetry is traceable throughout early development to the primitive streak stage ( Jones-Seaton, 1950a). Mouse and rat eggs show animal and vegetal poles which are indicated by the eccentric germinal vesicle or by the polar bodies ( Jones-Seaton, 1950b; Dalcq and Van Egmond, 1953; Dalcq, 1956).

The mouse ovary is almost completely enveloped by a bursa, and it is very unlikely that there can be any internal escape of the eggs into the abdominal cavity. No case of naturally occurring extrauterine pregnancy has been reported in the mouse.

The freshly ovulated mouse egg is enveloped by the vitelline membrane, a product of the egg; the zona pellucida, a product of the surrounding follicular cells; and by the cumulus oophorus composed of follicular cells and an intercellular matrix consisting of acidic mucopolysaccharide, hyaluronic acid, and protein ( Braden, 1962).

The zona pellucida is a mucoprotein. It serves to maintain the normal cleavage pattern and to prevent fusion of eggs ( Mintz, 1962). Eggs will adhere to one another when the zona is artificially removed with enzymes. This has served as the basis of methods for synthesizing genetic mosaics (Mintz, 1964, 1965). The zona is not necessary for development in mice ( Mintz, 1962). Nicholas and Hall ( 1942) separated the first two blastomeres of the rat egg, transplanted them to the uterus, and obtained implantation and egg cylinders of early embryos. Gwatkin ( 1964) removed the zona enzymatically and observed fusions of eggs. At the four-cell stage blastomeres separated easily and sometimes gave rise to miniature blastocysts. He concluded that the importance of the zona was in preventing loss of blastomeres and fusion of eggs. He also found that the zona in the mouse was not a complete barrier to virus infection of the ovum.

The innermost cells of the cumulus oophorus are tightly packed and oriented radially to form the corona radiata. Chang et al. ( 1962) have shown that the corona radiata plays a major role in the fertilization of rabbit ova or in the maintenance of their fertilizability. Sperm penetration may be dependent upon the corona. More probably the fertilizability of the ova is protected by the corona and would disappear shortly after its removal. The outside diameter of the zona pellucida of mature ovarian mouse eggs averages about 9 μ. Following fertilization the zona pellucida expands until its outer diameter becomes about 113 μ. This is just within the limits of visibility for the unaided eye ( Lewis and Wright, 1935). Within the vitellus is the egg nucleus, not clearly visible in living eggs such as the one shown in Figure 12-1, but easily seen in fixed and stained material.


The sperm head contains one complete set of chromosomes, and the middle piece contributes mitochondria. There is some evidence that Golgi material is also carried by the sperm ( Gresson, 1940; Huber, 1915; Lams and Doorme, 1907).

The acrosome is a caplike structure on the anterior part of the sperm. Its function is thought to be to carry hyaluronidase, which probably enables the sperm head to pass through the cumulus oophorus ( Austin and Bishop, 1958). The perforatorium lies under the acrosome and is thought to be a modification of the sperm nucleus. Austin and Bishop concluded that the acrosome becomes modified in sperm passing through the female reproductive tract and is detached before the sperm penetrates the zona pellucida. The perforatorium may carry a lysin capable of altering the zona in such a way as to permit penetration into the perivitelline space.

Dziuk and Runner ( 1960) found that sperm motility apparently influences fertilization. The percentage of blastocysts increases as the motility score of the sperm suspension increases. Cooled mouse sperm retains fertilizing capacity for at least 24 hours.

Numerous morphological and behavioral differences have been found in the sperm from eight inbred strains of mice ( Beatty and Sharma, 1960). Sharma ( 1960) investigated the phenotype of mouse spermatozoa in four inbred strains and their F1 crosses and found heterosis in F1 crosses.


The mean number of sperm ejaculated into rats is 58,060,000 ( Blandau and Money, 1944; Blandau and Odor, 1949). Within a few minutes after ejaculation the sperm of the rat are disseminated throughout the uterine corona. This rapid dispersion is brought about by the presence of large quantities of uterine fluid and marked muscular activity of the cornua. In some rats, within 15 minutes after the sperm have entered the cornua, some may be observed in the ampullae of the oviducts. In all animals examined 1 hour after ejaculation, sperm had reached the ampullae. These observations are similar to those of Lewis and Wright ( 1935) for the mouse.


Fertilization in the mouse occurs in the ampullae at the upper end of each oviduct where the eggs are found, usually in clumps, about 2 hours after ovulation. Sperm are usually present in the ampullae at the time of ovulation, but there is a delay of several hours before the sperm penetrates the egg (Braden, 1954, 1960; Edwards, 1955). Apparently the egg membranes, particularly the cumulus oophorus, need some form of maturation before sperm penetration can occur (Braden, 1960, 1962). Braden ( 1958) suggested that cells of the cumulus and corona radiata are less dense and loosen earlier in some strains of mice than in others. The delay between ovulation and penetration of eggs by sperm is considerably shorter (1 to 3 hours) in artificially ovulated mice than in mice ovulating naturally (3 to 5 hours).

Sperm penetration of the mouse egg normally takes place before dissolution of the cumulus ( Lewis and Wright, 1935). The number of sperm found at the site of fertilization is small, and it is generally considered that the hyaluronidase carried by the individual sperm enables it to make its own way through the cumulus to the zona pellucida ( Leonard et al., 1947; Braden, 1962). In the mouse the first sperm to arrive at the egg usually penetrates and fertilizes it.

Bishop and Tyler ( 1956) have demonstrated a sperm agglutinin of mammalian eggs, including eggs of mice, acting like fertilizins of invertebrate eggs. They tentatively indicated the zona pellucida as the source of mammalian "fertilizin."

According to Edwards and Gates ( 1959) mouse sperm penetrates the cumulus and zona in approximately 1 hour and remains in the perivitelline space for 50 minutes. Pronuclei are formed 4 ¼ hours later, and the first cleavage division occurs at about 25 hours after penetration of sperm. Usually only one sperm enters each egg. Almost immediately after entry, which may occur through any part of the egg's surface, the vitellus shrinks slightly in size and the zona pellucida expands, so that a perivitelline space forms between them ( Lewis and Wright, 1935; Pincus, 1936).

The zona pellucida increases in thickness and changes after the entry of the first sperm so that the chance of subsequent penetration is reduced ( Braden, 1958). This "zona reaction" is apparently stimulated in the vitelline cortex by the sperm head. Upon stimulation, the vitelline cortex may release a substance which traverses the perivitelline space and causes an alteration in the penetrability of the zona pellucida. But Austin and Braden ( 1956) suggested that the "zona reaction" is initiated in the zona itself at the point of entry of the first sperm and then is propagated through the zona.

In the rat and mouse the sperm passes rapidly through the thick zona and the head apparently adheres almost immediately to the surface of the vitellus ( Austin and Braden, 1956). Shortly after this the vitellus loses its adhesiveness and supplementary sperm remain free in the perivitelline space. Thus the attachment of the sperm head to the vitellus probably elicits both the "zona reaction" and the block to polyspermy at the surface of the vitellus. The head lies flat on the vitelline surface for about ½ hour. It then becomes engulfed by the vitellus and becomes the male pronucleus which moves toward the center of the egg. The female pronucleus is large than the male and is close to the second polar body. After fertilization, the pronuclei come together and fuse in approximately the center of the zygotes to form the zygote nucleus. At the first cleavage division the nuclear walls break down, the chromosomes split longitudinally, and one-half of each split chromosome is carried to each daughter cell. Hence at this division, as at all future somatic divisions, each cell receives a full complement of chromosomes from each parent.

Within 2 to 3 hours after fertilization the second polar body is formed and is everted into the perivitelline space. The first polar body has already been formed in the ovarian egg. Fertilization is discussed in Chapter 11.


Preimplantation development in many mammals may vary in some details, but in general a similar pattern exists throughout cleavage and the formation of the morula. Edwards ( 1957) has published a table of time intervals between cleavages up to the sixth division for the mouse.

Jones-Seaton ( 1950a) has shown that the dorsal regions of the rat, rabbit, mouse, and mole ova are marked by a cortical zone or ribonucleic acid and the ventral by a vacuolate cortex. The plane of the first cleavage in the rat has no set relation to the dorsoventral axis. When the early cleavages are unequal, the larger blastomeres divide more rapidly.

The first two cleavages in the mouse occur while the eggs are still in the oviduct. The first cleavage occurs about 24 hours after copulation and results in two cells almost equal in size ( Figure 12-2A). Following divisions occur somewhat more rapidly, giving rise to four-cell, eight-cell stages, etc., and are usually nearly synchronous. Occasionally, however, eggs are found with some divisions completed and others still incomplete, hence showing an odd number of cells ( Figure 12-2B). The actual act of division requires only 5 or 10 minutes; the interval between divisions lasts 12 hours. Eggs of 16 cells or more, in which no cavity has appeared, are called morulae. Eggs usually reach this stage about 60 hours after fertilization and pass from the oviduct, through which they have been gradually moving, into the uterus, some 6 to 12 hours later ( Lewis and Wright, 1935). This is subject to considerable variation, however, and in one study passage into the uterus at 4 days was found to be the rule ( Burckhard, 1901).

Whitten and Dagg ( 1961) demonstrated that some factor(s) contributed by the spermatozoa can influence the intermitotic times during early cleavages of the eggs of BALB/c mice. The nature of this influence is unknown.

Smith ( 1956) showed that the mutant gene t12 acts prior to blastocoel formation. Homozygous animals cease development just prior to blastocyst formation. This is the earliest action of any gene or chromosomal deficiency ( Geyer-Duszynska, 1964) known in the mouse. Mintz ( 1964) showed that blastocyst formation does not depend upon an early morphogenetic prepattern set before or during cleavage. Any normal cell, rather than certain ones, may be involved at first in regions that will become inner cell mass or trophoblast by other, epigenetic rather than prepatterned mechanisms. At the eight-cell stage and even beyond, the eggs is still entirely labile, and the decision for further regional differentiation is made much closer to the blastocyst stage. In spite of the regulative capacity of the egg, all of its cells continue to undergo progressive change in increased adhesiveness between cells and increased cell motility during cleavage. Pronounced modifications in ribonucleic acid and protein synthesis are also occurring ( Mintz, 1964).

Mintz ( 1965) has obtained genetic mosaic mice by aggregating, during cleavage stages, the blastomeres of two embryos of different genotype into a single cluster, and by transferring the developing aggregates to the uterus of a surrogate mother. These remarkable animals include hermaphrodites and mosaics of alleles at the H-2 and a loci.


Shortly after the egg has entered the uterus and usually sometime after it has reached the 32-cell stage, an eccentrically located fluid-filled cavity appears among the cells of the morula. This enlarges rapidly to produce the blastocoel ( Figure 12-3). The cavity is bounded by a single layer of cells except on one side where most of the cells are grouped to form a structure called the inner cell mass. Eggs in this stage are known as blastocysts. Blastocysts settle into the uterine crypts between 4 1/3 and 4 ½ days ( Eaton and Green, 1963).

It is not quite clear what part the uterine environment and what part the blastocyst itself plays in shedding the zona pellucida. In normal development these two factors probably contribute jointly ( Tarkowski, 1962). It is known that the uterine environment is more acid than the fallopian tubes and that lowering the pH of the medium dissolves the zona. However, the blastocysts liberate themselves from their zonae when transferred to extrauterine sites.

Oviducal segmented eggs of the mouse, transplanted beneath the kidney capsule, gave rise to trophoblast and extraembryonic membranes, but not to embryonic derivatives ( Kirby, 1962a). Uterine blastocysts (but not blastocysts locked in the fallopian tubes) similarly transplanted gave rise to embryonic structure. Kirby concluded that possibly the denial of a "uterine factor" renders mouse eggs incapable of developing structures other than extraembryonic ones. One of us(L.C.S.) transplanted segmenting ova of strain 129 mice to the adult testis and obtained disorganized teratoid growths composed of a wide variety of mature tissues indicating that a "uterine factor" is not necessary for development. Kirby ( 1962b) found a higher proportion of successful transplants of uterine blastocysts into the testis than for any other extrauterine organ tried.

In a study on the fine structure of the 5-day rat blastocyst, Schlafke and Enders ( 1963) found that the cells of the trophoblast form a single continuous layer of low cuboidal cells. Microvillae are irregularly distributed on the apical surfaces projecting toward the zona. At 7 days the microvillae are less numerous and less regular than previously. There is no evidence of differentiation within the inner cell mass.


The uterus in the mouse is duplex, consisting of two horns uniting just anterior to their junction with the vagina, and each attached to the dorsal body wall by a mesentery, the mesometrium ( Figure 12-4). There are two layers of muscle in each horn, an outer longitudinal layer and an inner circular layer. The uterine lumen is lined with epithelium. Between the epithelium and the muscle layers is the mucosa, a tissue forming the bulk of the uterine wall. The epithelium is indented by numerous small crypts.

Very shortly after entering the uterus the eggs become spaced more or less evenly throughout its length, and each egg finds its way into a uterine crypt on the ventral or antimesometrial side of the lumen, thereby coming into close contact with the uterine epithelium ( Figure 12-3). The presence of the blastocysts quickly sets up changes at the implantation site. Within a few hours the epithelium begins to loosen, and its nuclei show degenerative changes ( Figure 12-5). Within 15 hours it is sloughed entirely ( Figure 12-6). At the same time active growth commences in the mucosa, so that by 1 days after implantation (5 days after mating) there is an appreciable swelling in the uterus at the implantation site. The swollen mucosa at the implantation site is known as decidua.

Hollander and Strong ( 1950) observed that in the mouse the external appearance of pregnant uteri and the frequency of placental fusion suggest that blastocysts implant at random after being scattered throughout each uterine horn by its churning. Krehbiel ( 1962) studied the distribution of ova in the rat nucleus and reached a similar conclusion.

Dickerman and Noyes ( 1960) showed that on the fifth day of pregnancy in rats both the blastocyst and the endometrium have reached a specific stage of development favorable for implantation. The ovum must have reached the stage of development of the 5-day blastocyst, and the endometrium must be "ripe" but not overripe for implantation to begin. The uterine environment is unsuitable for implantation on both the fourth and sixth days of pseudopregnancy.

up to the time of implantation there has been no growth in size of the egg. Cleavage has resulted in a division of the egg, originally one large cell, into numerous smaller cells, but little if any new protoplasm has been formed in the process. Beginning with implantation, however, rapid growth commences. At first the blastocoel enlarges while the inner cell mass assumes a flattened cup shape with the concave face towards the cavity ( Figure 12-5). In the living condition the blastocoel is probably distended with fluid, and its walls are tightly pressed against the uterine epithelium, but in fixed material at this stage there is always some collapse. This initial expansion of the blastocoel requires only a few hours and is quickly followed by a growth of the inner cell mass down into the enlarged cavity ( Figure 12-6). The blastocoel and inner cell mass are called thereafter the yolk cavity and the egg cylinder. A comparison of Figures 12-7, 12-8, 12-10, and 12-12 will show the rapid growth of the eggs cylinder occurring during the next 2 ½ or 3 days.

Formation of endoderm

As the blastocoel begins to enlarge, the inner cell mass is composed to two types of cells ( Figure 12-5). Adjacent to the blastocoel is a single layer of darkly staining cells, the endoderm, one of the three primary germ layers. The rest of the blastocyst is composed of ectoderm, divided into the ectoderm of the inner cell mass, and the trophectoderm, a single-celled layer bounding the blastocoel ventrally and laterally. The trophectoderm (troph from the Greek word for nourishment) derives its name from its probable role in the nourishment of the young embryo. The mesoderm has not yet appeared.

Shortly after the first appearance of the endoderm, single cells or strands of cells grow out from its margin along the inner surface of the trophectoderm. At first these cells are few and widely separated (Figures 12-7, 12-8), but by 6 ½ days they lie evenly spaced and quite close together over the trophectoderm's entire inner surface ( Figure 12-10). The layer of cells thus formed is known as the distal (or parietal) endoderm. Meanwhile the inner cell mass has grown down into the yolk cavity to form the egg cylinder. This is composed of an inner mass of ectoderm cells, which represents considerably more than the ectoderm of the future embryo, and an outer layer of endoderm cells ( Figure 12-8). This layer of endoderm cells bounding the egg cylinder is known as the proximal (or visceral) endoderm. The endoderm is thus divided into two distinct parts, distal and proximal, lining the distal and proximal walls of the yolk cavity.

Embryonic and extraembryonic ectoderm

At about 4 ½ days, the egg cylinder ectoderm, first beginning to form, is divided into two parts: a dorsal, more darkly staining (when counterstained with congo red) region with elongated nuclei, and a ventral, more lightly staining portion with round nuclei ( Figure 12-6). The dorsal part gives rise to various extraembryonic structures and is called the extraembryonic ectoderm; the ventral part gives rise to the ectoderm of the embryo proper and is called the embryonic ectoderm. Although the difference in staining reaction and in the shape of the nuclei has disappeared by 5 ½ days, the division between the two regions is still quite distinct ( Figure 12-8). Strictly speaking the trophectoderm is also extraembryonic ectoderm, but as a matter of convenience the term will be used only for the extraembryonic ectoderm of the egg cylinder.

At about 5 days a cleft or cavity, the proamniotic cavity, appears in the embryonic ectoderm ( Figure 12-7). At this stage, according to Bonnevie ( 1950), there is evidence of bilateral symmetry. The proamniotic cavity is covered on what may be the anterior side of the embryo by a single layer of cells, whereas on the posterior side the cell layer is thicker. Bonnevie suggested that this thickening may represent a localized growth center within the posterior-median line of the inner cell mass. The formation of the proamniotic cavity in the embryonic ectoderm is followed very shortly by the appearance of a similar cleft in the extraembryonic ectoderm and by the fusion of these two, so that by 5 ½ days the egg cylinder contains a very narrow lumen ( Figure 12-8).

Ectoplacental cone

Beginning after 5 or 5 ½ days of active growth, the dorsal end of the extraembryonic ectoderm gives rise to a new structure, the ectoplacental cone, which joins the egg cylinder ventrally and extends dorsally towards the lumen of the uterus ( Figure 12-8). This develops rapidly, its cells showing numerous mitoses, and by 6 ½ days it composes almost one-half the total length of the embryo. Its structure, particularly at the upper extremity, is porous, and the interstices between the strands of cells that compose it soon become infiltrated with maternal blood ( Figure 12-10). In the rat the outermost cells of the cone become true giant cells on the seventh day. This suggests that the cone cells in the 6-day embryo represent undifferentiated trophoblast ( Bridgman, 1948). In later stages it becomes part of the placenta.

Inversion of germ layers

At 5 ½ days ( Figure 12-8) the egg cylinder consists of a double wall enclosing a narrow lumen, the inner layer of which is composed of ectoderm, the outer of endoderm. This relation of ectoderm and endoderm, found in the mouse, rat, rabbit, guinea pig, and their close relatives, proved very puzzling to early embryologists, being the reverse of that found in all other chordates. It has been called the inversion of the germ layers. Although at first sight it seems to indicate a drastic alteration in early development, actually there is no very fundamental change in the relations of the important structures. Those changes that are involved are easily understood from a comparison of the early development of the mouse with that of a primitive rodent, the thirteen-striped ground squirrel. Three comparable stages for each species are shown diagrammatically in Figure 12-9. Beginning students of embryology will want to refer again to this figure after completing the study of later stages in the mouse.

In primitive rodents, as represented by the thirteen-striped ground squirrel, the embryonic area (embryonic ectoderm and underlying endoderm) forms a disc that overlies an almost spherical yolk cavity. In the mouse, the embryonic area forms a deep cup pushed far down into the yolk cavity, which thereby is greatly reduced in size. During some period in the evolution of the mouse the embryonic area invaginated into the yolk cavity, the curvature of the embryonic area thereby being reversed and the relation of ectoderm and endoderm inverted. The change is comparable to that produced when a rubber ball has one side pushed in, being altered thereby from a sphere to a cup.

In Figure 12-9D the lumen of the egg cylinder is shown extending through the ectoplacental cone to the outside. This condition may be the exception rather than the rule, but it has been described by Sobotta ( 1911) and Melissinos ( 1907), and we have found it in a few cases in our material. It is significant evidence for the theory that the inversion of the germ layers is due to an invagination of the embryonic area. Bonnevie ( 1950) described a sloughing of cells from the lining of the proamniotic cavity and suggested that they may be expelled into the ectoplacenta through this lumen.

Further evidence is provided by the later development of the thirteen-striped ground squirrel ( Figure 12-9C). In this species the whole embryo sinks down into the yolk cavity, carrying the splanchnopleure with it. The splanchnopleure is thereby inverted, but no inversion of embryonic ectoderm and endoderm occurs because of the advanced development of the embryo at the time. However, if the sinking or invagination of the embryonic area were pushed back to an earlier period of development, the condition found in the mouse would result.

One interesting consequence of the inversion of the germ layers is a very compact form of early development. Much seemingly waste space in the yolk cavity is eliminated. Note that the drawings of the mouse embryos in Figure 12-9 are at a higher scale of magnification than those of the ground squirrel embryos. Actually, at comparable stages of early development, the total volume of a mouse embryo is perhaps one-fiftieth that of the total volume of an embryo of the ground squirrel. This reduction in total volume involves little if any reduction in the volume of the embryonic area proper.

Primitive streak and mesoderm formation

At 6 ½ days the middle germ layer or mesoderm makes its appearance ( Figure 12-10). The first mesoderm cells are formed at the posterior end of the embryo by delamination from a narrow strip of embryonic ectoderm extending from the line of junction of the embryonic and extraembryonic ectoderm anteriorly to about halfway to the tip of the egg cylinder. This strip of ectoderm is known as the primitive streak, and since it lies at the posterior end of the embryo an anteroposterior axis is at once established with its appearance.

In our material we have noted that from 5 to 5 ½ days the egg cylinder and more particularly the proamniotic cavity, instead of being round in cross-section, are slightly flattened along an axis perpendicular to the mesometrium. This is the same as the future anteroposterior axis. However, it cannot be determined until the appearance of the mesoderm which end of the axis is anterior and which end posterior. With the appearance of the mesoderm the flattening of the egg cylinder, if any, is along the opposite axis.

Mesoderm cells form a loose tissue of very characteristic appearance. They multiply rapidly, wedging their way laterally between ectoderm and endoderm toward the anterior margin of the egg cylinder ( Figure 12-14A). The forward growth is particularly rapid along the line marking the junction between embryonic and extraembryonic ectoderm, and in this line mesoderm may be found at the anterior margin of the egg cylinder about 12 hours after the first mesoderm cells appeared ( Figure 12-12). Elsewhere the two lateral wings of mesoderm do not penetrate to the mid-sagittal region until much later. Some mesoderm cells also push between the extraembryonic ectoderm and the adjacent endoderm, thus leaving the region of the embryo proper. These mesoderm cells, for the most part, are destined to take part in the formation of the yolk sac, an extraembryonic membrane which later envelops the embryo and which is discarded at birth.

Bennett and Dunn ( 1960) described the effects of the homozygous tw18/ tw18 genotype on mouse development. During the seventh day the main abnormality was a thickening of the wall of the egg cylinder in the region of the primitive streak, just at the time when the primitive streak begins to form. On the eighth day the overgrowth of the primitive streak begins to form. On the eighth day the overgrowth of the primitive streak produced a median bulge projecting dorsally into the proamniotic cavity. Since the swollen area was still covered with the neural material normally overlying the primitive streak, it produced, in effect, two neural grooves separated by the median mass of overgrown primitive-streak material.

Orientation of embryo in uterus

Since the primitive streak is at the posterior margin of the egg cylinder, its formation, heralded by the appearance of the mesoderm, establishes an anteroposterior axis in the embryo. It is appropriate at this point to consider how this axis and the other axes of the embryo are oriented in relation to the uterus.

At the time of implantation the embryo settles to the ventral or antimesometrial side of the uterus. When it first implants, the inner cell mass is up or towards the mesometrium, the blastocoel is down or away from the mesometrium ( Figure 12-4A). In terms of an older embryo, the ectoplacental cone is up and the embryonic portion of the egg cylinder is down ( Figure 12-4B). The dorsoventral axis of the embryo is thus parallel to the mesometrium and perpendicular to the long axis of the uterus ( Figure 12-11). The anteroposterior axis of the embryo likewise has a definite orientation with respect to the uterus, being as a rule perpendicular to the mesometrium. Departures from this orientation by as much as 45° may, however, occur. This orientation persists until about 8 or 8 ½ days when the embryo begins to shift its position in the uterus.

Amnion, chorion, and exocoelom

When mesoderm cells first appear between the ectoderm and endoderm at the posterior margin of the egg cylinder, they cause the ectoderm at the line of junction between its embryonic and extraembryonic portions to bulge into the proamniotic cavity. This bulge is the beginning of the posterior amniotic fold ( Figure 12-10). In like manner the lateral wings of mesoderm, progressing around the egg cylinder towards its anterior margin, give rise to folds along the sides of the cylinder. These are the lateral amniotic folds. Finally, when the mesoderm reaches the anterior margin of the egg cylinder, a small anterior fold is produced ( Figure 12-12). The posterior, lateral and anterior folds should be thought of as not separate structures, but as a continuous constriction about the middle of the egg cylinder drawn tighter and tighter as the folds develop. Because of the very precocious development of the posterior as compared with the anterior amniotic fold, the constriction is eccentric, the point of final closure being far towards the anterior margin of the egg cylinder. In the rat, the anterior amniotic fold is much better developed than in the mouse, and the constriction, therefore, less eccentric ( Jolly and Férester-Tadié, 1936).

In Figure 12-10 a second fold may be seen pushing into the proamniotic cavity just dorsal to the posterior amniotic fold. Sobotta ( 1911) showed this in his Figure 5 but interpreted it as an artifact. Our material indicates that it is regularly though briefly present. Its significance is unknown, but it is perhaps indicative of the very rapid growth that occurs in the whole posterior wall of the egg cylinder at the time of mesoderm formation.

Before the anterior fold forms, small cavities begin to appear in the mesoderm of the posterior and lateral folds (None of the embryos in our collection show this early stage in the formation of the exocoelom. This description is based on the observation of Jolly and Férester-Tadié ( 1936). Bonnevie ( 1950) also described the formation of the exocoelom in the mouse). These soon coalesce to form a single large cavity, the extraembryonic coelom, or, more concisely, the exocoelom ( Figure 12-12). The exocoelom at this stage and at all future stages is lined by mesoderm. The extraembryonic mesoderm lying between the posterior amniotic fold and the columnar endoderm appears infiltrated with fluid, and small lumina are formed. A large number of small lumina gradually unite into larger ones and the proamnion wall is reduced in thickness. For a short time a second cavity is present in the posterior amniotic fold between the mesoderm and ectoderm ( Figure 12-12), but this is a transitory structure of no particular significance.

In less than a day after the first appearance of the amniotic folds the girdle which they form has closed. The resulting condition is shown in the sagittal reproduced in Figure 12-13. As the amniotic folds coalesce, they become thin and delicate as a result of continued delamination of mesodermal cells from them ( Bonnevie, 1950).

Three cavities are now present in the egg cylinder in place of the single proamniotic cavity it formerly contained. The most ventral of these is the amniotic cavity, lined with embryonic ectoderm. At this stage it is cupshaped, as can be seen from Figures 12-13 and 12-14A, which show it in sagittal and cross section respectively. In the middle is the exocoelom, lined with mesoderm. At the dorsal extremity of the egg cylinder is the ectoplacental cavity lined with extraembryonic ectoderm. The ectoplacental cavity, the smallest of the three, gradually becomes narrower and finally disappears.

The membrane separating the amniotic cavity from the exocoelom is called the amnion. It is composed of two thin cellular layers, one of ectoderm, the other of mesoderm. Separating the exocoelom from the ectoplacental cavity is another membrane, the chorion, likewise composed of ectoderm and mesoderm.

Head process

Jolly and Férester-Tadié ( 1936) first described the head process for the mouse and rat; our observations are in accord with theirs. Bonnevie ( 1950) also described the formation of the head process in detail. In cross-sections the head process is found exactly opposite the primitive streak. It will be remembered that mesoderm is first proliferated by the primitive streak in embryos about 6 ½ days old. The growth is entirely from the lateral and caudal margins of the primitive streak; no mesoderm is proliferated from its cephalic extremity. Beginning at about 7 days, however, growth does occur in this region, but the structure formed shows greater affinity to the endoderm than to the mesoderm. It is known as the head process. The head process takes its origin from Hensen's node, a globular mass of mesoderm cells appearing at the cranial end of the primitive streak. In sagittal sections the head process first appears as a wedge-shaped group of cells between the ectoderm and endoderm at the ventral extremity of the egg cylinder ( Figure 12-12). The base of the wedge is attached to the cranial end of the primitive streak from which it takes its origin; the tip of the wedge points forward towards the anterior margin of the egg cylinder. Cells grow our rapidly from the margins of the wedge, forming a thin spreading sheet between ectoderm and endoderm.

The endoderm and the margins of the head process are so thin and close together at this stage that favorable conditions are necessary to distinguish them. In the section shown in Figure 12-12 there are several cells at the anterior limit of the head process that cannot be distinguished as either head process or endoderm. The division in the drawing in this region is somewhat arbitrary. When head process and mesoderm come into contact there is also possibility for confusion. However, in well-fixed preparations cut at a favorable angle, the division can almost always be precisely determined.

As shown in Figure 12-10 the endoderm over the ventral extremity of the egg cylinder is stretched and the cells flattened, but near the upper margin of the embryonic portion of the cylinder there is a sudden change to a higher type of cell. The transition is particularly abrupt at the anterior margin of the cylinder. The thin or flattened endoderm we shall refer to as squamous endoderm, the thick endoderm as columnar endoderm, the line of junction between the two as the transition line. According to Bonnevie ( 1950) the nutritive function of the columnar endoderm is evident from the small or large fluid drops within the cells. Chiquoine ( 1958) found that the apical cytoplasm of the embryonic endoderm of 5- to 8-day rat embryos contains striking amounts of periodic acid-Schiff material. The amount of this material varies with the height of the cells. The tall columnar cells contain the largest amounts. Note the precise location of the transition line in Figures 12-10 and 12-12. The flat squamous endodermal cells are characteristic of the embryonic part of the egg cylinder during further development. Bonnevie ( 1950) pointed out that the varying height of the columnar endodermal cells and their fluid contents are useful for indicating the seat of the most intensive nutritive activity.

The limits of the head process are as follows. Caudad, it begins at the anterior extremity of the primitive streak, that is, just a little above and caudad to the ventral tip of the egg cylinder. Cephalad, it extends to the transition line. Laterad, at its broadest point it may extend almost around the anterior half of the circumference of the egg cylinder ( Figure 12-14A), but mostly it is narrower than this, filling perhaps the anterior fifth of the egg cylinder's circumference.

When its forward growth brings it to the transition line, the head process fuses with the columnar endoderm with which it has thus newly come in contact ( Figure 12-13). The fusion is so complete that in later stages the line of junction is completely lost. Laterally, its outer margins fuse with the squamous endoderm. Meantime the squamous endoderm underlying the head process, already very thin, becomes increasingly attenuated, its nuclei become widely separated and very flat, and the cytoplasm almost disappears (Figures 12-14A, B). At 7 ½ days no trace of it remains.

In the course of the upward and laterad growth of the head process and the forward growth of the mesoderm the two cell layers come in contact and overlap ( Figure 12-14A). In the regions of overlapping, the head process stays adjacent to and advances over the surface of the endoderm, whereas the mesoderm remains next to the ectoderm. At 7 ½ days the development of the mesoderm has brought it between ectoderm and head process everywhere except for a strip along the midsagittal plane of the embryo. As we shall see later, the head process of this midsagittal strip gives rise to the notochord, whereas the remainder of the head process contributes to the lining of the gut.

Neural groove

It can be seen from Figure 12-14B, a cross-section of the anterior part of an egg cylinder of a 7 ¼-day embryo, that the ectoderm in the midsagittal plane forms a definite trough or truncated V. This trough extends forward in the midsagittal plane from the cephalic end of the primitive streak well towards the cephalic limit of the embryonic ectoderm. Developed between the 7 and 7 ¼-day stages ( Figures 12-14A, B), it is the beginning of the neural groove which later gives rise to the central nervous system. The further development of the neural groove will be discussed later; look ahead to Figures 12-20 and 12-21 showing the way in which it deepens and narrows and finally closes at the top to form the neural tube. The point to be emphasized is that the appearance of the neural groove establishes a clear anteroposterior axis throughout the length of the embryo. The neural groove anteriorly and the primitive streak posteriorly lie in the midplane.


At the same time that the neural groove is differentiating in the midsagittal region of the head process which immediately underlies it ( Figure 12-15). In this region the head process thickens and the oval nuclei become oriented in general perpendicular to the ectoderm. Elsewhere it forms a thin membrane with the nuclei oriented parallel to the plane of the membrane. The structure thus differentiated ventral to and in contact with the ectoderm of the neural groove is the notochord. It is the axis about which the vertebral column is later laid down. The remainder of the head process, together with a part of the endoderm to which it is fused, becomes the lining of the gut. This part of the head process will hereafter be referred to as gut endoderm. It seems likely that most or all of the midgut is lined by head process. Whether or not any of it enters into the formation of the fore- and hindguts is not clear. For a considerable period notochord and gut endoderm remain joined. Eventually, however, the two halves of the gut endoderm grow across the ventral surface of the notochord and unite in the midventral line, leaving the notochord as an axial, rodlike structure between ectoderm and endoderm.

Huber ( 1918) describes the head process in the guinea pig as giving rise to notochord only. Our material, however, confirms completely the contention of Jolly and Férester-Tadié ( 1936) that in the mouse at least some gut endoderm is also derived from the head process. As shown in Figure 12-14B, the head process extends laterally considerably beyond the limits of the differentiating notochord.

It is questionable whether the notochord should be classed as ectoderm, endoderm, or mesoderm ( Kingsbury, 1920). Since it is formed from the head process and since the very complete fusion of the margins of the head process with the endoderm indicates a close affinity between the two tissues, classification as endoderm would seem logical. If, however, head process is classed as endoderm, it must be remembered that its origin in time is quite different from that of all the other endoderm, and two separate stages of endodermal proliferation must be recognized. As to the place of origin, there is a certain similarity between the two tissues, one forming at the ventral margin of the inner cell mass, the other near the ventral tip of the egg cylinder, which is, so to speak, simply the inner cell mass grown up. Cell lineage studies might reveal a closer similarity in origin than is superficially apparent.


At 7 ½ days there is a broad depression in the rather thick base of the notochord adjacent to its junction with the primitive streak (Figures 12-15D, 12-16). The depression is a conspicuous landmark at this stage, but it is a transitory structure, the first signs of it appearing at 7 ¼ days and disappearance being complete about 12 hours later. It probably is homologous to the archenteron of the lower chordates.


Soon after the exocoelomic cavity becomes well established, the allantois begins to grow into it from the mesoderm at the caudal end of the primitive streak. The allantois ( Figure 12-13) is an extraembryonic mesodermal structure whose function is to convey blood vessels from the embryo to the placenta where they establish contact with the maternal circulation. In many vertebrates the allantois contains a cavity lined with endoderm and connected with the gut. There is no endoderm-lined cavity in mice, but there are numerous small cavities in the mesoderm giving the organ a porous structure.

After its first appearance at 7 ¼ days the allantois grows rapidly across the exocoelom in the direction of the ectoplacental cone (Figures 12-16, 12-17. Meanwhile the chorion becomes flattened against the base of the cone, constricting the ectoplacental cavity and finally eliminating it altogether. When the allantois makes contact with the chorion at about 8 days, a continuous structure is established connecting the posterior end of the primitive streak with the ectoplacental cone. In due course embryonic blood vessels will find their way along this pathway to make contact with the maternal blood supply. In the rat the allantoic vessels reach the chorion and the chorioallantoic placenta is established by the latter half of the ninth day ( Bridgman, 1948).

Foregut and hindgut

In the early stages of its formation the digestive tract consists of three parts; foregut, hindgut, and midgut. These appear in the order named. The foregut can be traced back to the 7-day stage when it consists merely of a deep notch in the thick or columnar endoderm at the anterior margin of the egg cylinder ( Figure 12-12). Six hours later there is little change ( Figure 12-13), but by 7 ½ days ( Figure 12-16) the notch has been replaced by a definite pocket in the endoderm, and the endoderm surrounding the pocket together with the overlying ectoderm form a bulge which projects into the amniotic cavity. From this stage on, growth of the foregut is exceedingly rapid, the pocket changing in a few hours into a deep pouch ( Figure 12-17). The process is due to a progressive drawing together in the midventral line of the folds of endoderm that bound the anterior intestinal portal (Figures 12-26C, 12-28) the portal thus being shifted toward the posterior end of the embryo. The process has been aptly described as a "zipper action."

The foregut is lined by endoderm which is surrounded by mesoderm and ectoderm. Thus in this region the process of invagination has reversed the inversion of the germ layers found in the early egg cylinder. This is the first of the steps by which the germ layers in mice are brought into the relation characteristic of the adult, i.e., endoderm on the inside, ectoderm on the outside, mesoderm in between.

The hindgut, less precocious than the foregut, appears at about 7 ¾ days as an invagination in the endoderm and overlying layers at the posterior end of the primitive streak ( Figure 12-17).

The open ends of the fore- and hindguts are eventually joined by the midgut, whose formation will be described in a later section. The blind ends of the two guts break through to the outside and give rise to mouth and anus. An early stage in the development of the mouth may be seen in 8-day embryos ( Figure 12-23). In the ectoderm of the head there is an invagination directed toward the anterior end of the foregut. This is the stomodeum. The wall between the stomodeum and the foregut is the oral plate. In the course of time this ruptures and the mouth opening is formed. The anus develops in a similar manner at a somewhat later stage.

Head Fold

The invagination of the foregut involves a pushing or folding of the adjacent tissues into the amniotic cavity. The structure thus produced is the head fold ( Figure 12-16). First appearing at about 7 ½ days, it becomes a large an conspicuous structure within less than 24 hours (Figures 12-23, 12-26A). The growth of the neural folds in this region is more rapid than elsewhere, presaging the formation of the brain. The heart, just ventral to the head fold, is also conspicuous by its rapid growth. In 8 or 8 ½ - day embryos the difference in size between the head region and the middle of the trunk is striking. The head fold is a region of particularly rapid growth ( Hyman, 1927).


Since the somites are mesodermal structures, it will be useful before discussing their development to review the distribution of the mesoderm at the 7 ½ - day stage just as somite formation begins. In the extraembryonic region, the entire exocoelom is lined with mesoderm. Also, the exocoelom contains the allantois, a wholly mesodermal structure. In the embryo proper there is little mesoderm in the midsagittal region. One small mass which will later contribute to the formation of the heart occurs anterior to the foregut ( Figure 12-16). The primitive streak in the midsagittal plane consists of a tissue which joins and, in structure, is intermediate between ectoderm and mesoderm. Whether or not this should be called mesoderm is a matter of definition. At the caudal end of the embryo its structure is essentially that of true mesoderm and there is mesoderm in the midsagittal plane in this region.

Even though the notochord blocks the entrance of mesoderm into most of the midsagittal area, there are well-developed sheets of mesoderm on each side. These lie between ectoderm and endoderm and are continuous laterally with the extraembryonic mesoderm ( Figure 12-18). Anterior to the primitive streak, there are two distinct areas in these mesodermal sheets: an area of paraxial mesoderm adjacent to the notochord and an area of lateral mesoderm adjacent to the extraembryonic coelom. The former gives rise to the somites, the latter to the mesoderm of the embryonic coelom. At 7 ½ days there is no visible division between the two areas ( Figure 12-20B), but beginning at about 7 ¾ days, coincident with the development of the somites, they are separated by a longitudinal cleft that becomes increasingly pronounced as the differentiation of the somites progresses ( Figure 12-18).

The somites are paired, segmental structure arising in the paraxial mesoderm (Figures 12-19, 12-26D). They are the first indication of metamerism in the developing embryo. The first pair forms a little anterior to the anterior end of the primitive streak. Each member of the pair appears as a localized denser area which grades off anteriorly into a loose mesoderm, and which posteriorly is separated by a cleft from the undifferentiated mesoderm. The second pair forms posterior to the first and is likewise separated by a cleft from the undifferentiated caudal mesoderm. Additional pairs of somites form at more or less regular intervals, each new pair differentiating just posterior to the last pair formed until in the rat a total of 65 pairs has appeared ( Butcher, 1929). The number of somites may be slightly different for the mouse, and there is some individual variation. Butcher states in his excellent paper on the somites of the rat that the first pair dedifferentiates and disappears at about the 7-somite stage. We found no evidence of such a dedifferentiation in our material. As may be seen from Figures 12-19A, B, and C, the first somite can be traced clearly to the 11-somite stage. Continued and rapid proliferation of the mesoderm on each side of the primitive streak maintains a supply of undifferentiated cells. These push forward to about the level of the base of the notochord where the new somites are segmented off in regular succession. The anterior somites are older and, at any given stage, more highly differentiated than the posterior somites.

Figure 12-19 shows how the "zipper action" by which the foregut is formed moves the location of the anterior intestinal portal steadily caudad. At the 2-somite stage the opening of the shallow foregut lies anterior to the first somite. At the 7-somite stage the anterior portal has moved posteriorly until it is about at the level of the first somite. At the 11-somite stage it has reached approximately the level of the sixth somite. (The portion of the gut shown in Figure 12-19C is midgut just caudal to the anterior intestinal portal.)

The number of somites in an embryo is a convenient means of stating its stage of development.

Primitive streak as a growth center

The primitive streak is remarkable as being a region in which the three germ layers meet ( Figure 12-16). It is continuous dorsally with the ectoderm, laterally and posteriorly with the mesoderm, and anteriorly with the head process (which is endodermal in nature and indistinguishably fused with the original proliferation of endoderm). Of these three, it gave rise to two: the mesoderm and the head process. It may be added that the proliferation of mesoderm cells that produced the allantois occurred at its posterior end. Its own cells are undifferentiated and cannot be classified as either ectoderm, endoderm, or mesoderm. The only structure in primitive chordates possessing these characteristics is the dorsal lip of the blastopore, and the primitive streak and the dorsal lip of the blastopore may be homologous. Besides being a point of origin for new tissues, it is the center of a region of rapid growth. The adjacent mesoderm is full of dividing cells which are continually pushed forward to give rise to somites anterior to the primitive streak. Much of the increase in length of the embryo is due to growth in this region. There are thus two regions of particularly active growth in the developing embryo: the primitive streak and the head fold ( Hyman, 1927). There is one point of contrast between these two: The tissues in the head region are well advanced in differentiation, whereas the tissues of the primitive streak region remain relatively undifferentiated.


Coincident with the formation of the somites in the paraxial mesoderm, the coelom or body cavity develops in the lateral mesoderm. It is formed by a division of this mesoderm into two layers, a dorsal or somatic layer adjacent to the ectoderm, and a ventral or splanchnic layer adjacent to the endoderm. The coelom is the space between the two ( Figure 12-18). Because the somatic mesoderm and the ectoderm are closely associated and undergo many foldings in common, it is convenient to designate the two layers together by the term somatopleure. For the same reasons splanchnic mesoderm and endoderm together are designated as splanchnopleure. The mesoderm and ectoderm of the somatopleure dorsal to the coelom are continuous with the similar layers in the amnion. The amnion, therefore, may also be classed as somatopleure. Similarly the mesoderm and endoderm of the splanchnopleure ventral to the coelom are continuous with the similar layers in a tissue which bounds the extraembryonic coelom laterally. This tissue, therefore, may also be classed as splanchnopleure.

It has been previously stated that there is a mass of mesoderm in the midsagittal plane anterior to the foregut. This extends to right and left across the front of the foregut and is continuous laterally with the lateral sheets of mesoderm. It thus forms the base of a U of which the lateral mesoderm forms the sides. By about the 4-somite stage or slightly later the coelom extends not only throughout the lateral mesoderm but also as a passage through this anterior mesoderm (Figures 12-17, 12-30). The coelom also is thus U-shaped. The whole posterior portion of the coelom opens laterally into the extraembryonic coelom (exocoel) (Figures 12-18, 12-29). The anterior part of the coelom on the other hand, forming the base of the U and extending as far posteriorly as the second somite, is separated from the extraembryonic coelom by a partition of mesoderm. Much of this anterior portion of the coelom becomes the pericardial coelom enclosing the heart. The connection between the anterior and the lateral parts of the coelom is called the pericardial-peritoneal canal ( Figure 12-29).

The relations of coelom and extraembryonic coelom can be studied from the series of sections of a seven-somite embryo shown in Figures 12-23A to G.

Reichert's membrane

The mouse embryo is protected during its development by three extraembryonic membranes: Reichert's membrane, the amnion, and the yolk sac. There is no essential difference between the amnion of rodents and of other mammals. The yolk sac in Rodentia has come to have rather unusual relations to other structures. The chorion, an important fetal membrane in most mammals, is present in the mouse but remains small and unimportant as a protective structure.

To follow the development of Reichert's membrane we must go back to the 5 ½ - day stage ( Figure 12-8). Except in the region of the ectoplacental cone, the embryo is bounded by the trophectoderm. This is continuous with the margins of the cone and is separated from the egg cylinder by the yolk cavity. Laterally its cells are in close contact with the maternal decidua, a contact so intimate in fact that in some cases it is impossible to tell whether a given cell is of embryonic or maternal origin. Ventrally it stretches across the remains of the uterine lumen, now filled with a degenerating mass of uterine epithelium. On the inner surface of the trophectoderm are a few widely separated endoderm cells.

A day later ( Figure 12-10) these endoderm cells have increased in number and form a uniform though not quite continuous layer over the inner surface of the trophectoderm. Between the two cell layers there soon begins to appear a thin, noncellular, pink-staining membrane called Reichert's membrane. The first signs of it are often visible at the ventral extremity of the egg where there is apt to be a cluster of endoderm cells. It soon becomes continuous over the entire inner surface of the trophectoderm. The fully developed membrane is of uniform thickness and, as can be demonstrated by dissection, surprisingly tough for so delicate a structure. Pierce et al. ( 1962) and Midgley and Pierce ( 1963) have shown that parietal (distal) endoderm cells of the mouse embryo secrete Reichert's membrane. This membrane has the ultrastructural and immunochemical attributes of a basement membrane. Late on the 16th day in the rat, Reichert's membrane and the decidua capsularis rupture ( Bridgman, 1948). This permits the placenta to shorten in diameter and become thicker.


The early stages of the development of the amnion have been described. Owing to the inversion of the germ layers, the amniotic folds have only a short distance to grow, and amnion formation is consequently precocious in the mouse as compared with most other mammals ( Figure 12-9B, E). For the same reason the area of the amnion at first is small. It expands rapidly, however, to accommodate the growing embryo and by 8 days it forms a large sac over the embryo's entire dorsal surface (Figures 12-23, 12-26C). In the later stages of development the embryo floats free in the amniotic cavity attached only by the umbilical cord.

Yolk sac

The mammalian ovum contains virtually no yolk. The mammals are, however, descended from reptilian ancestors in whose eggs yolk was abundant, and this long period in their evolutionary history has left an indelible imprint on mammalian development. Most striking, perhaps, is the development of a yolk sac so similar in many details to the reptilian yolk sac as to be unmistakably homologous. In the small rodents the yolk sac constitutes an important accessory nutritive organ during the whole period of gestation ( Böe, 1951). Yolk sac cells probably exercise both absorptive and secretory activities simultaneously ( Wislocki et al., 1946).

The yolk cavity of the mouse may be defined as the cavity derived from the original segmentation cavity or blastocoel and lying between the egg cylinder and the distal endoderm with its secretion, Reichert's membrane ( Figure 12-20A). The yolk sac is only a part of the boundary of this cavity; namely, that middle portion of the egg cylinder wall composed of mesoderm and endoderm or, in other words, of extraembryonic splanchnopleure. In many mammals, e.g., the pig, the allantois as well as the yolk sac are derived from splanchnopleure. This is not the case in the mouse. In this species the yolk sac, as we are using the term, and the extraembryonic splanchnopleure are identical.

At 7 ¼ days the extraembryonic splanchnopleure or yolk sac is a structure of limited area forming the central or ectoderm-free portion of the egg cylinder wall ( Figure 12-20A). Although small at first, it is an area of rapid growth and by 8 or 8 ½ days forms an extensive membrane enveloping the amnion and a greatly enlarged exocoelomic cavity ( Figure 12-20B, C). The whole embryo changes its shape in the process, the egg cylinder becoming an ovoid and the ovoid a sphere. At 8 days the yolk sac is still attached to the embryo along a band that runs anterior to the opening of the foregut and posterior to the opening of the hindgut so that most of the ventral surface of the embryo is outside it ( Figure 12-23). After the midgut has formed, however, this portion of the embryo, too, is enveloped by the yolk sac ( Figure 12-21). The details of this process will be described later.

Blood islands

Associated with the yolk sac splanchnopleure are structures known as the blood islands. These appear in the mouse at 7 ½ days as thickenings in the inner or mesodermal layer of the yolk sac about which they form an irregular girdle ( Figure 12-17). As the name implies, the blood islands give rise to part of the circulatory system. The peripheral cells differentiate to form the endothelium of a system of blood vessels encircling the yolk sac while the inner cells become primitive blood corpuscles that circulate in the embryonic blood stream. Block ( 1946) performed an experimental analysis of hematopoiesis in the rat yolk sac. Böe ( 1951) described the vascularization of the rat yolk sac.

Germ cells

The germ cells of the mouse make their appearance early in embryonic life ( Chiquoine, 1954; Mintz and Russell, 1957), and mitotic descendants of the same cell lineage persist to maturity in the adult in both males and females ( Mintz, 1960). All definitive germ cells are derived from fewer than 100 cells first seen in the endodermal yolk sac epithelium localized near the base of the allantois at 8 days of gestation. At 9 to 12 days, as the paired genital ridges make their appearance, the germ cells migrate—apparently actively and probably selectively—up the dorsal gut mesentery and into these ridges. The original few cells multiply to 5000 or more when the migratory period terminates ( Mintz and Russell, 1957). At 11 days of development the distribution of the germ cells in the genital ridges is peripheral in females and central in males.

All female germ cells enter meiosis before birth ( Mintz, 1960). Leptotene chromosomes are already present in some germ cells at 13 days. Pachytene oocytes are numerous at 16 to 17 days. They may continue through diplotene and enter the dictyate stage by birth. The dictyate stage persists until just before ovulation when the chromosome organization is again visible and anaphase and telophase of meiosis I are completed.

In at least two inbred strains of mice (strain 129 and A/HeJ) the male primordial germ cells may be experimentally induced to initiate development and form testicular teratomas ( Stevens, 1964). When genital ridges from 12 ½ - day (but not older) male fetuses are grafted to the adult testis, primordial germ cells in most of them begin to develop germ layers and subsequently many kinds of differentiated, but not organized, tissues.

Changes in the uterus

Implantation in the mouse is accompanied by a rapid growth of the uterine mucosa adjacent to the implantation site to produce a definitive swelling, the decidual swelling. For a while the uterine crypt containing the embryo maintains its connection with the uterine cavity, but by about 7 ½ days the growth of the decidua has blocked this off so that the cavity containing the embryo is separated from the main lumen ( Figure 12-4B). The bridge of tissue thus formed dorsal to the ectoplacental cone will later become part of the placenta. Further growth of the decidua constricts and finally, by about 8 days, completely closes the uterine lumen dorsal to the embryo except for one or more small isolated chambers ( Figure 12-22). On each side of the decidual swelling the uterine lumen remains open, but at this period in development there is no continuous passage throughout the length of the uterus. A little later a continuous lumen is reestablished, but the new lumen is on the opposite side of the decidual swelling from the old, passing ventral instead of dorsal to the embryo. An early stage in this reestablishment of the lumen may be seen at about 8 days ( Figure 12-22). The epithelium lining the lumen on each side of the decidual swelling has grown in between the muscle layers and the decidua ventral to the embryo. The extreme limits of this growth consist of a double but unsplit layer of epithelium. In the slightly older epithelium nearer the lumen the two layers have split so that two wedge-shaped spaces extend from the lumen between decidua and muscles on each side of the decidual swelling. in course of time the wedges penetrating from the two sides meet ventral to the embryo, thus completing the formation of the new lumen.

Besides the changes in the uterine lumen there are interesting changes in the histology of the decidua. Starting as a relatively homogeneous tissue, it differentiates into several different zones, each with its characteristic structure. As many as six zones can be distinguished ( Krehbiel, 1937), but we will note no more than three ( Figure 12-22). Ventrally there is an antimesometrial zone or decidua capsularis characterized by large bi-, tri-, or tetranucleate cells. In the rat this region is characterized by binucleate cells ( Krehbiel, 1937). The individual nuclei in this zone as well as the cells are larger than elsewhere in the decidua, and this together with the grouping of the nuclei gives the zone a very characteristic appearance. It will be noted that it lies between the embryo and the new uterine lumen. With the growth of the embryo it becomes stretched until, in the later part of the gestation period, it is hardly more than a thin membrane separating embryo and lumen. Dorsally there is a mesometrial zone, or decidua basalis, whose cells at 8 days still closely resemble those of the unaltered mucosa. It later contributes tot he formation of the placenta. Between the antimesometrial and mesometrial zones is an intermediate or vascular zone characterized by the presence of numerous irregular endothelial-lined blood spaces or sinusoids. Its cells tend to be multinucleate like those of the decidua capsularis.

Nourishment of the embryo

Prior to attachment of trophoblast cells the nutrient requirements of the blastocyst are thought to be supplied by the uterine milk ( Eaton and Green, 1963). The degenerating cells of the uterine epithelium that originally lined the implantation chamber may serve as a source of food. The epithelium is sloughed off and begins to undergo degenerative changes at just about the same time that the first real increase in size of the embryo occurs. At the mesometrial pole of the embryo at 4 ½ days may be seen droplets of secretion that may contain an enzyme concerned with the digestion of the epithelial cells ( Figure 12-6). This stage in the nourishment of the embryo is brief; by 5 ½ days only a remnant of the epithelial cells remains ( Figure 12-8).

At the same time a new source of nourishment makes its appearance. At 5 ½ days the blood-filled sinusoids of the intermediate zone of the decidua begin to rupture, pouring their contents into the lumen surrounding the embryo. In a very short time the embryo is completely bathed in maternal blood. This blood is not stagnant as was once supposed but remains part of the maternal circulation. In the rat Everett ( 1935) found that there is a complete replacement every 20 minutes.

The maternal blood is separated from the embryo by Reichert's membrane, by the yolk cavity, and in later stages by the yolk sac. Reichert's membrane probably plays an entirely passive role in the transportation of nutrient substances from the maternal blood to the embryo, acting simply as a semipermeable membrane. The yolk sac, on the other hand, probably actively absorbs nutrients. This is particularly true after the blood islands girdling the yolk sac have developed into a capillary network and after the embryonic circulation is established. By the time this occurs the yolk sac has become pressed against and partly fused with Reichert's membrane, obliterating the yolk cavity ( Figure 12-21B). The embryonic yolk sac circulation is thus brought very close to the maternal circulation, and the yolk sac is established as "an organ of exchange whose importance is not secondary to that of the allantoic placenta" ( Everett, 1935). It is interesting to observe that in rodents the yolk sac has thus recovered a role as an organ of absorption it possessed in the reptiles with, however, the important difference that the material absorbed comes from the maternal blood instead of from yolk deposited within the egg.

The sinusoids in the intermediate zone of the decidua extend from the decidual cavity containing the embryo to the periphery of the decidual swelling where this borders on the uterine lumen. Beginning at about 7 ½ or 8 days there is bleeding into the uterus from these peripheral sinusoids ( Stafford, 1930; Venable, 1939). At about 10 days some of this blood finds its way into the vagina, persisting there for 3 or 4 days ( Sato, 1936). It is a convenient early sign of pregnancy.

In the later stages of development the decidua basalis, the ectoplacental cone, the chorion, and parts of the allantois fuse to give rise to a true placenta which thereupon assumes a major role in transferring nutritive material to the embryo. Bridgman ( 1948) has described the history and cytology of the developing chorioallantoic placenta of the rat. Wislocki et al. ( 1946) performed a histochemical investigation of the placentas of several rodents, including the mouse.

Trophoblastic giant cells

A conspicuous feature in sections of mouse embryos of 6 to 14 days is the presence of certain remarkably large cells lying between Reichert's membrane and the decidua. These are the trophoblastic giant cells ( Figure 12-22). Fawcett et al. ( 1947) observed these giant cells in grafts of mouse ova to the anterior chamber of the eye and concluded that they were derived from trophectoderm and not from uterine elements. Transformation of cells to trophoblast giant cells is retarded in blastocysts undergoing delay of implantation so as to be coincident to implantation ( Dickson, 1963).

It is convenient to distinguish three types of giant cells. The first large and unmistakable giant cells to appear are at the ventral extremity of the embryo (Figures 12-6, 12-8). Already quite large at 5 ½ days they have become enormous by 7 days, at which time they have penetrated for some distance into the remains of the implantation cavity ventral to the embryo. These are primary giant cells. The trophectoderm cells lateral to the egg cylinder also give rise to similar though somewhat smaller primary giant cells. Although mitotic figures are seen in the earliest stages, cell division is not characteristic of these enlarged cells ( Alden, 1948).

A second and much more numerous group of trophoblastic giant cells is derived from the ectoplacental cone. At 5 days cells my be seen growing outside the trophectoderm from the region of the future cone ( Figure 12-7). Later, when the embryo is surrounded by maternal blood, these become long strands of cells extending down, within the blood or along the inner surface of the decidua, from the cone toward the ventral extremity of the egg cylinder. At first small, these cells increase in size and at 8 days form a loose meshwork of large cells whose long protoplasmic processes extend across the blood-filled space between Reichert's membrane and the decidua ( Figure 12-22). Other similar cells may be seen adjacent to the ectoplacental cone. These are the secondary giant cells. At 8 days their continuity with the cells of the ectoplacental cone is still quite obvious. It should be pointed out that the division between primary and secondary giant cells is partly arbitrary; the trophectoderm and the ectoplacental cone are continuous structures, and cells from near the line of junction might be said to give rise to either type. One obvious function of the giant cells is to anchor Reichert's membrane to the decidua.

The third class of giant cells appear multinucleate. They first appear in the decidua adjacent to the embryo at 7 or 7 ½ days. The number of nuclei per cell appears to be extraordinary, mounting into the dozens by 8 days, but Kirby and Malhorta ( 1964) reported that all stages of mouse trophoblast showed distinct delimiting cell membranes between nuclei.

The precise roles played by the trophoblastic giant cells and the uterus during implantation are not clear. Alden ( 1948) tentatively suggested that both may play an active role. The denudation of the mucosa at the very beginning of implantation may be attributed to the action of trophoblast cells or to edematous changes in the underlying mucosa or to both ( Boyd and Hamilton, 1952). Blandau ( 1949) showed that the decidual reaction in the rat starts before any apparent damage to uterine epithelium. Finn and Hinchliffe ( 1964) suggested that epithelial degeneration is an inherent property of the uterus during the period of implantation and is not directly caused by the activities of the giant cells of the trophoblast.

The uterine epithelial cells first struck by the trophoblastic giant cells are apparently completely ingested and, it seems, under this stimulus the primary giant cells become decidedly larger ( Alden, 1948) once a breach in the epithelium has been established, the succeeding stages of cell removal seem to be as follows: The cytoplasm of the giant cell commonly exhibits an advancing border of weakly staining hyaline material often seen in other motile cells. This ectoplasmic tongue flows around a group of epithelial cells, engulfing them.

Eaton and Green ( 1963) found that the lethal action of the Ay/ Ay genotype in mice is due to the fact that the trophoblastic giant cells fail to differentiate normally. These giant cells differentiate at the equatorial plane of the blastocyst and at the ventral pole. In normal animals the giant cells form filamentous pseudopodia which penetrate between cells of the intact uterine epithelium. Completion of penetration is followed by extension of giant cell pseudopodia between the uterine epithelium and basement membrane. The sloughed uterine epithelium is then phagocytized by giant cells making attachment of the trophoblast to the endometrium possible. Penetration and attachment of giant cells begins at 4 days and 20 hours after copulation. In homozygous yellow ( Ay/ Ay) animals, the giant cells do not differentiate from the trophectoderm and attachment fails to take place. Kirby and Malhotra ( 1964) reported that the mouse probably has the most invasive trophoblast.

Wilson ( 1963) described "primary invasive" cells in the early mouse embryo. When the blastocysts reach their implantation sites, certain cells within the inner cell mass apparently migrate around the inner surface of the ventral trophectoderm. Approximately 100 hours after copulation cells of this type can be found between the trophoblast cells but protruding from its surface and penetrating the uterine epithelium. The uterine epithelium starts degenerating some time after the first of these "primary invasive" cells have penetrated it. According to Wilson ( 1963) these primary invasive cells are characteristic and have not been described previously.

Seven-somite embryo

In embryos from genetically vigorous stock, the 7-somite stage is reached at about 8 days. Thereafter the embryo begins a series of important changes, and it will be useful to review here the development attained at this point (Figures 12-23, 12-24). In sagittal section the embryo is seen to form a letter S (facing to the left in Figure 12-23) with the head region convex, the trunk region concave towards the dorsal surface. In transverse section, whereas the embryo was formerly conspicuously cupshaped with the ectoderm on the inside, it has now flattened out: in the regions of the fore- and hindgut the endoderm has become the inner layer. The neural groove, deep and well developed, is still open dorsally, though in the midtrunk region the walls are quite close together. Anteriorly, precocious growth of certain parts of the neural grove ectoderm indicates early differentiation of the brain. The hindgut is small, but there is a deep foregut, and the heart, just anterior to the foregut, is a conspicuous structure. No midgut has formed. The allantois has almost reached the chorion; in fact in some embryos at this stage the allantois has already reached and fused with the chorion. The amnion and the yolk sac plus the chorion form a double arched roof over the whole dorsal surface of the embryo. The blood islands appear as a conspicuous hummocky band around the inner surface of the yolk sac. Blood vessels have begun to form within the embryo.

Tail fold

The hindgut appears much later than the foregut, but soon overtakes it in development. In 10-somite embryos the two are of approximately equal size ( Figure 12-30). A necessary concomitant of hindgut growth is the appearance of a tail fold; the gut endoderm pushes the overlying ectoderm and mesoderm ahead of it away from the yolk sac wall. A beginning of the this process can be seen in 6-somite embryos ( Figure 12-17), and in 10-somite embryos the tail fold is well developed ( Figure 12-27A). The process is comparable to the formation of the head fold except for one interesting difference; whereas the head fold lies entirely within the amniotic cavity, the tail fold lies only partially within it. The ventral surface of the tail fold is in exocoelom. This is because in its growth away from the yolk sac it pushes the base of the allantois and the adjacent margin of the amnion ahead of it. The amnion remains attached to its caudal and lateral walls, and only its dorsal surface is within the amniotic cavity ( Figure 12-25).

At about 9 days of gestation a transitory structure, the ventral ectodermal ridge of the tail ( Grüneberg, 1963 and earlier), appears on the ventral aspect of the tail tip. The ectoderm is thickened and columnar and covers unsegmented paraxial mesoderm. Histologically the ventral ectodermal ridge resembles the apical ectoderm ridge of the limb buds, which is widely believed to be closely associated with the outgrowth of the limbs. The ventral ectodermal ridge of the tail may bear a relation to the outgrowth of the tail similar to that of the apical ectodermal ridge to the outgrowth of the limb buds.

Turning of the embryo

Almost immediately after the 7-somite stage the embryo begins a process of turning which results in a reversal of the curvature of the whole trunk region. Thus instead of being S-shaped the embryo becomes C-shaped with the ventral surface everywhere on the inside of the C. The turning begins in the head and tail folds and consists of a rotation of each along its long axis or, in other words, on axes parallel to the fore- and hindguts (Figures 12-25, 12-26, 12-27). Viewing each fold from the cephalic toward its caudal end, the direction of rotation is clockwise in each case. Of course, both folds cannot be viewed in this direction from any one point because of the curvature of the embryo. Viewed from the mesometrial pole, in sections the turning of the head fold appears to be clockwise, of the tail fold counterclockwise ( Figure 12-25).

At first the turning is confined to the head and tail folds; the midtrunk region, still firmly attached to the yolk sac, remains in its original position. At about 8 ½ days and at about the 11- or 12-somite stage, the midtrunk region suddenly turns also. Transverse sections of the trunk region at about this period show it to be either turned or not turned (Figures 12-20C, 12-21A). It is quite possible that after the growth of the head and tail folds reduces sufficiently the attachment of the trunk region to the yolk sac, this region snaps over like a spring whose tension has come to exceed the forces holding it. Some time elapses after the turning of the midtrunk before the head and tail regions complete their rotation, which eventually amounts to a full 180°. Essentially, however, by about 9 days the embryo has become concave towards the ventral surface (Figures 12-27B, C).


The turning of the midtrunk region automatically results in the formation of the midgut. Prior to turning, the two sheets of embryonic splanchnopleure in the midtrunk region extend straight out from the sides of the embryo, forming a virtually plane surface ( Figure 12-20C). There is thus no indication of a midgut. When the midtrunk region turns suddenly towards its left side, the two sheets are pulled after it, forming between them a groove which is continuous anteriorly and posteriorly with the fore- and hindguts. This groove is the midgut ( Figure 12-21A). The two sheets of splanchnopleure rapidly draw closer together ( Figure 12-21B), and at the 19-somite stage, which may be reached as early as 8 ¾ days, have fused distally to form a closed tube.


In 7 ½-day embryos a small region of mesoderm anterior to the foregut ( Figure 12-16) forms the base of a U of which the two lateral sheets of mesoderm form the sides. Within this U the coelom develops and is, therefore, itself U-shaped. The base of the U and the two sides approximately as far caudad as the second pair of somites, contain that portion of the coelom which ultimately encloses the heart and which, therefore, is known as the pericardial coelom ( Figure 12-23). The curved shape of the pericardial coelom in cross section in Figure 12-23 should not be confused with the U-shape of the pericardial coelom as a whole.

The heart is derived from the splanchnic mesoderm forming the ventral wall of the pericardial coelom ( Figure 12-28). In 5-somite embryos this mesoderm has differentiated into to layers. Adjacent to the pericardial coelom is a thick continuous layer known as the epimyocardium because it will give rise both to the heavy muscular layer of the heart wall (myocardium) and to its outer covering (epicardium). Between the epimyocardium and the underlying endoderm are a number of irregular cavities which later fuse to form the cavity of the heart. The lining of these cavities is the endocardium.

Because of its relation to the U-shaped pericardial coelom, the heart is itself a U-shaped structure at this stage with the base of the U lying just cephalad to the anterior intestinal portal ( Figure 12-29). In many vertebrates the heart originates as two entirely distinct primordia which later fuse. As has been shown by Goss ( 1940)and by Burlingame and Long ( 1939), this is not the case in the rat. Our observations indicate that the condition in the mouse corresponds closely to that in the rat. As the intestinal portal moves caudad due to the "zipper action" causing the progressive folding together and fusion in the midventral line of the bounding endoderm, the sides of the U are likewise brought into approximation and fused together in the midventral line. The endocardium is thus transformed from a U-shaped structure into a single tube. At the 3-somite stage (in the rat) the different regions of the heart are not clearly set apart, though a slight constriction serves to mark the boundary between the atrium and the ventricle. As a result of subsequent foldings of the endocardial tube the different regions of the heart are clearly differentiated ( Figure 12-30).

Blood vessels

In 10-somite embryos, still in the process of turning, a number of blood vessels have become established ( Figure 12-30). The dorsal aorta at this stage is a paired vessel running the length of the trunk. It connects anteriorly with the heart by way of the aortic arches and the ventral aorta. Posteriorly its two halves fuse at the caudal extremity of the hindgut to form the single median omphalomesenteric artery. This runs cephalad for a short distance ventral to the hindgut and then turns away from the embryo towards the inner surface of the yolk sac on which it spreads out into a network of capillaries. These capillaries are derived from the blood islands. At this stage actual blood channels have not appeared in most of the blood islands, but when these are established, a capillary network is formed encircling the yolk sac. Blood is collected from this network anteriorly by the paired, omphalomesenteric veins which convey it back to the heart. When the heart starts beating, this system of blood vessels provides a generous circulation through the yolk sac serving at this time as the principal organ for the procurement of food from the mother.

Change in shape of yolk sac

When the embryo starts turning, the yolk sac and ectoplacental cone form a slightly flattened sphere ( Figure 12-20C). When turning has been completed, these bounding structures of the embryo shortly assume the form of a slightly saucered-out hemisphere ( Figure 12-21B). The ectoplacental cone becomes flattened and then dorsally concave, and the yolk sac adjacent to the cone pushes outward into the porous blood-filled vascular zone of the decidua. The embryo meanwhile, still attached to the yolk sac by the walls of the midgut, tips over so that it lies with its left side adjacent to the yolk sac, its right side facing the placenta.


After fertilization the mouse egg undergoes cleavage in the oviduct and forms a morula. During cleavage, the cells are labile and have not yet been determined. The morula enters the uterus and becomes a blastocyst: a fluid-filled sphere formed dorsally by the inner cell mass and laterally and ventrally by the trophectoderm. About 5 days after fertilization the blastocyst implants in the uterine mucosa. Endoderm delaminates from the ventral part of the inner cell mass and migrates around the inner surface of the trophectoderm to form the distal (parietal) endoderm. The inner cell mass is composed of ectoderm covered laterally and ventrally by a sheet of proximal endoderm. It grows down into the blastocoel or yolk cavity. The ectoderm of the inner cell mass becomes divided into a ventral region which will form the embryo proper and a dorsal region which is extraembryonic ectoderm. The dorsal pole of the extraembryonic ectoderm becomes the ectoplacental cone which will form trophoblastic giant cells and will become part of the placenta. Because the inner cell mass, composed of ectoderm covered with endoderm, grows down into the yolk cavity, the germ layers are temporarily inverted. The embryonic and extraembryonic ectoderm line the proamniotic cavity. At about 6 ½ days the mesoderm is proliferated from the posterior part of the embryonic ectoderm, called the primitive streak. Mesodermal cells migrate between the embryonic and extraembryonic ectoderm and the endoderm. At the posterior end of the primitive streak the proliferation of cells causes a bulge into the proamniotic cavity. This bulge of ectodermal and mesodermal cells is the primordium of the amnion separating the amniotic cavity from the dorsal cavity lined by extraembryonic ectoderm. The amniotic cavity closes by fusion of the amniotic folds and separates the embryo from the extraembryonic structures dorsal to it. The amnion is the floor of the exocoelom which forms in the mesoderm underlying the extraembryonic ectoderm. The chorion forms the roof of the exocoelom. The chorion fuses with the ectoplacental cone and becomes part of the placenta.

The head process is formed at the anterior end of the primitive streak, near the ventral pole of the embryo. It migrates anteriorly between the endoderm and ectoderm and gives rise to the notochord and part of the lining of the gut. The neural grove forms dorsal to the notochord. Soon after the exocoelom is formed, a mesodermal fingerlike process, the allantois, grows into it from the posterior end of the primitive streak. The allantois elongates and fuses with the chorion and ectoplacental cone. Embryonic blood vessels will form within the allantois and will connect the embryo to the maternal blood supply.

The foregut and hindgut form as pockets in the endoderm at the anterior and posterior ends of the embryo. There is a progressive drawing together in the midventral line, a "zipper-like" action, of the endodermal lining of the foregut and hindgut which contributes to the formation of the midgut. The invagination of the foregut pushes the endoderm and overlying mesoderm and ectoderm as a bulge into the amniotic cavity. This bulge is the head fold. There is rapid growth of neural folds in this region which will form the brain. The heart will develop from the mesoderm of the head fold.

At about 7 ½ days the somites begin to form as paired segmental structures in the paraxial mesoderm a little anterior to the primitive streak. Approximately 65 pairs are formed.

The primitive streak is a region of rapid growth of undifferentiated cells giving rise anteriorly to the head process and posteriorly to the mesoderm and allantois. Coincident with the formation of the somites in the paraxial mesoderm, the lateral mesoderm splits into two layers forming the coelom. The coelom extends anteriorly around the foregut where it will become the pericardial coelom.

The embryo is enveloped by three membranes: Reichert's membrane, the amnion, and the yolk sac. Reichert's membrane is a protective structure secreted by the distal endoderm and is anchored to the maternal decidua by the trophoblastic giant cells. The amnion is small at first since it merely covers the cavity formed by the U-shaped embryo. It expands rapidly as the embryo straightens and completely encloses it. The yolk sac is formed from extraembryonic endoderm and its underlying mesoderm. It grows rapidly and after the midgut is formed it envelops the entire embryo. Blood islands develop in the mesodermal layer of the yolk sac. The primordial germ cells are located in the yolk sac near the allantois of 8 ½-day embryos. These cells are the progenitors of all the definitive germ cells of both sexes.

Implantation occurs on the antimesometrial wall of the uterus with the ectoplacental cone projecting dorsally into the uterine lumen. The uterine mucosa (decidua) grows rapidly at the point of implantation so that it completely surrounds the embryo and extraembryonic structures. The decidua grows and contacts the mesometrial wall of the uterus and fuses with it to form part of the placenta. At the same time, it breaks away from the antimesometrial wall so that the embryo is now attached to the wall opposite to the implantation site.

The preimplanted embryo is thought to be nourished by the uterine milk. Early implantation stages may utilize digested uterine epithelial cells. After implantation the embryo is bathed in maternal blood from which the yolk sac probably actively absorbs nutrients. Later, the decidua basalis, the ectoplacental cone, the chorion, and parts of the allantois fuse to give rise to a true placenta which assumes a major role in transferring nutrients to the embryo. Trophoblastic giant cells are derived from the trophectoderm and the ectoplacental cone. They may play an active role in implantation and they anchor Reichert's membrane to the decidua.

At about 8 days, the 7-somite embryo becomes S-shaped, the foregut and hindgut are forming, the neural grove is well developed, and the heart develops in the mesoderm anterior to the foregut. The allantois has almost reached the chorion, the blood islands appear, and blood vessels have begun to form within the embryo.

In the 10-somite embryo the hindgut endoderm pushes the overlying mesoderm and ectoderm into the base of the allantois to form the tail fold.

Shortly after the 7-somite stage the embryo rotates 180° along its longitudinal axis and the embryo changes from being S-shaped to being C-shaped with the ventral surface on the inside of the C. This turning brings the walls of the midgut parallel to each other to form a groove. The walls then fuse to form a closed tube, the midgut.

In 10-somite embryos the dorsal aorta is paired and connects with the heart by way of the aortic arches and the ventral aorta. Posteriorly its two halves fuse to form the single median omphalomesenteric artery which spreads our over the yolk sac as capillaries. The omphalomesenteric veins convey blood from the yolk sac back to the heart.


Alden, R.H. 1948. Implantation of the rat egg. III. Origin and development of primary trophoblast giant cells. Amer. J. Anat. 83: 143-181.

Austin, C.R., and M.W.H. Bishop. 1958. Role of the rodent acrosome and perforatorium in fertilization. Proc. Roy. Soc. B 148: 241-248.
See also PubMed.

Austin, C.R., and A.W.H. Braden. 1956. Early reactions of the rodent egg to spermatozoan penetration. J. Exp. Biol. 33: 358-365.

Beatty, R.A., and K.N. Sharma. 1960. Genetics of gametes. III. Strain differences in spermatozoa from eight inbred strains of mice. Proc. Roy. Soc. Edinb. B 68: 25-53.
See also MGI.

Bennett, D., and L.C. Dunn. 1960. A lethal mutant (tw18) in the house mouse showing partial duplications. J. Exp. Zool. 1443: 203-219.
See also MGI.

Bishop, D.W., and A. Tyler. 1956. Fertilizin of mammalian eggs. J. Exp. Zool. 132: 575-602.

Blandau, R.J. 1949. Embryo-endometrial interrelationship in the rat and guinea pig. Anat. Rec. 104: 331-360.

Blandau, R.J., and W.L. Money. 1944. Observations on the rate of transport of spermatozoa in the female genital tract of the rat. Anat. Rec. 90: 255-260.

Blandau, R.J., and D.L. Odor. 1949. The total number of spermatozoa reaching various segments of the reproductive tract in the female albino rat at intervals after insemination. Anat. Rec. 103: 93-109.

Block, M. 1946. An experimental analysis of hematopoiesis in the rat yolk sac. Anat. Rec. 96: 289-304.

Böe, F. 1951. Studies on placental circulation in rats. III. Vascularization of the yolk sac. Acta Endocrinol. 7: 42-53.
See also PubMed.

Bonnevie, K. 1950. New facts on mesoderm formation and proamnion derivatives in the normal mouse embryo. J. Morphol. 86: 495-546.

Boyd, J.D., and W.J. Hamilton. 1952. Cleavage, early development and implantation of the egg, p. 1-126. In A.S. Parkes [ed.] Marshall's Physiology of Reproduction, 3rd ed. Vol 2. Longmans, Gree, London.

Braden, A.W.H. 1958. Variation between strains of mice in phenomena associated with sperm penetration and fertilization. J. Genet. 56: 37-47.

Braden, A.W.H. 1960. Genetic influences on the morphology and function of the gametes. J. Cell. Comp. Physiol. 56 (Suppl. 1): 17-29.

Braden, A.W.H. 1962. Spermatozoon penetration and fertilization in the mouse. Symp. Genet. Biol. Ital. 9: 1-8.

Braden, A.W.H., and C.R. Austin. 1954. The fertile life of mouse and rat eggs. Science 120: 610-611.
See also PubMed.

Bridgman, J. 1948. A morphological study of the development of the placenta of the rat. II. An histological and cytological study of the development of the chorioallantoic placenta of the white rat. J. Morphol. 83: 195-224.

Burckhard, G. 1901. Die implantation des Eies der Maus in die Uterusschleimhaout und die Umbildung derdelben zur Decidua. Arch. Mikroskop. Anat. 57: 528-569.

Burlingame, P.L., and J.A. Long. 1939. The development of the heart in the rat. Univ. Calif. Pub. Zool. 43: 249-320.

Butcher, E.O. 1929. The development of the somites in the white rat (Mus norvegicus albinus) and the fate of the myotomes, neural tube, and the gut in the tail. Amer. J. Anat. 44: 381-439.

Chang, M.C. 1962. Fertilizability of rabbit ova after removal of the corona radiata. Fertil. Steril. 13: 421-425.

Chiquoine, A.D. 1954. The identification, origin, and migration of the primordial germ cells in the mouse embryo. Anat. Rec. 118: 135-146.
See also PubMed.

Chiquoine, A.D. 1958. The distribution of polysaccharides during gastrulation and embryogenesis in the mouse embryo. Anat. Rec. 129: 495-516.
See also PubMed.

Dalcq, A.M. 1956. Effets du réactif de Schiff sur les oeufs en segmentation du rat et de la souris. Étude qualitative. Exp. Cell Res. 10: 99-119.
See also PubMed.

Dalcq, A., and M. Van Egmond. 1953. Effets de la centrifugation sur l'oocyte de trois Mammifères (Rat, Hamster, Taupe). Arch. Biol. 66: 312-397.

Dickerman, Z., and R.W. Noyes. 1960. The fate of ova transferred into the uterus of the rat. J. Reprod. Fertil. 1: 197-212.

Dickson, A.D. 1963. Trophoblastic giant cell transformation of mouse blastocyst. J. Reprod. Fertil. 6: 465-466.
See also PubMed.

Dziuk, P.J., and M.N. Runner. 1960. Recovery of blastocysts and induction of implantation following Artificial insemination of immature mice. J. Reprod. Fertil. 1: 321-331.
See also PubMed.

Eaton, G.J., and M.M. Green. 1963. Giant cell differentiation and lethality of homozygous Yellow mouse embryos. Genetica 34: 155-161.
See also MGI.

Edwards, R.G. 1957. The experimental induction of gynogenesis in the mouse. I. Irradiation of the sperm by X-rays. Proc. Roy. Soc. B 146: 469-487.
See also PubMed.

Edwards, R.G., and A.H. Gates. 1959. Timing of the stages of the maturation divisions, ovulation, fertilization and the first cleavage of eggs of adult mice treated with gonadotrophins. J. Endocrinol. 18: 292-304.
See also PubMed.

Everett, J.W. 1935. Morphological and physiological studies of the placenta in the albino rat. J. Exp. Zool. 70: 243-287.

Fawcett, D.W., G.B. Wislocki, and C.M. Waldo. 1947. The development of mouse ova in the anterior chamber of the eye and in the abdominal cavity. Amer. J. Anat. 81: 413-443.

Fekete, E., O. Bartholomew, and G.D. Snell. 1940. A technique for the preparation of sections of early mouse embryos. Anat. Rec. 76: 441-447.

Finn, C.A., and J.R. Hinchliffe. 1964. Reaction of the mouse uterus during implantation and deciduoma formation as demonstrated by changes in the distribution of alkaline phosphatase. J. Reprod. Fertil. 8: 331-338.
See also PubMed.

Geyer-Duszynska, I. 1964. Cytological investigations on the T-locus in Mus musculus L. Chromosoma 15: 478-502.
See also PubMed.

Goss, C.M. 1940. First contractions of the heart without cytological differentiation. Anat. Rec. 76: 19-27.

Graves, A.P. 1945. Development of the golden hamster, Crictus auratus Waterhouse during the first nine days. Amer. J. Anat. 77: 219-251.

Gresson, R.A.R. 1940. Presence of the sperm middle-piece in the fertilized egg of the mouse (Mus musculus). Nature 145: 425.

Grüneberg, H. 1943. The development of some external features in mouse embryos. J. Hered. 34: 89-92.

Grüneberg, H. 1963. The Pathology of Development. A Study of Inherited Skeletal Disorders in Animals. Wiley, New York. 309 p.

Gwatkin, R.B.I. 1964. Effect of enzymes and acidity on the zona pellicuda of the mouse egg before and after fertilization. J. Reprod. Fertil. 7: 99-105.
See also PubMed.

Hollander, W.F., and L.C. Strong. 1950. Intrauterine mortality and placental fusions in the mouse. J. Exp. Zool. 115: 131-150.

Huber, G.C. 1915. The development of the albino rat, Mus norvegicus albinus. I. From the pronuclear stage to the stage of mesoderm anlage; end of the first to the end of the ninth day. J. Morphol. 26: 1-114.

Huber, G.C. 1918. On the anlage and morphogenesis of the chorda dorsalis in mammalia, in particular the guinea pig (Cavia cobaya). Anat. Rec. 14: 217-263.

Hyman, L.H. 1927. The metabolic gradients of vertebrate embryos. III. The chick. Biol. Bull. 52: 1-38.

Jolly, J., and M. Férester-Tadié. 1936. Recherches sur l'oeuf du rat et de la souris. Arch. Anat. Microscop. 32: 323-390.

Jones-Seaton, A. 1950a. A study of cytoplasmic basophily in the egg of the rat and some other mammals. Ann. Soc. Roy. Zool. (Belgium) 80: 76-86.

Jones-Seaton, A. 1950b. Étude de l'organization cytoplasmique de l'oeuf des rongeurs principalment quant á la basophilie ribonucléique. Arch. Biol. 61: 291-444.
See also PubMed.

Kingsbury, B.F. 1920. The developmental origin of the notochord. Science 51: 190-193.

Kirby, D.R.S. 1962a. The influence of uterine environment on the development on the development of mouse eggs. J. Embryol. Exp. Morphol. 10: 496-506.
See also PubMed.

Kirby, D.R.S. 1962b. The developmental of mouse blastocysts transplanted to the scrotal and cryptorchid testis. J. Anat. 97: 119-130.
See also PubMed.

Kirby, D.R.S., and S.K. Malhotra. 1964. Cellular nature of the invasive mouse trophoblast. Nature 201: 520.
See also PubMed.

Krehbiel, R.H. 1937. Cytological studies of the decidual reaction in the rat during the early pregnancy and the production of deciduomata. Physiol. Zool. 10: 212-234.

Krehbiel, R.H. 1962. Distribution of ova in the rat uterus. Anat. Rec. 143: 239-241.
See also PubMed.

Lams, H., and J. Doorme. 1907. Nouvelles recherches sur la maturation et la fécondation de l'oeuf des Mammifères. Arch. Biol. 23: 259-365.

Leonard, S.L., P.L. Perlman, and R. Kurzrok. 1947. Relation between time of fertilization and follicle cell dispersal in rat ova. Proc. Soc. Exp. Biol. Med. 66: 517-518.

Lewis, W.H., and E.S. Wright. 1935. On the early development of the mouse egg. Carnegie Inst. Wash. Pub. No. 459: 133-144.

Mandl, A.M. 1963. Pre-ovulatory changes in the oocyte of the adult rat. Proc. Roy. Soc. B 158: 105-118.

Melissinos, K. 1907. Die Entwicklung des Eis der Mäuse von der ersten Furchungs-Phänomenen bis zur Festsetzung der Allantois an der Ectoplacentarplatte. Arch. Mikroscop. Anat. 70: 557-628.

Midgley, A.R., and G.B. Pierce. 1963. Immunohistochemical analysis of basement membranes of the mouse. Amer. J. Anat. 63: 929-944.
See also PubMed.

Mintz, B. 1960. Embryological phases of mammalian gametogenesis. J. Cell. Comp. Physiol. 56 (Suppl. 1): 31-47.
See also PubMed.

Mintz, B. 1962. Experimental study of the developng mammalian egg: removal of the zona pellucida. Science 138: 594-595.

Mintz, B. 1964. Formation of genetically mosaic mouse embryos, and early development of "lethal (t12/t12)-normal" mosaics. J. Exp. Zool. 157: 273-292.
See also PubMed.

Mintz, B. 1965. Genetic mosaicism in adult mice of quadriparental lineage. Science 148: 1232-1233.
See also PubMed.

Mintz, B., and E.S. Russell. 1957. Gene-induced embryological modifications of primordial germ cells in the mouse. J. Exp. Zool. 134: 207-238.
See also MGI.

Mossman, H.W., and L.A. Weisfeldt. 1939. The fetal membrances of a primitive rodent, the thirteen-striped ground squirrel. Amer. J. Anat. 64: 59-109.

Nicholas, J.S., and B.V. Hall. 1942. Experiments on developing rats. II. The development of isolated blastomeres and fused eggs. J. Exp. Zool. 90: 441-460.

Otis, E.M., and R. Brent, 1954. Equivalent ages in mouse and human embryos. Anat. Rec. 120: 33-64.
See also PubMed.

Pierce, G.B., A.R. Midgley, J. Sri Ram, and J.D. Feldman. 1962. Parietal yolk sac carcinoma: clue to the histogenesis of Rechert's membrane of the mouse embryo. Amer. J. Pathol. 41: 549-566.
See also PubMed.

Pincus, G. 1936. The Eggs of Mammals. Macmillan, New York. 161 p.

Rugh, R. 1964. The mouse: a discoid placentate, p. 236-303. In R. Rugh, Vertebrate embryology; the dynamics of development. Harcourt, New York.

Sato, K. 1936. Über die Entwicklungsgeschichte des Mäuseeies I. Die intratubare Entwicklung deselben. Okayama-Igakkai-Zasshi 48: 423-441.

Schlafke, S., and A.C. Enders. 1963. Observations on the fine structure of the rat blastocyst. J. Anat. 97: 353-360.
See also PubMed.

Sharma, K.N. 1960. Genetics of gametes. IV. The phenotype of mouse spermatozoa in four inbred strains and their F crosses. Proc. Roy. Soc. Edinb. B 68: 54-71.
See also MGI.

Smith, L.J. 1956. A morphological and histochemical investigation of a preimplantation lethal (t12) in the house mouse. J. Exp. Zool. 132: 51-84.

Snell, G.D. 1941. The early embryology of the mouse, p. 1-54. In G.D. Snell [ed.] Biology of the Laboratory Mouse. Blakiston, Philadelphia.

Sobotta, J. 1911. Die Entwicklung des Eies der Maus vom ersten Auftreten des Mesoderms an bis zen Ausbildung der Embryonalanlage und dem Auftreten der Allantois. Arch. Mikroskop. Anat. 78: 271-352.

Stafford, E.S. 1930. The origin of the blood of the "placental sign." Anat. Rec. 47: 43-57.

Stevens, L.C. 1964. Experimental production of testicular teratomas in mice. Proc. Nat. Acad. Sci. 52: 645-661.
See also PubMed.

Tarkowski, A.K. 1962. Interspecific transfers of eggs between rat and mouse. J. Embryol. Exp. Morphol. 10: 476-495.
See also PubMed.

Venable, J.W. 1939. Intra-uterine bleeding in the pregnant albino rat. The "placental sign." Anat. Rec. 74: 273-293.

Whitten, W.K., and C.P. Dagg. 1961. Influence of spermatozoa on the cleavage rate of mouse eggs. J. Exp. Zool. 148: 173-183.
See also MGI.

Wilson, I.B. 1963. A new factor associated with the implantation of the mouse egg. J. Reprod. Fertil. 5: 281-282.
See also PubMed.

Wislocki, G.B., H.W. Deane, and E.W. Dempsey. 1946. The histochemistry of the rodent's placenta. Amer. J. Anat. 72: 281-345.

Previous   Next