[26] The term articulation is used in this chapter to denote both “segmentation” and “articulation” in the ordinary sense.—Translator.

The vertebrate stem, to which our race belongs as one of the latest and most advanced outcomes of the natural development of life, is rightly placed at the head of the animal kingdom. This privilege must be accorded to it, not only because man does in point of fact soar far above all other animals, and has been lifted to the position of “lord of creation”; but also because the vertebrate organism far surpasses all the other animal-stems in size, in complexity of structure, and in the advanced character of its functions. From the point of view of both anatomy and physiology, the vertebrate stem outstrips all the other, or invertebrate, animals.

There is only one among the twelve stems of the animal kingdom that can in many respects be compared with the vertebrates, and reaches an equal, if not a greater, importance in many points. This is the stem of the articulates, composed of three classes: 1, the annelids (earth-worms, leeches, and cognate forms); 2, the crustacea (crabs, etc.); 3, the tracheata (spiders, insects, etc.). The stem of the articulates is superior not only to the vertebrates, but to all other animal-stems, in variety of forms, number of species, elaborateness of individuals, and general importance in the economy of nature.

When we have thus declared the vertebrates and the articulates to be the most important and most advanced of the twelve stems of the animal kingdom, the question arises whether this special position is accorded to them on the ground of a peculiarity of organisation that is common to the two. The answer is that this is really the case; it is their segmental or transverse articulation, which we may briefly call metamerism. In all the vertebrates and articulates the developed individual consists of a series of successive members (segments or metamera = “parts”); in the embryo these are called primitive segments or somites. In each of these segments we have a certain group of organs reproduced in the same arrangement, so that we may regard each segment as an individual unity, or a special “individual” subordinated to the entire personality.

The similarity of their segmentation, and the consequent physiological advance in the two stems of the vertebrates and articulates, has led to the assumption of a direct affinity between them, and an attempt to derive the former directly from the latter. The annelids were supposed to be the direct ancestors, not only of the crustacea and tracheata, but also of the vertebrates. We shall see later (Chapter XX) that this annelid theory of the vertebrates is entirely wrong, and ignores the most important differences in the organisation of the two stems. The internal articulation of the vertebrates is just as profoundly different from the external metamerism of the articulates as are their skeletal structure, nervous system, vascular system, and so on. The articulation has been developed in a totally different way in the two stems. The unarticulated chordula (Figs. 83–86), which we have recognised as one of the chief palingenetic embryonic forms of the vertebrate group, and from which we have inferred the existence of a corresponding ancestral form for all the vertebrates and tunicates, is quite unthinkable as the stem-form of the articulates.

All articulated animals came originally from unarticulated ones. This phylogenetic principle is as firmly established as the ontogenetic fact that every articulated animal-form develops from an unarticulated embryo. But the organisation of the embryo is totally different in the two stems. The chordula-embryo of all the vertebrates is characterised by the dorsal medullary tube, the neurenteric canal, which passes at the primitive mouth into the alimentary canal, and the axial chorda between the two. None of the articulates, either annelids or arthropods (crustacea and tracheata), show any trace of this type of organisation. Moreover, the development of the chief systems of organs proceeds in the opposite way in the two stems. Hence the segmentation must have arisen independently in each. This is not at all surprising; we find analogous cases in the stalk-articulation of the higher plants and in several groups of other animal stems.

The characteristic internal articulation of the vertebrates and its importance in the organisation of the stem are best seen in the study of the skeleton. Its chief and central part, the cartilaginous or bony vertebral column, affords an obvious instance of vertebrate metamerism; it consists of a series of cartilaginous or bony pieces, which have long been known as vertebrÆ (or spondyli). Each vertebra is directly connected with a special section of the muscular system, the nervous system, the vascular system, etc. Thus most of the “animal organs” take part in this vertebration. But we saw, when we were considering our own vertebrate character (in Chapter XI), that the same internal articulation is also found in the lowest primitive vertebrates, the acrania, although here the whole skeleton consists merely of the simple chorda, and is not at all articulated. Hence the articulation does not proceed primarily from the skeleton, but from the muscular system, and is clearly determined by the more advanced swimming-movements of the primitive chordonia-ancestors.

It is, therefore, wrong to describe the first rudimentary segments in the vertebrate embryo as primitive vertebrÆ or provertebrÆ; the fact that they have been so called for some time has led to much error and misunderstanding. Hence we shall give the name of “somites” or primitive segments to these so-called “primitive vertebrÆ.” If the latter name is retained at all, it should only be used of the sclerotom—i.e., the small part of the somites from which the later vertebra does actually develop.

Articulation begins in all vertebrates at a very early embryonic stage, and this indicates the considerable phylogenetic age of the process. When the chordula (Figs. 83–86) has completed its characteristic composition, often even a little earlier, we find in the amniotes, in the middle of the sole-shaped embryonic shield, several pairs of dark square spots, symmetrically distributed on both sides of the chorda (Figs. 131–135).Transverse sections (Fig. 93 uw) show that they belong to the stem-zone (episoma) of the mesoderm, and are separated from the parietal zone (hyposoma) by the lateral folds; in section they are still quadrangular, almost square, so that they look something like dice. These pairs of “cubes” of the mesoderm are the first traces of the primitive segments or somites, the so-called “protovertebrÆ.” (Figs. 153–155 uw).

Among the mammals the embryos of the marsupials have three pairs of somites (Fig. 131) after sixty hours, and eight pairs after seventy-two hours (Fig. 135). They develop more slowly in the embryo of the rabbit; this has three somites on the eighth day (Fig. 132), and eight somites a day later (Fig. 134). In the incubated hen’s egg the first somites make their appearance thirty hours after incubation begins (Fig. 153). At the end of the second day the number has risen to sixteen or eighteen (Fig. 155). The articulation of the stem-zone, to which the somites owe their origin, thus proceeds briskly from front to rear, new transverse constrictions of the “protovertebral plates” forming continuously and successively. The first segment, which is almost half-way down in the embryonic shield of the amniote, is the foremost of all; from this first somite is formed the first cervical vertebra with its muscles and skeletal parts. It follows from this, firstly, that the multiplication of the primitive segments proceeds backwards from the front, with a constant lengthening of the hinder end of the body; and, secondly, that at the beginning of segmentation nearly the whole of the anterior half of the sole-shaped embryonic shield of the amniote belongs to the later head, while the whole of the rest of the body is formed from its hinder half. We are reminded that in the amphioxus (and in our hypothetic primitive vertebrate, Figs. 98–102) nearly the whole of the fore half corresponds to the head, and the hind half to the trunk.

The number of the metamera, and of the embryonic somites or primitive segments from which they develop, varies considerably in the vertebrates, according as the hind part of the body is short or is lengthened by a tail. In the developed man the trunk (including the rudimentary tail) consists of thirty-three metamera, the solid centre of which is formed by that number of vertebrÆ in the vertebral column (seven cervical, twelve dorsal, five lumbar, five sacral, and four caudal). To these we must add at least nine head-vertebrÆ, which originally (in all the craniota) constitute the skull. Thus the total number of the primitive segments of the human body is raised to at least forty-two; it would reach forty-five to forty-eight if (according to recent investigations) the number of the original segments of the skull is put at twelve to fifteen. In the tailless or anthropoid apes the number of metamera is much the same as in man, only differing by one or two; but it is much larger in the long-tailed apes and most of the other mammals. In long serpents and fishes it reaches several hundred (sometimes 400).

In order to understand properly the real nature and origin of articulation in the human body and that of the higher vertebrates, it is necessary to compare it with that of the lower vertebrates, and bear in mind always the genetic connection of all the members of the stem. In this the simple development of the invaluable amphioxus once more furnishes the key to the complex and cenogenetically modified embryonic processes of the craniota. The articulation of the amphioxus begins at an early stage—earlier than in the craniotes. The two coelom-pouches have hardly grown out of the primitive gut (Fig. 156 c) when the blind fore part of it (farthest away from the primitive mouth, u) begins to separate by a transverse fold (s): this is the first primitive segment. Immediately afterwards the hind part of the coelom-pouches begins to divide into a series of pieces by new transverse folds (Fig. 157). The foremost of these primitive segments (us1) is the first and oldest; in Figs. 124 and 157 there are already five formed. They separate so rapidly, one behind the other, that eight pairs are formed within twenty-four hours of the beginning of development, and seventeen pairs twenty-four hours later. The number increases as the embryo grows and extends backwards, and new cells are formed constantly (at the primitive mouth) from the two primitive mesodermic cells (Figs. 159–160).

This typical articulation of the two coelom-sacs begins very early in the lancelet, before they are yet severed from the primitive gut, so that at first each segment-cavity (us) still communicates by a narrow opening with the gut, like an intestinal gland. But this opening soon closes by complete severance, proceeding regularly backwards. The closed segments then extend more, so that their upper half grows upwards like a fold between the ectoderm (ak) and neural tube (n), and the lower half between the ectoderm and alimentary canal (ch; Fig. 82 d, left half of the figure). Afterwards the two halves completely separate, a lateral longitudinal fold cutting between them (mk, right half of Fig. 82). The dorsal segments (sd) provide the muscles of the trunk the whole length of the body (159): this cavity afterwards disappears. On the other hand, the ventral parts give rise, from their uppermost section, to the pronephridia or primitive-kidney canals, and from the lower to the segmental rudiments of the sexual glands or gonads. The partitions of the muscular dorsal pieces (myotomes) remain, and determine the permanent articulation of the vertebrate organism. But the partitions of the large ventral pieces (gonotomes) become thinner, and afterwards disappear in part, so that their cavities run together to form the metacoel, or the simple permanent body-cavity.

The articulation proceeds in substantially the same way in the other vertebrates, the craniota, starting from the coelom-pouches. But whereas in the former case there is first a transverse division of the coelom-sacs (by vertical folds) and then the dorso-ventral division, the procedure is reversed in the craniota; in their case each of the long coelom-pouches first divides into a dorsal (primitive segment plates) and a ventral (lateral plates) section by a lateral longitudinal fold. Only the former are then broken up into primitive segments by the subsequent vertical folds; while the latter (segmented for a time in the amphioxus) remain undivided, and, by the divergence of their parietal and visceral plates, form a body-cavity that is unified from the first. In this case, again, it is clear that we must regard the features of the younger craniota as cenogenetically modified processes that can be traced palingenetically to the older acrania.

We have an interesting intermediate stage between the acrania and the fishes in these and many other respects in the cyclostoma (the hag and the lamprey, cf. Chapter XXI).

Among the fishes the selachii, or primitive fishes, yield the most important information on these and many other phylogenetic questions (Figs. 161 and 162). The careful studies of RÜckert, Van Wijhe, H. E. Ziegler, and others, have given us most valuable results. The products of the middle germinal layer are partly clear in these cases at the period when the dorsal primitive segment cavities (or myocoels, h) are still connected with the ventral body-cavity (lh; Fig. 161). In Fig. 162, a somewhat older embryo, these cavities are separated. The outer or lateral wall of the dorsal segment yields the cutis-plate (cp), the foundation of the connective corium. From its inner or median wall are developed the muscle-plate (mp, the rudiment of the trunk-muscles) and the skeletal plate, the formative matter of the vertebral column (sk).

In the amphibia, also, especially the water-salamander (Triton), we can observe very clearly the articulation of the coelom-pouches and the rise of the primitive segments from their dorsal half (cf. Fig. 91, A, B, C). A horizontal longitudinal section of the salamander-embryo (Fig. 163) shows very clearly the series of pairs of these vesicular dorsal segments, which have been cut off on each side from the ventral side-plates, and lie to the right and left of the chorda.

The metamerism of the amniotes agrees in all essential points with that of the three lower classes of vertebrates we have considered; but it varies considerably in detail, in consequence of cenogenetic disturbances that are due in the first place (like the degeneration of the coelom-pouches) to the large development of the food-yelk. As the pressure of this seems to force the two middle layers together from the start, and as the solid structure of the mesoderm apparently belies the original hollow character of the sacs, the two sections of the mesoderm, which are at that time divided by the lateral fold—the dorsal segment-plates and ventral side-plates—have the appearance at first of solid layers of cells (Figs. 94–97). And when the articulation of the somites begins in the sole-shaped embryonic shield, and a couple of protovertebrÆ are developed in succession, constantly increasing in number towards the rear, these cube-shaped somites (formerly called protovertebrÆ, or primitive vertebrÆ) have the appearance of solid dice, made up of mesodermic cells (Fig. 93). Nevertheless, there is for a time a ventral cavity, or provertebral cavity, even in these solid “protovertebrÆ” (Fig. 143 uwh). This vesicular condition of the provertebra is of the greatest phylogenetic interest; we must, according to the coelom theory, regard it as an hereditary reproduction of the hollow dorsal somites of the amphioxus (Figs. 156–160) and the lower vertebrates (Fig. 161–163). This rudimentary “provertebral cavity” has no physiological significance whatever in the amniote-embryo; it soon disappears, being filled up with cells of the muscular plate.

The innermost median part of the primitive segment plates, which lies immediately on the chorda (Fig. 145 ch) and the medullary tube (m), forms the vertebral column in all the higher vertebrates (it is wanting in the lowest); hence it may be called the skeleton plate. In each of the provertebrÆ it is called the “sclerotome” (in opposition to the outlying muscular plate, the “myotome”). From the phylogenetic point of view the myotomes are much older than the sclerotomes. The lower or ventral part of each sclerotome (the inner and lower edge of the cube-shaped provertebra) divides into two plates, which grow round the chorda, and thus form the foundation of the body of the vertebra (wh). The upper plate presses between the chorda and the medullary tube, the lower between the chorda and the alimentary canal (Fig. 137 C). As the plates of two opposite provertebral pieces unite from the right and left, a circular sheath is formed round this part of the chorda. From this develops the body of a vertebra—that is to say, the massive lower or ventral half of the bony ring, which is called the “vertebra” proper and surrounds the medullary tube (Figs. 164–166). The upper or dorsal half of this bony ring, the vertebral arch (Fig. 145 wb), arises in just the same way from the upper part of the skeletal plate, and therefore from the inner and upper edge of the cube-shaped primitive vertebra. As the upper edges of two opposing somites grow together over the medullary tube from right and left, the vertebra-arch becomes closed.

The whole of the secondary vertebra, which is thus formed from the union of the skeletal plates of two provertebral pieces and encloses a part of the chorda in its body, consists at first of a rather soft mass of cells; this afterwards passes into a firmer, cartilaginous stage, and finally into a third, permanent, bony stage. These three stages can generally be distinguished in the greater part of the skeleton of the higher vertebrates; at first most parts of the skeleton are soft, tender, and membranous; they then become cartilaginous in the course of their development, and finally bony.

At the head part of the embryo in the amniotes there is not generally a cleavage of the middle germinal layer into provertebral and lateral plates, but the dorsal and ventral somites are blended from the first, and form what are called the “head-plates” (Fig. 148 k). From these are formed the skull, the bony case of the brain, and the muscles and corium of the body. The skull develops in the same way as the membranous vertebral column. The right and left halves of the head curve over the cerebral vesicle, enclose the foremost part of the chorda below, and thus finally form a simple, soft, membranous capsule about the brain. This is afterwards converted into a cartilaginous primitive skull, such as we find permanently in many of the fishes. Much later this cartilaginous skull becomes the permanent bony skull with its various parts. The bony skull in man and all the other amniotes is more highly differentiated and modified than that of the lower vertebrates, the amphibia and fishes. But as the one has arisen phylogenetically from the other, we must assume that in the former no less than the latter the skull was originally formed from the sclerotomes of a number of (at least nine) head-somites.

While the articulation of the vertebrate body is always obvious in the episoma or dorsal body, and is clearly expressed in the segmentation of the muscular plates and vertebrÆ, it is more latent in the hyposoma or ventral body. Nevertheless, the hyposomites of the vegetal half of the body are not less important than the episomites of the animal half. The segmentation in the ventral cavity affects the following principal systems of organs: 1, the gonads or sex-glands (gonotomes); 2, the nephridia or kidneys (nephrotomes); and 3, the head-gut with its gill-clefts (branchiotomes).

The metamerism of the hyposoma is less conspicuous because in all the craniotes the cavities of the ventral segments, in the walls of which the sexual products are developed, have long since coalesced, and formed a single large body-cavity, owing to the disappearance of the partition. This cenogenetic process is so old that the cavity seems to be unsegmented from the first in all the craniotes, and the rudiment of the gonads also is almost always unsegmented. It is the more interesting to learn that, according to the important discovery of RÜckert, this sexual structure is at first segmental even in the actual selachii, and the several gonotomes only blend into a simple sexual gland on either side secondarily.

Amphioxus, the sole surviving representative of the acrania, once more yields us most interesting information; in this case the sexual glands remain segmented throughout life. The sexually mature lancelet has, on the right and left of the gut, a series of metamerous sacs, which are filled with ova in the female and sperm in the male. These segmental gonads are originally nothing else than the real gonotomes, separate body-cavities, formed from the hyposomites of the trunk.

The gonads are the most important segmental organs of the hyposoma, in the sense that they are phylogenetically the oldest. We find sexual glands (as pouch-like appendages of the gastro-canal system) in most of the lower animals, even in the medusÆ, etc., which have no kidneys. The latter appear first (as a pair of excretory tubes) in the platodes (turbellaria), and have probably been inherited from these by the articulates (annelids) on the one hand and the unarticulated prochordonia on the other, and from these passed to the articulated vertebrates. The oldest form of the kidney system in this stem are the segmental pronephridia or prorenal canals, in the same arrangement as Boveri found them in the amphioxus. They are small canals that lie in the frontal plane, on each side of the chorda, between the episoma and hyposoma (Fig. 102 n); their internal funnel-shaped opening leads into the various body-cavities, their outer opening is the lateral furrow of the epidermis. Originally they must have had a double function, the carrying away of the urine from the episomites and the release of the sexual cells from the hyposomites.

The recent investigations of Ruckert and Van Wijhe on the mesodermic segments of the trunk and the excretory system of the selachii show that these “primitive fishes” are closely related to the amphioxus in this further respect. The transverse section of the shark-embryo in Fig. 161 shows this very clearly.

In other higher vertebrates, also, the kidneys develop (though very differently formed later on) from similar structures, which have been secondarily derived from the segmental pronephridia of the acrania. The parts of the mesoderm at which the first traces of them are found are usually called the middle or mesenteric plates. As the first traces of the gonads make their appearance in the lining of these middle plates nearer inward (or the middle) from the inner funnels of the nephro-canals, it is better to count this part of the mesoderm with the hyposoma.

The chief and oldest organ of the vertebrate hyposoma, the alimentary canal, is generally described as an unsegmented organ. But we could just as well say that it is the oldest of all the segmented organs of the vertebrate; the double row of the coelom-pouches grows out of the dorsal wall of the gut, on either side of the chorda. In the brief period during which these segmental coelom-pouches are still openly connected with the gut, they look just like a double chain of segmented visceral glands. But apart from this, we have originally in all vertebrates an important articulation of the fore-gut, that is wanting in the lower gut, the segmentation of the branchial (gill) gut.

The gill-clefts, which originally in the older acrania pierced the wall of the fore-gut, and the gill-arches that separated them, were presumably also segmental, and distributed among the various metamera of the chain, like the gonads in the after-gut and the nephridia. In the amphioxus, too, they are still segmentally formed. Probably there was a division of labour of the hyposomites in the older (and long extinct) acrania, in such wise that those of the fore-gut took over the function of breathing and those of the after-gut that of reproduction. The former developed into gill-pouches, the latter into sex-pouches. There may have been primitive kidneys in both. Though the gills have lost their function in the higher animals, certain parts of them have been generally maintained in the embryo by a tenacious heredity. At a very early stage we notice in the embryo of man and the other amniotes, at each side of the head, the remarkable and important structures which we call the gill-arches and gill-clefts (Figs. 167–170 f). They belong to the characteristic and inalienable organs of the amniote-embryo, and are found always in the same spot and with the same arrangement and structure. There are formed to the right and left in the lateral wall of the fore-gut cavity, in its foremost part, first a pair and then several pairs of sac-shaped inlets, that pierce the whole thickness of the lateral wall of the head. They are thus converted into clefts, through which one can penetrate freely from without into the gullet. The wall thickens between these branchial folds, and changes into an arch-like or sickle-shaped piece—the gill, or gullet-arch. In this the muscles and skeletal parts of the branchial gut separate; a blood-vessel arch rises afterwards on their inner side (Fig. 98 ka). The number of the branchial arches and the clefts that alternate with them is four or five on each side in the higher vertebrates (Fig. 170 d, f, f, f). In some of the fishes (selachii) and in the cyclostoma we find six or seven of them permanently.

These remarkable structures had originally the function of respiratory organs—gills. In the fishes the water that serves for breathing, and is taken in at the mouth, still always passes out by the branchial clefts at the sides of the gullet. In the higher vertebrates they afterwards disappear. The branchial arches are converted partly into the jaws, partly into the bones of the tongue and the ear. From the first gill-cleft is formed the tympanic cavity of the ear.

There are few parts of the vertebrate organism that, like the outer covering or integument of the body, are not subject to metamerism. The outer skin (epidermis) is unsegmented from the first, and proceeds from the continuous horny plate. Moreover, the underlying cutis is also not metamerous, although it develops from the segmental structure of the cutis-plates (Figs. 161, 162 cp). The vertebrates are strikingly and profoundly different from the articulates in these respects also.

Further, most of the vertebrates still have a number of unarticulated organs, which have arisen locally, by adaptation of particular parts of the body to certain special functions. Of this character are the sense-organs in the episoma, and the limbs, the heart, the spleen, and the large visceral glands—lungs, liver, pancreas, etc.—in the hyposoma. The heart is originally only a local spindle-shaped enlargement of the large ventral blood-vessel or principal vein, at the point where the subintestinal passes into the branchial artery, at the limit of the head and trunk (Figs. 170, 171). The three higher sense-organs—nose, eye, and ear—were originally developed in the same form in all the craniotes, as three pairs of small depressions in the skin at the side of the head.

The organ of smell, the nose, has the appearance of a pair of small pits above the mouth-aperture, in front of the head (Fig. 169 n). The organ of sight, the eye, is found at the side of the head, also in the shape of a depression (Figs. 169 l, 170 b), to which corresponds a large outgrowth of the foremost cerebral vesicle on each side. Farther behind, at each side of the head, there is a third depression, the first trace of the organ of hearing (Fig. 169 g). As yet we can see nothing of the later elaborate structure of these organs, nor of the characteristic build of the face.

When the human embryo has reached When the human embryo has reached this stage of development, it can still scarcely be distinguished from that of any other higher vertebrate. All the chief parts of the body are now laid down: the head with the primitive skull, the rudiments of the three higher sense-organs and the five cerebral vesicles, and the gill-arches and clefts; the trunk with the spinal cord, the rudiment of the vertebral column, the chain of metamera, the heart and chief blood-vessels, and the kidneys. At this stage man is a higher vertebrate, but shows no essential morphological difference from the embryos of the mammals, the birds, the reptiles, etc. This is an ontogenetic fact of the utmost significance. From it we can gather the most important phylogenetic conclusions.

There is still no trace of the limbs. Although head and trunk are separated and all the principal internal organs are laid down, there is no indication whatever of the “extremities” at this stage; they are formed later on. Here again we have a fact of the utmost interest. It proves that the older vertebrates had no feet, as we find to be the case in the lowest living vertebrates (amphioxus and the cyclostoma). The descendants of these ancient footless vertebrates only acquired extremities—two fore-legs and two hind-legs—at a much later stage of development. These were at first all alike, though they afterwards vary considerably in structure—becoming fins (of breast and belly) in the fishes, wings and legs in the birds, fore and hind legs in the creeping animals, arms and legs in the apes and man. All these parts develop from the same simple original structure, which forms secondarily from the trunk-wall (Figs. 172, 173). They have always the appearance of two pairs of small buds, which represent at first simple roundish knobs or plates. Gradually each of these plates becomes a large projection, in which we can distinguish a small inner part and a broader outer part. The latter is the rudiment of the foot or hand, the former that of the leg or arm. The similarity of the original rudiment of the limbs in different groups of vertebrates is very striking.

How the five fingers or toes with their blood-vessels gradually differentiate within the simple fin-like structure of the limbs can be seen in the instance of the lizard in Fig. 174. They are formed in just the same way in man: in the human embryo of five weeks the five fingers can clearly be distinguished within the fin-plate (Fig. 175).

The careful study and comparison of human embryos with those of other vertebrates at this stage of development is very instructive, and reveals more mysteries to the impartial student than all the religions in the world put together. For instance, if we compare attentively the three successive stages of development that are represented, in twenty different amniotes we find a remarkable likeness. When we see that as a fact twenty different amniotes of such divergent characters develop from the same embryonic form, we can easily understand that they may all descend from a common ancestor.

In the first stage of development, in which the head with the five cerebral vesicles is already clearly indicated, but there are no limbs, the embryos of all the vertebrates, from the fish to man, are only incidentally or not at all different from each other. In the second stage, which shows the limbs, we begin to see differences between the embryos of the lower and higher vertebrates; but the human embryo is still hardly distinguishable from that of the higher mammals. In the third stage, in which the gill-arches have disappeared and the face is formed, the differences become more pronounced. These are facts of a significance that cannot be exaggerated.^[27]

[27] Because they show how the most diverse structures may be developed from a common form. As we actually see this in the case of the embryos, we have a right to assume it in that of the stem-forms. Nevertheless, this resemblance, however great, is never a real identity. Even the embryos of the different individuals of one species are usually not really identical. If the reader can consult the complete edition of this work at a library, he will find six plates illustrating these twenty embryos.

If there is an intimate causal connection between the processes of embryology and stem-history, as we must assume in virtue of the laws of heredity, several important phylogenetic conclusions follow at once from these ontogenetic facts. The profound and remarkable similarity in the embryonic development of man and the other vertebrates can only be explained when we admit their descent from a common ancestor. As a fact, this common descent is now accepted by all competent scientists; they have substituted the natural evolution for the supernatural creation of organisms.