Under the guidance of the biogenetic law, and on the basis of the evidence we have obtained, we now turn to the interesting task of determining the series of man’s animal ancestors. Phylogeny us a whole is an inductive science. From the totality of the biological processes in the life of plants, animals, and man we have gathered a confident inductive idea that the whole organic population of our planet has been moulded on a harmonious law of evolution. All the interesting phenomena that we meet in ontogeny and paleontology, comparative anatomy and dysteleology, the distribution and habits of organisms—all the important general laws that we abstract from the phenomena of these sciences, and combine in harmonious unity—are the broad bases of our great biological induction.

But when we come to the application of this law, and seek to determine with its aid the origin of the various species of organisms, we are compelled to frame hypotheses that have essentially a deductive character, and are inferences from the general law to particular cases. But these special deductions are just as much justified and necessitated by the rigorous laws of logic as the inductive conclusions on which the whole theory of evolution is built. The doctrine of the animal ancestry of the human race is a special deduction of this kind, and follows with logical necessity from the general inductive law of evolution.

I must point out at once, however, that the certainty of these evolutionary hypotheses, which rest on clear special deductions, is not always equally strong. Some of these inferences are now beyond question; in the case of others it depends on the knowledge and the competence of the inquirer what degree of certainty he attributes to them. In any case, we must distinguish between the absolute certainty of the general (inductive) theory of descent and the relative certainty of special (deductive) evolutionary hypotheses. We can never determine the whole ancestral series of an organism with the same confidence with which we hold the general theory of evolution as the sole scientific explanation of organic modifications. The special indication of stem-forms in detail will always be more or less incomplete and hypothetical. This is quite natural. The evidence on which we build is imperfect, and always will be imperfect; just as in comparative philology.

The first of our documents, paleontology, is exceedingly incomplete. We know that all the fossils yet discovered are only an insignificant fraction of the plants and animals that have lived on our planet. For every single species that has been preserved for us in the rocks there are probably hundreds, perhaps thousands, of extinct species that have left no trace behind them. This extreme and very unfortunate incompleteness of the paleontological evidence, which cannot be pointed out too often, is easily explained. It is absolutely inevitable in the circumstances of the fossilisation of organisms. It is also due in part to the incompleteness of our knowledge in this branch. It must be borne in mind that the great majority of the stratified rocks that compose the crust of the earth have not yet been opened. We have only a few specimens of the innumerable fossils that are buried in the vast mountain ranges of Asia and Africa. Only a part of Europe and North America has been investigated carefully. The whole of the fossils known to us certainly do not amount to a hundredth part of the remains that are really buried in the crust of the earth. We may, therefore, look forward to a rich harvest in the future as regards this science. However, our paleontological evidence will (for reasons that I have fully explained in the sixteenth chapter of the History of Creation) always be defective.

The second chief source of evidence, ontogeny, is not less incomplete. It is the most important source of all for special phylogeny; but it has great defects, and often fails us. We must, above all, clearly distinguish between palingenetic and cenogenetic phenomena. We must never forget that the laws of curtailed and disturbed heredity often make the original course of development almost unrecognisable. The recapitulation of phylogeny by ontogeny is only fairly complete in a few cases, and is never wholly complete. As a rule, it is precisely the earliest and most important embryonic stages that suffer most from alteration and condensation. The earlier embryonic forms have had to adapt themselves to new circumstances, and so have been modified. The struggle for existence has had just as profound an influence on the freely moving and still immature young forms as on the adult forms. Hence in the embryology of the higher animals, especially, palingenesis is much restricted by cenogenesis; it is to-day, as a rule, only a faded and much altered picture of the original evolution of the animal’s ancestors. We can only draw conclusions from the embryonic forms to the stem-history with the greatest caution and discrimination. Moreover, the embryonic development itself has only been fully studied in a few species.

Finally, the third and most valuable source of evidence, comparative anatomy, is also, unfortunately, very imperfect; for the simple reason that the whole of the living species of animals are a mere fraction of the vast population that has dwelt on our planet since the beginning of life. We may confidently put the total number of these at more than a million species. The number of animals whose organisation has been studied up to the present in comparative anatomy is proportionately very small. Here, again, future research will yield incalculable treasures. But, for the present, in view of this patent incompleteness of our chief sources of evidence, we must naturally be careful not to lay too much stress in human phylogeny on the particular animals we have studied, or regard all the various stages of development with equal confidence as stem-forms.

In my first efforts to construct the series of man’s ancestors I drew up a list of, at first ten, afterwards twenty to thirty, forms that may be regarded more or less certainly as animal ancestors of the human race, or as stages that in a sense mark off the chief sections in the long story of evolution from the unicellular organism to man. Of these twenty to thirty stages, ten to twelve belong to the older group of the Invertebrates and eighteen to twenty to the younger division of the Vertebrates.

In approaching, now, the difficult task of establishing the evolutionary succession of these thirty ancestors of humanity since the beginning of life, and in venturing to lift the veil that covers the earliest secrets of the earth’s history, we must undoubtedly look for the first living things among the wonderful organisms that we call the Monera; they are the simplest organisms known to us—in fact, the simplest we can conceive. Their whole body consists merely of a simple particle or globule of structureless plasm or plasson. The discoveries of the last four decades have led us to believe with increasing certainty that wherever a natural body exhibits the vital processes of nutrition, reproduction, voluntary movement, and sensation, we have the action of a nitrogenous carbon-compound of the chemical group of the albuminoids; this plasm (or protoplasm) is the material basis of all vital functions. Whether we regarded the function, in the monistic sense, as the direct action of the material substratum, or whether we take matter and force to be distinct things in the dualistic sense, it is certain that we have not as yet found any living organism in which the exercise of the vital functions is not inseparably bound up with plasm.

The soft slimy plasson of the body of the moneron is generally called “protoplasm,” and identified with the cellular matter of the ordinary plant and animal cells. But we must, to be accurate, distinguish between the plasson of the cytodes and the protoplasm of the cells. This distinction is of the utmost importance for the purposes of evolution. As I have often said, we must recognise two different stages of development in these “elementary organisms,” or plastids (“builders”), that represent the ultimate units of organic individuality. The earlier and lower stage are the unnucleated cytodes, the body of which consists of only one kind of albuminous matter—the homogeneous plasson or “formative matter.” The later and higher stage are the nucleated cells, in which we find a differentiation of the original plasson into two different formative substances—the caryoplasm of the nucleus and the cytoplasm of the body of the cell (cf. pp. 37 and 42).

The Monera are permanent cytodes. Their whole body consists of soft, structureless plasson. However carefully we examine it with our finest chemical reagents and most powerful microscopes, we can find no definite parts or no anatomic structure in it. Hence, the Monera are literally organisms without organs; in fact, from the philosophic point of view they are not organisms at all, since they have no organs. They can only be called organisms in the sense that they are capable of the vital functions of nutrition, reproduction, sensation, and movement. If we were to try to imagine the simplest possible organism, we should frame something like the moneron.

The Monera that we find to-day in various forms fall into two groups according to the nature of their nutrition—the Phytomonera and the Zoomonera; from the physiological point of view, the former are the simplest specimens of the plant (phyton) kingdom, and the latter of the animal (zoon) world. The Phytomonera, especially in their simplest form, the Chromacea (Phycochromacea or Cyanophycea), are the most primitive and the oldest of living organisms. The typical genus Chroococcus (Fig. 226) is represented by several fresh-water species, and often forms a very delicate bluish-green deposit on stones and wood in ponds and ditches. It consists of round, light green particles, from 1/7000 to 1/2500 of an inch in diameter.

The whole life of these homogeneous globules of plasm consists of simple growth and reproduction by cleavage. When the tiny particle has reached a certain size by the continuous assimilation of inorganic matter, it divides into two equal halves, by a constriction in the middle. The two daughter-monera that are thus formed immediately begin a similar vital process. It is the same with the brown Procytella primordialis (formerly called the Protococcus marinus); it forms large masses of floating matter in the arctic seas. The tiny plasma-globules of this species are of a greenish-brown colour, and have a diameter of 1/10,000 to 1/5000 of an inch. There is no membrane discoverable in the simplest Chroococcacea, but we find one in other members of the same family; in Aphanocapsa (Fig. 227) the enveloping membranes of the social plastids combine; in Gloecapsa they are retained through several generations, so that the little plasma-globules are enfolded in many layers of membrane.

Next to the Chromacea come the Bacteria, which have been evolved from them by the remarkable change in nutrition which gives us the simple explanation of the differentiation of plant and animal in the protist kingdom. The Chromacea build up their plasm directly from inorganic matter; the Bacteria feed on organic matter. Hence, if we logically divide the protist kingdom into plasma-forming Protophyta and plasma-consuming Protozoa, we must class the Bacteria with the latter; it is quite illogical to describe them—as is still often done—as Schizomycetes, and class them with the true fungi. The Bacteria, like the Chromacea, have no nucleus. As is well-known, they play an important part in modern biology as the causes of fermentation and putrefaction, and of tuberculosis, typhus, cholera, and other infectious diseases, and as parasites, etc. But we cannot linger now to deal with these very interesting features; the Bacteria have no relation to man’s genealogical tree.

We may now turn to consider the remarkable Protamoeba, or unnucleated Amoeba. I have, in the first volume, pointed out the great importance of the ordinary Amoeba in connection with several weighty questions of general biology. The tiny ProtamoebÆ, which are found both in fresh and salt water, have the same unshapely form and irregular movements of their simple naked body as the real AmoebÆ; but they differ from them very materially in having no nucleus in their cell-body. The short, blunt, finger-like processes that are thrust out at the surface of the creeping Protamoeba serve for getting food as well as for locomotion. They multiply by simple cleavage (Fig. 228).

The next stage to the simple cytode-forms of the Monera in the genealogy of mankind (and all other animals) is the simple cell, or the most rudimentary form of the cell which we find living independently to-day as the Amoeba. The earliest process of inorganic differentiation in the structureless body of the Monera led to its division into two different substances—the caryoplasm and the cytoplasm. The caryoplasm is the inner and firmer part of the cell, the substance of the nucleus. The cytoplasm is the outer and softer part, the substance of the body of the cell. By this important differentiation of the plasson into nucleus and cell-body, the organised cell was evolved from the structureless cytode, the nucleated from the unnucleated plastid. That the first cells to appear on the earth were formed from the Monera by such a differentiation seems to us the only possible view in the present condition of science. We have a direct instance of this earliest process of differentiation to-day in the ontogeny of many of the lower Protists (such as the GregarinÆ).

The unicellular form that we have in the ovum has already been described as the reproduction of a corresponding unicellular stem-form, and to this we have ascribed the organisation of an Amoeba (cf. Chapter VI). The irregular-shaped Amoeba, which we find living independently to-day in our fresh and salt water, is the least definite and the most primitive of all the unicellular Protozoa (Fig. 16). As the unripe ova (the protova that we find in the ovaries of animals) cannot be distinguished from the common AmoebÆ, we must regard the Amoeba as the primitive form that is reproduced in the embryonic stage of the amoeboid ovum to-day, in accordance with the biogenetic law. I have already pointed out, in proof of the striking resemblance of the two cells, that the ova of many of the sponges were formerly regarded as parasitic AmoebÆ (Figure 1.18). Large unicellular organisms like the AmoebÆ were found creeping about inside the body of the sponge, and were thought to be parasites. It was afterwards discovered that they were really the ova of the sponge from which the embryos were developed. As a matter of fact, these sponge-ova are so much like many of the AmoebÆ in size, shape, the character of their nucleus, and movement of the pseudopodia, that it is impossible to distinguish them without knowing their subsequent development.

Our phylogenetic interpretation of the ovum, and the reduction of it to some ancient amoeboid ancestral form, supply the answer to the old problem: “Which was first, the egg or the chick?” We can now give a very plain answer to this riddle, with which our opponents have often tried to drive us into a corner. The egg came a long time before the chick. We do not mean, of course, that the egg existed from the first as a bird’s egg, but as an indifferent amoeboid cell of the simplest character. The egg lived for thousands of years as an independent unicellular organism, the Amoeba. The egg, in the modern physiological sense of the word, did not make its appearance until the descendants of the unicellular Protozoon had developed into multicellular animals, and these had undergone sexual differentiation. Even then the egg was first a gastrÆa-egg, then a platode-egg, then a vermalia-egg, and chordonia-egg; later still acrania-egg, then fish-egg, amphibia-egg, reptile-egg, and finally bird’s egg. The bird’s egg we have experience of daily is a highly complicated historical product, the result of countless hereditary processes that have taken place in the course of millions of years.

The earliest ancestors of our race were simple Protophyta, and from these our protozoic ancestors were developed afterwards. From the morphological point of view both the vegetal and the animal Protists were simple organisms, individualities of the first order, or plastids. All our later ancestors are complex organisms, or individualities of a higher order—social aggregations of a plurality of cells. The earliest of these, the MorÆada, which represent the third stage in our genealogy, are very simple associations of homogeneous, indifferent cells—undifferentiated colonies of social AmoebÆ or Infusoria. To understand the nature and origin of these protozoa-colonies we need only follow step by step the first embryonic products of the stem-cell. In all the Metazoa the first embryonic process is the repeated cleavage of the stem-cell, or first segmentation-cell (Fig. 229). We have already fully considered this process, and found that all the different forms of it may be reduced to one type, the original equal or primordial segmentation (cf. Chapter VIII). In the genealogical tree of the Vertebrates this palingenetic form of segmentation has been preserved in the Amphioxus alone, all the other Vertebrates having cenogenetically modified forms of cleavage. In any case, the latter were developed from the former, and so the segmentation of the ovum in the Amphioxus has a great interest for us (cf. Fig. 38). The outcome of this repeated cleavage is the formation of a round cluster of cells, composed of homogeneous, indifferent cells of the simplest character (Fig. 230). This is called the morula (= mulberry-embryo) on account of its resemblance to a mulberry or blackberry.

It is clear that this morula reproduces for us to-day the simple structure of the multicellular animal that succeeded the unicellular amoeboid form in the early Laurentian period. In accordance with the biogenetic law, the morula recalls the ancestral form of the MorÆa, or simple colony of Protozoa. The first cell-communities to be formed, which laid the early foundation of the higher multicellular body, must have consisted of homogeneous and simple amoeboid cells. The oldest AmoebÆ lived isolated lives, and even the amoeboid cells that were formed by the segmentation of these unicellular organisms must have continued to live independently for a long time. But gradually small communities of AmoebÆ arose by the side of these eremitical Protozoa, the sister-cells produced by cleavage remaining joined together. The advantages in the struggle for life which these communities had over the isolated cells favoured their formation and their further development. We find plenty of these cell-colonies or communities to-day in both fresh and salt water. They belong to various groups both of the Protophyta and Protozoa.

To have some idea of those ancestors of our race that succeeded phylogenetically to the MorÆada, we have only to follow the further embryonic development of the morula. We then see that the social cells of the round cluster secrete a sort of jelly or a watery fluid inside their globular body, and they themselves rise to the surface of it (Fig. 29 F, G). In this way the solid mulberry-embryo becomes a hollow sphere, the wall of which is composed of a single layer of cells. We call this layer the blastoderm, and the sphere itself the blastula, or embryonic vesicle.

This interesting blastula is very important. The conversion of the morula into a hollow ball proceeds on the same lines originally in the most diverse stems—as, for instance, in many of the zoophytes and worms, the ascidia, many of the echinoderms and molluscs, and in the amphioxus. Moreover, in the animals in which we do not find a real palingenetic blastula the defect is clearly due to cenogenetic causes, such as the formation of food-yelk and other embryonic adaptations. We may, therefore, conclude that the ontogenetic blastula is the reproduction of a very early phylogenetic ancestral form, and that all the Metazoa are descended from a common stem-form, which was in the main constructed like the blastula. In many of the lower animals the blastula is not developed within the foetal membranes, but in the open water. In those cases each blastodermic cell begins at an early stage to thrust out one or more mobile hair-like processes; the body swims about by the vibratory movement of these lashes or whips (Fig. 29 F).

We still find, both in the sea and in fresh water, various kinds of primitive multicellular organisms that substantially resemble the blastula in structure, and may be regarded in a sense as permanent blastula-forms—hollow vesicles or gelatinous balls, with a wall composed of a single layer of ciliated homogeneous cells. There are “blastÆads” of this kind even among the Protophyta—the familiar Volvocina, formerly classed with the infusoria. The common Volvox globator is found in the ponds in the spring—a small, green, gelatinous globule, swimming about by means of the stroke of its lashes, which rise in pairs from the cells on its surface. In the similar HalosphÆra viridis also, which we find in the marine plancton (floating matter), a number of green cells form a simple layer at the surface of the gelatinous ball; but in this case there are no cilia.

Some of the infusoria of the flagellata-class (Signura, MagosphÆra, etc.) are similar in structure to these vegetal clusters, but differ in their animal nutrition; they form the special group of the Catallacta. In September, 1869, I studied the development of one of these graceful animals on the island of Gis-Oe, off the coast of Norway (MagosphÆra planula), Figures 2.231 and 2.232). The fully-formed body is a gelatinous ball, with its wall composed of thirty-two to sixty-four ciliated cells; it swims about freely in the sea. After reaching maturity the community is dissolved. Each cell then lives independently for some time, grows, and changes into a creeping amoeba. This afterwards contracts, and clothes itself with a structureless membrane. The cell then looks just like an ordinary animal ovum. When it has been in this condition for some time the cell divides into two, four, eight, sixteen, thirty-two, and sixty-four cells. These arrange themselves in a round vesicle, thrust out vibratory lashes, burst the capsule, and swim about in the same magosphÆra-form with which we started. This completes the life-circle of the remarkable and instructive animal.

If we compare these permanent blastulÆ with the free-swimming ciliated larvÆ or blastulÆ, with similar construction, of many of the lower animals, we can confidently deduce from them that there was a very early and long-extinct common stem-form of substantially the same structure as the blastula. We may call it the BlastÆa. Its body consisted, when fully formed, of a simple hollow ball, filled with fluid or structureless jelly, with a wall composed of a single stratum of ciliated cells. There were probably many genera and species of these blastÆads in the Laurentian period, forming a special class of marine protists.

It is an interesting fact that in the plant kingdom also the simple hollow sphere is found to be an elementary form of the multicellular organism. At the surface and below the surface (down to a depth of 2000 yards) of the sea there are green globules swimming about, with a wall composed of a single layer of chlorophyll-bearing cells. The botanist Schmitz gave them the name of HalosphÆra viridis in 1879.

The next stage to the BlastÆa, and the sixth in our genealogical tree, is the GastrÆa that is developed from it. As we have already seen, this ancestral form is particularly important. That it once existed is proved with certainty by the gastrula, which we find temporarily in the ontogenesis of all the Metazoa (Fig. 29 J, K). As we saw, the original, palingenetic form of the gastrula is a round or oval uni-axial body, the simple cavity of which (the primitive gut) has an aperture at one pole of its axis (the primitive mouth). The wall of the gut consists of two strata of cells, and these are the primary germinal layers, the animal skin-layer (ectoderm) and vegetal gut-layer (entoderm).

The actual ontogenetic development of the gastrula from the blastula furnishes sound evidence as to the phylogenetic origin of the GastrÆa from the BlastÆa. A pit-shaped depression appears at one side of the spherical blastula (Fig. 29 H). In the end this invagination goes so far that the outer or invaginated part of the blastoderm lies close on the inner or non-invaginated part (Fig. 29 J). In explaining the phylogenetic origin of the gastrÆa in the light of this ontogenetic process, we may assume that the one-layered cell-community of the blastÆa began to take in food more largely at one particular part of its surface. Natural selection would gradually lead to the formation of a depression or pit at this alimentary spot on the surface of the ball. The depression would grow deeper and deeper. In time the vegetal function of taking in and digesting food would be confined to the cells that lined this hole; the other cells would see to the animal functions of locomotion, sensation, and protection. This was the first division of labour among the originally homogeneous cells of the blastÆa.

The effect, then, of this earliest histological differentiation was to produce two different kinds of cells—nutritive cells in the depression and locomotive cells on the surface outside. But this involved the severance of the two primary germinal layers—a most important process. When we remember that even man’s body, with all its various parts, and the body of all the other higher animals, are built up originally out of these two simple layers, we cannot lay too much stress on the phylogenetic significance of this gastrulation. In the simple primitive gut or gastric cavity of the gastrula and its rudimentary mouth we have the first real organ of the animal frame in the morphological sense; all the other organs were developed afterwards from these. In reality, the whole body of the gastrula is merely a “primitive gut.” I have shown already (Chapters VIII and XIX) that the two-layered embryos of all the Metazoa can be reduced to this typical gastrula. This important fact justifies us in concluding, in accordance with the biogenetic law, that their ancestors also were phylogenetically developed from a similar stem-form. This ancient stem-form is the gastrÆa.

The gastrÆa probably lived in the sea during the Laurentian period, swimming about in the water by means of its ciliary coat much as free ciliated gastrulÆ do to-day. Probably it differed from the existing gastrula only in one essential point, though extinct millions of years ago. We have reason, from comparative anatomy and ontogeny, to believe that it multiplied by sexual generation, not merely asexually (by cleavage, gemmation, and spores), as was no doubt the case with the earlier ancestors. Some of the cells of the primary germ-layers probably became ova and others fertilising sperm. We base these hypotheses on the fact that we do to-day find the simplest form of sexual reproduction in some of the living gastrÆads and other lower animals, especially the sponges.

The fact that there are still in existence various kinds of gastrÆads, or lower Metazoa with an organisation little higher than that of the hypothetical gastrÆa, is a strong point in favour of our theory. There are not very many species of these living gastrÆads; but their morphological and phylogenetic interest is so great, and their intermediate position between the Protozoa and Metazoa so instructive, that I proposed long ago (1876) to make a special class of them. I distinguished three orders in this class—the Gastremaria, Physemaria, and Cyemaria (or Dicyemida). But we might also regard these three orders as so many independent classes in a primitive gastrÆad stem.

The Gastremaria and Cyemaria, the chief of these living gastrÆads, are small Metazoa that live parasitically inside other Metazoa, and are, as a rule, 1/50 to 1/25 of an inch long, often much less (Fig. 233, 1–15). Their soft body, devoid of skeleton, consists of two simple strata of cells, the primary germinal layers; the outer of these is thickly clothed with long hair-like lashes, by which the parasites swim about in the various cavities of their host. The inner germinal layer furnishes the sexual products. The pure type of the original gastrula (or archigastrula, Fig. 29 I) is seen in the Pemmatodiscus gastrulaceus, which Monticelli discovered in the umbrella of a large medusa (Pilema pulmo) in 1895; the convex surface of this gelatinous umbrella was covered with numbers of clear vesicles, of 1/25 to 1/8 inch in diameter, in the fluid contents of which the little parasites were swimming. The cup-shaped body of the Pemmatodiscus (Fig. 233, 1) is sometimes rather flat, and shaped like a hat or cone, at other times almost curved into a semi-circle. The simple hollow of the cup, the primitive gut (g), has a narrow opening (o). The skin layer (e) consists of long slender cylindrical cells, which bear long vibratory hairs; it is separated by a thin structureless, gelatinous plate (f) from the visceral or gut layer (i), the prismatic cells of which are much smaller and have no cilia. Pemmatodiscus propagates asexually, by simple longitudinal cleavage; on this account it has recently been regarded as the representative of a special order of gastrÆads (Mesogastria).

Probably a near relative of the Pemmatodiscus is the Kunstleria Gruveli (Fig. 233, 2). It lives in the body-cavity of Vermalia (Sipunculida), and differs from the former in having no lashes either on the large ectodermic cells (e) or the small entodermic (i); the germinal layers are separated by a thick, cup-shaped, gelatinous mass, which has been called the “clear vesicle” (f). The primitive mouth is surrounded by a dark ring that bears very strong and long vibratory lashes, and effects the swimming movements.

Pemmatodiscus and Kunstleria may be included in the family of the Gastremaria. To these gastrÆads with open gut are closely related the Orthonectida (Rhopalura, Fig. 233, 3–5). They live parasitically in the body-cavity of echinoderms (Ophiura) and vermalia; they are distinguished by the fact that their primitive gut-cavity is not empty, but filled with entodermic cells, from which the sexual cells are developed. These gastrÆads are of both sexes, the male (Fig. 3) being smaller and of a somewhat different shape from the oval female (Fig. 4).

The somewhat similar Dicyemida (Fig. 6) are distinguished from the preceding by the fact that their primitive gut-cavity is occupied by a single large entodermic cell instead of a crowded group of sexual cells. This cell does not yield sexual products, but afterwards divides into a number of cells (spores), each of which, without being impregnated, grows into a small embryo. The Dicyemida live parasitically in the body-cavity, especially the renal cavities, of the cuttle-fishes. They fall in several genera, some of which are characterised by the possession of special polar cells; the body is sometimes roundish, oval, or club-shaped, at other times long and cylindrical. The genus Conocyema (Figs. 7–15) differs from the ordinary Dicyema in having four polar pimples in the form of a cross, which may be incipient tentacles.

The classification of the Cyemaria is much disputed; sometimes they are held to be parasitic infusoria (like the Opalina), sometimes platodes or vermalia, related to the suctorial worms or rotifers, but having degenerated through parasitism. I adhere to the phylogenetically important theory that I advanced in 1876, that we have here real gastrÆads, primitive survivors of the common stem-group of all the Metazoa. In the struggle for life they have found shelter in the body-cavity of other animals.

The small Coelenteria attached to the floor of the sea that I have called the Physemaria (Haliphysema and Gastrophysema) probably form a third order (or class) of the living gastrÆads. The genus Haliphysema (Figs. 234, 235) is externally very similar to a large rhizopod (described by the same name in 1862) of the family of the Rhabdamminida, which was at first taken for a sponge. In order to avoid confusion with these, I afterwards gave them the name of Prophysema. The whole mature body of the Prophysema is a simple cylindrical or oval tube, with a two-layered wall. The hollow of the tube is the gastric cavity, and the upper opening of it the mouth (Fig. 235 m).

The two strata of cells that form the wall of the tube are the primary germinal layers. These rudimentary zoophytes differ from the swimming gastrÆads chiefly in being attached at one end (the end opposite to the mouth) to the floor of the sea.

In Prophysema the primitive gut is a simple oval cavity, but in the closely related Gastrophysema it is divided into two chambers by a transverse constriction; the hind and smaller chamber above furnishes the sexual products, the anterior one being for digestion.

The simplest sponges (Olynthus, Fig. 238) have the same organisation as the Physemaria. The only material difference between them is that in the sponge the thin two-layered body-wall is pierced by numbers of pores. When these are closed they resemble the Physemaria. Possibly the gastrÆads that we call Physemaria are only olynthi with the pores closed. The Ammoconida, or the simple tubular sand-sponges of the deep-sea (Ammolynthus, etc.), do not differ from the gastrÆads in any important point when the pores are closed. In my Monograph on the Sponges (with sixty plates) I endeavoured to prove analytically that all the species of this class can be traced phylogenetically to a common stem-form (Calcolynthus).

The lowest form of the Cnidaria is also not far removed from the gastrÆads. In the interesting common fresh-water polyp (Hydra) the whole body is simply an oval tube with a double wall; only in this case the mouth has a crown of tentacles. Before these develop the hydra resembles an ascula (Figs. 236, 237). Afterwards there are slight histological differentiations in its ectoderm, though the entoderm remains a single stratum of cells. We find the first differentiation of epithelial and stinging cells, or of muscular and neural cells, in the thick ectoderm of the hydra.

In all these rudimentary living coelenteria the sexual cells of both kinds—ova and sperm cells—are formed by the same individual; it is possible that the oldest gastrÆads were hermaphroditic. It is clear from comparative anatomy that hermaphrodism—the combination of both kinds of sexual cells in one individual—is the earliest form of sexual differentiation; the separation of the sexes (gonochorism) was a much later phenomenon. The sexual cells originally proceeded from the edge of the primitive mouth of the gastrÆad.