Talk:Russell1930 7

VII WILHELM ROUX AND THE MECHANICS OF DEVELOPMENT[edit] T HE experimental study of development arose almost contemporaneously with the cytological studies that culminated in the theory of the germ-plasm, and it owed more to William Roux than to any other man. But one forerunner, W. His, deserves passing mention. In 1874, 1 in revolt against the prevailing fashion of phylogenetic speculation, he pointed out that the problems of development were essentially physiological, and he went some way towards indicating the proper line of attack upon them. His principle, for instance, of organ-forming areas in the germ 2 was later to prove a valuable and important one, and his ‘principle of unequal growth’ was also fertile.

His attitude was that of the physiologists of his time who based their work on the mechanistic hypothesis. Each stage of development was regarded by His as the sufficient cause of the next, which followed mechanically from it, and it was the business of embryology to trace out these causal connexions.

‘Embryology’, he wrote, ‘is, in essence, a physiological science; it has not only to describe the building up of every single form from the egg, according to its different phases, but to trace it back in such a way that every stage of development with all its peculiarities appears as the necessary result of those immediately preceding’ (p. 2). Note the contrast with Baer’s view (supra, p. 35) that each stage conditions the next stage in development, but is not in any full sense its cause.

1 W. His, Unsere Korperform und das pbysiologiscbe Problem ibrer Entstebung y Leipzig, 1874.

2 ‘I call principle of organ-forming germinal areas the principle according to which the germinal disc contains the plan of the organs laid out in superficial extension, each point of the germinal disc corresponding to one of the later developed organs’ (P- > 9 )

9 6 WILHELM ROUX AND THE

His was a bitter opponent of Haeckel’s ‘biogenetic law’ of recapitulation, and made the interesting suggestion, later developed by O. Hertwig, 1 that many of the early and simple stages of ontogeny are necessary results of the progress from simplicity to complexity, and may be explained without reference to ancestral history (p. 210). In general he accepted Baer’s law of development, but emphasized the fact that specific distinctions are often observable in the very early stages of ontogeny.

Although there were experimentalists before Roux who tackled the problems of form, he may be regarded as the real founder of causal morphology, by reason of his pioneer work on functional adaptation and experimental embryology, his foundation of the Archiv fur Entwickelungsmechanik (1895), the immense amount of research which he stimulated, and the hard thinking he put into the interpretation of his results. 2 For our special purpose here, the study of method, he is particularly important, in that he attempted to steer a middle course between a ‘too simple’ materialism and a ‘metaphysical’ vitalism — Scylla and Charybdis as he calls them in his Introduction to the Archiv — though remaining fully convinced that materialism formed the only sound basis for exact science.

We may consider first this introductory paper, 3 in which perhaps the clearest account of Roux’s conception of biological method is to be found. The choice of the word Entwickelungsmechanik — the mechanics of development — is deliberate and significant. With Kant and Spinoza he considers that for the purpose of exact science causality is synonymous with mechanical necessity. Hence all science is in the long run mechanics, and the science of the production

1 Handbuch vergl. exper . Entwickelungslehre der Wirbeltiere, iii, 3, pp. 149-80, Jena, 1906.

1 The early history and development of causal morphology is outlined in Form and Function (1916), Ch. XVIII, where also many details of Roux’s work are given which need not be reproduced here.

3 ‘Einleitung’. Archiv fur Entwickelungsmechanik , i, 1895, pp. 1-41. English translation by W. M. Wheeler under the title ‘The Problems, Methods and Scope of Developmental Mechanics’ in Wood’s Holl Biol. Lectures for 18Q4 , Boston, 1895, pp. 149-90.

MECHANICS OF DEVELOPMENT 97

of form is properly designated the mechanics of development. Physics and chemistry attempt to reduce the most diverse phenomena to movements of mat ter ; developmental mechanics must in the measure open to it attempt the same reduction. It need not push its analysis below the level of physicochemical processes, and it may not at present be able to get even so far. This reduction to physico-chemical processes — to ‘simple components’ as Roux calls them — must, however, remain its ultimate aim.

But in practice we find that ‘organic structure is mainly due to the operation of components which at present are so complicated as to exceed the limits of our observation’. These Roux calls ‘complex components’. Although they depend in the last analysis upon simple components, ‘nevertheless the complexity of their composition lends them attributes which often differ so widely from those of inorganic modi operandi (‘Wirkungsweise’) that they are not only very dissimilar but even appear to contradict in part the functions of these same inorganic modi operandi [simple components]’ (Wheeler’s trans., p. 153). The proximate task of developmental mechanics must be then to analyse the process of formproduction into its complex components. These complex components may be mysterious in the sense that they cannot yet be fully explained in terms of their simple components, but in so far as they are constant and produce the same result under the same conditions they are of great scientific value in explaining development.

The complex components which Roux distinguishes are, first, the elementary cell-functions — assimilation, dissimilation, movement, division (as a special case of movement). To these must be added the more complicated functions of the cell, such as its powers of typical form-production and of qualitative self-differentiation. Underlying them all is the fundamental power of self-regulation without which living organisms could not persist. If we take into account the phylogenetic history of living things we must provisionally recognize still more complicated components, such as variation (adaptation) and heredity.

3727 o

98 WILHELM ROUX AND THE

The result of such analysis applied to the development of any complicated organism is that

‘all the extremely diverse structures of multicellular organisms may be traced back to the few modi operandi of cell-growth, of cell-evanescence (Zellenschwund), cell-division, cell-migration, active cell-formation, cell-elimination, and the qualitative metamorphosis of cells; certainly, in appearance at least, a very simple derivation. But the infinitely more difficult problem remains not only to ascertain the special role which each of these processes performs in the individual structure, but also to decompose these complex components themselves into more and more subordinate components’ (p. 152).

It will be noticed that the method is purely analytical, and that its first result, as Roux himself admits, is complication rather than simplification. Simplification will only appear after analysis has been applied to many complex processes and has disclosed again and again the same simple components. There is no real attempt made towards a re-synthesis of the components distinguished; the concept of the organism as a unitary functional whole is disregarded by Roux; the complex components which his analysis disentangles are not joined up again and combined into the individualized activity of the whole from which by abstraction they were originally separated out.

Leaving this question of the unitary conception of the organism for further discussion later, we may note as characteristic of Roux’s point of view that he considers the first business of developmental mechanics to be the discovery of the complex components of form-production and formmaintenance — the analysis of these into their simple components can wait. ‘Developmental mechanics should’, he writes, ‘cultivate the analysis of formative processes into constant “complex components” to a greater extent, if anything, than the ascertainment of simple components’(p. 170). We must be content for many years to come with an analysis into complex components. Here then is the point of Roux’s remark about Scylla and Charybdis to which reference was made at the beginning of the chapter. We may quote the passage in full:

‘ “ Incidit in Scyllam , qui vult vitare Charybdim ” is particularly

MECHANICS OF DEVELOPMENT 99

applicable to the investigator in the field of developmental mechanics. The too simply mechanical and the metaphysical conception represent the Scylla and the Charybdis, to steer one’s course between which is indeed a difficult task, a task which few have hitherto accomplished. It cannot, however, be denied that the seductiveness of the latter views has been increasing with the increase in our knowledge’ (p. 172).

Roux was no crude and unthinking materialist. While he accepted with full conviction the view that all material happenings, including the phenomena of organic development and activity, were mechanically determined and were in ultimate resort to be explained in terms of matter in motion, he was by no means blind to the complexity and mystery of organic activities, and he saw clearly that the reduction of life to ultimate mechanism was a goal so remote as to be almost unattainable; meanwhile the sound line of advance was to analyse the complex phenomena of life into simpler, more elementary functions, common to the majority of cells, tissues, and organs. He rejected dualistic vitalism on the ground that it inhibited research, by its insistence on the hopelessness of the task of arriving at a mechanistic explanation of life. He preferred to continue the struggle towards a possibly unattainable ideal.

His attitude to the vitalism of Driesch and others is clearly expressed in an important paper 1 written towards the close of his life, in which he looks back upon his life-work in the field of biological method. It is worth our while to reproduce the general argument of this paper, at least in outline, for the light it throws upon Roux’s final position.

Living things, he declares, can be defined only in terms of function , 2 and he distinguishes nine fundamental properties or functions essential to life. These are much the

1 ‘Die Selbstregulation, ein characteristisches und nicht notwendig vitalistisches Vermogen aller Lebewesen’. Nova Acta, Abb , K, Leop. -Carol, Deutscben Akad . Naturf, y vol. c, Nr. 2, Halle, 1914, 91 pp. (The complete volume is dated 1915).

a See also his book Die Entwickelungsmecbanik , ein neuer Zweig der biologiscben Wissenscbafty Vortrage und Aufsatze uber Entwickelungsmechanik, Heft i, Leipzig, 1905, where he states the ‘functional minimum-definition’ of life — ‘Living beings are bodies which change through causes inherent in themselves, but — also through inherent causes — have the power of remaining for a certain time relatively unaltered in spite of interchange of matter with the external environment’ (p. 107).

ioo WILHELM ROUX AND THE

same as he enumerated in his paper of 1895 (see above, p. 97) as ‘complex components’. Abstraction is made of all psychical qualities of living things, though the existence and purposive nature of such qualities is not denied. To the nine functional properties there must be added one fundamental faculty before the definition is complete; this is the power of selfregulation, which is shown in the exercise of the nine main functions. ‘From this general faculty there results direct adaptability to the change of external relations, as also protection against the effects of this change, and at the same time the power of self-maintenance and the stability of the organism are greatly increased’ (p. 86).

Living things carry to a very marked degree the causes of their activity and development within themselves. For this reason Roux speaks always of the ‘s^lf-activity’ of the organism and qualifies all the nine functions with the prefix ‘Self’. This independence of environment is of course not an absolute one, but for the most part the environment supplies not the determining but the conditioning factors of organic activity. The main causes of form-production are to be sought inside, not outside, the organism.

The possession of these functional activities, together with the fundamental power of self-regulation, and the fact that these properties are ‘self-determined’, owing little or nothing of their essential nature and mode of manifestation to the external environment, mark off living things as in some measure different from all inorganic objects. ‘The “totality” of these characteristic vital functions ( Autoergasien ) makes the living being into something different from all inorganic bodies, and invests it with a so-called inwardness ( Innerlich keiif (p. 1 1). This is one of the passages in which Roux gets nearest to accepting the functional unity of the organism as something distinguishing it from all other material bodies (see also below, p. 105). The power of self-regulation, he admits, looks very much as if it depended upon a purposive agent, as the vitalists contend ; but this power, as well as the other vital functions, might conceivably have arisen at the very dawn of life by selection of fortuitous variations

MECHANICS OF DEVELOPMENT ioi

(p. 88) 1 and thus be in principle mechanically explicable. He concludes therefore that there is no proof that a mechanistic explanation of these specifically vital activities is impossible (p. 87). Hence we should reject vitalism, and continue to work steadily and doggedly along mechanistic lines.

He considers in some detail Driesch’s supposed ‘proofs’ of the autonomy of vital processes and criticizes them acutely. Of very special methodological interest is the alternative hypothesis to Driesch’s entelechy which he adopts as an explanation of the differentiation and regulation of structure in development. Driesch had objected that a complex machine-structure cannot possibly be divided in such a way that two replicas of half-size are formed, and yet on mechanistic principles this is what must happen at every cell-division throughout the course of development. Roux’s reply is worth quoting at some length:

‘A self-propagation of the “developed living creature”, i.c. of a complete machine ready to function as such, does not take place at all in sexual propagation. But the “generative germ-plasm” proliferates, and from it there develops subsequently the new being. Who, however, in the present state of biology would maintain that this development took place therefore without the determining co-operation of the germ-plasm, in this case the somatic germ-plasm? In any case it is simpler to bring in the germ-plasm — the substance generally responsible for reproduction and heredity — rather than an entelechy. The actual reproduction as such of a machine, constructed for a specific purpose, need not be considered, since a specific reproductive substance contained therein, though not concerned with the actual functioning of the machine, can produce and further this reproduction. Reproduction can accordingly take place everywhere, corresponding to Weismann’s continuity of the germ-plasm, by means of a substance evolved solely for this object, and adapted to it — the germplasm. The real problem of the reproduction of living things depends actually only upon the morphological assimilation of the germ-plasm. But that this substance produces that for which and through which

1 In his voluminous writings and in particular in his early work Der Kampf der Tbeile im Organismus (Leipzig, 1881). Roux devotes much space to showing the possibility of the 'natural’ origin of the fundamental vital functions. See also below, p. 103.

102 _ WILHELM ROUX AND THE

it has been evolved is not metaphysical, and requires no entelechy* (pp. 68-9).

It will be remembered how in discussing Weismann’s theory we remarked that the germ-plasm played as it were the part of a ‘materialistic entelechy’ (see above, p. 50). Here we see that Roux actually puts in place of Driesch’s entelechy the concept of an active form-producing germplasm. The one explanation is in fact the equivalent of the other. This point is a very important one for an understanding of the modern attitude towards the problems of development and heredity, and it will occupy our attention again in a later chapter.

It leads us also, to consider some further aspects of Roux’s conception of development, upon which we have not hitherto touched. Roux was a firm adherent of the germplasm doctrine, though not in the elaborate form given to it by Weismann. 1 He expressed it much more in physiological than in morphological terms. Thus he attributed to the germ-plasm a high power of self-regulation, to which its constancy from generation to generation was due. ‘The constancy of the species is merely the ultimate product of the self-regulating mechanism of the germ-plasm, acquired only very slowly, but finally brought to a wonderful state of perfection. Thus in spite of great changes in the needs of its life externally, it remains relatively unchanged and propagates’ (p. 38). Differentiation he considered to be due to the interaction of the germ-plasm of the somatic cells with the cell-body, and as early as 1903 he suggested that the cytoplasm acts as an activating and differentiating influence upon the uniform nuclear germ-plasm (p. 42, f. n.).

But speaking generally, the germ-plasm theory plays no very great part in what is distinctive in his thought. It is worth noting, however, that like most upholders of the germplasm doctrine he postulated the existence of ultra-microscopic units, after the fashion of Weismann’s biophors,

1 He was doubtful whether it was necessary to ascribe to the egg a highly complex germinal structure, and he believed much more in epigenesis than did Weismann (1905, pp. 101-3).

MECHANICS OF DEVELOPMENT 103

accepting thus the principle of ‘biological atomism’. It is not necessary to enter into detail regarding the hierarchy of units which he distinguished 1 — isoplassons, autokineons, automerizons, idioplassons, and cell-organs — but one point may be brought out. The lowest conceivable unit is the isoplasson which exhibits metabolism, but need not have any elaborate morphological structure, for a flame, as Roux points out, also shows metabolism. An autokineon is an isoplasson possessing the faculty of movement. A certain degree of selfregulation characterizes both the isoplasson and autokineon, which may be regarded as phylogenetically the precursors of living beings, but not as being themselves possessed of the full qualities or properties of life. True living beings of the lowest order are the automerizons, which have the power of self-division. They are more or less equivalent to Weismann’s biophors, de Vries’ pangens, and other hypothetical units (1905, p. 1 15). A higher step is represented by the idioplassons, which have a definite morphological structure handed down by division. They appear to correspond with Weismann’s determinants. At this stage there appears the all-important faculty of ‘morphological assimilation’, i.e. the building up and maintenance through metabolism of characteristic structure ; on this depends ‘morphological heredity’, or the handing on of typical structure, together with the power of maintaining that structure. The lower orders of vital units depend upon ‘chemical assimilation’ for their continuance, and show only ‘chemical heredity’. These grades of units have for us only a theoretical interest, and it is easy to see that Roux arrived at them merely by means of progressive abstraction; they are not necessarily ‘real’. But Roux laid some stress on this line of thought, for it seemed to him to indicate how living beings, with all their apparently marvellous properties, could have evolved step by step from the isoplasson upwards, adding now one quality and now another, as these arose through variation and were perpetuated by selection. He sought in this way an escape from the conclusion that living beings show powers which are not

1 See Roux, 1905, pp. 108-27.

io 4 WILHELM ROUX AND THE

explicable mechanically. We shall not follow out his argument, which is ingenious but unconvincing.

What is more important from our point of view is Roux’s insistence on metabolism as the foundation of the other qualities of the organism. In morphological assimilation Roux saw ‘the most general, most essential, and most characteristic formative activity of life’. 1 Through morphological assimilation form arises and form is maintained; assimilation is thus the basis of growth, development, and heredity.

‘Assimilation in its different forms is the basis of this faculty of selfpreservation and the preliminary condition of the self-formation of the individual, as of the continuity of the living thing, for the substance which produces this effect must in all its specific structure be renewed in metabolism and proliferate as well. Accordingly only the variations which are capable of assimilation are heritable. It can be said with wider significance that assimilation effects a kind of material memory, or the transference of the law of persistence from simple movements to those processes of form-production which are connected with metabolism’ (1905, p. 108).

The principle here enunciated, that only those variations can be inherited which fall into, or share in, the metabolic cycle, is clearly a pregnant one.

Roux may perhaps best be described as a physiological morphologist. He was physiologist in that he insisted upon the importance of metabolism as the foundation of all formproduction, all form-maintenance, and all form-change; he was morphologist in that he was interested primarily in form and not in physiological chemistry. His aim was to build up a new branch of biology, a science of the production of form, and this took him beyond the scope and methods of ordinary physiology, which is mainly concerned with the mode of action of the formed body, considered as a physico-chemical mechanism.

Although Roux clung firmly to the mechanistic theory as the only sound basis for exact science, he did attempt to take

1 ‘Ueber die Selbstregulation der Lebewesen’, Arch. Ent.-Mecb xiii, 1902 (pp. 610-50), p. 631.

MECHANICS OF DEVELOPMENT 105

up a position differing from that of the purely mechanistic physiologist. This position was not really intermediate, as he believed, between simple materialism and dualistic vitalism. He seems all the time to be striving after a position which should be distinctly biological and yet not in contradiction with the fundamental mechanistic standpoint which he adopts. But he is never quite successful in this attempt, and indeed the effect is foredoomed to failure, for the strict mechanistic point of view is incompatible with any real ‘organismal’ theory of living things.

As we have already seen (p. 98, above), Roux does not actually reach the conception of organism as something unitary and individual ; he treats of vital properties in terms of metabolic processes; living substance regarded as a metabolic system is his ultimate concept, not the living organism, or even the living cell. The way was open to him when he seized the importance of ‘complex components’ — which are essentially the functions or activities of cells and organisms — but he did not take it. The conviction (arising from his mechanistic philosophy) that complex components must be analysed into simple components prevented him from seeing that a satisfactory middle position could be reached by linking up his complex components, which are functions, with the activity of the organism regarded as a unitary whole, and basing a distinctively biological method upon this conception of unity. He preferred to take the other course — that of progressive analysis, leading away from the conception of synthetic unity and ‘wholeness’.

It must, however, be recognized that Roux was not far off recognizing the individuality of the living thing as something unique and different from anything in the inorganic sphere. We have already quoted a passage in which he speaks of the living being as something marked off by the ‘Gesamtheit’ of its functions from all inorganic bodies, and as possessing a certain ‘Innerlichheit’ (see p. 100, above). Another passage may be adduced in which he emphasizes the fact that living things are in no way artificial systems, but show a unique degree of autonomy and independence —

3727

P

io6 WILHELM ROUX AND THE

‘Living things are not systems arbitrarily separated by us conceptually, but systems which demarcate themselves, and reproduce themselves through division and development, in which processes all, or almost all that is essential, is selfdetermined’ (1905, p. 182). But it seems clear from a study of Roux’s work as a whole that he considered organic systems to differ in degree and not in kind from inorganic systems, and both to be explicable in the long run on mechanistic principles.

We may turn now, after this discussion of Roux’s general standpoint, to consider some of his more positive contributions to the understanding of development and heredity, of which we shall single out one or two of the most important. As I have tried to show in a previous work 1 Roux’s greatest service to biology was his insistence on the importance of function and functional activity. He made a close study of functional adaptation and analysed the facts with much success. He was the first to realize and to demonstrate the great role that function plays in the development of form, and he established a valuable generalization regarding the relation of form to function at different stages of development. This was expressed first in his book Die Karnff der Theile (Leipzig, 1881), in which he wrote, ‘There must be distinguished in the life of all the parts two periods, an embryonic in the broad sense, during which the parts develop, differentiate, and grow of themselves, and a period of completer development, during which growth, and in many cases also the balance of assimilation over dissimilation, can come about only under the influence of stimuli’ (p. 180). There is thus a period of self-differentiation in which the organs are roughly formed in anticipation of functioning, and a period of functional development, in which the organs are perfected through functioning, and only through functioning. The two periods cannot be sharply separated from one another, nor does the transition from the one to the other occur at the same time in the different tissues and organs.

1 Form and Function , 1916, pp. 316-29.

MECHANICS OF DEVELOPMENT 107

The conception is more fully expressed by Roux in 1905 as follows:

This separation (of development into two periods) is intended only as a first beginning. The first period I called the embryonic period kclt ££oxrjv 9 or the period of organ rudiments. It includes the “directly inherited” structures, i. e. the structures which are directly predetermined in the structure of the germ-plasm, as, for instance, the first differentiation of the germ, segmentation, the formation of the germ-layers and the organ-rudiments, as well as the next stage of “further differentiation”, and of independent growth and maintenance, that is, of growth and maintenance that take place without the functioning of the organs.

This is accordingly the period of direct fashioning through the activity of the formative mechanism implicit in the germ-plasm, also the period of the self-conservation of the formed parts without active functioning.

The second period is the period of “functional form-development”. It includes the further differentiation and the maintenance in their typical form of the organs laid down in the first period; and this is brought about by the exercise of the specific functions of the organs. This period adds the finishing touches to the finer functional differentiation of the organs and so brings to pass the “finer functional harmony” of all organs with the whole. The formative activity displayed during this period depends upon the circumstance that the functional stimulus, or rather the exercise by the organs of their specific functions, is accompanied by a subsidiary formative activity, which acts partly by producing new form and partly by maintaining that which is already formed. . . . Between the two periods lies presumably a transition period, an intermediary stage of varying duration in the different organs, in which both classes of causes are concerned in the further building-up of the already formed, those of the first period in gradually decreasing measure, those of the second in an increasing degree* (1905, pp. 94 -6).

Roux had in his earlier researches paid much attention to the process which is characteristic of second-stage development, namely, functional adaptation. He had shown, for instance, that the later development of the blood vessels comes about in direct response to functional requirements. Thus from the rudiments formed in the first period of development, before the system is properly functional, there

io8 WILHELM ROUX AND THE

sprout out and grow the definitive blood-vessels supplying tne tissues which have need of them. The size, direction, and intimate structure of these blood-vessels come to be accurately adjusted through functional adaptation to the part they play in the economy of the whole, and this adjustment or adaptation is brought about by the active response of the various tissues of which the blood-vessels are composed. 1

The relation between the two stages of development naturally suggests the question — is it possible that characters which have been developed as second-stage characters in direct response to functional requirements can appear in later generations as first-stage characters, in advance of functioning ? Roux was, at least in his early days, strongly inclined to the view that this transformation of second-stage characters into first-stage did take place in the course of many generations, and on this hypothesis he explained the transmissibility of acquired characteristics.

Speaking of the formative stimuli which are active in second-period development, Roux wrote in 1881: ‘These stimuli can also produce new structure, which if it is constantly formed throughout many generations finally becomes hereditary, i.e. develops in the descendants in the absence of the stimuli, becomes in our sense embryonic’ (1881, p. 180). Again, ‘form-characteristics which were originally acquired in post-embryonic life through functional adaptation may be developed in the embryo without the functional stimulus, and may in later development become more or less completely differentiated, and retain this differentiation without functional activity or with a minimum of it. But in the continued absence of functional activity they become atrophied . . . and in the end disappear’ (1881, p. 201).

The transmission of acquired characters certainly formed an element in Roux’s earlier conceptions of heredity and development ; he was of opinion that such transmission takes place in small degree and gradually, and that many generations are required before a new character can become

1 Sec A. Oppel, Ueber die gestaltlicbe Anpassung der Blutgefasse (containing a long section by W. Roux), Roux’s For tr age und Aufsdtze , x, Leipzig, 1910.

MECHANICS OF DEVELOPMENT 109

hereditary. It should be noted however that these views date from before the establishment of the germ-plasm doctrine.

Roux paid much attention not only to normal development, particularly in its functional aspect, but also to the response of the developing organism to abnormal conditions, whether external or internal — to the problem of autonomy and regulation. It was one of the great services of experimental embryology that it showed how widespread and important such regulatory processes are — and how difficult to explain.

Roux distinguished two main kinds of development — typical and regulatory. Typical development is found when the hereditary equipment of the germ and the conditions of development are both ‘normal’. It is the production of typical form in the typical way. Regulatory development on the other hand builds up, in spite of atypical germinal beginnings or abnormal environmental conditions, the typical form of the organism (1905, p. 76). It is shown very clearly in all cases of restitution and regeneration. Typical development as a matter of fact never actually takes place, for the least departure from the typical course of development calls forth into activity the mechanisms of self-regulation. Much of normal development is in reality regulatory, for instance, most processes of functional adaptation.

Regulatory development is much more difficult to explain mechanistically than typical development. In typical development there is much that is automatic and mechanical, and therefore susceptible of strict analysis. In the ‘first period’ at least, typical development takes a more or less habitual and unchanging course, and each part shows a considerable degree of independence and a certain power of self-differentiation. Regulatory development on the other hand shows a disconcerting spontaneity, and is governed in a mysterious manner by the condition of the organism as a whole or at least of large parts of it (1905, p. 72). Regulatory development is much more a living, responsive activity of the organism than is typical development, and it is to this degree

no

WILHELM ROUX AND THE less easy to explain. It was the extraordinary purposiveness of regulatory phenomena that led many investigators, e.g. G. Wolff and fi. Driesch, to assume the intervention in development of a vitalistic agent.

Roux, as we have seen, did not accept this vitalistic conclusion. He held that the riddle is not quite so insoluble as it seems at first sight. He pointed out that no new tissuequalities are implied in regulatory processes, beyond those that may be conceived to have arisen through natural selection, in the course of the evolution of the primal organisms from isoplassons and idioplassons. Many cases of regeneration could, he thought, be explained on the hypothesis that reserve idioplasm is stored in the chromosomes of the somatic cells ready to regenerate lost parts.

Even from the slight sketch of Roux’s work which we have given it will be apparent that there were few of the great problems of development and heredity which he did not touch upon and illuminate. We have seen with what skill and assiduity he treated the problems of differentiation, bringing to light the very important distinction between first-stage and second-stage development, and emphasizing the role that function plays in the moulding of form. We have seen that although he accepted in its main lines the germ-plasm theory as an adequate explanation of heredity, he tried to interpret it physiologically rather than morphologically, recognizing that the germinal substance must show the fundamental characteristics of morphological assimilation and selfregulation, and not be merely a static configuration passively handed down. Its self-identity or continuity was not something fixed and rigid, but a state constantly destroyed and constantly restored through metabolism and regulation. His physiological point of view puts him in advance of Weismann, whose treatment of the problems of development and heredity was, as we have seen, schematic and formal. In his early theory of the phylogenetic relation between the two stages of development he opened the way to a rational explanation of the fact of recapitulation. Methodologically, he came near to a functional or organismal treatment of

MECHANICS OF DEVELOPMENT hi

living things, but was held back from full acceptance of this point of view by his deep-rooted belief in mechanism.

The last twenty years of the nineteenth century saw an immense amount of work carried out on the problems of experimental embryology, a great accession of new facts, and a lively interest in the philosophical questions which they inevitably raised. Typical of the new experimental school was its insistence on the importance of environment; its main interest was the study of the developing organism regarded as actively responding to its surroundings , 1 and not as a static mechanism such as morphology had conceived it.

It is not my intention to treat this period historically; all we need note here is that the majority of workers held fast to mechanistic views and attempted to apply physiological methods of explanation; some of these views we have discussed in the last chapter. There was, however, an important minority who felt that a mechanistic explanation of the new facts was in principle impossible, and fell back accordingly upon a dualistic vitalism. Noteworthy among these were G. Wolff 2 and H. Driesch. I have elsewhere 3 given my reasons for rejecting the vitalism of Driesch, and do not propose to discuss his theories 4 in detail in this book; they do not seem to lead in practice to a method of attack upon the problems in any way different from the mechanistic, and they have already been fully treated both by Driesch and by his critics. The controversy between mechanism and vitalism, having raged fiercely for years, has now died down. Let us not revive it !

1 See particularly C. Herbst, Formative Reize in der tieriseben Ontogenese , Leipzig, 1901.

2 Arcb.f. Ent.-Mcch. i, pp. 380-90, 1895; Beitrage zur Kritik der Darwinscben Lebre , Leipzig, 1898.

3 T be Study of Living Things , 1924, pp. 20-6*

4 The Science and Philosophy of the Organism (Gifford Lectures), 2 vols., London, 1908. New edition in one vol., 1929. In German as Philosophic des Organiscben , Leipzig, 1909, 2nd edit., 1921. Ordnungslehre , Jena, 1912, 2nd edit., 1923. Driesch’s earlier ‘analytic theory of organic development* (1894) is dealt with adequately by J. W. Jenkinson, Experimental Embryology , Oxford, 1909, pp. 280-6.