CHAPTER 1.13. DORSAL BODY AND VENTRAL BODY.

The earliest stages of the human embryo are, for the reasons already

given, either quite unknown or only imperfectly known to us. But as

the subsequent embryonic forms in man behave and develop just as they

do in all the other mammals, there cannot be the slightest doubt that

the preceding stages also are similar. We have been able to see in the

coelomula of the human embryo (Figure 1.97), by transverse sections

through its primitive mouth, that its two coelom-pouches are developed

in just the same way as in the rabbit (Figure 1.96); moreover, the

peculiar course of the gastrulation is just the same.

(FIGURE 1.120. Embryonic vesicle of a seven-days-old rabbit with oval

embryonic shield (ag).

A seen from above, B from the side. (From Kolliker.) ag dorsal shield

or embryonic spot. In B the upper half of the vesicle is made up of

the two primary germinal layers, the lower (up to ge) only from the

outer layer.)

The germinative area forms in the human embryo in the same way as in

the other mammals, and in the middle part of this we have the

embryonic shield, the purport of which we considered in Chapter 1.12.

The next changes in the embryonic disk, or the "embryonic spot," take

place in corresponding fashion. These are the changes we are now going

to consider more closely.

The chief part of the oval embryonic shield is at first the narrow

hinder end; it is in the middle line of this that the primitive streak

appears (Figure 1.121 ps). The narrow longitudinal groove in it--the

so-called "primitive groove"--is, as we have seen, the primitive mouth

of the gastrula. In the gastrula-embryos of the mammals, which are

much modified cenogenetically, this cleft-shaped prostoma is

lengthened so much that it soon traverses the whole of the hinder half

of the dorsal shield; as we find in a rabbit embryo of six to eight

days (Figure 1.122 pr). The two swollen parallel borders that limit

this median furrow are the side lips of the primitive mouth, right and

left. In this way the bilateral-symmetrical type of the vertebrate

becomes pronounced. The subsequent head of the amniote is developed

from the broader and rounder fore-half of the dorsal shield.

In this fore-half of the dorsal shield a median furrow quickly makes

its appearance (Figure 1.123 rf). This is the broader dorsal furrow or

medullary groove, the first beginning of the central nervous system.

The two parallel dorsal or medullary swellings that enclose it grow

together over it afterwards, and form the medullary tube. As is seen

in transverse sections, it is formed only of the outer germinal layer

(Figures 1.95 and 1.136). The lips of the primitive mouth, however,

lie, as we know, at the important point where the outer layer bends

over the inner, and from which the two coelom pouches grow between the

primary germinal layers.

(FIGURE 1.121. Oval embryonic shield of the rabbit (A of six days

eighteen hours, B of eight days). (From Kolliker.) ps primitive

streak, pr primitive groove, arg area germinalis, sw sickle-shaped

germinal growth.

FIGURE 1.122. Dorsal shield (ag) and germinative area of a

rabbit-embryo of eight days. (From Kolliker.) pr primitive groove, rf

dorsal furrow.

FIGURE 1.123. Embryonic shield of a rabbit of eight days. (From Van

Beneden.) pr primitive groove, cn canalis neurentericus, nk nodus

neurentericus (or "Hensen's ganglion"), kf head-process (chorda).

FIGURE 1.124. Longitudinal section of the coelomula of amphioxus (from

the left). i entoderm, d primitive gut, cn medullary duct, n nerve

tube, m mesoderm, s first primitive segment, c coelom-pouches. (From

Hatschek.))

Thus the median primitive furrow (pr) in the hind-half and the median

medullary furrow (rf) in the fore-half of the oval shield are totally

different structures, although the latter seems to a superficial

observer to be merely the forward continuation of the former. Hence

they were formerly always confused. This error was the more pardonable

as immediately afterwards the two grooves do actually pass into each

other in a very remarkable way. The point of transition is the

remarkable neurenteric canal (Figure 1.124 cn). But the direct

connection which is thus established does not last long; the two are

soon definitely separated by a partition.

The enigmatic neurenteric canal is a very old embryonic organ, and of

great phylogenetic interest, because it arises in the same way in all

the chordonia (both tunicates and vertebrates). In every case it

touches or embraces like an arch the posterior end of the chorda,

which has been developed here in front out of the middle line of the

primitive gut (between the two coelom-folds of the sickle groove)

("head-process," Figure 1.123 kf). These very ancient and strictly

hereditary structures, which have no physiological significance

to-day, deserve (as "rudimentary organs") our closest attention. The

tenacity with which the useless neurenteric canal has been transmitted

down to man through the whole series of vertebrates is of equal

interest for the theory of descent in general, and the phylogeny of

the chordonia in particular.

The connection which the neurenteric canal (Figure 1.123 cn)

establishes between the dorsal nerve-tube (n) and the ventral gut-tube

(d) is seen very plainly in the amphioxus in a longitudinal section of

the coelomula, as soon as the primitive mouth is completely closed at

its hinder end. The medullary tube has still at this stage an opening

at the forward end, the neuroporus (Figure 1.83 np). This opening also

is afterwards closed. There are then two completely closed canals over

each other--the medullary tube above and the gastric tube below, the

two being separated by the chorda. The same features as in the acrania

are exhibited by the related tunicates, the ascidiae.

Again, we find the neurenteric canal in just the same form and

situation in the amphibia. A longitudinal section of a young tadpole

(Figure 1.125) shows how we may penetrate from the still open

primitive mouth (x) either into the wide primitive gut-cavity (al) or

the narrow overlying nerve-tube. A little later, when the primitive

mouth is closed, the narrow neurenteric canal (Figure 1.126 ne)

represents the arched connection between the dorsal medullary canal

(mc) and the ventral gastric canal.

(FIGURE 1.125. Longitudinal section of the chordula of a frog. (From

Balfour.) nc nerve-tube, x canalis neurentericus, al alimentary canal,

yk yelk-cells, m mesoderm.

FIGURE 1.126. Longitudinal section of a frog-embryo. (From Goette.) m

mouth, l liver, an anus, ne canalis neurentericus, mc medullary-tube,

pn pineal body (epiphysis), ch chorda.

FIGURES 1.127 AND 1.128. Dorsal shield of the chick. (From Balfour.)

The medullary furrow (me), which is not yet visible in Figure 1.130,

encloses with its hinder end the fore end of the primitive groove (pr)

in Figure 1.131.)

In the amniotes this original curved form of the neurenteric canal

cannot be found at first, because here the primitive mouth travels

completely over to the dorsal surface of the gastrula, and is

converted into the longitudinal furrow we call the primitive groove.

Hence the primitive groove (Figure 1.128 pr), examined from above,

appears to be the straight continuation of the fore-lying and younger

medullary furrow (me). The divergent hind legs of the latter embrace

the anterior end of the former. Afterwards we have the complete

closing of the primitive mouth, the dorsal swellings joining to form

the medullary tube and growing over it. The neurenteric canal then

leads directly, in the shape of a narrow arch-shaped tube (Figure

1.129 ne), from the medullary tube (sp) to the gastric tube (pag).

Directly in front of it is the latter end of the chorda (cli).

While these important processes are taking place in the axial part of

the dorsal shield, its external form also is changing. The oval form

(Figure 1.117) becomes like the sole of a shoe or sandal, lyre-shaped

or finger-biscuit shaped (Figure 1.130). The middle third does not

grow in width as quickly as the posterior, and still less than the

anterior third; thus the shape of the permanent body becomes somewhat

narrow at the waist. At the same time, the oval form of the

germinative area returns to a circular shape, and the inner pellucid

area separates more clearly from the opaque outer area (Figure 1.131

a). The completion of the circle in the area marks the limit of the

formation of blood-vessels in the mesoderm.

(FIGURE 1.129. Longitudinal section of the hinder end of a chick.

(From Balfour.) sp medullary tube, connected with the terminal gut

(pag) by the neurenteric canal (ne), ch chorda, pr neurenteric (or

Hensen's) ganglion, al allantois, ep ectoderm, hy entoderm, so

parietal layer, sp visceral layer, an anus-pit, am amnion.)

The characteristic sandal-shape of the dorsal shield, which is

determined by the narrowness of the middle part, and which is compared

to a violin, lyre, or shoe-sole, persists for a long time in all the

amniotes. All mammals, birds, and reptiles have substantially the same

construction at this stage, and even for a longer or shorter period

after the division of the primitive segments into the coelom-folds has

begun (Figure 1.132). The human embryonic shield assumes the

sandal-form in the second week of development; towards the end of the

week our sole-shaped embryo has a length of about one-twelfth of an

inch (Figure 1.133).

The complete bilateral symmetry of the vertebrate body is very early

indicated in the oval form of the embryonic shield (Figure 1.117) by

the median primitive streak; in the sandal-form it is even more

pronounced (Figures 1.131 to 1.135). In the lateral parts of the

embryonic shield a darker central and a lighter peripheral zone become

more obvious; the former is called the stem-zone (Figure 1.134 stz),

and the latter the parietal zone (pz); from the first we get the

dorsal and from the second the ventral half of the body-wall. The

stem-zone of the amniote embryo would be called more appropriately the

dorsal zone or dorsal shield; from it develops the whole of the dorsal

half of the later body (or permanent body)--that is to say, the dorsal

body (episoma). Again, it would be better to call the "parietal zone"

the ventral zone or ventral shield; from it develop the ventral

"lateral plates," which afterwards separate from the embryonic vesicle

and form the ventral body (hyposoma)--that is to say, the ventral half

of the permanent body, together with the body-cavity and the gastric

canal that it encloses.

(FIGURE 1.130. Germinal area or germinal disk of the rabbit, with

sole-shaped embryonic shield, magnified about ten times. The clear

circular field (d) is the opaque area. The pellucid area (c) is

lyre-shaped, like the embryonic shield itself (b). In its axis is seen

the dorsal furrow or medullary furrow (a). (From Bischoff.))

The sole-shaped germinal shields of all the amniotes are still, at the

stage of construction which Figure 1.134 illustrates in the rabbit and

Figure 1.135 in the opossum, so like each other that we can either not

distinguish them at all or only by means of quite subordinate

peculiarities in the size of the various parts. Moreover, the human

sandal-shaped embryo cannot at this stage be distinguished from those

of other mammals, and it particularly resembles that of the rabbit. On

the other hand, the outer form of these flat sandal-shaped embryos is

very different from the corresponding form of the lower animals,

especially the acrania (amphioxus). Nevertheless, the body is just the

same in the essential features of its structure as that we find in the

chordula of the latter (Figures 1.83 to 1.86), and in the embryonic

forms which immediately develop from it. The striking external

difference is here again due to the fact that in the palingenetic

embryos of the amphioxus (Figures 1.83 and 1.84) and the amphibia

(Figures 1.85 and 1.86) the gut-wall and body-wall form closed tubes

from the first, whereas in the cenogenetic embryos of the amniotes

they are forced to expand leaf-wise on the surface owing to the great

extension of the food-yelk.

(FIGURE 1.131. Embryo of the opossum, sixty hours old, one-sixth of an

inch in diameter. (From Selenka) b the globular embryonic vesicle, a

the round germinative area, b limit of the ventral plates, r dorsal

shield, v its fore part, u the first primitive segment, ch chorda, chr

its fore-end, pr primitive groove (or mouth).

FIGURE 1.132. Sandal-shaped embryonic shield of a rabbit of eight

days, with the fore part of the germinative area (ao opaque, ap

pellucid area). (From Kolliker.) rf dorsal furrow, in the middle of

the medullary plate, h, pr primitive groove (mouth), stz dorsal (stem)

zone, pz ventral (parietal) zone. In the narrow middle part the first

three primitive segments may be seen.)

It is all the more notable that the early separation of dorsal and

ventral halves takes place in the same rigidly hereditary fashion in

all the vertebrates. In both the acrania and the craniota the dorsal

body is about this period separated from the ventral body. In the

middle part of the body this division has already taken place by the

construction of the chorda between the dorsal nerve-tube and the

ventral canal. But in the outer or lateral part of the body it is only

brought about by the division of the coelom-pouches into two

sections--a dorsal episomite (dorsal segment or provertebra) and a

ventral hyposomite (or ventral segment) by a frontal constriction. In

the amphioxus each of the former makes a muscular pouch, and each of

the latter a sex-pouch or gonad.

These important processes of differentiation in the mesoderm, which we

will consider more closely in the next chapter, proceed step by step

with interesting changes in the ectoderm, while the entoderm changes

little at first. We can study these processes best in transverse

sections, made vertically to the surface through the sole-shaped

embryonic shield. Such a transverse section of a chick embryo, at the

end of the first day of incubation, shows the gut-gland layer as a

very simple epithelium, which is spread like a leaf over the outer

surface of the food-yelk (Figure 1.92). The chorda (ch) has separated

from the dorsal middle line of the entoderm; to the right and left of

it are the two halves of the mesoderm, or the two coelom-folds. A

narrow cleft in the latter indicates the body-cavity (uwh); this

separates the two plates of the coelom-pouches, the lower (visceral)

and upper (parietal). The broad dorsal furrow (Rf) formed by the

medullary plate (m) is still wide open, but is divided from the

lateral horn-plate (h) by the parallel medullary swellings, which

eventually close.

(FIGURE 1.133. Human embryo at the sandal-stage, one-twelfth of an

inch long, from the end of the second week, magnified twenty-five

times. (From Count Spee.)

FIGURE 1.134. Sandal-shaped embryonic shield of a rabbit of nine days.

(From Kolliker.) (Back view from above.) stz stem-zone or dorsal

shield (with eight pairs of primitive segments), pz parietal or

ventral zone, ap pellucid area, af amnion-fold, h heart, ph

pericardial cavity, vo omphalo-mesenteric vein, ab eye-vesicles, vh

fore brain, mh middle brain, hh hind brain, uw primitive segments (or

vertebrae).)

During these processes important changes are taking place in the outer

germinal layer (the "skin-sense layer"). The continued rise and growth

of the dorsal swellings causes their higher parts to bend together at

their free borders, approach nearer and nearer (Figure 1.136 w), and

finally unite. Thus in the end we get from the open dorsal furrow, the

upper cleft of which becomes narrower and narrower, a closed

cylindrical tube (Figure 1.137 mr). This tube is of the utmost

importance; it is the beginning of the central nervous system, the

brain and spinal marrow, the medullary tube. This embryonic fact was

formerly looked upon as very mysterious. We shall see presently that

in the light of the theory of descent it is a thoroughly natural

process. The phylogenetic explanation of it is that the central

nervous system is the organ by means of which all intercourse with the

outer world, all psychic action and sense-perception, are

accomplished; hence it was bound to develop originally from the outer

and upper surface of the body, or from the outer skin. The medullary

tube afterwards separates completely from the outer germinal layer,

and is surrounded by the middle parts of the provertebrae and forced

inwards (Figure 1.146). The remaining portion of the skin-sense layer

(Figure 1.93 h) is now called the horn-plate or horn-layer, because

from it is developed the whole of the outer skin or epidermis, with

all its horny appendages (nails, hair, etc.).

(FIGURE 1.135. Sandal-shaped embryonic shield of an opossum

(Didelphys), three days old. (From Selenka.) (Back view from above.)

stz stem-zone or dorsal shield (with eight pairs of primitive

segments), pz parietal or ventral zone, ap pellucid area, ao opaque

area, hh halves of the heart, v fore-end, h hind-end. In the median

line we see the chorda (ch) through the transparent medullary tube

(m). u primitive segment, pr primitive streak (or primitive mouth).)

A totally different organ, the prorenal (primitive kidney) duct (ung),

is found to be developed at an early stage from the ectoderm. This is

originally a quite simple, tube-shaped, lengthy duct, or straight

canal, which runs from front to rear at each side of the provertebrae

(on the outer side, Figure 1.93 ung). It originates, it seems, out of

the horn-plate at the side of the medullary tube, in the gap that we

find between the provertebral and the lateral plates. The prorenal

duct is visible in this gap even at the time of the severance of the

medullary tube from the horn-plate. Other observers think that the

first trace of it does not come from the skin-sense layer, but the

skin-fibre layer.

The inner germinal layer, or the gut-fibre layer (Figure 1.93 dd),

remains unchanged during these processes. A little later, however, it

shows a quite flat, groove-like depression in the middle line of the

embryonic shield, directly under the chorda. This depression is called

the gastric groove or furrow. This at once indicates the future lot of

this germinal layer. As this ventral groove gradually deepens, and its

lower edges bend towards each other, it is formed into a closed tube,

the alimentary canal, in the same way as the medullary groove grows

into the medullary tube. The gut-fibre layer (Figure 1.137 f), which

lies on the gut-gland layer (d), naturally follows it in its folding.

Moreover, the incipient gut-wall consists from the first of two

layers, internally the gut-gland layer and externally the gut-fibre

layer.

The formation of the alimentary canal resembles that of the medullary

tube to this extent--in both cases a straight groove or furrow arises

first of all in the middle line of a flat layer. The edges of this

furrow then bend towards each other, and join to form a tube (Figure

1.137). But the two processes are really very different. The medullary

tube closes in its whole length, and forms a cylindrical tube, whereas

the alimentary canal remains open in the middle, and its cavity

continues for a long time in connection with the cavity of the

embryonic vesicle. The open connection between the two cavities is

only closed at a very late stage, by the construction of the navel.

The closing of the medullary tube is effected from both sides, the

edges of the groove joining together from right and left. But the

closing of the alimentary canal is not only effected from right and

left, but also from front and rear, the edges of the ventral groove

growing together from every side towards the navel. Throughout the

three higher classes of vertebrates the whole of this process of the

construction of the gut is closely connected with the formation of the

navel, or with the separation of the embryo from the yelk-sac or

umbilical vesicle.

In order to get a clear idea of this, we must understand carefully the

relation of the embryonic shield to the germinative area and the

embryonic vesicle. This is done best by a comparison of the five

stages which are shown in longitudinal section in Figures 1.138 to

1.142. The embryonic shield (c), which at first projects very slightly

over the surface of the germinative area, soon begins to rise higher

above it, and to separate from the embryonic vesicle. At this point

the embryonic shield, looked at from the dorsal surface, shows still

the original simple sandal-shape (Figures 1.133 to 1.135). We do not

yet see any trace of articulation into head, neck, trunk, etc., or

limbs. But the embryonic shield has increased greatly in thickness,

especially in the anterior part. It now has the appearance of a thick,

oval swelling, strongly curved over the surface of the germinative

area. It begins to sever completely from the embryonic vesicle, with

which it is connected at the ventral surface. As this severance

proceeds, the back bends more and more; in proportion as the embryo

grows the embryonic vesicle decreases, and at last it merely hangs as

a small vesicle from the belly of the embryo (Figure 1.142 ds). In

consequence of the growth-movements which cause this severance, a

groove-shaped depression is formed at the surface of the vesicle, the

limiting furrow, which surrounds the vesicle in the shape of a pit,

and a circular mound or dam (Figure 1.139 ks) is formed at the outside

of this pit by the elevation of the contiguous parts of the germinal

vesicle.

(FIGURE 1.136. Transverse section of the embryonic disk of a chick at

the end of the first day of incubation, magnified about twenty times.

The edges of the medullary plate (m), the medullary swellings (w),

which separate the medullary from the horn-plate (h), are bending

towards each other. At each side of the chorda (ch) the primitive

segment plates (u) have separated from the lateral plates (sp). A

gut-gland layer. (From Remak.))

In order to understand clearly this important process, we may compare

the embryo to a fortress with its surrounding rampart and trench. The

ditch consists of the outer part of the germinative area, and comes to

an end at the point where the area passes into the vesicle. The

important fold of the middle germinal layer that brings about the

formation of the body-cavity spreads beyond the borders of the embryo

over the whole germinative area. At first this middle layer reaches as

far as the germinative area; the whole of the rest of the embryonic

vesicle consists in the beginning only of the two original limiting

layers, the outer and inner germinal layers. Hence, as far as the

germinative area extends the germinal layer splits into the two plates

we have already recognised in it, the outer skin-fibre layer and the

inner gut-fibre layer. These two plates diverge considerably, a clear

fluid gathering between them (Figure 1.140 am). The inner plate, the

gut-fibre layer, remains on the inner layer of the embryonic vesicle

(on the gut-gland layer). The outer plate, the skin-fibre layer, lies

close on the outer layer of the germinative area, or the skin-sense

layer, and separates together with this from the embryonic vesicle.

From these two united outer plates is formed a continuous membrane.

This is the circular mound that rises higher and higher round the

whole embryo, and at last joins above it (Figures 1.139 to 1.142 am).

To return to our illustration of the fortress, we must imagine the

circular rampart to be extraordinarily high and towering far above the

fortress. Its edges bend over like the combs of an overhanging wall of

rock that would enclose the fortress; they form a deep hollow, and at

last join together above. In the end the fortress lies entirely within

the hollow that has been formed by the growth of the edges of this

large rampart.

(FIGURE 1.137. Three diagrammatic transverse sections of the embryonic

disk of the higher vertebrate, to show the origin of the tubular

organs from the bending germinal layers. In Figure A the medullary

tube (n) and the alimentary canal (a) are still open grooves. In

Figure B the medullary tube (n) and the dorsal wall are closed, but

the alimentary canal (a) and the ventral wall are open; the prorenal

ducts (u) are cut off from the horn-plate (h) and internally connected

with segmental prorenal canals. In Figure C both the medullary tube

and the dorsal wall above and the alimentary canal and ventral wall

below are closed. All the open grooves have become closed tubes; the

primitive kidneys are directed inwards. The letters have the same

meaning in all three figures: h skin-sense layer, n medullary tube, u

prorenal ducts, x axial rod, s primitive-vertebra, r dorsal wall, b

ventral wall, c body-cavity or coeloma, f gut-fibre layer, t primitive

artery (aorta), v primitive vein (subintestinal vein), d gut-fibre

layer, a alimentary canal.)

As the two outer layers of the germinative area thus rise in a fold

about the embryo, and join above it, they come at last to form a

spacious sac-like membrane about it. This envelope takes the name of

the germinative membrane, or water-membrane, or amnion (Figure 1.142

am). The embryo floats in a watery fluid, which fills the space

between the embryo and the amnion, and is called the amniotic fluid

(Figures 1.141 and 1.142 ah). We will deal with this remarkable

formation and with the allantois later on (Chapter 1.15). In front of

the allantois the yelk-sac or umbilical vesicle (ds), the remainder of

the original embryonic vesicle, starts from the open belly of the

embryo (Figure 1.138 kh). In more advanced embryos, in which the

gastric wall and the ventral wall are nearly closed, it hangs out of

the navel-opening in the shape of a small vesicle with a stalk

(Figures 1.141 and 1.142 ds). The more the embryo grows, the smaller

becomes the vitelline (yelk) sac. At first the embryo looks like a

small appendage of the large embryonic vesicle. Afterwards it is the

yelk-sac, or the remainder of the embryonic vesicle, that seems a

small pouch-like appendage of the embryo (Figure 1.142 ds). It ceases

to have any significance in the end. The very wide opening, through

which the gastric cavity at first communicates with the umbilical

vesicle, becomes narrower and narrower, and at last disappears

altogether. The navel, the small pit-like depression that we find in

the developed man in the middle of the abdominal wall, is the spot at

which the remainder of the embryonic vesicle (the umbilical vesicle)

originally entered into the ventral cavity, and joined on to the

growing gut.

(FIGURES 1.138 TO 1.142. Five diagrammatic longitudinal sections of

the maturing mammal embryo and its envelopes. In Figures 1.138 to

1.141 the longitudinal section passes through the sagittal or middle

plane of the body, dividing the right and left halves; in Figure 1.142

the embryo is seen from the left side. In Figure 1.138 the tufted it

prochorion (dd apostrophe) encloses the germinal vesicle, the wall of

which consists of the two primary layers. Between the outer (a) and

inner (i) layer the middle layer (m) has been developed in the region

of the germinative area. In Figure 1.139 the embryo (e) begins to

separate from the embryonic vesicle (ds), while the wall of the

amnion-fold rises about it (in front as head-sheath, ks, behind as

tail-sheath, ss). In Figure 1.140 the edges of the amniotic fold (am)

rise together over the back of the embryo, and form the amniotic

cavity (ah); as the embryo separates more completely from the

embryonic vesicle (ds) the alimentary canal (dd) is formed, from the

hinder end of which the allantois grows (al). In Figure 1.141 the

allantois is larger; the yelk-sac (ds) smaller. In Figure 1.142 the

embryo shows the gill-clefts and the outline of the two legs; the

chorion has formed branching villi (tufts.) In all four figures e =

embryo, a outer germinal layer, m middle germinal layer, i inner

germinal layer, am amnion (ks head-sheath, ss tail-sheath), ah

amniotic cavity, as amniotic sheath of the umbilical cord, kh

embryonic vesicle, ds yelk-sac (umbilical vesicle), dg vitelline duct,

df gut-fibre layer, dd gut-gland layer, al allantois, vl = hh place of

heart, d vitelline membrane (ovolemma or prochorion), d apostrophe

tufts or villi of same, sh serous membrane (serolemma), sz tufts of

same, ch chorion, chz tufts or villi, st terminal vein, r pericoelom

or serocoelom (the space, filled with fluid, between the amnion and

chorion). (From Kolliker.))

The origin of the navel coincides with the complete closing of the

external ventral wall. In the amniotes the ventral wall originates in

the same way as the dorsal wall. Both are formed substantially from

the skin-fibre layer, and externally covered with the horn-plate, the

border section of the skin-sense layer. Both come into existence by

the conversion of the four flat germinal layers of the embryonic

shield into a double tube by folding from opposite directions; above,

at the back, we have the vertebral canal which encloses the medullary

tube, and below, at the belly, the wall of the body-cavity which

contains the alimentary canal (Figure 1.137).

We will consider the formation of the dorsal wall first, and that of

the ventral wall afterwards (Figures 1.143 to 1.147). In the middle of

the dorsal surface of the embryo there is originally, as we already

know, the medullary (mr) tube directly underneath the horn-plate (h),

from the middle part of which it has been developed. Later, however,

the provertebral plates (uw) grow over from the right and left between

these originally connected parts (Figures 1.145 and 1.146). The upper

and inner edges of the two provertebral plates push between the

horn-plate and medullary tube, force them away from each other, and

finally join between them in a seam that corresponds to the middle

line of the back. The coalescence of these two dorsal plates and the

closing in the middle of the dorsal wall take place in the same way as

the medullary tube, which is henceforth enclosed by the vertebral

tube. Thus is formed the dorsal wall, and the medullary tube takes up

a position inside the body. In the same way the provertebral mass

grows afterwards round the chorda, and forms the vertebral column.

Below this the inner and outer edge of the provertebral plate splits

on each side into two horizontal plates, of which the upper pushes

between the chorda and medullary tube, and the lower between the

chorda and gastric tube. As the plates meet from both sides above and

below the chorda, they completely enclose it, and so form the tubular,

outer chord-sheath, the sheath from which the vertebral column is

formed (perichorda, Figure 1.137 C, s; Figures 1.145 uwh, 1.146).

(FIGURES 1.143 TO 1.146. Transverse sections of embryos (of chicks).

Figure 1.143 of the second, Figure 1.144 of the third, Figure 1.145 of

the fourth, and Figure 1.146 of the fifth day of incubation. Figures

1.143 to 1.145 from Kolliker, magnified about 100 times; Figure 1.146

from Remak, magnified about twenty times. h horn-plate, mr medullary

tube, ung prorenal duct, un prorenal vesicles, hp skin-fibre layer, m

= mu = mp muscle-plate, uw provertebral plate (wh cutaneous rudiment

of the body of the vertebra, wb of the arch of the vertebra, wq the

rib or transverse continuation), uwh provertebral cavity, ch axial rod

or chorda, sh chorda-sheath, bh ventral wall, g hind and v fore root

of the spinal nerves, a = af = am amniotic fold, p body-cavity or

coeloma, df gut-fibre layer, ao primitive aortas, sa secondary aorta,

vc cardinal veins, d = dd gut-gland layer, dr gastric groove. In

Figure 1.143 the larger part of the right half, in Figure 1.144 the

larger part of the left half, of the section is omitted. Of the

yelk-sac or remainder of the embryonic vesicle only a small piece of

the wall is indicated below.)

We find in the construction of the ventral wall precisely the same

processes as in the formation of the dorsal wall (Figure 1.137 B,

Figure 1.144 hp, Figure 1.146 bh). It is formed on the flat embryonic

shield of the amniotes from the upper plates of the parietal zone. The

right and left parietal plates bend downwards towards each other, and

grow round the gut in the same way as the gut itself closes. The outer

part of the lateral plates forms the ventral wall or the lower wall of

the body, the two lateral plates bending considerably on the inner

side of the amniotic fold, and growing towards each other from right

and left. While the alimentary canal is closing, the body-wall also

closes on all sides. Hence the ventral wall, which encloses the whole

ventral cavity below, consists of two parts, two lateral plates that

bend towards each other. These approach each other all along, and at

last meet at the navel. We ought, therefore, really to distinguish two

navels, an inner and an outer one. The internal or intestinal navel is

the definitive point of the closing of the gut wall, which puts an end

to the open communication between the ventral cavity and the cavity of

the yelk-sac (Figure 1.105). The external navel in the skin is the

definitive point of the closing of the ventral wall; this is visible

in the developed body as a small depression.

(FIGURE 1.147. Median longitudinal section of the embryo of a chick

(fifth day of incubation), seen from the right side (head to the

right, tail to the left). Dorsal body dark, with convex outline. d

gut, o mouth, a anus, l lungs, h liver, g mesentery, v auricle of the

heart, k ventricle of the heart, b arch of the arteries, t aorta, c

yelk-sac, m vitelline (yelk) duct, u allantois, r pedicle (stalk) of

the allantois, n amnion, w amniotic cavity, s serous membrane. (From

Baer.))

With the formation of the internal navel and the closing of the

alimentary canal is connected the formation of two cavities, which we

call the capital and the pelvic sections of the visceral cavity. As

the embryonic shield lies flat on the wall of the embryonic vesicle at

first, and only gradually separates from it, its fore and hind ends

are independent in the beginning; on the other hand, the middle part

of the ventral surface is connected with the yelk-sac by means of the

vitelline or umbilical duct (Figure 1.147 m). This leads to a notable

curving of the dorsal surface; the head-end bends downwards towards

the breast and the tail-end towards the belly. We see this very

clearly in the excellent old diagrammatic illustration given by Baer

(Figure 1.147), a median longitudinal section of the embryo of the

chick, in which the dorsal body or episoma is deeply shaded. The

embryo seems to be trying to roll up, like a hedgehog protecting

itself from its pursuers. This pronounced curve of the back is due to

the more rapid growth of the convex dorsal surface, and is directly

connected with the severance of the embryo from the yelk-sac. The

further bending of the embryo leads to the formation of the

"head-cavity" of the gut (Figure 1.148 above D) and a similar one at

the tail, known as its "pelvic cavity."

As a result of these processes the embryo attains a shape that may be

compared to a wooden shoe, or, better still, to an overturned canoe.

Imagine a canoe or boat with both ends rounded and a small covering

before and behind; if this canoe is turned upside down, so that the

curved keel is uppermost, we have a fair picture of the canoe-shaped

embryo (Figure 1.147). The upturned convex keel corresponds to the

middle line of the back; the small chamber underneath the fore-deck

represents the capital cavity, and the small chamber under the

rear-deck the pelvic chamber of the gut (cf. Figure 1.140).

The embryo now, as it were, presses into the outer surface of the

embryonic vesicle with its free ends, while it moves away from it with

its middle part. As a result of this change the yelk-sac becomes

henceforth only a pouch-like outer appendage at the middle of the

ventral wall. The ventral appendage, growing smaller and smaller, is

afterwards called the umbilical (navel) vesicle. The cavity of the

yelk-sac or umbilical vesicle communicates with the corresponding

visceral cavity by a wide opening, which gradually contracts into a

narrow and long canal, the vitelline (yelk) duct (ductus vitellinus,

Figure 1.147 m). Hence, if we were to imagine ourselves in the cavity

of the yelk-sac, we could get from it through the yelk-duct into the

middle and still wide open part of the alimentary canal. If we were to

go forward from there into the head-part of the embryo, we should

reach the capital cavity of the gut, the fore-end of which is closed

up.

The reader will ask: "Where are the mouth and the anus?" These are not

at first present in the embryo. The whole of the primitive gut-cavity

is completely closed, and is merely connected in the middle by the

vitelline duct with the equally closed cavity of the embryonic vesicle

(Figure 1.140). The two later apertures of the alimentary canal--the

anus and the mouth--are secondary constructions, formed from the outer

skin. In the horn-plate, at the spot where the mouth is found

subsequently, a pit-like depression is formed, and this grows deeper

and deeper, pushing towards the blind fore-end of the capital cavity;

this is the mouth-pit. In the same way, at the spot in the outer skin

where the anus is afterwards situated a pit-shaped depression appears,

grows deeper and deeper, and approaches the blind hind-end of the

pelvic cavity; this is the anus-pit. In the end these pits touch with

their deepest and innermost points the two blind ends of the primitive

alimentary canal, so that they are now only separated from them by

thin membranous partitions. This membrane finally disappears, and

henceforth the alimentary canal opens in front at the mouth and in the

rear by the anus (Figures 1.141 and 1.147). Hence at first, if we

penetrate into these pits from without, we find a partition cutting

them off from the cavity of the alimentary canal, which gradually

disappears. The formation of mouth and anus is secondary in all the

vertebrates.

(FIGURE 1.148. Longitudinal section of the fore half of a chick-embryo

at the end of the first day of incubation (seen from the left side). k

head-plates, ch chorda. Above it is the blind fore-end of the ventral

tube (m); below it the capital cavity of the gut. d gut-gland layer,

df gut-fibre layer, h horn plate, hh cavity of the heart, hk

heart-capsule, ks head-sheath, kk head-capsule. (From Remak.))

During the important processes which lead to the formation of the

navel, and of the intestinal wall and ventral wall, we find a number

of other interesting changes taking place in the embryonic shield of

the amniotes. These relate chiefly to the prorenal ducts and the first

blood-vessels. The prorenal (primitive kidney) ducts, which at first

lie quite flat under the horn-plate or epiderm (Figure 1.93 ung), soon

back towards each other in consequence of special growth movements

(Figures 1.143 to 1.145 ung). They depart more and more from their

point of origin, and approach the gut-gland layer. In the end they lie

deep in the interior, on either side of the mesentery, underneath the

chorda, (Figure 1.145 ung). At the same time, the two primitive aortas

change their position (cf. Figures 1.138 to 1.145 ao); they travel

inwards underneath the chorda, and there coalesce at last to form a

single secondary aorta, which is found under the rudimentary vertebral

column (Figure 1.145 ao). The cardinal veins, the first venous

blood-vessels, also back towards each other, and eventually unite

immediately above the rudimentary kidneys (Figures 1.145 vc, 152 cav).

In the same spot, at the inner side of the fore-kidneys, we soon see

the first trace of the sexual organs. The most important part of this

apparatus (apart from all its appendages) is the ovary in the female

and the testicle in the male. Both develop from a small part of the

cell-lining of the body-cavity, at the spot where the skin-fibre layer

and gut-fibre layer touch. The connection of this embryonic gland with

the prorenal ducts, which lie close to it and assume most important

relations to it, is only secondary.

(FIGURE 1.149. Longitudinal section of a human embryo of the fourth

week, one-fifth of an inch long, magnified fifteen times. Showing:

bend of skull, yelk-sac, umbilical cord, terminal gut, rudimentary

kidneys, mesoderm, head-gut (with gill-clefts), primitive lungs,

liver, stomach, pancreas, mesentery, primitive kidneys, allantoic

duct, rectum. (From Kollmann.)

FIGURE 1.150. Transverse section of a human embryo of fourteen days.

mr medullary tube, ch chorda. vu umbilical vein, mt myotome, mp middle

plate, ug prorenal duct, lh body-cavity, e ectoderm, bh ventral skin,

hf skin-fibre layer, df gut-fibre layer. (From Kollmann.)

FIGURE 1.151. Transverse section of a shark-embryo (or young

selachius). mr medullary tube, ch chorda, a aorta, d gut, vp principal

(or subintestinal) vein, mt myotome, mm muscular mass of the

provertebra, mp middle plate, ug prorenal duct, lh body-cavity, e

ectoderm of the rudimentary extremities, mz mesenchymic cells, z point

where the myotome and nephrotome separate. (From H.E. Ziegler.)

FIGURE 1.152. Transverse section of a duck-embryo with twenty-four

primitive segments. (From Balfour.) From a dorsal lateral joint of the

medullary tube (spc) the spinal ganglia (spg) grow out between it and

the horn-plate. ch chorda, ao double aorta, hy gut-gland layer, sp

gut-fibre layer, with blood-vessels in section, ms muscle plate, in

the dorsal wall of the myocoel (episomite). Below the cardinal vein

(cav) is the prorenal duct (wd) and a segmental prorenal canal (st).

The skin-fibre layer of the body-wall (so) is continued in the

amniotic fold (am). Between the four secondary germinal layers and the

structures formed from them there is formed embryonic connective

matter with stellate cells and vascular structures (Hertwig's

"mesenchym").)

The Evolution of Man — Volume 1 · 第 33 / 35 章