X. THE INFLUENCE OF DARWIN ON THE STUDY OF ANIMAL EMBRYOLOGY. By A. Sedgwick, M.A., F.R.S.
Professor of Zoology and Comparative Anatomy in the University of Cambridge.
The publication of "The Origin of Species" ushered in a new era in the study of Embryology. Whereas, before the year 1859 the facts of anatomy and development were loosely held together by the theory of types, which owed its origin to the great anatomists of the preceding generation, to Cuvier, L. Agassiz, J. Muller, and R. Owen, they were now combined together into one organic whole by the theory of descent and by the hypothesis of recapitulation which was deduced from that theory. The view (First clearly enunciated by Fritz Muller in his well-known work, "Fur Darwin", Leipzig, 1864; (English Edition, "Facts for Darwin", 1869).) that a knowledge of embryonic and larval histories would lay bare the secrets of race-history and enable the course of evolution to be traced, and so lead to the discovery of the natural system of classification, gave a powerful stimulus to morphological study in general and to embryological investigation in particular. In Darwin's words: "Embryology rises greatly in interest, when we look at the embryo as a picture, more or less obscured, of the progenitor, either in its adult or larval state, of all the members of the same great class." ("Origin" (6th edition), page 396.) In the period under consideration the output of embryological work has been enormous. No group of the animal kingdom has escaped exhaustive examination and no effort has been spared to obtain the embryos of isolated and out of the way forms, the development of which might have an important bearing upon questions of phylogeny and classification. Marine zoological stations have been established, expeditions have been sent to distant countries, and the methods of investigation have been greatly improved. The result of this activity has been that the main features of the developmental history of all the most important animals are now known and the curiosity as to developmental processes, so greatly excited by the promulgation of the Darwinian theory, has to a considerable extent been satisfied.
To what extent have the results of this vast activity fulfilled the expectations of the workers who have achieved them? The Darwin centenary is a fitting moment at which to take stock of our position. In this inquiry we shall leave out of consideration the immense and intensely interesting additions to our knowledge of Natural History. These may be said to constitute a capital fund upon which philosophers, poets and men of science will draw for many generations. The interest of Natural History existed long before Darwinian evolution was thought of and will endure without any reference to philosophic speculations. She is a mistress in whose face are beauties and in whose arms are delights elsewhere unattainable. She is and always has been pursued for her own sake without any reference to philosophy, science, or utility.
Darwin's own views of the bearing of the facts of embryology upon questions of wide scientific interest are perfectly clear. He writes ("Origin" (6th edition), page 395.):
"On the other hand it is highly probable that with many animals the embryonic or larval stages show us, more or less completely, the condition of the progenitor of the whole group in its adult state. In the great class of the Crustacea, forms wonderfully distinct from each other, namely, suctorial parasites, cirripedes, entomostraca, and even the malacostraca, appear at first as larvae under the nauplius-form; and as these larvae live and feed in the open sea, and are not adapted for any peculiar habits of life, and from other reasons assigned by Fritz Muller, it is probable that at some very remote period an independent adult animal, resembling the Nauplius, existed, and subsequently produced, along several divergent lines of descent, the above-named great Crustacean groups. So again it is probable, from what we know of the embryos of mammals, birds, fishes, and reptiles, that these animals are the modified descendants of some ancient progenitor, which was furnished in its adult state with branchiae, a swim-bladder, four fin-like limbs, and a long tail, all fitted for an aquatic life.
"As all the organic beings, extinct and recent, which have ever lived, can be arranged within a few great classes; and as all within each class have, according to our theory, been connected together by fine gradations, the best, and, if our collections were nearly perfect, the only possible arrangement, would be genealogical; descent being the hidden bond of connexion which naturalists have been seeking under the term of the Natural System. On this view we can understand how it is that, in the eyes of most naturalists, the structure of the embryo is even more important for classification than that of the adult. In two or more groups of animals, however much they may differ from each other in structure and habits in their adult condition, if they pass through closely similar embryonic stages, we may feel assured that they all are descended from one parent-form, and are therefore closely related. Thus, community in embryonic structure reveals community of descent; but dissimilarity in embryonic development does not prove discommunity of descent, for in one of two groups the developmental stages may have been suppressed, or may have been so greatly modified through adaptation to new habits of life, as to be no longer recognisable. Even in groups, in which the adults have been modified to an extreme degree, community of origin is often revealed by the structure of the larvae; we have seen, for instance, that cirripedes, though externally so like shell-fish, are at once known by their larvae to belong to the great class of crustaceans. As the embryo often shows us more or less plainly the structure of the less modified and ancient progenitor of the group, we can see why ancient and extinct forms so often resemble in their adult state the embryos of existing species of the same class. Agassiz believes this to be a universal law of nature; and we may hope hereafter to see the law proved true. It can, however, be proved true only in those cases in which the ancient state of the progenitor of the group has not been wholly obliterated, either by successive variations having supervened at a very early period of growth, or by such variations having been inherited at an earlier stage than that at which they first appeared. It should also be borne in mind, that the law may be true, but yet, owing to the geological record not extending far enough back in time, may remain for a long period, or for ever, incapable of demonstration. The law will not strictly hold good in those cases in which an ancient form became adapted in its larval state to some special line of life, and transmitted the same larval state to a whole group of descendants; for such larvae will not resemble any still more ancient form in its adult state."
As this passage shows, Darwin held that embryology was of interest because of the light it seems to throw upon ancestral history (phylogeny) and because of the help it would give in enabling us to arrive at a natural system of classification. With regard to the latter point, he quotes with approval the opinion that "the structure of the embryo is even more important for classification than that of the adult." What justification is there for this view? The phase of life chosen for the ordinary anatomical and physiological studies, namely, the adult phase, is merely one of the large number of stages of structure through which the organism passes. By far the greater number of these are included in what is specially called the developmental or (if we include larvae with embryos) embryonic period, for the developmental changes are more numerous and take place with greater rapidity at the beginning of life than in its later periods. As each of these stages is equal in value, for our present purpose, to the adult phase, it clearly follows that if there is anything in the view that the anatomical study of organisms is of importance in determining their mutual relations, the study of the organism in its various embryonic (and larval) stages must have a greater importance than the study of the single and arbitrarily selected stage of life called the adult.
But a deeper reason than this has been assigned for the importance of embryology in classification. It has been asserted, and is implied by Darwin in the passage quoted, that the ancestral history is repeated in a condensed form in the embryonic, and that a study of the latter enables us to form a picture of the stages of structure through which the organism has passed in its evolution. It enables us on this view to reconstruct the pedigrees of animals and so to form a genealogical tree which shall be the true expression of their natural relations.
The real question which we have to consider is to what extent the embryological studies of the last 50 years have confirmed or rendered probable this "theory of recapitulation." In the first place it must be noted that the recapitulation theory is itself a deduction from the theory of evolution. The facts of embryology, particularly of vertebrate embryology, and of larval history receive, it is argued, an explanation on the view that the successive stages of development are, on the whole, records of adult stages of structure which the species has passed through in its evolution. Whether this statement will bear a critical verbal examination I will not now pause to inquire, for it is more important to determine whether any independent facts can be alleged in favour of the theory. If it could be shown, as was stated to be the case by L. Agassiz, that ancient and extinct forms of life present features of structure now only found in embryos, we should have a body of facts of the greatest importance in the present discussion. But as Huxley (See Huxley's "Scientific Memoirs", London, 1898, Vol. I. page 303: "There is no real parallel between the successive forms assumed in the development of the life of the individual at present, and those which have appeared at different epochs in the past." See also his Address to the Geological Society of London (1862) 'On the Palaeontological Evidence of Evolution', ibid. Vol. II. page 512.) has shown and as the whole course of palaeontological and embryological investigation has demonstrated, no such statement can be made. The extinct forms of life are very similar to those now existing and there is nothing specially embryonic about them. So that the facts, as we know them, lend no support to theory.
But there is another class of facts which have been alleged in favour of the theory, viz. the facts which have been included in the generalisation known as the Law of v. Baer. The law asserts that embryos of different species of animals of the same group are more alike than the adults and that, the younger the embryo, the greater are the resemblances. If this law could be established it would undoubtedly be a strong argument in favour of the "recapitulation" explanation of the facts of embryology. But its truth has been seriously disputed. If it were true we should expect to find that the embryos of closely similar species would be indistinguishable from one another, but this is notoriously not the case. It is more difficult to meet the assertion when it is made in the form given above, for here we are dealing with matters of opinion. For instance, no one would deny that the embryo of a dogfish is different from the embryo of a rabbit, but there is room for difference of opinion when it is asserted that the difference is less than the difference between an adult dogfish and an adult rabbit. It would be perfectly true to say that the differences between the embryos concern other organs more than do the differences between the adults, but who is prepared to affirm that the presence of a cephalic coelom and of cranial segments, of external gills, of six gill slits, of the kidney tubes opening into the muscle-plate coelom, of an enormous yolk-sac, of a neurenteric canal, and the absence of any trace of an amnion, of an allantois and of a primitive streak are not morphological facts of as high an import as those implied by the differences between the adults? The generalisation undoubtedly had its origin in the fact that there is what may be called a family resemblance between embryos and larvae, but this resemblance, which is by no means exact, is largely superficial and does not extend to anatomical detail.
It is useless to say, as Weismann has stated ("The Evolution Theory", by A. Weismann, English Translation, Vol. II. page 176, London, 1904.), that "it cannot be disputed that the rudiments [vestiges his translator means] of gill-arches and gill-clefts, which are peculiar to one stage of human ontogeny, give us every ground for concluding that we possessed fish-like ancestors." The question at issue is: did the pharyngeal arches and clefts of mammalian embryos ever discharge a branchial function in an adult ancestor of the mammalia? We cannot therefore, without begging the question at issue in the grossest manner, apply to them the terms "gill-arches" and "gill-clefts". That they are homologous with the "gill-arches" and "gill-clefts" of fishes is true; but there is no evidence to show that they ever discharged a branchial function. Until such evidence is forthcoming, it is beside the point to say that it "cannot be disputed" that they are evidence of a piscine ancestry.
It must, therefore, be admitted that one outcome of the progress of embryological and palaeontological research for the last 50 years is negative. The recapitulation theory originated as a deduction from the evolution theory and as a deduction it still remains.
Let us before leaving the subject apply another test. If the evolution theory and the recapitulation theory are both true, how is it that living birds are not only without teeth but have no rudiments of teeth at any stage of their existence? How is it that the missing digits in birds and mammals, the missing or reduced limb of snakes and whales, the reduced mandibulo-hyoid cleft of elasmobranch fishes are not present or relatively more highly developed in the embryo than in the adult? How is it that when a marked variation, such as an extra digit, or a reduced limb, or an extra segment, makes its appearance, it is not confined to the adult but can be seen all through the development? All the clear evidence we can get tends to show that marked variations, whether of reduction or increase, of organs are manifest during the whole of the development of the organ and do not merely affect the adult. And on reflection we see that it could hardly be otherwise. All such evidence is distinctly at variance with the theory of recapitulation, at least as applied to embryos. In the case of larvae of course the case will be different, for in them the organs are functional, and reduction in the adult will not be accompanied by reduction in the larva unless a change in the conditions of life of the larva enables it to occur.
If after 50 years of research and close examination of the facts of embryology the recapitulation theory is still without satisfactory proof, it seems desirable to take a wider sweep and to inquire whether the facts of embryology cannot be included in a larger category.
As has been pointed out by Huxley, development and life are co-extensive, and it is impossible to point to any period in the life of an organism when the developmental changes cease. It is true that these changes take place more rapidly at the commencement of life, but they are never wholly absent, and those which occur in the later or so-called adult stages of life do not differ in their essence, however much they may differ in their degree, from those which occur during the embryonic and larval periods. This consideration at once brings the changes of the embryonic period into the same category as those of the adult and suggests that an explanation which will account for the one will account for the other. What then is the problem we are dealing with? Surely it is this: Why does an organism as soon as it is established at the fertilisation of the ovum enter upon a cycle of transformations which never cease until death puts an end to them? In other words what is the meaning of that cycle of changes which all organisms present in a greater or less degree and which constitute the very essence of life? It is impossible to give an answer to this question so long as we remain within the precincts of Biology—and it is not my present purpose to penetrate beyond those precincts into the realms of philosophy. We have to do with an ultimate biological fact, with a fundamental property of living matter, which governs and includes all its other properties. How may this property be stated? Thus: it is a property of living matter to react in a remarkable way to external forces without undergoing destruction. The life-cycle, of which the embryonic and larval periods are a part, consists of the orderly interaction between the organism and its environment. The action of the environment produces certain morphological changes in the organism. These changes enable the organism to come into relation with new external forces, to move into what is practically a new environment, which in its turn produces further structural changes in the organism. These in their turn enable, indeed necessitate, the organism to move again into a new environment, and so the process continues until the structural changes are of such a nature that the organism is unable to adapt itself to the environment in which it finds itself. The essential condition of success in this process is that the organism should always shift into the environment to which its new structure is suited—any failure in this leading to the impairment of the organism. In most cases the shifting of the environment is a very gradual process (whether consisting in the very slight and gradual alteration in the relation of the embryo as a whole to the egg-shell or uterine wall, or in the relations of its parts to each other, or in the successive phases of adult life), and the morphological changes in connection with each step of it are but slight. But in some cases jumps are made such as we find in the phenomena known as hatching, birth, and metamorphosis.
This property of reacting to the environment without undergoing destruction is, as has been stated, a fundamental property of organisms. It is impossible to conceive of any matter, to which the term living could be applied, being without it. And with this property of reacting to the environment goes the further property of undergoing a change which alters the relation of the organism to the old environment and places it in a new environment. If this reasoning is correct, it necessarily follows that this property must have been possessed by living matter at its first appearance on the earth. In other words living matter must always have presented a life-cycle, and the question arises what kind of modification has that cycle undergone? Has it increased or diminished in duration and complexity since organisms first appeared on the earth? The current view is that the cycle was at first very short and that it has increased in length by the evolutionary creation of new adult phases, that these new phases are in addition to those already existing and that each of them as it appears takes over from the preceding adult phase the functional condition of the reproductive organs. According to the same view the old adult phases are not obliterated but persist in a more or less modified form as larval stages. It is further supposed that as the life-history lengthens at one end by the addition of new adult phases, it is shortened at the other by the abbreviation of embryonic development and by the absorption of some of the early larval stages into the embryonic period; but on the whole the lengthening process has exceeded that of shortening, so that the whole life-history has, with the progress of evolution, become longer and more complicated.
Now there can be no doubt that the life-history of organisms has been shortened in the way above suggested, for cases are known in which this can practically be seen to occur at the present day. But the process of lengthening by the creation of new stages at the other end of the life-cycle is more difficult to conceive and moreover there is no evidence for its having occurred. This, indeed, may have occurred, as is suggested below, but the evidence we have seems to indicate that evolutionary modification has proceeded by ALTERING and not by SUPERSEDING: that is to say that each stage in the life-history, as we see it to-day, has proceeded from a corresponding stage in a former era by the modification of that stage and not by the creation of a new one. Let me, at the risk of repetition, explain my meaning more fully by taking a concrete illustration. The mandibulo-hyoid cleft (spiracle) of the elasmobranch fishes, the lateral digits of the pig's foot, the hind-limbs of whales, the enlarged digit of the ostrich's foot are supposed to be organs which have been recently modified. This modification is not confined to the final adult stage of the life-history but characterises them throughout the whole of their development. A stage with a reduced spiracle does not proceed in development from a preceding stage in which the spiracle shows no reduction: it is reduced at its first appearance. The same statement may be made of organs which have entirely disappeared in the adult, such as bird's teeth and snake's fore-limbs: the adult stage in which they have disappeared is not preceded by embryonic stages in which the teeth and limbs or rudiments of them are present. In fact the evidence indicates that adult variations of any part are accompanied by precedent variations in the same direction in the embryo. The evidence seems to show, not that a stage is added on at the end of the life-history, but only that some of the stages in the life-history are modified. Indeed, on the wider view of development taken in this essay, a view which makes it coincident with life, one would not expect often to find, even if new stages are added in the course of evolution, that they are added at the end of the series when the organism has passed through its reproductive period. It is possible of course that new stages have been intercalated in the course of the life-history, though it is difficult to see how this has occurred. It is much more likely, if we may judge from available evidence, that every stage has had its counterpart in the ancestral form from which it has been derived by descent with modification. Just as the adult phase of the living form differs, owing to evolutionary modification, from the adult phase of the ancestor from which it has proceeded, so each larval phase will differ for the same reason from the corresponding larval phase in the life-history of the ancestor. Inasmuch as the organism is variable at every stage of its independent existence and is exposed to the action of natural selection there is no reason why it should escape modification at any stage.
If there is any truth in these considerations it would seem to follow that at the dawn of life the life-cycle must have been, either in posse or in esse, at least as long as it is at the present time, and that the peculiarity of passing through a series of stages in which new characters are successively evolved is a primordial quality of living matter.
Before leaving this part of the subject, it is necessary to touch upon another aspect of it. What are these variations in structure which succeed one another in the life-history of an organism? I am conscious that I am here on the threshold of a chamber which contains the clue to some of our difficulties, and that I cannot enter it. Looked at from one point of view they belong to the class of genetic variations, which depend upon the structure or constitution of the protoplasm; but instead of appearing in different zygotes (A zygote is a fertilised ovum, i.e. a new organism resulting from the fusion of an ovum and a spermatozoon.), they are present in the same zygote though at different times in its life-history. They are of the same order as the mutational variations of the modern biologist upon which the appearance of a new character depends. What is a genetic or mutational variation? It is a genetic character which was not present in either of the parents. But these "growth variations" were present in the parents, and in this they differ from mutational variations. But what are genetic characters? They are characters which must appear if any development occurs. They are usually contrasted with "acquired characters," using the expression "acquired character" in the Lamarckian sense. But strictly speaking they ARE acquired characters, for the zygote at first has none of the characters which it subsequently acquires, but only the power of acquiring them in response to the action of the environment. But the characters so acquired are not what we technically understand and what Lamarck meant by "acquired characters." They are genetic characters, as defined above. What then are Lamarck's "acquired characters"? They are variations in genetic characters caused in a particular way. There are, in fact, two kinds of variation in genetic characters depending on the mode of causation. Firstly, there are those variations consequent upon a variation in the constitution of the protoplasm of a particular zygote, and independent of the environment in which the organism develops, save in so far as this simply calls them forth: these are the so-called genetic or mutational variations. Secondly, there are those variations which occur in zygotes of similar germinal constitution and which are caused solely by differences in the environment to which the individuals are respectively exposed: these are the "acquired characters" of Lamarck and of authors generally. In consequence of this double sense in which the term "acquired characters" may be used, great confusion may and does occur. If the protoplasm be compared to a machine, and the external conditions to the hand that works the machine, then it may be said that, as the machine can only work in one way, it can only produce one kind of result (genetic character), but the particular form or quality (Lamarckian "acquired character") of the result will depend upon the hand that works the machine (environment), just as the quality of the sound produced by a fiddle depends entirely upon the hand which plays upon it. It would be improper to apply the term "mutation" to those genetic characters which are not new characters or new variants of old characters, but such genetic characters are of the same nature as those characters to which the term mutation has been applied. It may be noticed in passing that it is very questionable if the modern biologist has acted in the real interests of science in applying the term mutation in the sense in which he has applied it. The genetic characters of organisms come from one of two sources: either they are old characters and are due to the action of what we call inheritance or they are new and are due to what we call variation. If the term mutation is applied to the actual alteration of the machinery of the protoplasm, no objection can be felt to its use; but if it be applied, as it is, to the product of the action of the altered machine, viz. to the new genetic character, it leads to confusion. Inheritance is the persistence of the structure of the machine; characters are the products of the working of the machine; variation in genetic characters is due to the alteration (mutation) in the arrangement of the machinery, while variation in acquired characters (Lamarckian) is due to differences in the mode of working the machinery. The machinery when it starts (in the new zygote) has the power of grinding out certain results, which we call the characters of the organism. These appear at successive intervals of time, and the orderly manifestation of them is what we call the life-history of the organism. This brings us back to the question with which we started this discussion, viz. what is the relation of these variations in structure, which successively appear in an organism and constitute its life-history, to the mutational variations which appear in different organisms of the same brood or species. The question is brought home to us when we ask what is a bud-sport, such as a nectarine appearing on a peach-tree? From one point of view, it is simply a mutation appearing in asexual reproduction; from another it is one of these successional characters ("growth variations") which constitute the life-history of the zygote, for it appears in the same zygote which first produces a peach. Here our analogy of a machine which only works in one way seems to fail us, for these bud-sports do not appear in all parts of the organism, only in certain buds or parts of it, so that one part of the zygotic machine would appear to work differently to another. To discuss this question further would take us too far from our subject. Suffice it to say that we cannot answer it, any more than we can this further question of burning interest at the present day, viz. to what extent and in what manner is the machine itself altered by the particular way in which it is worked. In connection with this question we can only submit one consideration: the zygotic machine can, by its nature, only work once, so that any alteration in it can only be ascertained by studying the replicas of it which are produced in the reproductive organs.
It is a peculiarity that the result which we call the ripening of the generative organs nearly always appears among the final products of the action of the zygotic machine. It is remarkable that this should be the case. What is the reason of it? The late appearance of functional reproductive organs is almost a universal law, and the explanation of it is suggested by expressing the law in another way, viz. that the machine is almost always so constituted that it ceases to work efficiently soon after the reproductive organs have sufficiently discharged their function. Why this should occur we cannot explain: it is an ultimate fact of nature, and cannot be included in any wider category. The period during which the reproductive organs can act may be short as in ephemerids or long as in man and trees, and there is no reason to suppose that their action damages the vital machinery, though sometimes, as in the case of annual plants (Metschnikoff), it may incidentally do so; but, long or short, the cessation of their actions is always a prelude to the end. When they and their action are impaired, the organism ceases to react with precision to the environment, and the organism as a whole undergoes retrogressive changes.
It has been pointed out above that there is reason to believe that at the dawn of life the life-cycle was, EITHER IN ESSE OR IN POSSE, at least as long as it is at the present time. The qualification implied by the words in italics is necessary, for it is clearly possible that the external conditions then existing were not suitable for the production of all the stages of the potential life-history, and that what we call organic evolution has consisted in a gradual evolution of new environments to which the organism's innate capacity of change has enabled it to adapt itself. We have warrant for this possibility in the case of the Axolotl and in other similar cases of neoteny. And these cases further bring home to us the fact, to which I have already referred, that the full development of the functional reproductive organs is nearly always associated with the final stages of the life-history.
On this view of the succession of characters in the life-history of organisms, how shall we explain the undoubted fact that the development of buds hardly ever presents any phenomena corresponding to the embryonic and larval changes? The reason is clearly this, that budding usually occurs after the embryonic stage is past; when the characters of embryonic life have been worked out by the machine. When it takes place at an early stage in embryonic life, as it does in cases of so-called embryonic fission, the product shows, either partly or entirely, phenomena similar to those of embryonic development. The only case known to me in which budding by the adult is accompanied by morphological features similar to those displayed by embryos is furnished by the budding of the medusiform spore-sacs of hydrozoon polyps. But this case is exceptional, for here we have to do with an attempt, which fails, to form a free-swimming organism, the medusa; and the vestiges which appear in the buds are the umbrella-cavity, marginal tentacles, circular canal, etc., of the medusa arrested in development.
But the question still remains, are there no cases in which, as implied by the recapitulation theory, variations in any organ are confined to the period in which the organ is functional and do not affect it in the embryonic stages? The teeth of the whalebone whales may be cited as a case in which this is said to occur; but here the teeth are only imperfectly developed in the embryo and are soon absorbed. They have been affected by the change which has produced their disappearance in the adult, but not to complete extinction. Nor are they now likely to be extinguished, for having become exclusively embryonic they are largely protected from the action of natural selection. This consideration brings up a most important aspect of the question, so far as disappearing organs are concerned. Every organ is laid down at a certain period in the embryo and undergoes a certain course of growth until it obtains full functional development. When for any cause reduction begins, it is affected at all stages of its growth, unless it has functional importance in the larva, and in some cases its life is shortened at one or both ends. In cases, as in that of the whale's teeth, in which it entirely disappears in the adult, the latter part of its life is cut off; in others, the beginning of its life may be deferred. This happens, for instance, with the spiracle of many Elasmobranchs, which makes its appearance after the hyobranchial cleft, not before it as it should do, being anterior to it in position, and as it does in the Amniota in which it shows no reduction in size as compared with the other pharyngeal clefts. In those Elasmobranchs in which it is absent in the adult but present in the embryo (e.g. Carcharias) its life is shortened at both ends. Many more instances of organs, of which the beginning and end have been cut off, might be mentioned; e.g. the muscle-plate coelom of Aves, the primitive streak and the neurenteric canal of amniote blastoderms. In yet other cases in which the reduced organ is almost on the verge of disappearance, it may appear for a moment and disappear more than once in the course of development. As an instance of this striking phenomenon I may mention the neurenteric canal of avine embryos, and the anterior neuropore of Ascidians. Lastly the reduced organ may disappear in the developing stages before it does so in the adult. As an instance of this may be mentioned the mandibular palp of those Crustacea with zoaea larvae. This structure disappears in the larva only to reappear in a reduced form in later stages. In all these cases we are dealing with an organ which, we imagine, attained a fuller functional development at some previous stage in race-history, but in most of them we have no proof that it did so. It may be, and the possibility must not be lost sight of, that these organs never were anything else than functionless and that though they have been got rid of in the adult by elimination in the course of time, they have been able to persist in embryonic stages which are protected from the full action of natural selection. There is no reason to suppose that living matter at its first appearance differed from non-living matter in possessing only properties conducive to its well-being and prolonged existence. No one thinks that the properties of the various forms of inorganic matter are all strictly related to external conditions. Of what use to the diamond is its high specific gravity and high refrangibility, and to gold of its yellow colour and great weight? These substances continue to exist in virtue of other properties than these. It is impossible to suppose that the properties of living matter at its first appearance were all useful to it, for even now after aeons of elimination we find that it possesses many useless organs and that many of its relations to the external world are capable of considerable improvement.
In writing this essay I have purposely refrained from taking a definite position with regard to the problems touched. My desire has been to write a chapter showing the influence of Darwin's work so far as Embryology is concerned, and the various points which come up for consideration in discussing his views. Darwin was the last man who would have claimed finality for any of his doctrines, but he might fairly have claimed to have set going a process of intellectual fermentation which is still very far from completion.