The structural features that distinguish the vertebrates from the invertebrates are so prominent that there was the greatest difficulty in the earlier stages of classification in determining the affinity of these two great groups. When scientists began to speak of the affinity of the various animal groups in more than a figurative—in a genealogical—sense, this question came at once to the front, and seemed to constitute one of the chief obstacles to the carrying-out of the evolutionary theory. Even earlier, when they had studied the relations of the chief groups, without any idea of real genealogical connection, they believed they had found here and there among the invertebrates points of contact with the vertebrates: some of the worms, especially, seemed to approach the vertebrates in structure, such as the marine arrow-worm (Sagitta). But on closer study the analogies proved untenable. When Darwin gave an impulse to the construction of a real stem-history of the animal kingdom by his reform of the theory of evolution, the solution of this problem was found to be particularly difficult. When I made the first attempt in my General Morphology (1866) to work out the theory and apply it to classification, I found no problem of phylogeny that gave me so much trouble as the linking of the vertebrates with the invertebrates.

But just at this time the true link was discovered, and at a point where it was least expected. Towards the end of 1866 two works of the Russian zoologist, Kowalevsky, who had lived for some time at Naples, and studied the embryology of the lower animals, were issued in the publications of the St. Petersburg Academy. A fortunate accident had directed the attention of this able observer almost simultaneously to the embryology of the lowest vertebrate, the Amphioxus, and that of an invertebrate, the close affinity of which to the Amphioxus had been least suspected, the Ascidia. To the extreme astonishment of all zoologists who were interested in this important question, there turned out to be the utmost resemblance in structure from the commencement of development between these two very different animals—the lowest vertebrate and the mis-shaped, sessile invertebrate. With this undeniable identity of ontogenesis, which can be demonstrated to an astounding extent, we had, in virtue of the biogenetic law, discovered the long-sought genealogical link, and definitely identified the invertebrate group that represents the nearest blood-relatives of the vertebrates.

[ 192 ]

The discovery was confirmed by other zoologists, and there can no longer be any doubt that of all the classes of invertebrates that of the Tunicates is most closely related to the vertebrates, and of the Tunicates the nearest are the Ascidiæ. We cannot say that the vertebrates are descended from the Ascidiæ—and still less the reverse—but we can say that of all the invertebrates it is the Tunicates, and, within this group, the Ascidiæ, that are the nearest blood-relatives of the ancient stem-form of the vertebrates. We must assume as the common ancestral group of both stems an extinct family of the extensive vermalia-stem, the Prochordonia or Prochordata (“primitive chorda-animals”).

In order to appreciate fully this remarkable fact, and especially to secure the sound basis we seek for the genealogical tree of the vertebrates, it is necessary to study thoroughly the embryology of both these animals, and compare the individual development of the Amphioxus step by step with that of the Ascidia. We begin with the ontogeny of the Amphioxus.

From the concordant observations of Kowalevsky at Naples and Hatschek at Messina, it follows, firstly, that the ovum-segmentation and gastrulation of the Amphioxus are of the simplest character. They take place in the same way as we find them in many of the lower animals of different invertebrate stems, which we have already described as original or primordial; the development of the Ascidia is of the same type. Sexually mature specimens of the Amphioxus, which are found in great quantities at Messina from April or May onwards, begin as a rule to eject their sexual products in the evening; if you catch them about the middle of a warm night and put them in a glass vessel with seawater, they immediately eject through the mouth their accumulated sexual products, in consequence of the disturbance. The males give out masses of sperm, and the females discharge ova in such quantity that many of them stick to the fibrils about their mouths. Both kinds of cells pass first into the mantle-cavity after the opening of the gonads, proceed through the gill-clefts into the branchial gut, and are discharged from this through the mouth.

The ova are simply round cells. They are only 1/250 of an inch in diameter, and thus are only half the size of the mammal ova, and have no distinctive features. The clear protoplasm of the mature ovum is made so turbid by the numbers of dark granules of food-yelk or deutoplasm scattered in it that it is difficult to follow the process of fecundation and the behaviour of the two nuclei during it (p. 51). The active elements of the male sperm, the cone-shaped spermatozoa, are similar to those of most other animals (cf. Fig. 20). Fecundation takes place when these lively ciliated cells of the sperm approach the ovum, and seek to penetrate into the yelk-matter or the cellular substance of the ovum with their head-part—the thicker part of the cell that encloses the nucleus. Only one spermatozoon can bore its way into the yelk at one pole of the ovum-axis; its head or nucleus coalesces with the female nucleus, which remains after the extrusion of the directive bodies from the germinal vesicle. Thus is formed the “stem-nucleus,” or the nucleus of the “stem-cell” (cytula, Fig. 2). This now undergoes total segmentation, dividing into two, four, eight, sixteen, thirty-two cells, and so on. In this way we get the spherical, mulberry-shaped body, which we call the morula.

The segmentation of the Amphioxus is not entirely regular, as was supposed after the first observations of Kowalevsky (1866). It is not completely equal, but a little unequal. As Hatschek afterwards found (1879), the segmentation-cells only remain equal up to the morula-stage, the spherical body of which consists of thirty-two cells. Then, as always happens in unequal segmentation, the more sluggish vegetal cells are outstripped in the cleavage. At the lower or vegetal pole of the ovum a crown of eight large entodermic cells remains for a long time unchanged, while the other cells divide, owing to the formation of a series of horizontal circles, into an increasing number of crowns of sixteen cells each. Afterwards the segmentation-cells get more or less irregularly displaced, while the segmentation-cavity enlarges in the centre of the morula; in the end the former all lie on the surface of the latter, so that the fœtus attains the familiar blastula shape and forms a hollow ball, the wall of which consists of a single stratum of cells (Fig. 38 A–C). This layer is the blastoderm, the simple epithelium from the cells of which all the tissues of the body proceed.

[ 193 ]

These important early embryonic processes take place so quickly in the Amphioxus that four or five hours after fecundation, or about midnight, the spherical blastula is completed. A pit-like depression is then formed at the vegetal pole of it, and in consequence of this the hollow sphere doubles on itself (Fig. 38 D). This pit becomes deeper and deeper (Fig. 38 E, F); at last the invagination (or doubling) is complete, and the inner or folded part of the blastula-wall lies on the inside of the outer wall. We thus get a hollow hemisphere, the thin wall of which is made up of two layers of cells (Fig. 38 E). From hemispherical the body soon becomes almost spherical once more, and then oval, the internal cavity enlarging considerably and its mouth growing narrower (Fig. 213). The form which the Amphioxus-embryo has thus reached is a real “cup-larva” or gastrula, of the original simple type that we have previously described as the “bell-gastrula” or archigastrula (Figs. 29–35).

As in all the other animals that form an archigastrula, the whole body is nothing but a simple gastric sac or stomach; its internal cavity is the primitive gut (progaster or archenteron, Fig. 38 g, 35 d), and its aperture the primitive mouth (prostoma or blastoporus, o). The wall is at once gut-wall and body-wall. It is composed of two simple cell-layers, the familiar primary germinal layers. The inner layer or the invaginated part of the blastoderm, which immediately encloses the gut-cavity is the entoderm, the inner or vegetal germ-layer, from which develop the wall of the alimentary canal and all its appendages, the cœlom-pouches, etc. (Figs. 35, 36 i). The outer stratum of cells, or the non-invaginated part of the blastoderm, is the ectoderm, the outer or animal germ-layer, which provides the outer skin (epidermis) and the nervous system (e). The cells of the entoderm are much larger, darker, and more fatty than those of the ectoderm, which are clearer and less rich in fatty particles. Hence before and during invagination there is an increasing differentiation of the inner from the outer layer. The animal cells of the outer layer soon develop vibratory hairs; the vegetal cells of the inner layer do so much later. A thread-like process grows out of each cell, and effects continuous vibratory movements. By the vibrations of these slender hairs the gastrula of the Amphioxus swims about in the sea, when it has pierced the thin ovolemma, like the gastrula of many other animals (Fig. 36). As in many other lower animals, the cells have only one whip-like hair each, and so are called flagellate (whip) cells (in contrast with the ciliated cells, which have a number of short lashes or cilia).

In the further course of its rapid development the roundish bell-gastrula becomes elongated, and begins to flatten on one side, parallel to the long axis. The flattened side is the subsequent dorsal side; the opposite or ventral side remains curved. The latter grows more quickly than the former, with the result that the primitive mouth is forced to the dorsal side (Fig. 39). In the middle of the dorsal surface a shallow longitudinal groove or furrow is formed (Fig. 79), and the edges of the body rise up on each side of this groove in the shape of two parallel swellings. This groove is, of course, the dorsal furrow, and the swellings are the dorsal or medullary swellings; they form the first structure of the central nervous system, the medullary tube. The medullary swellings now rise higher; the groove between them becomes deeper and deeper. The edges of the parallel swellings curve towards each other, and at last unite, and the medullary tube is formed (Figs. 83 m, 84 m). Hence the formation of a medullary tube out of the outer skin takes place in the naked dorsal surface of the free-swimming larva of the Amphioxus in just the same way as we have found in the embryo of man and the higher animals within the fœtal membranes.

Simultaneously with the construction of the medullary tube we have in the Amphioxus-embryo the formation of the chorda, the cœlom-pouches, and the mesoderm proceeding from their wall. These processes also take place with characteristic simplicity and clearness, so that they are very instructive to compare with the vermalia on the one hand and with the higher vertebrates on the other. While the medullary groove is sinking in the middle line of the flat dorsal side of the oval embryo, and its parallel edges unite to form the ectodermic neural tube, the single chorda is formed directly underneath them, and on each side of this a parallel longitudinal fold, from the dorsal wall of the primitive gut. These longitudinal folds of the entoderm proceed from the primitive mouth, or from its lower

[ 194 ]

and hinder edge. Here we see at an early stage a couple of large entodermic cells, which are distinguished from all the others by their great size, round form, and fine-grained protoplasm; they are the two promesoblasts, or polar cells of the mesoderm (Fig. 83 p). They indicate the original starting-point of the two cœlom-pouches, which grow from this spot between the inner and outer germinal layers, sever themselves from the primitive gut, and provide the cellular material for the middle layer.

Immediately after their formation the two cœlom-pouches of the Amphioxus are divided into several parts by longitudinal and transverse folds. Each of the primary pouches is divided into an upper dorsal and a lower ventral section by a couple of lateral longitudinal folds (Fig. 82). But these are again divided by several parallel transverse folds into a number of successive sacs, the primitive segments or somites (formerly called by the unsuitable name of “primitive vertebræ”). They have a different future above and below. The upper or dorsal segments, the episomites, lose their cavity later on, and form with their cells the muscular plates of the trunk. The lower or ventral segments, the hyposomites, corresponding to the lateral plates of the craniote-embryo, fuse together in the upper part owing to the disappearance of their lateral walls, and thus form the later body-cavity (metacœl); in the lower part they remain separate, and afterwards form the segmental gonads.

In the middle, between the two lateral cœlom-folds of the primitive gut, a single central organ detaches from this at an early stage in the middle line of its dorsal wall. This is the dorsal chorda (Figs. 83, 84 ch). This axial rod, which is the first foundation of the later vertebral column in all the vertebrates, and is the only representative of it in the Amphioxus, originates from the entoderm.

In consequence of these important folding-processes in the primitive gut, the simple entodermic tube divides into four different sections:— I, underneath, at the ventral side, the permanent alimentary canal or permanent gut; II, above, at the dorsal side, the axial rod or chorda; and III, the two cœlom-sacs, which immediately sub-divide into two structures:—IIIA, above, on the dorsal side, the episomites, the double row of primitive or muscular segments; and IIIB, below, on each side of the gut, the hyposomites, the two lateral plates that give rise to the sex-glands, and the cavities of which partly unite to form the body-cavity. At the same time, the neural or medullary tube is formed above the chorda, on the dorsal surface, by the closing of the parallel medullary swellings. All these processes, which outline the typical structure of the vertebrate, take place with astonishing rapidity in the embryo of the Amphioxus; in the afternoon of the first day, or twenty-four hours after fertilisation, the young vertebrate, the typical embryo, is formed; it then has, as a rule, six to eight somites.

The chief occurrence on the second day of development is the construction of the two permanent openings of the gut—the mouth and anus. In the earlier stages the alimentary tube is found to be entirely closed, after the closing of the primitive mouth; it only communicates behind by the neurenteric canal with the medullary tube. The permanent mouth is a secondary formation, at the opposite end. Here, at the end of the second day, we find a pit-like depression in the outer skin, which penetrates inwards into the closed gut. The anus is formed behind in the same way a few hours later (in the vicinity of the additional gastrula-mouth). In man and the higher vertebrates also the mouth and anus are formed, as we have seen, as flat pits in the outer skin; they then penetrate inwards, gradually becoming connected with the blind ends of the closed gut-tube. During the second day the Amphioxus-embryo undergoes few other changes. The number of primitive segments increases, and generally amounts to fourteen, some forty-eight to fifty hours after impregnation.

Almost simultaneously with the formation of the mouth the first gill-cleft breaks through in the fore section of the Amphioxus-embryo (generally forty hours after the commencement of development). It now begins to nourish itself independently, as the food material stored up in the ovum is completely used up. The further development of the free larvæ takes place very slowly, and extends over several months. The body becomes much longer, and is compressed at the sides, the head-end being broadened in a sort of triangle. Two rudimentary sense-organs are developed in it. Inside we find the first blood-vessels, an upper or dorsal vessel, corresponding to the aorta, between the gut and the dorsal cord, and a lower or ventral

[ 195 ]

vessel, corresponding to the subintestinal vein, at the lower border of the gut. Now, the gills or respiratory organs also are formed at the fore-end of the alimentary canal. The whole of the anterior or respiratory section of the gut is converted into a gill-crate, which is pierced trellis-wise by numbers of branchial-holes, as in the ascidia. This is done by the foremost part of the gut-wall joining star-wise with the outer skin, and the formation of clefts at the point of connection, piercing the wall and leading into the gut from without. At first there are very few of these branchial clefts; but there are soon a number of them—first in one, then in two, rows. The foremost gill-cleft is the oldest. In the end we have a sort of lattice work of fine gill-clefts, supported on a number of stiff branchial rods; these are connected in pairs by transverse rods.

Figs. 222–224—Transverse sections of young Amphioxus-larvæ (diagrammatic, from Ralph.) (Cf. also Fig. 216.) In Fig. 222 there is free communication from without with the gut-cavity (D) through the gill-clefts (K). In Fig. 223 the lateral folds of the body-wall, or the gill-covers, which grow downwards, are formed. In Fig. 224 these lateral folds have united underneath and joined their edges in the middle line of the ventral side (R seam). The respiratory water now passes from the gut-cavity (D) into the mantle-cavity (A). The letters have the same meaning throughout: N medullary tube, Ch chorda, M lateral muscles, Lh body-cavity, G part of the body-cavity in which the sexual organs are subsequently formed. D gut-cavity, clothed with the gut-gland layer (a). A mantle-cavity, K gill-clefts, b=E epidermis, E1 the same as visceral epithelium of the mantle-cavity, E2 as parietal epithelium of the mantle-cavity.

At an early stage of embryonic development the structure of the Amphioxus-larva is substantially the same as the ideal picture we have previously formed of the “Primitive Vertebrate” (Figs. 98–102). But the body afterwards undergoes various modifications, especially in the fore-part. These modifications do not concern us, as they depend on special adaptations, and do not affect the hereditary vertebrate type. When the free-swimming Amphioxus-larva is three months old, it abandons its pelagic habits and changes into the young animal that lives in the sand. In spite of its smallness (one-eighth of an inch), it has substantially the same structure as the adult. As regards the remaining organs of the Amphioxus, we need only mention that the gonads or sexual glands are developed very late, immediately out of the inner cell-layer of the

[ 196 ]

body-cavity. Although we can find afterwards no continuation of the body-cavity (Fig. 216 U) in the lateral walls of the mantle-cavity, in the gill-covers or mantle-folds (Fig. 224 U), there is one present in the beginning (Fig. 224 Lh). The sexual cells are formed below, at the bottom of this continuation (Fig. 224 S). For the rest, the subsequent development into the adult Amphioxus of the larva we have followed is so simple that we need not go further into it here.

We may now turn to the embryology of the Ascidia, an animal that seems to stand so much lower and to be so much more simply organised, remaining for the greater part of its life attached to the bottom of the sea like a shapeless lump. It was a fortunate accident that Kowalevsky first examined just those larger specimens of the Ascidiæ that show most clearly the relationship of the vertebrates to the invertebrates, and the larvæ of which behave exactly like those of the Amphioxus in the first stages of development. This resemblance is so close in the main features that we have only to repeat what we have already said of the ontogenesis of the Amphioxus.

The ovum of the larger Ascidia (Phallusia, Cynthia, etc.) is a simple round cell of 1/250 to 1/125 of an inch in diameter. In the thick fine-grained yelk we find a clear round germinal vesicle of about 1/750 of an inch in diameter, and this encloses a small embryonic spot or nucleolus. Inside the membrane that surrounds the ovum, the stem-cell of the Ascidia, after fecundation, passes through just the same metamorphoses as the stem-cell of the Amphioxus. It undergoes total segmentation; it divides into two, four, eight, sixteen, thirty-two cells, and so on. By continued total cleavage the morula, or mulberry-shaped cluster of cells, is formed. Fluid gathers inside it, and thus we get once more a globular vesicle (the blastula); the wall of this is a single stratum of cells, the blastoderm. A real gastrula (a simple bell-gastrula) is formed from the blastula by invagination, in the same way as in the amphioxus.

Up to this there is no definite ground in the embryology of the Ascidiæ for bringing them into close relationship with the Vertebrates; the same gastrula is formed in the same way in many other animals of different stems. But we now find an embryonic process that is peculiar to the Vertebrates, and that proves irrefragably the affinity of the Ascidiæ to the Vertebrates. From the epidermis of the gastrula a medullary tube is formed on the dorsal side, and, between this and the primitive gut, a chorda; these are the organs that are otherwise only found in Vertebrates. The formation of these very important organs takes place in the Ascidia-gastrula in precisely the same way as in that of the Amphioxus. In the Ascidia (as in the other case) the oval gastrula is first flattened on one side—the subsequent dorsal side. A groove or furrow (the medullary groove) is sunk in the middle line of the flat surface, and two parallel longitudinal swellings arise on either side from the skin layer. These medullary swellings join together over the furrow, and form a tube; in this case, again, the neural or medullary tube is at first open in front, and connected with the primitive gut behind by the neurenteric canal. Further, in the Ascidia-larva also the two permanent apertures of the alimentary canal only appear later, as independent and new formations. The permanent mouth does not develop from the primitive mouth of the gastrula; this primitive mouth closes up, and the later anus is formed near it by invagination from without, on the hinder end of the body, opposite to the aperture of the medullary tube.

During these important processes, that take place in just the same way in the Amphioxus, a tail-like projection grows out of the posterior end of the larva-body, and the larva folds itself up within the round ovolemma in such a way that the dorsal side is curved and the tail is forced on to the ventral side. In this tail is developed—starting from the primitive gut—a cylindrical string of cells, the fore end of which pushes into the body of the larva, between the alimentary canal and the neural canal, and is no other than the chorda dorsalis. This important organ had hitherto been found only in the Vertebrates, not a single trace of it being discoverable in the Invertebrates. At first the chorda only consists of a single row of large entodermic cells. It is afterwards composed of several rows of cells. In the Ascidia-larva, also, the chorda develops from the dorsal middle part of the primitive gut, while the two cœlom-pouches detach themselves from it on both sides. The simple body-cavity is formed by the coalescence of the two.

When the Ascidia-larva has attained

[ 197 ]

this stage of development it begins to move about in the ovolemma. This causes the membrane to burst. The larva emerges from it, and swims about in the sea by means of its oar-like tail. These free-swimming larvæ of the Ascidia have been known for a long time. They were first observed by Darwin during his voyage round the world in 1833. They resemble tadpoles in outward appearance, and use their tails as oars, as the tadpoles do. However, this lively and highly-developed condition does not last long. At first there is a progressive development; the foremost part of the medullary tube enlarges into a brain, and inside this two single sense-organs are developed, a dorsal auditory vesicle and a ventral eye. Then a heart is formed on the ventral side of the animal, or the lower wall of the gut, in the same simple form and at the same spot at which the heart is developed in man and all the other vertebrates. In the lower muscular wall of the gut we find a weal-like thickening, a solid, spindle-shaped string of cells, which becomes hollow in the centre; it begins to contract in different directions, now forward and now backward, as is the case with the adult Ascidia. In this way the sanguineous fluid accumulated in the hollow muscular tube is driven in alternate directions into the blood-vessels, which develop at both ends of the cardiac tube. One principal vessel runs along the dorsal side of the gut, another along its ventral side. The former corresponds to the aorta and the dorsal vessel in the worms. The other corresponds to the subintestinal vein and the ventral vessel of the worms.

Fig. 225—An Appendicaria (Copelata), seen from the left. m mouth, k branchial gut, o gullet, v stomach, a anus, n brain (ganglion above the gullet), g auditory vesicle, f ciliated groove under the gills, h heart, t testicles, e ovary, c chorda, s tail.

With the formation of these organs the progressive development of the Ascidia comes to an end, and degeneration sets in. The free-swimming larva sinks to the floor of the sea, abandons its locomotive habits, and attaches itself to stones, marine plants, mussel-shells, corals, and other objects; this is done with the part of the body that was foremost in movement. The attachment is effected by a number of out-growths, usually three, which can be seen even in the free-swimming larva. The tail is lost, as there is no further use for it. It undergoes a fatty degeneration, and disappears with the chorda dorsalis. The tailless body changes into an unshapely tube, and, by the atrophy of some parts and the modification of others, gradually assumes the appearance we have already described.

Among the living Tunicates there is a very interesting group of small animals that remain throughout life at the stage of development of the tailed, free Ascidia-larva, and swim about briskly in the sea by means of their broad oar-tail. These are the remarkable Copelata (Appendicaria and Vexillaria, Fig. 225). They are the only living Vertebrates that have throughout life a chorda dorsalis and a neural string above it; the latter must be regarded as the prolongation of the cerebral ganglion and the equivalent of the medullary tube. Their branchial gut also opens directly outwards by a pair of

[ 198 ]

branchial clefts. These instructive Copelata, comparable to permanent Ascidia-larvæ, come next to the extinct Prochordonia, those ancient worms which we must regard as the common ancestors of the Tunicates and Vertebrates. The chorda of the Appendicaria is a long, cylindrical string (Fig. 225 c), and serves as an attachment for the muscles that work the flat oar-tail.

Among the various modifications which the Ascidia-larva undergoes after its establishment at the sea-floor, the most interesting (after the loss of the axial rod) is the atrophy of one of its chief organs, the medullary tube. In the Amphioxus the spinal marrow continues to develop, but in the Ascidia the tube soon shrinks into a small and insignificant nervous ganglion that lies above the mouth and the gill-crate, and is in accord with the extremely slight mental power of the animal. This insignificant relic of the medullary tube seems to be quite beyond comparison with the nervous centre of the vertebrate, yet it started from the same structure as the spinal cord of the Amphioxus. The sense-organs that had been developed in the fore part of the neural tube are also lost; no trace of which can be found in the adult Ascidia. On the other hand, the alimentary canal becomes a most extensive organ. It divides presently into two sections—a wide fore or branchial gut that serves for respiration, and a narrower hind or hepatic gut that accomplishes digestion. The branchial or head-gut of the Ascidia is small at first, and opens directly outwards only by a couple of lateral ducts or gill-clefts—a permanent arrangement in the Copelata. The gill-clefts are developed in the same way as in the Amphioxus. As their number greatly increases we get a large gill-crate, pierced like lattice work. In the middle line of its ventral side we find the hypobranchial groove. The mantle or cloaca-cavity (the atrium) that surrounds the gill-crate is also formed in the same way in the Ascidia as in the Amphioxus. The ejection-opening of this peribranchial cavity corresponds to the branchial pore of the Amphioxus. In the adult Ascidia the branchial gut and the heart on its ventral side are almost the only organs that recall the original affinity with the vertebrates.

The further development of the Ascidia in detail has no particular interest for us, and we will not go into it. The chief result that we obtain from its embryology is the complete agreement with that of the Amphioxus in the earliest and most important embryonic stages. They do not begin to diverge until after the medullary tube and alimentary canal, and the axial rod with the muscles between the two, have been formed. The Amphioxus continues to advance, and resembles the embryonic forms of the higher vertebrates; the Ascidia degenerates more and more, and at last, in its adult condition, has the appearance of a very imperfect invertebrate.

If we now look back on all the remarkable features we have encountered in the structure and the embryonic development of the Amphioxus and the Ascidia, and compare them with the features of man’s embryonic development which we have previously studied, it will be clear that I have not exaggerated the importance of these very interesting animals. It is evident that the Amphioxus from the vertebrate side and the Ascidia from the invertebrate form the bridge by which we can span the deep gulf that separates the two great divisions of the animal kingdom. The radical agreement of the lancelet and the sea-squirt in the first and most important stages of development shows something more than their close anatomic affinity and their proximity in classification; it shows also their real blood-relationship and their common origin from one and the same stem-form. In this way, it throws considerable light on the oldest roots of man’s genealogical tree.

Title and Contents
Vol. II Title and Contents
Glossary
Chapter XVI
Chapter XVIII
Figs. 1–209
Figs. 210–408