The _via media_ adopted by Roux is the a.n.a.lysis of development, not directly into simple physico-chemical processes, but into more complex organic processes dependent upon the fundamental properties of living matter. The aim of _Entwicklungsmechanik_ is defined by Roux to be the reduction of developmental events to the fewest and simplest _Wirkungsweisen_, or causal processes.[483] Two cla.s.ses of causal processes may be distinguished, as "complex components" and "simple components" of development. The latter are directly explicable by the laws of physics and chemistry; the former, while in essence physico-chemical, are yet so very complicated that they cannot at present be reduced to physico-chemical terms. The ultimate aim of _Entwicklungsmechanik_ is to reduce development to its "simple components," but its main task at the present day and for many years to come is the a.n.a.lysis of development into its "complex components."

These complex components must be accepted as having much of the validity of physical and chemical laws. They are mysterious in the sense that they cannot yet be explained mechanistically, but they are constant in their action, and under the same conditions produce always the same effect--hence they may be made the subject of strictly scientific study.

They represent biological generalisations, in their way of equal validity with the generalisations of physics and chemistry.

The princ.i.p.al "complex components" which Roux recognises are somewhat as follows:--First come the elementary cell-functions of a.s.similation and dissimilation, growth, reproduction and heredity, movement and self-division (as a special co-ordination of cell-movements). Then at a somewhat higher level, self-differentiation, and the trophic reaction to functional stimuli. Components of even greater complexity may also be distinguished, as, for instance, the biogenetic law. The various tropisms exhibited in development may be regarded as "directive" complex components. There must be added, not as being itself a component, but rather as a mode or peculiar property of all functioning, the omnipresent faculty of self-regulation.

It will be noticed that Roux's "complex components" are simply the general properties or functions of organised matter.

Expressing Roux's thought in another way, we might say that life can only be defined functionally, _i.e._, by an enumeration of the "complex components" or elementary functions which all living beings manifest, even down to the very simplest. "Living beings," writes Roux, "can at present be defined with any approach to completeness only functionally, that is to say, through characterisation of their activities, for we have an adequate acquaintance with their functions in a general way, though our knowledge of particulars is by no means complete" (p. 105, 1905). Defined in the most general and abstract way, living things are material objects which persist in spite of their metabolism, and, by reason of their power of self-regulation, in spite also of the changes of the environment. This is the "functional minimum-definition of life"

(pp. 106-7, 1905).

We may now go on to consider the relation of function to form throughout the course of development. Roux distinguishes in all development two periods, in the first of which the organ is formed prior to and independent of its function, while in the second the differentiation and growth of the organ are dependent on its functioning. Latterly (1906 and 1910) Roux has distinguished three periods, counting as the second the transition period when form is partly self-determined, partly determined by functioning. As this conception of Roux's is of the greatest importance we shall follow it out in some detail.

The idea was first elaborated in the _Kampf der Theile_ (1881), where 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 a.s.similation 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 antic.i.p.ation 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.

The conception is more fully expressed 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 [Greek: kat'

exochen] 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 which 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 pa.s.s the 'finer functional harmony' of all organs with the whole. The formative activity displayed during this period depends upon the circ.u.mstance 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 cla.s.ses 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" (pp. 94-6, 1905).

In the first period the organ forms or determines the function, in the second period the function forms the organ, or at least completes its differentiation. It is characteristic that in the first period functionally adapted structure appears in the complete absence of the functional stimulus.

The explanation of the difference between the two periods is to be found in the different evolutionary history of the characters formed during each. First-period characters are _inherited_ characters, and taken together const.i.tute the historical basis of the organism's form and activity; second-period characters are those of later acquirement which have not yet become incorporated in the racial heritage.

Inherited characters appear in development in the absence of the stimulus that originally called them forth; acquired characters are those that have not yet freed themselves from this dependence upon the functional stimulus. First-period characters were originally, like second-period characters, entirely dependent for their development upon the functional stimuli in response to which they arose, and only gradually in the course of generations did they gain that independence of the functional stimulus which stamps them as true inherited characters. Speaking of the formative stimuli which are active in second-period development, Roux writes:--"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" (p. 180, 1881). 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" (p. 201, 1881).

This conception of the nature of hereditary transmission is an important one, and const.i.tutes the first big step towards a real understanding of the historical element in organic form and activity. It supplies a practical criterion for the distinguishing of "heritage" characters from acquired characters, of palingenetic from cenogenetic--a criterion which descriptive morphology was unable to find.[484] The introduction of a functional moment into the concept of heredity was a methodological advance of the first importance, for it linked up in an understandable way the problems of embryology, and indirectly of all morphology, with the problem of hereditary transmission, and gave form and substance to the conception of the organism as an historical being.

It is this element in Roux's theories that puts them so far in advance of those of Weismann. Weismann did not really tackle the big problem of the relation of form to function, and he left no place in his mechanical system of preformation for functional or second-period development; he conceived all development to be in Roux's sense embryonic, and due to the automatic unpacking of a complex germinal organisation. Roux himself was to a certain extent a preformationist, for the development of his first-period characters is conditioned by the inherited organisation of the germ-plasm, and is purely automatic. It was indeed his experiments on the frog's egg (1888) that supplied some of the strongest evidence in favour of the mosaic theory of development. The number of _Anlagen_ which he postulates in the germ is however small, and the germ-plasm in his conception of it has a relatively simple structure (p. 103, 1905).

The transmission of acquired characters forms, of course, an integral part of Roux's conception of heredity and development, for without this transmission second-stage characters could not be transformed into first-stage characters. He discusses this difficult question at some length in the _Kampf der Theile_, coming to the conclusion that such transmission takes place in small degree and gradually, and that many generations are required before a new character can become hereditary.

He thinks that acquired characters are probably transmitted at the chemical level. It is conceivable that acquired form-changes are dependent on chemical changes, or are correlative with such, and that, since the germ-cells stand in close metabolic relations with the soma, these chemical changes may soak through to the germ-cells and so modify them that a predisposition will appear in the descendants towards similar form-changes.[485] From this point of view the problem of transmission might be merged in the broader problem of the production of form through chemical processes--the central problem of all development.

Inherited characters develop by an automatic process of self-differentiation, and the separate parts of the embryo show during this first period a surprising functional independence of one another.

But this state of things changes progressively as the second period is reached, until finally all form-production and maintenance and all correlation depend upon functioning. It is in the first period of automatic development through internal "determining" factors that the "developmental" functions in the strict sense, _e.g._ automatic growth, division and self-differentiation, are most clearly shown. In the second or "functional" period the formative influence of function upon structure comes into play, and development becomes largely a matter of "functional adaptation" to functional requirements.

All structure, according to Roux, is either functional or non-functional. The former includes all structure that is adapted to subserve some function. "Such 'functional structures' are, for example, the composition of striated muscle fibres out of fibrillae and these out of muscle-prisms, or again the length and thickness of the muscles, the static structure of the bones, the composition of the stomach and the blood-vessels out of longitudinal and circular fibres, the external shape of the vertebral centra and of the cuneiform bones of the foot"

(p. 73, 1910). Indeed, as Cuvier had already pointed out, practically every organ in the body shows a functional structure which is accurately and minutely adjusted to the function it is intended to perform. Thus, to take some further examples, the arteries are admirably adapted as regards size of lumen, elasticity of wall, direction of branching, to conduct the blood to all parts of the body with the least possible waste of the propelling power through frictional resistance. So, too, the spongy substance of the long bones is arranged in lamellae which take the direction of the princ.i.p.al stresses and strains which fall upon the bones in action.

Functional structure may be formed either in the first or in the second period of development, may be either inherited or acquired, but it reaches its full differentiation only in the second period, _i.e._, under the influence of functioning. Practically speaking, functional structure is directly dependent for its full development and for its continued conservation upon the exercise of the particular function which it serves. In the second period, but not in the first, increased use leads to hypertrophy of the functional structure, disuse to atrophy.

From functional structure is to be distinguished nonfunctional structure, which has no relation to the bodily functions--is neither adapted to perform any of these, nor has arisen as a by-product of functional activity. "To this category belong, for example, among typical structures, the triangular form of the cross-section of the tibia, the dolicocephalic or brachycephalic shape of the skull, most of the external characters distinguishing genera and species, many of the external features of the embryo which change in the course of development, besides most of the abnormal forms shown by monstrosities, tumours, etc." (p. 74, 1910). Non-functional structure is not affected by functional adaptation, and may accordingly be left out of consideration here.

Now the influence of functioning upon the form and structure of an organ is twofold. There is first the immediate change brought about by the very act of functioning--for example, the shortening and thickening of skeletal muscles when they act. This is a purely temporary change, for the organ at once returns to its normal quiescent state as soon as it ceases to function. Such temporary functional change, brought about in the moment of functioning, is usually dependent for its initiation upon some neuro-muscular mechanism, though it may be elicited also by a chemical stimulus. It is thus always a phenomenon of "behaviour." "From such temporary changes are sharply to be distinguished all permanent alterations which first appear in perceptible fashion through oft-repeated or long-continued, enhanced functional activity. These produce a new and lasting internal equilibrium of the organ, consisting in an insertion of new molecules or a rearrangement of old. For this reason they outlast the periods of functional form-change, or, if as in the case of the muscles they themselves alter during functional activity, they regain their state when the organ ceases to function" (p.

72, 1910). "Oft-repeated exercise or heightened exercise of the specific functions, or repeated action of the functional stimuli which determine them, produces, as we have said before, true form-changes as a by-product. These are of two kinds. In so far as these form-changes facilitate the repet.i.tion of the specific functions, I have called them _functional adaptations_.... Such as do not improve the functioning of the organ are indeed by-products of functioning, but without adaptive character; they do not belong to the cla.s.s of functional adaptations at all" (p. 75, 1910).

We may now enquire in what way functional adaptations can arise as by-products of functioning.

It is clear that natural selection in the sense of individual or "personal" selection cannot adequately explain the origin of functional structure and the functional harmony of structure, for thousands of cells would have to vary together in a purposive way before any real advantage could be gained in the struggle for existence, and it is in the highest degree unlikely that this should come about by chance variation.[486] The development of purposive internal structure is only to be explained by the properties of the tissues concerned.

In ill.u.s.tration and proof of the statement that functional adaptation is due to the properties of the tissues we may adduce the development and regulation of the blood-vascular system, which has been thoroughly studied from this point of view by Roux and Oppel (1910).

It appears that only the very first rudiments of the vascular system are laid down in the short first period of automatic non-functional development. All the subsequent growth and differentiation of the blood-vessels falls into the second period, and is due wholly or in great part to direct functional adaptation to the requirements of the tissues. Thus from the rudiments formed in the first period there sprout out the definitive vessels in direct adaptation to the food-consumption of the tissues they are to supply. The size, direction and intimate structure of these vessels are accurately adjusted to the part they play in the economy of the whole, and this adjustment is brought about in virtue of the peculiar properties or reaction-capabilities of the different tissues of which the blood-vessels are composed.

The properties which Roux finds himself compelled to postulate in the vascular tissues, after a thorough-going a.n.a.lysis of the different kinds of functional adaptation shown by the blood-vessels, are summarised by him as follows:--

"(1) The faculty--depending on a direct sensibility possessed by the endothelium and perhaps also by the other layers of the intima--of yielding to the impact of the blood, so far as the external relations of the vessel permit. In this way the wall adapts itself to the haemodynamically conditioned 'natural' shape of the blood-stream, and reaches this shape as nearly as possible." Through this faculty of the lining tissue of the blood-vessels, the size of the lumen and the direction of branching are so regulated as to oppose the least possible resistance to the flow of the blood.

"(2) The faculty possessed by the endothelium of the capillaries of each organ of adapting itself qualitatively to the particular metabolism of the organ." This adaptedness of the capillaries is, however, more usually an inherited state, _i.e._, brought about in the first period of development.

"(3) The faculty possessed by the capillary walls of being stimulated to sprout out and branch by increased functioning, _i.e._, by increased diffusion, and their power to exhibit a chemically conditioned cytotropism, which causes the sprouts to find one another and unite. A similar process can be directly observed in isolated segmentation-cells, which tend to unite in consequence of a power of mutual attraction.

"(4) The faculty of developing normal arterial walls in response to strong intermittent pressure, and normal venous walls in response to continuous lesser pressure." It has been shown, for instance, by Fischer and Schmieden that in dogs a section of vein transplanted into an artery takes on an arterial structure, at least as regards the circular musculature, which doubles in thickness.

"(5) The power to regulate the normal[487] length of the arteries and veins, in adaptation to the growth of the surrounding tissues, in such a way that the stretching action of the blood-stream brings the vessel to its proper functional length.

"(6) The power to form, in response to slight increases in longitudinal tension, new structural parts which take their place alongside the existing longitudinal fibres.

"(7) The power to regulate the width of the circular musculature according to the degree of food-consumption by the tissues, in response to nerve impulses initiated in these tissues.

"(8) The power possessed by the circular musculature of responding to such continuous functional widening, by the formation of new structural parts in the circular musculature, and so of widening the vessel permanently or by this new formation of muscular fibres thickening the circular musculature.

"(9) The faculty of being stimulated by increased blood-pressure to produce the same structural changes as mentioned in par. 8, though here the response is otherwise conditioned" (pp. 126-7, 1910).

It is by virtue of the tissue-properties detailed above that the complex functional adaptations of the blood-vessels come about.

The development of the vascular system is no mere automatic and mechanical production of form, apart from and independent of functioning; it implies a living and co-ordinated activity of the tissues and organs concerned, a power of active response to foreseen and unforeseen contingencies. Form is then not something fixed and congealed--it is the ever-changing manifestation of functional activity.

"Since most of the structure and form of the blood-vessels arises in direct adaptation to function, the vessels of adult men and animals are no fixed structures, which, once formed, retain their form and structural build unchanged throughout life; on the contrary, they require even for their continued existence the stimulus of functional activity.... The fully formed blood-vessels are no static structures, such as they appear to be according to the teaching of normal histology, and such as they have long been taken to be. Observation and description of normal development never shows us anything but the visible side of organic happenings, the _products_ of activity, and leaves us ignorant of the real processes of form-development and form-conservation, and of their causes" (p. 125, 1910).

The real thing in organisation is not form but activity. It is in this return to the Cuvierian or functional att.i.tude to the problems of form that we hold Roux's greatest service to biology to consist. The att.i.tude, however, seems to smack of vitalism, and Roux, as we have seen, is no vitalist. He holds that the marvellous and apparently purposive tissue-qualities which underlie all processes of functional adaptation have arisen "naturally," in the course of evolution, by the action of natural selection upon the various properties, useful and useless, which appeared fortuitously in the primary living organisms. He is, moreover, deeply imbued with the materialistic philosophy of his youth, and it is indeed one of the chief characteristics of his system that he states the fundamental properties or qualities of life in terms of metabolism. A vital quality is for Roux a special process or mode of a.s.similation. The faculty of "morphological a.s.similation" whereby form is imposed upon formless chemical processes is the ultimate term of Roux's a.n.a.lysis--"the most general, most essential, and most characteristic formative activity of life" (p. 631, 1902).

We have now to consider very briefly the early results achieved by Roux's fellow-workers in the field of causal morphology. As D. Barfurth points out,[488] the years 1880-90 saw a general awakening of interest in experimental morphology, and it is hard to say whether Roux's work was cause or consequence. "There fall into this period," writes Barfurth, "the experimental investigations by Born and Pfluger on the s.e.xual difference in frogs (1881), by Pfluger on the parthenogenetic segmentation of Amphibian ova, on crossing among the Amphibia, and on other important subjects (1882). In the following year (1883) appeared two papers of fundamental importance, by E. Pfluger and W. Roux: Pfluger publishing his researches on 'the influence of gravity on cell-division,' Roux his experimental investigations on 'the time of the determination of the chief planes in the frog-embryo.'... In the same year appeared A. Rauber's experimental studies 'on the influence of temperature, atmospheric pressure, and various substances on the development of animal ova,' which have brought many similar works in their train. The following year (1884) saw a lively controversy on Pfluger's gravity-experiments with animal eggs, in which took part Pfluger, Born, Roux, O. Hertwig and others, and in this year appeared work by Roux dealing with the experimental study of development, and in particular giving the results of the first definitely localised p.r.i.c.king-experiments on the frog's egg (in the _Schles. Gesell. f.

vaterl. Kultur_, 15th Feb. 1884), also the important researches of M.

Nussbaum and Gruber (followed up later by Verworn, Hofer and Balbiani) on Protozoa, and other experimental work" (pp. xi.-xii.).

In 1888 appeared a famous paper by W. Roux,[489] in which he described how he had succeeded in killing by means of a hot needle one of the two first blastomeres of the frog's egg, and how a half-embryo had developed from the uninjured cell. Some years before[490] he had enunciated, at about the same time as Weismann, the view that development was brought about by a qualitative division of the germ-plasm contained in the nucleus, and that the complicated process of karyokinetic or mitotic division of the nucleus was essentially adapted to this end. He conceived that development proceeded by a mosaic-like distribution of potencies to the segmentation-cells, that, for instance, the first segmentation furrow separated off the material and potencies for the right half of the embryo from those for the left half. He had tried to show experimentally that the first furrow in the frog's egg coincided with the sagittal plane of the embryo,[491] and his later success in obtaining a half-embryo from one of the first two blastomeres seemed to establish the "mosaic theory" conclusively.

Roux's needle-experiment aroused much interest, especially as Weismann's theory of heredity was then being keenly discussed. Chabry had published in 1887 some interesting results on the Ascidian egg,[492] which strongly supported the Roux-Weismann theory. Considerable astonishment was therefore caused by Driesch's announcement in 1891[493] that he had obtained complete larvae from single blastomeres of the sea-urchin's egg isolated at the two-celled stage. He followed this up in the next year[493] by showing that whole embryos could be produced from one or more blastomeres isolated at the four-cell stage. Similar or even more striking results were obtained by E. B. Wilson on _Amphioxus_,[494] and Zoja on medusae.[495] Driesch succeeded also in disturbing the normal course and order of segmentation by compressing the eggs of the sea-urchin between gla.s.s plates, and yet obtained normal embryos.

Similar pressure-experiments were carried out on the frog by O.

Hertwig,[496] and on _Nereis_ by E. B. Wilson,[497] with a.n.a.logous results.

In 1895 O. Schultze[498] showed that if the frog's egg is held between two plates and inverted at the two-celled stage there are formed two embryos instead of one. In the same year T. H. Morgan[499] repeated Roux's fundamental experiment of destroying one of the two blastomeres, but inverted the egg immediately after the operation--a whole embryo of half size resulted. A year or two later Herlitzka[500] found that if the first two blastomeres of the newt's egg were separated by constriction, two normal embryos of rather more than half normal size were formed.

The main result of the first few years' work on the development of isolated blastomeres was to show that the mosaic theory was not strictly true, and that the hypothesis of a qualitative division of the nucleus was on the whole negatived by the facts.

Evidence soon acc.u.mulated that the cytoplasm of the egg stood for much in the differentiation of the embryo. A number of years previously Chun had made the discovery that single blastomeres of the Ctenoph.o.r.e egg, isolated at the two-celled stage, gave half-embryos. This was in the main confirmed by Driesch and Morgan in 1896,[501] and they made the further interesting discovery that the same defective larvae could be obtained by removing from the unsegmented egg a large amount of cytoplasm. Conclusive proof of the importance of the cytoplasm was obtained soon after by Crampton,[502] who removed the anucleate "yolk-lobe" from the egg of the mollusc _Ilyana.s.sa_ at the two-celled stage, and obtained larvae which lacked a mesoblast. This result was brilliantly confirmed and extended some years later by E. B. Wilson,[503]