Another example. The lungs of a mammal and the gills of a fish are analogous organs, since they have the same function of aëration of the blood. But they are not at all homologous: they are not built on the same plan; by no effort of the mind can we imagine that the former could have come out of the latter by modification. On the contrary, we have positive proof that it did not so come. But there is an organ in the fish which is homologous with the mammalian lung, viz., the air-bladder, or swim-bladder. We know it—1. Because we can trace in the taxonomic series all the gradations from the one to the other. In most fishes the air-bladder is wholly cut off from the gullet, and only very feebly supplied with blood. It is used and can be used only for flotation. In others, as the gar-pike, the swim-bladder is quite vascular and opens by a tube into the throat. Through this opening air is gulped down from time to time into the bladder, and again from time to time expelled. In other words, this fish supplements its gill-breathing by an imperfect lung-breathing. We have here the beginning of a lung. In still other fishes, viz., the Dipnoi (lepidosiren and ceratodus, Fig. 2), the air-bladder becomes a more perfect lung—i. e., a very vascular sacculated sac; and there is not only an opening into the throat, but also from the throat to the snout. In other words, we have for the first time nostrils. These fishes completely combine gill-breathing with lung-breathing. The step from these to the lowest amphibian reptiles is so small, that some have classed the lepidosiren among amphibians instead of fishes. The siredon or axolotl of New Mexico, the necturus or menobranchus of our Northern lakes, and the siren of our Southern swamps, have both gills and lungs, and breathe both air and water; but the lung is very imperfect, being only a sacculated sac, like the air-bladder of the ceratodus and lepidosiren. No one doubts that the air-breathing organ of an amphibian is a true lung; yet we have traced all the gradations between it and the air-bladder of a fish. We conclude, therefore, that if there be any such thing as transmutation of organic forms, the lung of higher animals must have been formed by the process above described.20

But we know it still more certainly—2. Because we can trace the change from the one to the other in the ontogenic series. In the life-history of the individual we can actually see the one thing change into the other. The frog, as is well known, when first hatched, is a tadpole. It has no legs, but locomotes by means of a vertically-expanded tail. It has no lungs, but breathes water instead of air, by means of gills. It is in all respects, therefore, a fish, and would be classed as such if it remained in this condition. But it does not; it gradually loses its tail and gills, and acquires legs and lungs, and breathes air only. Now in this change whence came the lungs? From the gills by modification? No; but from an organ similar in character and position to the air-bladder of a ceratodus, or a lepidosiren. This organ has gradually developed into a lung. The steps of the change are briefly as follow: First, the breathing is wholly water-breathing by gills. Next, by the development of this other organ, it is partly water-breathing by gills, and partly air-breathing by lungs. Lastly, the gills gradually dry up, and the lungs develop more and more, until the breathing is wholly by lungs.

We have dwelt somewhat upon this example, because it is an excellent example of what we mean by homology, and also because we will have occasion to use it again. But so important, for all that follows in this part, is a clear idea on the subject of homology, that it will be best to familiarize the mind of the reader with it by means of a few examples drawn from plants.

A potato is analogous to a root—a tuberous root like that of a dahlia or a sweet-potato—but is not at all homologous with these. On the contrary, it is homologous with a stem. It is essentially an underground, leafless branch, which has thickened enormously at the point by accumulation of starch. The evidence of this is found in the fact that it has rudimentary leaves (scales) arranged in regular spiral order of phylotaxis, each with its axillary bud (eyes). It is still more clearly shown by the fact that buds above-ground which, if let alone, would form leafy branches, may be made to become tubers by covering them with earth or dead leaves, and thus excluding the light; and, conversely, underground buds which, if let alone, would form tubers, may be made to grow into leafy branches by exposing them to the light.

Take another example: The broad, flat, elliptical, green masses so characteristic of the cactus family, and usually called their leaves, are indeed analogous to leaves in color, form, and function; for they are green and flat, and assimilate carbonic acid and water (CO₂ and H₂O) like leaves. But they are not, in truth, leaves, but modified stems, for they have the essential structure of stems, with their pith, wood, medullary rays, and bark, and may be traced through all gradations into the ordinary cylindrical form of stems. Where are their leaves, then? Their spines are their abortive leaves. These are arranged spirally like leaves, and bear buds in their axils like leaves. They are, in truth, leaves, modified to perform the function of defensive armor; while their function has been delegated to the stem flattened for this purpose.

One more example: The acacias, of which there are fifteen to twenty species in California, introduced from Australia, form two groups having extremely different styles of leaves. We will call them the feather-leaved and the simple-leaved acacias. In the former, the leaves are very finely bipinnate, and the general aspect of the foliage is extremely feathery and graceful. In the latter the leaves are simple, ovate, and, curiously enough, set on edge; and the general aspect of the tree is therefore rather stiff. It seems at first incredible that leaves so different and aspects so diverse should belong to plants of the same genus. But a little close examination shows that, as usual, the botanists are right and the popular judgment wrong. The plumose-leaf is the normal leaf-form for this genus. The simple leaf is not only abnormal, but in a homological sense is not a leaf at all—i. e., it does not correspond to the part called the blade in ordinary simple leaves of other trees. In the seedling of the simple-leaved acacias, and sometimes for a considerable time in the young tree, the leaves are all plumose. As the tree matures it gradually changes its dress and puts on its toga virilis. The gradual change from the one form to the other may easily be traced in the same tree, and even often in the same branch (Fig. 3). The steps of the change (a, b, c, and d) are shown in the following figure, drawn from nature. It is seen, by bare inspection of the figure, that the so-called leaf, d, of the simple-leaved acacias, is really the vertically-expanded leaf-stalk, l, s, the true leaf or blade being wholly aborted. The whole structure of this so-called leaf is different from that of a true blade. For example, its style of ribbing is parallel, its position is edgewise to the sky, its palisade cells are on both sides alike, etc. To emphasize this difference, botanists call such an apparent leaf a phyllodium, or phyllode.

After these illustrations we now repeat the definitions in different words. Analogy has reference to general resemblance of form determined by similarity of function, however different the origins of the parts compared may be. Homology has reference to community of origin, however obscured to the superficial observer such common origin may be by modifications necessary to adapt to different functions. Observe, then, there are two ideas here which must be kept distinct. One is common origin, always shown by deep-lying, essential identity of structure; the other is adaptive modification for function. Organs of the most diverse origin may resemble by adaptive modification for the same function. This is analogy. Organs of the same origin may assume very different appearance by adaptive modifications for different functions. This is homology. In the latter case, which is the one that concerns us, a profound study of essential structure and structural relations to other parts, and especially extensive comparison in the taxonomic and ontogenic series, will usually detect the homology, or common origin, in spite of the obscurations produced by adaptive modifications. It is seen, also, that analogy is a superficial resemblance, easily detected by the popular eye, and therefore embodied in popular language; while homology is a deep-seated and essential resemblance, detected often only by profound study and extensive comparison. Now, one of the strongest proofs of the truth of evolution is taken from the homologies of animal structure. Common origin completely explains homology. Every other explanation is transcendental, and therefore unscientific.

Primary Divisions of the Animal Kingdom.—Now, the animal kingdom consists of several primary divisions, called sub-kingdoms or departments. The animals in these groups differ so essentially from one another in their plan of structure, that it is difficult, if not impossible, to trace any structural relation between them—to imagine how the members of one could have been derived from those of another—or conceive the common stem from which they all separated. In other words, it is impossible, in the present state of knowledge, to trace homology with any certainty from one group to another. But within the limits of each primary group the homology is easy. Some naturalists—Agassiz and Cuvier—have made four or five of these primary groups. Some—Huxley—have made eight. Some make nine or ten.21 We will not trouble ourselves to settle this question; for all agree to make vertebrata and articulata or arthropoda two of them, and all our illustrations will be drawn from these. Other groups are too unfamiliar to the general reader to serve our purpose.

Now, as already stated, homology can not be traced with any certainty between the primary groups, but within the limits of each group it may be traced with ease and beauty. Analogy, however, being connected with function, and function being universal, can be traced throughout the animal kingdom. While, therefore, it is probable, nay, almost certain, that all animals have had a common origin, we can not yet trace these great departments by homology to that common origin. But the common origin of each department is quite clear. For example, the structure of all vertebrate animals is precisely such as would be the case if all came from one primal vertebrate, variously modified to adapt to various modes of life. Also, the structure of all arthropods is precisely such as would be if all came from one primal arthropod, which, from generation to generation, became gradually modified in different directions, in order to adapt itself to various modes of life. But between arthropods and vertebrates we can not yet clearly see a common origin, although there doubtless was such.

These great departments may, therefore, be compared to natural styles of animal architecture. As there are various styles of human architecture—Oriental, Egyptian, Greek, Gothic—each of which may be variously modified to adapt it to all the different purposes for which buildings are made, without destroying, though perhaps obscuring, the integrity of the style; so the different primary groups or departments may be regarded as different styles of animal structure, each of which may be and has been modified in many ways to adapt it to various habits and modes of life, obscuring but not destroying the general style. Or they may be compared to natural machines. As a steam-engine, by modification, may be adapted to many kinds of purposes, obscuring, perhaps, but not destroying the essential identity of structure; even so the vertebrate machine by modification may be, and has been, adapted to many kinds of purposes, and thus become a swimming-machine, a crawling-machine, a flying-machine, a running-and leaping-machine, without destroying, although obscuring, the essential identity of structure. As in architecture, æsthetic principles of form may be traced through each style, but not from style to style, while the mechanical principles of construction run through all alike; so also in animal architecture, the laws of form and styles of structure are traceable with ease only within the limits of each primary group, while the laws of function are traceable through all groups alike. Or, again, and finally: Each of these departments may be compared to a tree, with branches, twigs, and spray, all obviously coming from one common stem, but each stem seems separate. They are, indeed, probably, themselves only great branches of one common trunk, but their connection is too remote and obscure to be made out clearly by means of homology. Other evidences, however, drawn from other sources, as we shall see hereafter, are not wholly wanting.

CHAPTER V.
PROOFS FROM HOMOLOGIES OF THE VERTEBRATE SKELETON.

3. All vertebrates, and none other, have a number of their anterior vertebral joints enlarged and consolidated into a box to form the skull,22 in order to inclose and protect a similar enlargement of the nervous center, viz., the brain; and also usually, but not always, a number of posterior joints, enlarged and consolidated to form the pelvis, to serve as a firm support to the hind-limbs.

4. All vertebrates, and none other, have two cavities, inclosed and protected by the skeleton, viz., the neural cavity above, and the visceral or body cavity below, the vertebral column; so that a cross-section of the body is diagrammatically represented by Fig. 4.

5. All vertebrates, with few exceptions, and no other animals, have two and only two pair of limbs. The exceptions are of two kinds, viz.: a, some lowest fishes, amphioxus and lampreys, which probably represent the vertebrate condition before limbs were acquired; and b, degenerate forms like snakes and some lizards, which have lost their limbs by disuse.

So much concerns the general plan of skeletal structures, and is strongly suggestive of—in fact, is inexplicable without—common origin. But much more remains which is not only suggestive, but demonstrative of such origin. By extensive comparison in the taxonomic and ontogenic series, the whole vertebrate structure in all its details in different animals may be shown to be modifications one of another. Sometimes a piece is enlarged, sometimes diminished, or even becomes obsolete; sometimes several pieces are consolidated into one; but, in spite of all these obscurations, corresponding parts may usually be made out. This is the main subject of this chapter.

Special Homology of Vertebrate Limbs.—It would lead us much too far into unfamiliar technicalities to take up the whole skeleton. We select the limbs, both because their general structure is more familiar, and because in them the two fundamental ideas of essential identity and of adaptive modification are both admirably illustrated. The reason of this is, that it is by the limbs that the organism chiefly reacts on the environment, and is modified by it.

Fore-limbs.—In the accompanying figures (Figs. 5–18) we have represented, side by side, the fore-limbs of many vertebrates, taken from all the classes—mammals, birds, reptiles, and fishes. For convenience of comparison, the corresponding parts are similarly lettered in all. Also, in order to identify easily certain important corresponding segments, we have drawn through them a continuous dotted line. In man, nearly all the parts are present, and his limbs, therefore, may be taken as a term of comparison; for man’s structure, except his brain, is far less modified than that of many animals.

Note, then, the following points: 1. The collar-bone (clavicle) is associated with wide separation of the shoulders, and the free use of the fore-limb for prehension or for flight, but is gradually lost in proportion as the fore-limb is brought nearer together and used for support, because it is no longer wanted. I say gradually, for all the steps of the passing away may be found. The useless rudimentary condition is not uncommon.

2. The coracoid (c), it is seen, is a small, beak-like process of the blade-bone (scapula) in man and mammals; but in birds (Fig. 11) and reptiles (Figs. 14, 18) it is a separate bone as large as the blade-bone itself, jointed with the latter at the shoulder and with the breast-bone (sternum) in front, thus making together a strong shoulder-girdle for the attachment of the fore-limb. This was undoubtedly the condition in the original or earliest walking animal, viz., reptiles. It was inherited and retained by birds, because necessary for powerful action of the wings in flight. In mammals it gradually dwindled and became united with the blade-bone as a process. In one mammal, the lowest and most reptilian living—the ornithorhynchus—the coracoid is much like that of reptiles—a large, flat bone, separated from the blade-bone and articulated with the breast-bone. It is a significant fact that, in the mammalian embryo, it is first developed as a separate bone and afterward united with the scapula.

3. In man, monkeys, bears, and some other mammals, the limb is fairly free from the body and the elbow half-way down the limb; while in herbivores (Figs. 8, 9), such as the horse, ox, and deer, etc., the elbow is high on the side of the body, and the limb is free only from the elbow downward. Perhaps in these cases most observers do not recognize it as an elbow at all. All gradations between these extremes are easily traced. The free condition of the limb is evidently the original one, the condition in herbivores being an extreme modification associated with another modification mentioned under 5.

4. In man and in many mammals, and in all reptiles and birds, there are two bones in the forearm (radius and ulna). In the more specialized forms of hoofed animals (ungulates), such as horse and ruminants (Figs. 8, 9), there is apparently but one. Two is the normal and original number; but one of them, the ulna, has gradually become smaller and smaller, and finally is reduced to a short splint, and consolidated with the radius as a process extending backward to form the point of the elbow. In the horse family every step of this reduction and consolidation may be traced in the course of its geological history.

5. The wrist of many mammals and all birds differs in structure from that of man, chiefly in containing a smaller number of bones. The normal number, as in man, seems to be eight. The decrease takes place mainly by consolidation of two or more into one. In such cases usually the embryo will show the bones still separate, thus revealing the ancestral condition. Again, the position of the wrist is noteworthy. In man, monkeys, the bear family, and several other mammalian families, and in all reptiles, the hand bends forward at the wrist, so that the tread is on the whole palm (palmigrade). But, in all the most specialized mammals, the wrist can not bend in this direction, and therefore this joint can not be brought to the ground. The tread is therefore on the toes (digitigrade), and the wrist is high up above the ground. In the horse (Fig. 9), the ox, and many other mammals, for example, the wrist is so high that it is not usually recognized as a wrist, and is often called the fore-knee. Now, homologous parts ought to have the same scientific name; but to use the word “hand” in the case of lower animals might produce confusion and misconception. Therefore it has been agreed among comparative anatomists to use instead the Latin word “manus” for all that corresponds, in any animal, to the hand of man—i. e., all from the wrist downward. The manus of a horse is about fifteen inches long. The manus of a pterodactyl, such as that found by Marsh in the cretaceous strata of the West, with an expanse of wings of twenty-five feet, was probably not less than seven or eight feet long.

6. The number of palm-bones (metapodal) and toes deserves special notice. In fishes, and in some extinct swimming reptiles, these are or were very numerous, but in the earliest land-animals they became five. This is the number now in nearly all reptiles, and in all the more generalized mammals. It may be called the normal number for a walking animal. In very many mammals, such, for example, as the dog family, they are reduced to four, though the fifth often remains as a useless, rudimentary splint and dew-claw (Fig. 6), thus showing the process of dwindling in the ancestry. In hoofed animals the process of gradual diminution is shown even in existing forms, and still better in extinct forms. Confining ourselves, now, only to existing forms, in the elephant there are five palm-bones and toes, and in the hippopotamus there are four, all functional. In the hog (Fig. 7) there are still four, but two are behind the others and much smaller, and do not touch the ground—are not functional unless in soft ground. In the cow, deer, etc., the palm-bones are reduced to two, and these are consolidated into one (canon-bone), and the toes are reduced to two efficient and two useless rudiments. In the sheep and the goat (Fig. 8) these useless rudiments are dropped, and there are two only. Finally, in the horse (Fig. 9), the toes are reduced to one, although the palm-bones are still three, two of them, however, being reduced to rudimentary splints.

How is it with birds? Have these also palm-bones and fingers? Yes, in birds (Fig. 11) there are three palm-bones and three fingers (the fourth and fifth being wanting); one of them—the thumb—is free, and sometimes carries a claw. In the earliest known and most reptilian bird, the archæopteryx (Fig. 12), all the three fingers are free, have the full number of joints, and all of them carry claws. In the embryo of living birds the fingers are all free, as in the archæopteryx.

7. Observe, finally, as an admirable illustration of different adaptative modifications for the same purpose—flight—the structure of the manus of flying animals. In the bat (Fig. 10), the flat flying-plane is made by enormous elongation of the palm-bones and finger-bones, their wide separation and the stretching of a thin membrane between them. In the pterosaurs, or extinct flying reptiles (Fig. 13), one finger only is greatly enlarged and elongated, and the flying-membrane is stretched between it and the hind-leg (Fig. 19), while the other three fingers are free and provided with claws. If it be asked which finger is it that is so greatly enlarged in this animal, we answer, it is the little finger. In birds, on the contrary, the manus is consolidated to the last degree, to form a strong basis for attachments for the quills which form the flying-plane, and which are themselves extreme modifications of the scales of reptiles. But throughout all these extreme modifications the same essential structure is detectable.

It is perhaps unnecessary to dwell upon the still greater modifications of limbs for swimming, as in the whale (Fig. 16), the ichthyosaur, mosasaur (Fig. 18), and the fish (Fig. 17). A careful inspection of the figures, after what we have said, will be sufficient to explain them. In the fish alone the upper segments of the limb, viz., shoulder-girdle and humerus, are wanting, not being yet introduced, and the manus is not yet differentiated into palm-bones and fingers, and the fingers are indefinitely multiplied. All these characters are indications of low position in the scale of evolution. The earliest vertebrates were fishes. Limbs were not yet completely formed. In embryos of higher animals, also, the outer segments are first formed.

Hind-Limbs.—Figs. 20 to 24 represent, in a similar way, the hind-limbs of several animals—in this case all mammals. As before, corresponding parts are similarly lettered, and a dotted line is carried through certain prominent parts, especially the knee, heel, instep, and toes. By careful inspection the figures explain themselves. Nevertheless, it will be well to draw special attention to several of the more important points:

1. See, then, the position of the knee. The thigh-bone in man, monkeys, bears, and several other families of mammals, and all reptiles, is free from the body, and the knee is far removed and half-way down the limb (Figs. 20, 21). This is undoubtedly the original and normal condition of land-animals. But in all the more highly specialized and swifter animals the knee is brought nearer and nearer to the body, until, in the swiftest of all, such as the ruminants and the horse (Figs. 23, 24), it is high up on the side of the body, in the middle of what is usually called the thigh but which really includes the thigh and the upper part of the lower leg or shank.

2. See, again, the position of the heel. In man, monkey, bear, and many other mammals, and all living reptiles, the heel is on the ground, the tread is on the whole foot, plantigrade; while in all the more specialized and agile animals, and especially in the swiftest of all, such as the horse, the deer, etc., the heel is high in the air, and the tread is digitigrade.

3. Observe, again: there are two degrees of digitigradeness. The one we find in carnivorous or clawed digitigrades, the other in herbivores or hoofed digitigrades. In the one the tread is on the whole length of the toes to the balls, as in man when he tip-toes; in the other the tread is on the tip of the last joint alone. All that in any animal corresponds to the foot of a man—i. e., from the hamstring and heel downward—is called, in comparative anatomy, the “pes.” The pes, or foot of a horse, is eighteen inches long. It is easy to see what spring and activity this mode of treading gives to an animal. Think how helpless a horse would be if he trod on the whole foot, heel down!

4. Observe, again, the number of toes. In the process of specialization there is a tendency for these to become fewer and stronger.23 The normal number, as already seen, is five. All the earliest mammals, and many orders of mammals still living, have five; but in the most specialized orders, such as the ungulates or hoofed animals, they were steadily reduced in number in the course of evolution. In the elephant there are still five, in the hippopotamus there are four, in the rhinoceros three, in the goat two, in the horse one. Still more the order of the dropping is regular. If an animal have but four toes, it is usually the first, or great toe, or thumb, that is wanting, or may be rudimentary. If, as in the rhinoceros, there are only three, then No. 5, or little toe, is also wanting, and the existing toes are Nos. 2, 3, and 4. If an animal has only two toes, as the goat, these are Nos. 3 and 4; and if only one, as the horse, it is the third or middle toe. Or, to put it more definitely: hoofed animals are divided into two groups, even-toed (artiodactyl) and odd-toed (perissodactyl). The even-toed may have four, as in the hippopotamus; or two, as in the goat. The odd-toed may have three, as in the rhinoceros; or but one, as in the horse. Now, both of these orders came by differentiation, far back in the Eocene Tertiary, from a five-toed plantigrade ancestor. After dropping No. 1 (thumb or great toe) it is not yet decided, so far as number of toes is concerned, whether the resulting four-toed animal shall become artiodactyl or perissodactyl. If the former, then the two side-toes (Nos. 2 and 5) become shortened up, as in the hog; then rudimentary, as in the ox and deer; and finally pass away entirely, as in the goat. If, on the other hand, the four-toed animal is on the line of perissodactyl evolution, it becomes first a three-toed animal by dropping No. 5. Now, the two side-toes (Nos. 2 and 4) shorten up more and more, and the middle toe increases in size, until finally, in the modern horse, only the greatly enlarged middle toe (No. 3) remains. We look with wonder and admiration at the danseuse pirouetting on the point of one toe. The horse is performing this feat all the time. Yes, the one toe of a horse has all the three joints like ours. The coffin-bone is the last joint, and the hoof is the nail.

Genesis of the Horse.—Every step of this process on the perissodactyl line may be traced in the history of the genesis of the horse. The beautiful form and structure of this animal were not made at once, but by a slow process of integration of small changes from generation to generation, and from epoch to epoch of the earth’s history. The horse (as in fact did all ungulates) came from a five-toed plantigrade ancestor, but we are not able to trace the direct line of genesis quite so far. The earliest stage that we can trace with certainty, in this line of descent, is found in the eohippus of Marsh. This was a small animal, no bigger than a fox, with three toes behind and four serviceable toes in front, with an additional fifth palm-bone (splint), and perhaps a rudimentary fifth toe like a dew-claw. This was in early Eocene times. Then, in later Eocene, came the orohippus, which differs from the last chiefly in the disappearance of the rudimentary fifth toe and splint. (See Fig. 25.) Next, in the Miocene, came the mesohippus and miohippus. These were larger animals (about the size of a sheep), and had three serviceable toes all around; but in the former the rudiment of a fourth splint in the fore-limb yet remained. Then, in the Miocene, came the protohippus and pliohippus. These were still larger animals, being about the size of an ass. In the former the two side-toes were shortening up and the middle toe becoming larger. In the latter the two side-toes have become splints. Lastly, only in the Quaternary comes the genus Equus, or true horse. The size of the animal is become greater, the middle toe stronger, the side-splints smaller; but in the side-splints of the modern horse we have still remaining the evidence of its three-toed ancestor.

Similar gradual changes may be traced in the two leg-bones, which have gradually consolidated into one; in the teeth, which have become progressively longer and more complex in structure, and therefore a better grinder; in the position of the heel and wrist, which have become higher above-ground; in the general form, which has become more graceful and agile; and, lastly, in the brain, which has become progressively larger and more complex in its convolutions—to give greater battery-power, to make a more powerful dynamo—to work the improved skeletal machine. See, then, how long it has taken Nature to produce that beautiful finished article we call the horse!

We have taken only limbs as examples of what is true of the whole skeleton. To the superficial observer the bodies of animals of different classes seem to differ fundamentally in plan—to be entirely different machines, made each for its own purposes, at once, out of hand. Extensive comparison, on the contrary, shows them to be the same, although the essential identity is obscured by adaptive modifications. The simplest, in fact the only scientific, explanation of the phenomena of vertebrate structure is the idea of a primal vertebrate, modified more and more through successive generations by the necessities of different modes of life.

See, then, in conclusion, the difference between man’s mode of working and Nature’s. A man having made a steam-engine, and desiring to use it for a different purpose from that for which it was first designed and used, will nearly always be compelled to add new parts not contemplated in the original machine. Nature rarely makes new parts—never, if she can avoid it—but, on the contrary, adapts an old part to the new function. It is as if Nature were not free to use any and every device to accomplish her end, but were conditioned by her own plans of structure; as, indeed, she must be according to the derivation theory. For example: In early Devonian times fishes were the only representatives of the vertebrate type of structure. The vertebrate machine was then a swimming-machine. In the course of time, when all was ready and conditions were favorable, reptiles were introduced. Here, then, is a new function—that of locomotion on land. We want a walking-machine. Shall we have a new organ for this new function? No: the old swimming-organ is modified so as to adapt it for walking. Time went on, until the middle Jurassic, and birds were introduced. Here is a new and wonderful function, that of flying in the air. We want a flying-machine. We know how man would have done this; for we have the result of his imagination in angels of Christian art and griffins of Greek mythology. He would have added wings to already existing parts, and this would have necessitated the alteration of the whole plan of structure, both skeletal and muscular. Nature only modifies the fore-limbs for this new purpose. If we must have wings, we must sacrifice fore-legs. We can not have both without violating the laws of morphology. Finally, ages again passed, and, when time was fully ripe, man was introduced. Now we want some part to perform a new and still more wonderful function. We want a hand, the willing and efficient servant of a rational mind. We know, again, how man would have done this, for we have the result in the centaurs of Greek mythology, in which man’s chest, and arms, and head are added to the body of a quadruped. But natural laws must not be violated, even for man. If we want hands, we must sacrifice feet. Again, therefore, the fore-limbs are modified for this new and exquisite function. Thus, in the fin of a fish, the fore-paw of a reptile or a mammal, the wing of a bird, and the arm and hand of a man, we have the same part, variously modified for many purposes.

Many other illustrations might be taken from the skeleton and from other systems, especially the muscular and nervous. But in the muscular system the modifications have been so extreme that homology is much more difficult to trace, and therefore requires more extensive knowledge than we yet possess, and more extended comparison than has yet been attempted. It has been traced with some success through mammals, and probably will be through air-breathing vertebrates—i. e., also through birds, reptiles, and amphibians; but to trace it through fishes seems almost hopeless. In the case of the nervous system, and especially of the brain, it is again distinct; but this had better be taken up under another head, viz., proofs from ontogeny, Chapter VI.

In the visceral organs homology is very plain, in fact too plain. There is not modification enough in most cases even to obscure it, because function is the same in all animals. These organs do not, therefore, furnish good illustrations of that essential identity in the midst of adaptive modification which constitutes the argument for the derivative origin of structure. It is the organs of animal life that show this most perfectly, because it is these that take hold on the environment and are modified by it. There are, however, a few striking illustrations to be found among the visceral organs, especially the blood-system. This, however, had better also be deferred to the chapter on ontogeny.

CHAPTER VI.
HOMOLOGIES OF THE ARTICULATE SKELETON.

Here, then, we have an entirely different plan of structure—a different style of architecture and different mechanical principles of machinery. Instead of a skeleton within and muscles acting on the outside, we have the skeleton on the outside, and muscles acting from within. Instead of two cavities, a neural and visceral, the skeleton forms but one cavity, in which all organs are inclosed and protected. Instead of finding the nerve-axis on the dorsal aspect of the body, we find it on the ventral aspect.