Popular Science Monthly/Volume 61/October 1902/Origin of the Fins of Fishes
| ORIGIN OF THE FINS OF FISHES. |
By President DAVID STARR JORDAN,
LELAND STANFORD JR. UNIVERSITY.
ONE of the most interesting problems in vertebrate morphology, and one of the most important from its wide-reaching relations, is that of the derivation of the fins of fishes. This resolves itself at once into two problems, the origin of the median fins, which appear in the lancelets, at the very bottom of the fish-like series, and the origin of the paired fins or limbs, which are much more complex, and which first appear with the primitive sharks.
In this study the problem is to ascertain not what theoretically should happen, but what, as a matter of fact, has happened in the early history of the fish-like groups. That these structures, with the others in the fish body, have sprung from simple origins, growing more complex with the demands of varied conditions, and then at times again simple, through degeneration, there can be no doubt. It is also certain that each structure must have had some element of usefulness in all its stages. In such studies we have, as Haeckel has expressed it, 'three ancestral documents, paleontology, morphology and ontogeny,'—the actual history as shown by fossil remains, the side-light derived from comparison of structures, and the evidence of the hereditary influences shown in the development of the individual. As to the first of these ancestral documents, the evidence of paleontology is conclusive where it is complete. But in very few cases are we sure of any series of details. The records of geology are like a book with half its leaves torn out, the other half confused, displaced and blotted.
The evidence of comparative anatomy is most completely secured, but it is often indecisive as to relative age and primitiveness of origin among structures. As to ontogeny, it is, of course, true that through heredity, 'the life history of the individual is an epitome of the life history of the race.' 'Ontogeny repeats phytogeny,' and phylogeny, or line of descent of organisms and structures, is what we are seeking. But here the repetition is never perfect, never so perfect in fact as Haeckel and his followers expected to find it. The demands of natural selection may lead to the lengthening, shortening, or distortion of phases of growth, just as they may modify adult conditions. The conditions of the individual development may, therefore, furnish evidence in favor of certain theories of origins, but they cannot alone furnish the absolute proof.
Origin of the Median or Vertical Fins.
In the process of development the median or vertical fins are doubtless older than the paired fins or limbs, whatever be the origin of the latter. They arise in a dermal keel which is developed in a web fitting and accentuating the undulatory motion of the body. In the embryo of the fish the continuous vertical fin from the head along the back and around the tail precedes any trace of the paired fins.
In this elementary fin-fold slender supports, the rudiments of finrays, tend to appear at intervals. These are called by Eyder ray-hairs or actinotrichia. They are the prototype of fin-rays in the embryo fish, and doubtless similarly preceded the latter in geological time. In the development of fishes, the caudal fin becomes more and more the seat of propulsion. The fin-rays are strengthened, and their basal supports are more and more specialized.
That the vertical fins, dorsal, anal and caudal, have their origin in a median fold of the skin admits of no very serious question. In the lowest forms which bear fins these structures are dermal folds, being supported by very feeble rays. Doubtless, at first the vertical fins formed a continuous fold, extending around the tail, this fold ultimately broken by atrophy into distinct dorsal, anal and caudal fins. In the lower fishes, as in the earlier sharks, there is an approach to this condition of primitive continuity, and in the embryos of almost all fishes the same condition occurs. Dr. John A. Eyder points out the fact that there are certain unexplained exceptions to this rule. The sea-horse, pipe-fish and other highly modified forms do not show this unbroken fold, and it is wanting in the embryo of the top-minnow, Gambusia affinis. Nevertheless, the existence of a continuous vertical fold in the embryo is the rule, almost universal. The codfish with three dorsals, the Spanish mackerel with dorsal and anal finlets, the herring with one dorsal, the stickleback with a highly modified one, all show this character and we may well regard it as a certain trait of the primitive fish. This fold springs from the ectoblast or external series of cells in the embryo. The fin-rays and bony supports of the fins spring from the mesoblast or middle series of cells, being thrust upward from the skeleton as supports for the fin-fold.
Origin of the Paired Fins.
The question of the origin of the paired fins is much more difficult and is still far from settled, although the majority of recent writers have favored the theory that these are parts of a once continuous lateral fold of skin, corresponding to the vertical fold which forms the dorsal, anal and caudal. In this view the lateral fold is soon atrophied in the middle, while at either end it is highly specialized, at first into an organ of direction, then into fan-shaped and later paddle-shaped organs of locomotion. Finally, from the jointed paddle, which Gegenbaur has called the archipterygium, there has developed, on the one hand, the rayed pectoral and ventral fins of ordinary fishes, and on the other, in land-creeping animals, jointed legs and arms. As to this the evidence of paleontology is conflicting. An early shark of the Devonian, Cladoselache,
has fan-shaped paired fins so formed and placed. Another shark almost as old, Pleuracanthus, of the Carboniferous, has fins which fit best a totally different theory of origin. Its jointed or archipterygial fin has no resemblance to a fold of skin, but accords better with Gegenbaur's theory that the pectoral limb was at first a modified septum or gill arch. Sharks still older than either (Heterodontidæ) in the Silurian, so far as we can judge by their teeth, are closely related to forms bearing the more specialized type of fin found in the typical
sharks of to-day. Evidently none of these three, as seen in the rocks, represents the real beginning of paired fins in the life of the past. As we shall see, the evidence of comparative anatomy may be consistent with either of these theories, while that of ontogeny or embryology is apparently inconclusive, and that of paleontology seems contradictory.
Development of the Paired Fins in the Embryo.
According to Dr. John A. Ryder ('Embryography of Osseous Fishes,' 1882) "the paired fins in Teleostei, like the limbs of the higher vertebrates, arise locally, not, however, as blunt processes but as short longitudinal folds, with perhaps a few exceptions. The pectorals of Lepisosteus originate in the same way. Of the paired fins, the pectoral or anterior pair seems to be the first to be developed, the ventral or pelvic pair often making its appearance until after the absorption of the yolk-sac has been completed, in other cases, before that event as in Salmo and in Gambusia. The ventral undergoes less alteration of position during its evolution than the posterior pair."
In the codfish (Gadus callarias) the pectoral fin-fold "appears as a slight longitudinal elevation of the skin on either side of the body of the embryo a little way behind the auditory vesicles, and shortly after the tail of the embryo begins to bud out. At the very first, it appears to be merely a dermal fold, and in some forms, a layer of cells extends out underneath it from the sides of the body but does not ascend into it. It begins to develop as a very low fold, hardly noticeable, and as growth proceeds, its base does not expand antero-posteriorly but tends rather to become narrowed, so that it has a pedunculated form. With the progress of this process the margin of the fin-fold also becomes thinner 
Pectoral Fin of Hepterodontus philippi. From Nature. at its distal border, and at the basal part mesodermal cells make their appearance more noticeably within the inner contour line. In some species I am quite assured that there is a mesodermal tract or plate of cells developed just behind the auditory vesicles, just outside the source of the mesodermal cells which are carried up into the pectoral fin-fold. This is developed at about the time of the closure of the blastoderm and these lateral mesodermal folds of tissue may be called the pectoral plates. The free border of the fin-fold grows out laterally and longitudinally, expanding the portion outside of the inner contour line of the fin into fan-shape. This distal thinner portion is at first without any evidence of rays; further than that there is a manifest tendency to a radial disposition of the histological elements of the fin."
The next point of interest is found in the change of position of the pectoral fin by a rotation on its base. This is associated with changes in the development of the fish itself. The ventral fin is also, in most fishes, a short horizontal fold and just above the preanal part of the median vertical fold which becomes anal, caudal and dorsal. But in the top-minnow (Gambusia), of the order Haplomi, the ventral first appears as 'a little papilla and not as a fold, where the body walls join the hinder upper portion of the yolk-sac, a very little way in front of the vent.' 'These two modes of origin,' observes Dr. Eyder, 'are therefore in striking contrast and well calculated to impress us with the protean character of the means at the disposal of Nature to achieve one and the same end.'
Current Theories.
There are three chief theories as to the morphology and origin of the paired fins:
The earliest is that of Gegenbaur, supported by numerous others, that these fins are derived from modified gill-arches or septa between the gill-openings. According to this theory, the skeletal arrangements of the vertebrate limb are derived from modifications of one primitive form, a structure made up of successive joints, with a series of fin-rays on each side of it. To this structure, Gegenbaur gives the name of archipterygium. It is found in the shark, Pleuracanthus, and in all the Dipnoan and Crossopterygian fishes, its primitive form being still retained in the Australian genus of Dipnoans, Neoceratodus. This biserial archipterygium with its limb girdle is derived from a series of gill rays attached to a branchial arch.
Professor J. Graham Kerr observes:
"The Gegenbaur theory of the morphology of vertebrate limbs thus consists of two very distinct portions. The first, that the archipterygium is the ground-form from which all other forms of presently existing fin skeletons are derived, concerns us only indirectly as we are dealing here only with the origin of the limbs, i. e., their origin from other structures that were not limbs.
"It is the second part of the view that we have to do with, that deriving the archipterygium, the skeleton of the primitive paired fin, from a series of gill-rays and involving the idea that the limb itself is derived from the septum between two gill clefts.
'This view is based on the skeletal structures within the fin. It rests upon: (1) The assumption that the archipterygium is the primitive type of fin, and (2) the fact that amongst the Selachians is found a tendency for one branchial ray to become larger than the others, and when this has happened, for the base of attachment of neighboring rays to show a tendency to migrate from the branchial arch on to the base of the larger or, as we may call it, primary ray; a condition coming about which, were the process to continue rather further than it is known to do in actual fact, would obviously result in a structure practically identical with the archipterygium. Gegenbaur suggests that the archipterygium actually has arisen in this way in phylogeny.
The theory of Balfour, adopted by Dohrn, Wiedersheim, Thacher, Mivart, Byder, Dean, Boulenger and others, and now generally accepted by most morphologists as plausible, is this: that "The paired limbs are persisting and exaggerated portions of a fin-fold once continuous, which stretched along each side of the body and to which they bear an exactly similar phylogenetic relation as do the separate dorsal and anal fins to the once continuous median fin-fold."
"This view, in its modern form, was based by Balfour on his observation that in the embryos of certain Elasmobranchs the rudiments of the pectoral and pelvic fins are at a very early period connected together by a longitudinal ridge of thickened epiblast—of which indeed they are but exaggerations. In Balfour's own words referring to these observations: 'If the account just given of the development of the limb is an accurate record of what really takes place, it is not possible to deny that some light is thrown by it upon the first origin of the vertebrate limbs. The facts can only bear one interpretation, viz., that the limbs are the remnants of continuous lateral fins.'
"A similar view to that of Balfour was enunciated almost synchronously by Thacher and a little later by Mivart—in each case based on anatomical investigation of Selachians—mainly relating to the remarkable similarity of the skeletal arrangements in the paired and unpaired fins."
A third theory is suggested by Mr. J. Graham Kerr (Cambridge Philos. Trans., 1899), who has given us the best recent summary of the theories on this subject. Mr. Kerr agrees with Gegenbaur as to the primitive nature of the archipterygium, but believes that it is derived, not from the gill-septum but from an external gill. Such a gill is well developed in the young of all the living sharks, Dipnoans and Crossopterygians, and in the latter types of fishes it has a form strikingly similar to that of the archipterygium, although without bony or cartilaginous axis.
We may now take up the evidence in regard to each of the different theories, using largely the language of Kerr, the paragraphs in quotation marks being taken from his paper. We may first consider Balfour's theory of the lateral fold.
Balfour's Theory of the Lateral Fold.
"The evidence in regard to this view may be classed under three heads, as Ontogenetic, Comparative Anatomical, and Paleontological. The ultimate fact on which it was founded was Balfour 's discovery that in certain Elasmobranch embryos, but especially in Torpedo (parcoatis), the fin rudiments were, at an early stage, connected by a ridge of epiblast. I am not able to make out what were the other forms in which Balfour found this ridge, but subsequent research, in particular by Mollier, a supporter of the lateral-fold view, is to the effect that it does not occur in such ordinary sharks as Pristiurus and Mustelus, while it is to be gathered from Balfour himself that it does not occur in Scyllium (Scylliorhinus).
"It appears to me that the knowledge we have now that the longitudinal ridge is confined to the rays and absent in the less highly specialized sharks, greatly diminishes its security as a basis on which to rest a theory. In the rays, in correlation with their peculiar mode of life, the paired fins have undergone (in secondary development) enormous extension along the sides of the body, and their continuty in the embryo may well be a mere foreshadowing of this.
"An apparently powerful support from the side of embryology came in Dohrn and Rabl's discoveries that in Pristiurus all the interpterygial myotomes produce muscle buds. This, however, was explained away by the Gegenbaur school as being merely evidence of the backward migration of the hind limb—successive myotomes being taken up and left behind again as the limb moved further back. As either explanation seems an adequate one, I do not think we can lay stress upon this body of facts as supporting either one view or the other. The facts of the development of the skeleton can not be said to support the fold view; according to it we should expect to find a series of metameric supporting rays produced which later on become fused at their bases. Instead of this we find a longitudinal bar of cartilage developing quite continuously, the rays forming as projections from its outer side.
"The most important evidence for the fold view from the side of comparative anatomy is afforded by: (1) The fact that the limb derives its nerve supply from a large number of spinal nerves, and (2) the extraordinary resemblance met with between the skeletal arrangements of paired and unpaired fins. The believers in the branchial-arch hypothesis have disposed of the first of these in the same way as they did the occurrence of interpterygial myotomes, by looking on the nerves received from regions of the spinal cord anterior to the attachment of the limb as forming a kind of trail marking the backward migration of the limb.
"The similarity in the skeleton is indeed most striking, though its weight as evidence has been recently greatly diminished by the knowledge that the apparently metameric segmentation of the skeletal and muscular tissues of the paired fins is quite secondary and does not at all agree with the metamery of the trunk. What resemblance there is may well be of a homoplastic character when we take into account the similarity in function of the median and unpaired fins, especially in such forms as Raja where the anatomical resemblances are especially striking. There is a surprising dearth of paleontological evidence in favor of this view."
Objections to the Theory of the Lateral Fold.
The objection to the first view is its precarious foundation. Such lateral folds are found only in certain rays, in. which they may be developed as a secondary modification in connection with the peculiar form of these fishes. Professor Kerr observes that this theory must be looked upon and judged:
"Just as any other view at the present time regarding the nature of the vertebrate limb, rather as a speculation, brilliant and suggestive though it be, than as a logically constructed theory of the now known facts. It is, I think, on this account allowable to apply to it a test of a character which is admittedly very apt to mislead, that of 'common sense.'
"If there is any soundness in zoological speculation at all I think it must be admitted that the more primitive vertebrates were creatures possessing a notochordal axial skeleton near the dorsal side, with the main nervous axis above it, the main viscera below it, and the great mass of muscle lying in myotomes along its sides. Now such a creature is well adapted to movements of the character of lateral flexure, and not at all for movements in the sagittal plane—which would be not only difficult to achieve but would tend to alternately compress and extend its spinal cord and its viscera. Such a creature would swim through the water as does a Cyclostome, or a Lepidosiren, or any other elongated vertebrate without special swimming organs. Swimming like this, specialization for more and more rapid movement, would mean flattening of the tail region and its extension into an at first not separately mobile median tail fold. It is extremely difficult to my mind to suppose that a new purely swimming arrangement should have arisen involving up and down movement, and which, at its first beginnings, while useless as a swimming organ itself, must greatly detract from the efficiency of that which already existed."
Objections to Gegenbaur's Theory.
We now return to the Gegenbaur view—that the limb is a modified gill septum.
"Resting on Gegenbaur's discovery already mentioned that the gillrays in certain cases assume an arrangement showing great similarity to that of the skeletal elements of the archipterygium, it has, so far as I am aware, up to the present time received no direct support whatever cf a nature comparable with that found for the rival view in the fact that, in certain forms at all events, the limbs actually do arise in the individual in the way that the theory holds they did in phylogeny. No one has produced either a form in which a gill septum becomes the limb during ontogeny, or the fossil remains of any form which shows an intermediate condition.
"The portion of Gegenbaur's view which asserts that the biserial archipterygial fin is of an extremely primitive character is supported by a large body of anatomical facts, and is rendered further probable by the great frequency with which fins apparently of this character occur amongst the oldest known fishes. On the lateral fold view we should have to regard these as independently evolved, which would imply that fins of this type are of a very perfect character, and in that case we may be indeed surprised at their so complete disappearance in the more highly developed forms, which followed later on."
As to Gegenbaur's theory it is urged that no form is known in which a gill septum develops into a limb during the growth of the individual. The main thesis, according to Professor Kerr, "that the archipterygium was derived from gill-rays, is supported only by evidence of an indirect character. Gegenbaur in his very first suggestion of his theory pointed cut, as a great difficulty in the way of its acceptance, the position of the
limbs, especially of the pelvic limbs, in a position far removed from that of the branchial arches. This difficulty has been entirely removed by the brilliant work of Gegenbaur's followers, who have shown from the facts of comparative anatomy and embryology that the limbs, and the hind limbs especially, actually have undergone, and in ontogeny do undergo, an extensive backward migration. In some cases Braus has been able to find traces of this migration as far forward as a point just behind the branchial arches. Now, when we consider the numbers, the enthusiasm, and the ability of Gegenbaur's disciples, we cannot help being struck by the fact that the only evidence in favor of this derivation of the limbs has been that which tends to show that a migration of the limbs backwards has taken place from a region somewhere near the last branchial arch, and that they have failed utterly to discover any intermediate steps between gill-rays and archipterygial fin. And if for a moment we apply the test of common sense we cannot but be impressed by the improbability of the evolution of a gill septum, which in all the lower forms of fishes is fixed firmly in the body wall, and beneath its surface, into an organ of locomotion.
"May I express the hope that what I have said is sufficient to show in what a state of uncertainty our views are regarding the morphological nature of the paired fins, and upon what an exceedingly slender basis rest both of the two views which at present hold the field?"
Kerr's Theory of Modified External Gills.
"It is because I feel that in the present state of our knowledge neither of the two views I have mentioned has a claim to any higher rank than that of extremely suggestive speculations that I venture to say a few words for the third view, which is avowedly a mere speculation.
"Before proceeding with it I should say that I assume the serial homology of fore-and hind-limbs to be beyond dispute. The great and deep-seated resemblances between them are such as to my mind seem not to be adequately explicable except on this assumption.
"In the Urodela (salamanders) the external gills are well-known structures—serially arranged projections from the body wall near the upper ends of certain of the branchial arches. When one considers the ontogenetic development of these organs, from knob-like outgrowth from the outer face of the branchial arch, covered with ectoderm and possessing a mesoblastic core, and which frequently if not always appear before the branchial clefts are open, one cannot but conclude that they are morphologically projections of the outer skin and that they have nothing whatever to do with the gill pouches of the gut wall. Amongst the Urodela one such gill projects from each of the first three branchial arches. In Lepidosiren there is one on each of branchial arches I.-IV. In Polypterus and Calamoichthys (Erpetoichthys) there is one on the hyoid arch. Finally, in many Urodelan larvæ we have present at the same time as the external gills a pair of curious structures called balancers. At an early stage of my work on Lepidosiren, while looking over other vertebrate embryos and larvae for purposes of comparison, my attention was arrested by these structures, and further examinations by section and otherwise, convinced me that they were serial homologues of the external gills, situated on the mandibular arch. On then looking up the literature, I found that I was by no means first in this view. Eusconi had long ago noticed the resemblance, and in more recent times both Orr and Maurer had been led to the same conclusion as I had been. Three different observers having been independently led to exactly the same conclusions, we may, I think, fairly enough regard the view I have mentioned of the morphological nature of the balancers as probably a correct one.
"Here then, we have a series of homologous structures projecting from each of the series of visceral arches. They crop up on the Crossopterygii, the Dipnoi and the Urodela, i. e., in three of the most archaic of the groups of Gnathatomata. But we may put it in another way. The groups in which they do not occur are those whose young possess a very large yolk-sac (or which are admittedly derived from such forms). Now wherever we have a large yolk-sac we have developed on its surface a rich network of blood vessels for purposes of nutrition. But such a network must necessarily act as an extraordinarily efficient organ of respiration, and did we not know the facts we might venture to prophesy that in forms possessing it any other small skin organ of respiration would tend to disappear.
"No doubt these external gills are absent also in a few of the admittedly primitive forms such as, e. g., (neo-) Ceratodus. But I would ask that in this connection one should bear in mind one of the marked characteristics of external gills—their great regenerative power. This involves their being extremely liable to injury and consequently a source of danger to their possessor. Their absence, therefore, in certain cases may well have been due to natural selection. On the other hand, the presence in so many lowly forms of these organs, the general close similarity in structure that runs through them in different forms and the exact correspondence in their position and relations to the body can, it seems to me, only be adequately explained by looking on them as being homologous structures inherited from a common ancestor and consequently of great antiquity in the vertebrate stem."
As to the third theory, Professor Kerr suggests tentatively that the external gill, as developed in salamanders and in the young Lepidosiren, may be the structure modified to form the paired limbs. Of the homology of fore and hind limbs and consequently of their like origin there can be no doubt.
The general gill structures have, according to Kerr, "the primary function of respiration. They are also, however, provided with an elaborate muscular apparatus comprising elevators, depressors and adductors, and larva? possessing them may be seen every now and then to give them a sharp backward twitch. They are thus potentially motor organs. In such a Urodele as Amblystoma their homologues on the mandibular arch are used as supporting structures against a solid substratum exactly as are the limbs of the young Lepidosiren.
"I have, therefore, to suggest that the more ancient Gnathostomata possessed a series of potentially motor, potentially supporting structures projecting from their visceral arches; it was inherently extremely probable that these should be made use of when actual supporting and motor appendages had to be developed in connection with clambering about a solid substratum. If this had been so we should look upon the limb ab a modified external gill; the limb girdle, with Gegenbaur, as a modified branchial arch.
"This theory of the vertebrate paired limb seems to me, I confess, to be a more plausible one on the face of it than either of the two which at present hold the field. If untrue it is so dangerously plausible as to surely deserve more consideration than it appears to have had. One of the main differences between it and the other two hypotheses is that, instead of deriving the swimming fin from the walking and supporting limb, it goes the other way about. That this is the safer line to take seems to me to be shown by the consideration that a very small and rudimentary limb could only be of use if provided with a fixed point d'appui. Also on this view, the pentadactyle limb and the swimming fin would probably be evolved independently from a simple form of limb. This would evade the great difficulties which have beset those who have endeavored to establish the homologies of the elements of the pentadactyle limb with those of any type of fully-formed fin."
Uncertain Conclusions.
In conclusion, we may say that the evidence of embryology in this matter is inadequate, that of morphology is inconclusive and perhaps the final answer may be given by paleontology. If the records of the rocks were complete they would be decisive. At present we have to decide which is the more primitive of two forms of pectoral fin actually known among fossils. That of Cladoselache is a low, horizontal fold of skin, with feeble rays, called by Cope ptychopterygium. That of Pleuracanthus is a jointed paddle-shaped appendage with a fringe of rays on either side. In the theory of Gegenbaur and Kerr Pleuracanthus must be, so far as the limbs are concerned, the form nearest the primitive limb-bearing vertebrate. In Balfour's theory Cladoselache is nearest the primitive type from which the other and with it the archipterygium of later forms may be derived.
Boulenger and others question even this, believing that the archipterygium in Pleuracanthus and that in Neo-Ceratodus and its Dipnoan and Crossopterygian allies have been derived independently from the archipterygium in the primitive sharks. In the one theory, the type of Pleuracanthus would be ancestral to the other sharks, on the one hand, and to Crossopterygians and all higher vertebrates on the other. With the theory of the origin of the pectoral from a lateral fold, Pleuracanthus would be merely a curious specialized offshoot from the primitive sharks, without descendants and without special significance in phylogeny.
As elements bearing on this decision we may note that the tapering unspecialized diphycercal tail of Pleuracanthus seems very primitive in comparison with the short heterocercal tail of Cladoselache. This evidence, perhaps deceptive, is balanced by the presence on the head of Pleuracanthus of a highly specialized serrated spine, evidence of a far from primitive structure.