[518] _Presidential Address to the British a.s.sociation_, 1908.

CHAPTER XX

THE CLa.s.sICAL TRADITION IN MODERN MORPHOLOGY

To write a history of contemporary movements from a purely objective standpoint is well recognised to be an impossible task. It is difficult for those in the stream to see where the current is carrying them: the tendencies of the present will only become clear some twenty years in the future.

I propose, therefore, in this concluding chapter to deal only with certain characteristics of modern work on the problems of form which seem to me to be derived directly from the older cla.s.sical tradition of Cuvier and von Baer.

The present time is essentially one of transition. Complete uncertainty reigns as to the main principles of biology. Many of us think that the materialistic and simplicist method has proved a complete failure, and that the time has come to strike out on entirely different lines. Just in what direction the new biology will grow out is hard to see at present, so many divergent beginnings have been made--the materialistic vitalism of Driesch, the profound intuitionalism of Bergson, the psychological biology of Delpino, France, Pauly, A. Wagner and W.

Mackenzie. But if any of these are destined to give the future direction to biology, they will in a measure only be bringing biology back to its pre-materialistic tradition, the tradition of Aristotle, Cuvier, von Baer and J. Muller. It may well be that the intransigent materialism of the 19th century is merely an episode, an aberration rather, in the history of biology--an aberration brought about by the over-rapid development of a materialistic and luxurious civilisation, in which man's material means have outrun his mental and moral growth.

Two movements seem significant in the morphology of the last decade or so of the 19th century--first, the experimental study of form, and second, the criticism of the concepts or prejudices of evolutionary morphology.

The period was characterised also by the great interest taken in cytology, following upon the pioneer work of Hertwig, van Beneden and others on the behaviour of the nuclei in fertilisation and maturation.[519] This line of work gained added importance in connection with contemporary research and speculation on the nature of hereditary transmission, and it has in quite recent years received an additional stimulus from the re-discovery of Mendelian inheritance. Its importance, however, seems to lie rather in its possible relation to the problems of heredity than in any meaning it may have for the problems of form. More significant is the revolt against the cell-theory started by Sedgwick[520]

and Whitman,[521] on the ground that the organism is something more than an aggregation of discrete, self-centred cells.

The experimental work on the causes of the production and restoration of form infused new life into morphology. It opened men's eyes to the fact that the developing organism is very much a living, active, responsive thing, quite capable of relinquishing at need the beaten track of normal development which its ancestors have followed for countless generations, in order to meet emergencies with an immediate and purposive reaction.

It was cases of this kind, cases of active regulation in development and regeneration, that led men like G. Wolff and H. Driesch to cast off the bonds of dogmatic Darwinism and declare boldly for vitalism and teleology.

There was the famous case of the regeneration of the lens in Amphibia from the edge of the iris--an entirely novel mode of origin, not occurring in ontogeny. The fact seems to have been discovered first by Colucci in 1891, and independently by G. Wolff in 1895.[522] The experiment was later repeated and confirmed by Fischel and other workers. Wolff drew from this and other facts the conclusion that the organism possesses a faculty of "primary purposiveness" which cannot have arisen through natural selection.[523] And, as is well known, Driesch derived one of his most powerful arguments in favour of vitalism from the extraordinary regenerative processes shown by _Tubularia_ and _Clavellina_ in the course of which the organism actually demolishes and rebuilds a part or the whole of its structure. But under the influence of physiologists like Loeb many workers held fast to materialistic methods and conceptions.

The great variety of regulative response of which the organism showed itself capable made it very difficult for the morphologist to uphold the generalisations which he had drawn from the facts of normal undisturbed development. The germ-layer theory was found inadequate to the new facts, and many reverted to the older criterion of h.o.m.ology based on destiny rather than origin. The trend of opinion was to reject the ontogenetic criterion of h.o.m.ology, and to refuse any morphological or phylogenetic value to the germ-layers.[524]

The biogenetic law came more and more into disfavour, as the developing organism more and more showed itself to be capable of throwing off the dead-weight of the past, and working out its own salvation upon original and individual lines.[525] A. Giard in particular called attention to a remarkable group of facts which went to show that embryos or larvae of the same or closely allied species might develop in most dissimilar ways according to the conditions in which they found themselves.[526] His cla.s.sical case of "poecilogeny" was that of the shrimp _Palaemonetes varians_, the fresh-water form of which develops in an entirely different way from the salt-water form.

Experimental workers indeed were inclined to rule the law out of account, to disregard completely the historical element in development, and this was perhaps the chief weakness of the neo-vitalist systems which took their origin in this experimental work.

From the side also of descriptive morphology the biogenetic law underwent a critical revision. It was studied as a fact of embryology and without phylogenetic bias by men like Oppel, Keibel, Mehnert, O.

Hertwig and Vialleton,[527] and they arrived at a critical estimate of it very similar to that of von Baer.

Theoretical objections to the biogenetic law had been raised from time to time by many embryologists, but the positive testing of it by the comparison of embryos in respect of the degree of development of their different organs starts with Oppel's work of 1891.[528] He studied a large number of embryos of different species at different stages of their development, and determined the relative time of appearance of the princ.i.p.al organs and their relative size. His results are summarised in tabular form and have reference to all the more important organs. He was led to ascribe a certain validity to the biogenetic law, but he drew particular attention to the very considerable anomalies in the time of appearance which are shown by many organs, anomalies which had been cla.s.sed by Haeckel under the name of heterochronies.

Oppel's main conclusions were as follows:--"There are found in the developmental stages of different Vertebrates 'similar ontogenetic series,' that is to say, Vertebrates show at definite stages similarities with one another in the degree of development of the different organs. Early stages resemble one another, so also do later stages; equivalent stages of closely allied species resemble one another, and older stages of lower animals resemble younger stages of higher animals; young stages are more alike than old stages.... The differences which these similar series show (for which reason they cannot be regarded as identical) may be designated as temporal disturbances in the degree of development of the separate organs or organ-systems. Some organs show very considerable temporal dislocations, others a moderate amount, others again an inconsiderable amount. Among the developmental stages of various higher animals can be found some which correspond to the ancestral forms and also to the lower types which resemble these ancestral forms. On the basis of the tabulated data here given there can be distinguished with certainty in the ontogeny of Amniotes a pro-fish stage, a fish-stage, a land-animal stage, a pro-amniote stage, and following on these a fully developed reptile, bird or mammal stage."[529]

Oppel's methods were employed by Keibel[530] in his investigations on the development of the pig, which formed the model for the well-known series of _Normentafeln_ of the ontogeny of Vertebrates which were issued in later years under Keibel's editorship. Keibel was more critical of the biogenetic law than Oppel, and he held that the ancestral stages distinguished by Oppel could not be satisfactorily established. He suggested an interesting explanation of heterochrony in development, according to which the premature or r.e.t.a.r.ded appearance of organs in ontogeny stands in close relation with the time of their entering upon functional activity. Thus in many mammals the mesodermal part of the allantois often appears long before the endodermal part, though this is phylogenetically older. This Keibel ascribes to the fact that the endodermal part is almost functionless. "One can directly affirm," he writes, "that the time of appearance of an organ depends in an eminent degree upon the time when it has to enter upon functional activity. This moment is naturally dependent upon the external conditions. Among the highest Vertebrates, the mammals, the traces of phylogeny shown in ontogeny are to a great extent obliterated through the adaptation of ontogeny to the external conditions, and through the modifications which the germs of more highly organised animals necessarily exhibit from the very beginning as compared with germs which do not reach such a high level of development" (p. 754, 1897).

Study of individual variation in the time of appearance of the organs in embryos of the same species was prosecuted with interesting results by Bonnet,[531] Mehnert,[532] and Fischel.[533] Fischel found that variability was greatest among the younger embryos, and became progressively less in later stages. Like von Baer (_supra_, p. 114) he inferred that regulatory processes were at work during development which brought divergent organs back to the normal and enabled them to play their part as correlated members of a functional whole.

Important theoretical views were developed by Mehnert[534] in a series of publications appearing from 1891 to 1898. Like Keibel, Mehnert emphasised the importance of function in determining the late or early appearance of organs, but he conceived the influence of function to be exerted not only in ontogeny, but also throughout the whole course of phylogeny, by reason of the transmission to descendants of the effects of functioning in the individual life.

In his paper of 1897 Mehnert details the results of an extensive examination of the development of the extremities throughout the Amniote series. He finds that in all cases a pentadactylate rudiment is formed, even in those forms in which only a few of the elements of the hand or foot come to full development. But whereas in forms with a normally developed hand, _e.g._ the tortoise and man, all the digits develop and differentiate at about the same rate, in forms which have in the adult reduced digits, _e.g._ the ostrich and the pig, these vestigial digits undergo a very slow and incomplete differentiation, while the others develop rapidly and completely. He draws a general distinction between organs that are phylogenetically progressive and such as are phylogenetically regressive, and seeks to prove that progressive organs show an ontogenetic acceleration and regressive organs a r.e.t.a.r.dation.[535]

The acceleration or r.e.t.a.r.dation affects not only the ma.s.s-growth of the organs, but also their histological differentiation.

Now between progression and functioning and between regression and functional atrophy there is obviously a close connection. Loss of function is well known to be one of the chief causes of the degeneration of organs in the individual life, and on the other hand, as Roux has pointed out, all post-embryonic development is ruled and guided by functioning. It is thus in the long run functioning that brings about phylogenetic progression, absence of functional activity that causes phylogenetic regression. This comes about through the transmission of acquired functional characters, a transmission which Mehnert conceives to be extraordinarily accurate and complete.

In general Mehnert adopts the functional standpoint of Cuvier, von Baer, and Roux. His considered judgment as to the phylogenetic value of the biogenetic law closely resembles that formed by von Baer, for he admits recapitulation only as regards the single organs, not as regards the organism as a whole. He has, however, much more sympathy with the law than either Keibel or Oppel, though he agrees that it cannot be used for the construction of ancestral trees. But he ascribes to it as a fact of development considerable importance. The following pa.s.sage gives a good summary of his view as to the scope and validity of the law. "The biogenetic law has not been shaken by the attacks of its opponents. The a.s.sertion is still true that individual organogenesis is exclusively dependent on phylogeny. But we must not expect to find that all the stages in the development of the separate organs, which coexisted in any member of the phylogenetic series, appear _at the same time_ in the individual ontogeny of the descendants, because each organ possesses its own specific rate of development. In this way it comes about naturally that organs which become differentiated rapidly, as, for example, the medullary tube, as a rule dominate earlier periods of ontogeny than do the organs of locomotion. For the same reason the cerebral hemispheres of man are almost as large in youth as in maturity. The picture which an embryo gives is not a repet.i.tion in detail of one and the same phylogenetic stage; it consists rather of an a.s.semblage of organs, some of which are at a phyletically early stage of development, while others are at a phyletically older stage."[536]

A different line of attack was that adopted by O. Hertwig in a series of papers, which contain also what is perhaps the best critical estimate of the present position and value of descriptive morphology.[537]

It had not escaped the notice of many previous observers that quite early embryos not infrequently show specific characters even before the characters proper to their cla.s.s, order and genus are developed--in direct contradiction of the law of von Baer. Thus L. Aga.s.siz[538] had remarked in 1859 that specific characteristics were often developed precociously. "The Snapping Turtle, for instance, exhibits its small crosslike sternum, its long tail, its ferocious habits, even before it leaves the egg, before it breathes through lungs, before its derm is ossified to form a bony shield, etc.; nay, it snaps with its gaping jaws at anything brought near, when it is still surrounded by its amnion and allantois, and its yolk still exceeds in bulk its whole body" (p. 269).

Wilhelm His,[539] in the course of an acute and damaging criticism of the biogenetic law as enunciated by Haeckel, showed clearly that by careful examination the very earliest embryos of a whole series of Vertebrates could be distinguished with certainty from one another. "An ident.i.ty in external form of different animal embryos, despite the common affirmation to the contrary, does not exist. Even at early stages in their development embryos possess the characters of their cla.s.s and order, nay, we can hardly doubt, of their species and s.e.x, and even their individual characteristics" (201).

This specificity of embryos was affirmed with even greater confidence by Sedgwick in a paper critical of von Baer's law.[540] He wrote:--"If v.

Baer's law has any meaning at all, surely it must imply that animals so closely allied as the fowl and duck would be indistinguishable in the early stages of development; and that in two species so closely similar that I was long in doubt whether they were distinct species, viz., _Peripatus capensis_ and _Balfouri_, it would be useless to look for embryonic differences; yet I can distinguish a fowl and a duck embryo on the second day by the inspection of a single transverse section through the trunk, and it was the embryonic differences between the Peripatuses which led me to establish without hesitation the two separate species.... I need only say ... that a species is distinct and distinguishable from its allies from the very earliest stages all through the development, although these embryonic differences do not necessarily implicate the same organs as do the adult differences" (p.

39).

Hertwig interprets this fact of the specific distinctness of closely allied embryos in the light of the preformistic conception of heredity.

According to this view the whole adult organisation is represented in the structure of the germ-plasm contained in the fertilised ovum, from which it follows that the ova of two different species, and also their embryos at every stage of development, must be as distinct from one another as are the adults themselves, even though the differences may not be so obvious. If this be the case there can be no real recapitulation in ontogeny of the phylogeny of the race, for the egg-cell represents not the first term in phylogeny, but the last. The egg-cell _is_ the organism in an undeveloped state; it has a vastly more complicated structure than was possessed by the primordial cell from which its race has sprung, and it can in no way be considered the equivalent of this ancestral cell.

Hertwig puts this vividly when he says that "the hen's egg is no more the equivalent of the first link in the phylogenetic chain than is the hen itself" (p. 160, 1906, b).

If ontogeny is not a recapitulation of phylogeny, how is it that the early embryonic stages are so alike, even in animals of widely different organisation? Hertwig's answer to this is very interesting. He takes the view that many of the processes characterising early embryonic development are the means necessarily adopted for attaining certain ends. Such are the processes of segmentation, the formation of a blastula, of cell-layers, of medullary folds where the nervous system is a closed tube, the formation of the notochord as a necessary condition of the development of the vertebral column, and so on. "Looked at from this standpoint it cannot surprise us that in all animal phyla the earliest embryonic processes take place in similar fashion, so that we observe the occurrence both in Vertebrates and Invertebrates of a segmentation-process, a morula-stage, a blastula and a gastrula. If now these developmental processes do not depend on chance, but, on the contrary, are rooted in the nature of the animal cell itself, we have no reason for inferring from the recurrence of a similar segmentation-process, morula, blastula, and gastrula in all cla.s.ses of the animal kingdom the common descent of all animals from one blastula-like or gastrula-like ancestral form. We recognise rather in the successive early stages of animal development only the manifestation of special laws, by which the shaping of animal forms (as distinct from plant forms) is brought about" (p. 178, 1906, b).

"The princ.i.p.al reason why certain stages recur in ontogeny with such constancy and always in essentially the same manner is that they provide under all circ.u.mstances the necessary pre-conditions through which alone the later and higher stages of ontogeny can be realised. The unicellular organism can by its very nature transform itself into a multicellular organism only by the method of cell-division. Hence, in all Metazoa, ontogeny must start with a segmentation-process, and a similar statement could be made with regard to all the later stages" (p. 57, 1906, a).

Similarities in early development are therefore no evidence of common descent, and in the same way the resemblances of adult animals, subsumed under the concepts of h.o.m.ology and the unity of plan, are not necessarily due to community of descent, but may also be brought about by the similarity or ident.i.ty of the laws which govern the evolution of these animals. In the absence, therefore, of positive evidence as to the actual lines of descent (to be obtained only from palaeontology), h.o.m.ological resemblance cannot be taken as proof of blood relationship, for h.o.m.ology is a wider concept than h.o.m.ogeny. The only valid definition of h.o.m.ology is that adopted in pre-evolutionary days, when those organs were considered h.o.m.ologous "which agree up to a certain point in structure and composition, in position, arrangement, and relation to the neighbouring organs, and accordingly possess identical functions and uses in the organism" (p. 151, 1906, b).

The concept of h.o.m.ology has thus a value quite independent of any evolutionary interpretation which may be superadded to it. "h.o.m.ology is a mental concept obtained by comparison, which under all circ.u.mstances retains its validity, whether the h.o.m.ology finds its explanation in common descent or in the common laws that rule organic development" (p.

151, 1906, b). As A. Braun long ago pointed out, "It is not descent which decides in matters of morphology, but, on the contrary, morphology which has to decide as to the possibility of descent."[541]

Hertwig, in a word, reverts to the pre-evolutionary conception of h.o.m.ology. "We see in h.o.m.ology," he writes, "only the expression of regularities (_Gesetzma.s.sigkeiten_) in the organisation of the animals showing it, and we regard the question, how far this h.o.m.ology can be explained by common descent and how far by other principles, as for the present an open one, requiring for its solution investigations specially directed towards its elucidation" (p. 179, 1906, b).

Holding, as he does, that no definite conclusions can be drawn from the facts of comparative anatomy and embryology as to the probable lines of descent of the animal kingdom, Hertwig accords very little value to phylogenetic speculation. It is, he admits, quite probable that the archetype of a cla.s.s represents in a general sort of way the ancestral form, but this does not, in his opinion, justify us in a.s.suming that such generalised types ever existed and gave origin to the present-day forms. "It is not legitimate to picture to ourselves the ancestral forms of the more highly organised animals in the guise of the lower animals of the present day--and that is just what we do when we speak of Proselachia, Proamphibia and Proreptilia" (p. 155, 1906, b).

He rejects on the same general grounds the evolutionary dogma of monophyletic or almost monophyletic descent, and admits with Kolliker, von Baer, Wigand, Naegeli and others that evolution may quite well have started many times and from many different primordial cells.

There is indeed a great similarity between the views developed by O.

Hertwig and those held by the older critics of Darwinism--von Baer, Kolliker, Wigand, E. von Hartmann and others. It is true the philosophical standpoint is on the whole different, for while many of that older generation were vitalists Hertwig belongs to the mechanistic school.

But both Hertwig and the older school agree in pointing out the _pet.i.tio principii_ involved in the a.s.sumption that the archetype represents the ancestral form; both reject the simplicist conception of a monophyletic evolution (which may be likened to the "one animal" idea of the transcendentalists); both admit the possibility that evolution has taken place along many separate and parallel lines, and explain the correspondences shown by these separate lines by the similarity of the intrinsic laws of evolution; finally, both emphasise the fact that we know nothing of the actual course of evolution save the few indications that are furnished by palaeontology, and both insist upon the unique importance of the palaeontological evidence.[542]

It was a curious but very typical characteristic of evolutionary morphology that its devotees paid very little attention to the positive evidence acc.u.mulated by the palaeontologists,[543] but shut themselves up in their tower of ivory and went on with their work of constructing ideal genealogies. It was perhaps fortunate for their peace of mind that they knew little of the advances made by palaeontology, for the evidence acquired through the study of fossil remains was distinctly unfavourable to the pretty schemes they evolved.

As Neumayr, Zittel, Deperet, Steinmann and others have pointed out, the palaeontological record gives remarkably little support to the ideal genealogies worked out by morphologists. There is, for instance, a striking absence of transition forms between the great cla.s.sificatory groups. A few types are known which go a little way towards bridging over the gaps--the famous _Archaeopteryx_, for example--but these do not always represent the actual phylogenetic links. There is an almost complete absence of the archetypal ancestral forms which are postulated by evolutionary morphology. Amphibia do not demonstrably evolve from an archetypal Proamphibian, nor do mammals derive from a single generalised Promammalian type. Few of the hypothetical ancestral types imagined by Haeckel have ever been found as fossils. The great cla.s.sificatory groups are almost as distinct in early fossiliferous strata as they are at the present day. As Deperet says in his admirable book,[544] in the course of a presentation of the matured views of the great Karl von Zittel, "We cannot forget that there exist a vast number of organisms which are not connected by any intermediate links, and that the relations between the great divisions of the animal and vegetable kingdoms are much less close than the theory [of evolution] demands. Even the Archaeopteryx, the discovery of which made so much stir and appeared to establish a genetic relation between cla.s.ses so distinct as Birds and Reptiles, fills up the gap only imperfectly, and does not indicate the point of bifurcation of these two cla.s.ses. Intermediate links are lacking between Amphibia and Reptiles. Mammals, too, occupy an isolated position, and no zoologist can deny that they are clearly demarcated from other Vertebrates; indeed, no fossil mammal is certainly known which comes nearer to the lower Vertebrates than does Ornithorhynchus at the present day" (p.

115).

To take a parallel from the Invertebrata, B. B. Woodward,[545] after discussing the phylogeny of the Mollusca as worked out by the morphologists and comparing it with the probable actual course of the evolution of the group, as evidenced by fossil sh.e.l.ls, sums up as follows:--"The lacunae in our knowledge of the interrelationships of the members of the various families and orders of Mollusca are slight however, compared with the blank caused by the total absence from palaeontological history of any hint of pa.s.sage forms between the cla.s.ses themselves, or between the Mollusca and their nearest allies. Nor is this hiatus confined to the Molluscan phylum; it is the same for all branches of the animal kingdom. There is circ.u.mstantial evidence that transitional forms must have existed, but of actual proof none whatever.

All the cla.s.ses of Mollusca appear fully fledged, as it were. No form has as yet been discovered of which it could be said that it in any way approached the hypothecated prorhipidoglossate mollusc, still less one linking all the cla.s.ses" (p. 79).

Pointing in the same direction as the absence of transitional forms is the undeniable fact that all the great groups of animals appear with all their typical characters at a very early geological epoch. Thus, in the Silurian age a very rich fauna has already developed, and representatives are found of all the main Invertebrate groups--sponges, corals, hydroid colonies, five types of Echinoderms, Bryozoa, Brachiopods, Worms, many types of Mollusca and Arthropoda. Of Vertebrates, at least two types of fish are present--Ganoids and Elasmobranchs. In the very earliest fossiliferous rocks of all, the Precambrian formation, there are remains of Molluscs, Trilobites and Gigantostraca, similar to those which flourished in Cambrian and Silurian times.

The contributions of palaeontology to the solution of the problems of descent posed by morphology are, however, not all of this negative character. The law of recapitulation is in some well-controlled cases triumphantly vindicated by palaeontology. Thus Hyatt and others found that in Ammonites the first formed coils of the sh.e.l.l often reproduce the characters belonging to types known to be ancestral, and what is more they have demonstrated the actual occurrence of the phenomenon known as acceleration or tachygenesis, often postulated by speculative morphologists.[546] This is the tendency universally shown by embryos to reproduce the characters of their ancestors at earlier and earlier stages in their development.