The Principles of Biology Vol. I/Chapter II.10a

§ 97a. Since the foregoing four chapters were written, thirty-four years ago, the topics with which they deal have been widely discussed and many views propounded. Ancient hypotheses have been abandoned, and other hypotheses, referring tacitly or avowedly to the cell-doctrine, have been set forth. Before proceeding it will be well to describe the chief among these.

Most if not all of them proceed on the assumption, shown in § 66 to be needful, that the structural characters of organisms are determined by the special natures of units which are intermediate between the chemical units and the morphological units—between the invisible molecules of proteid-substances and the visible tissue-components called cells.

Four years after the first edition of this volume was published, appeared Mr. Darwin's work, The Variation of Animals and Plants under Domestication; and in this he set forth his doctrine of Pangenesis. Referring to the doctrine of physiological units which the preceding chapters work out, he at first expressed a doubt whether his own was or was not the same, but finally concluded that it was different. He was right in so concluding. Throughout my argument the implication everywhere is that the physiological units are all of one kind; whereas Mr. Darwin regards his component units, or "gemmules," as being of innumerable unlike kinds. He supposes that every cell of every tissue gives off gemmules special to itself, and capable of developing into similar cells. We may here, in passing, note that this view implies a fundamental distinction between unicellular organisms and the component cells of multicellular organisms, which are otherwise homologous with them. For while in their essential structures, their essential internal changes, and their essential processes of division, the Protozoa and the component units of the Metazoa are alike, the doctrine of Pangenesis implies that though the units when separate do not give off invisible gemmules the grouped units do.

Much more recently have been enunciated the hypotheses of Prof. Weismann, differing from the foregoing hypotheses in two respects. In the first place it is assumed that the fragment of matter out of which each organism arises consists of two portions—one of them, the germ-plasm, reserved within the generative organ of the incipient individual, representing in its components the structure of the species, and gives origin to the germs of future individuals; and the other of them, similarly representative of the specific structure, giving origin to the rest of the body, or soma, but contains in its components none of those latent powers possessed by those of the germ-plasm. In the second place the germ-plasm, in common with the soma-plasm, consists of multitudinous kinds of units portioned out to originate the various organs. Of these there are groups, sub-groups, and sub-sub-groups. The largest of them, called "idants," are supposed each to contain a number of "ids"; within each id there are numerous "determinants"; and each determinant is made up of many "biophors"—the smallest elements possessing vitality. Passing over details, the essential assumption is that there exists a separate determinant for each part of the organism capable of independent variation; and Prof. Weismann infers that while there may be but one for the blood and but one for a considerable area of skin (as a stripe of the zebra) there must be a determinant for each scale on a butterfly's wing: the number on the four wings being over two hundred thousand. And then each cluster of biophors composing a determinant has to find its way to the place where there is to be formed the part it represents.

Here it is needless to specify the modifications of these hypotheses espoused by various biologists—all of them assuming that the structural traits of each species are expressed in certain units intermediate between morphological units and chemical units.

§ 97b. A true theory of heredity must be one which recognizes the relevant phenomena displayed by all classes of organism. We cannot assume two kinds of heredity, one for plants and another for animals. Hence a theory of heredity may be first tested by observing whether it is equally applicable to both kingdoms of living things. Genesis, heredity, and variation, as seen in plants, are simpler and more accessible than as seen in animals. Let us then note what these imply.

Already in § 77 I have illustrated the power which some plants possess of developing new individuals from mere fragments of leaves and even from detached scales. Striking as are the facts there instanced, they are scarcely more significant than some which are familiar. The formation of cauline buds, presently growing into shoots, shows us a kind of inheritance which a true theory must explain. As described by Kerner, such buds arise in Pimpernel, Toad-flax, etc., below the seed-leaves, even while yet there are no axils in which buds usually grow; and in many plants they arise from intermediate places on the stem: that is, without definite relations to pre-existing structures. How fortuitous is their origin is shown when a branch is induced to bud by keeping it wrapped round with a wet cloth. Even still better proved is the absence of any relation between cauline buds and normal germs by the frequent growth of them out of "callus"—the tissue which spreads over wounds and the cut ends of branches. It is not easy to reconcile these facts with Mr. Darwin's hypothesis of gemmules. We have to assume that where a cauline bud emerges there are present in due proportions gemmules of all the parts which will presently arise from it—leaves, stipules, bracts, petals, stamens, anthers, etc. We have to assume this though, at the time the bud originates, sundry of these organs, as the parts of flowers, do not exist on the plant or tree. And we have to assume that the gemmules of such parts are duly provided in a portion of adventitious callus, far away from the normal places of fructification. Moreover, the resulting shoot may or may not produce all the parts which the gemmules represent; and when, perhaps after years, flowers are produced on its side shoots, there must exist at each point the needful proportion of the required gemmules; though there have been no cells continually giving them off.

Still less does the hypothesis of Prof. Weismann harmonize with the evidence as plants display it. Plant-embryogeny yields no sign of separation between germ-plasm and soma-plasm; and, indeed, the absence of such separation is admitted. After instancing cases among certain of the lower animals, in which no differentiation of the two arises in the first generation resulting from a fertilized ovum, Prof. Weismann continues:—

"The same is true as regards the higher plants, in which the first shoot arising from the seed never contains germ-cells, or even cells which subsequently become differentiated into germ cells. In all these last-mentioned cases the germ-cells are not present in the first person arising by embryogeny as special cells, but are only formed in much later cell-generations from the offspring of certain cells of which this first person was composed." (Germ-Plasm, p. 185.)

How this admission consists with the general theory it is difficult to understand. The units of the soma-plasm are here recognized as having the same generative powers as the units of the germ-plasm. In so far as one organic kingdom and a considerable part of the other are concerned the doctrine is relinquished. Relinquishment is, indeed, necessitated even by the ordinary facts, and still more by the facts just instanced. Defence of it involves the assertion that where buds arise, normal or cauline, there exist in due proportion the various ids with their contained determinants—that these are diffused throughout the growing part of the soma; and this implies that the somatic tissue does not differ in generative power from the germ-plasm.

The hypothesis of physiological units, then, remains outstanding. For cauline buds imply that throughout the plant-tissue, where not unduly differentiated, the local physiological units have a power of arranging themselves into the structure of the species.

But this hypothesis, too, as it now stands, is inadequate. Under the form thus far given to it, it fails to explain some accompanying facts. For if the branch just instanced as producing a cauline bud be cut off and its end stuck in the ground, or if it be bent down and a portion of it covered with earth, there will grow from it rootlets and presently roots. The same portion of tissue which otherwise would have produced a shoot with all its appendages, constituting an individual, now produces only a special part of an individual.

§ 97c. Certain kindred facts of animal development may now be considered. Similar insufficiencies are disclosed.

The often-cited reproduction of a crab's lost claw or a lizard's tail, Mr. Darwin thought explicable by his hypothesis of diffused gemmules, representing all organs or their component cells. But though, after simple amputation, regrowth of the proximate part of the tail is conceivable as hence resulting, it is not easy to understand how the remoter part, the components of which are now absent from the organism, can arise afresh from gemmules no longer originated in due proportion. Prof. Weismann's hypothesis, again, implies that there must exist at the place of separation, a ready-provided supply of determinants, previously latent, able to reproduce the missing tail in all its details—nay, even to do this several times over: a strong supposition! The hypothesis of physiological units, as set forth in preceding chapters, appears less incompetent: reproduction of the lost part would seem to be a normal result of the proclivity towards the form of the entire organism. But now what are we to say when, instead of being cut off transversely, the tail is divided longitudinally and each half becomes a complete tail? What are we to say when, if these two tails are similarly dealt with, the halves again complete themselves; and so until as many as sixteen tails have been formed? Here the hypothesis of physiological units appears to fail utterly; for the tendency it implies is to complete the specific form, by reproducing a single tail only.

Various annulose animals display anomalies of development difficult to explain on any hypothesis. We have creatures like the fresh-water Nais which, though it has advanced structures, including a vascular system, branchiæ, and a nervous cord ending with cephalic ganglia, nevertheless shows us an ability like that of the Hydra to reproduce the whole from a small part: nearly forty pieces into which a Nais was cut having severally grown into complete animals. Again we have, in the order Polychætæ, types like Myrianida, in which by longitudinal budding a string of individuals, sometimes numbering even thirty, severally develop certain segments into heads, while increasing their segments in number. In yet other types there occurs not longitudinal gemmation only, but lateral gemmation: a segment will send out sideways a bud which presently becomes a complete worm. Once more, Syllis ramosa is a species in which the individual worms growing from lateral buds, while remaining attached to the parent, themselves give origin to buds; and so produce a branched aggregate of worms. How shall we explain the reparative and reproductive powers thus exemplified? It seems undeniable that each portion has an ability to produce, according to circumstances, the whole creature or a missing part of the creature. When we read of Sir J. Dalyell that he "cut a Dasychone into three pieces; the hindermost produced a head, the anterior piece developed an anus, and the middle portion formed both a head and a tail" we are not furnished with an explanation by the hypothesis of gemmules or by the hypothesis of determinants; for we cannot arbitrarily assume that wherever a missing organ has to be produced there exists the needful supply of gemmules or of determinants representing that organ. The hypothesis that physiological units have everywhere a proclivity towards the organic form of the species, appears more congruous with the facts; but even this does not cover the cases in which a new worm grows from a lateral bud. The tendency to complete the individual structure might be expected rather to restrain this breaking of the lines of complete structure.

Still less explicable in any way thus far proposed are certain remedial actions seen in animals. An example of them was furnished in § 67, where "false joints" were described—joints formed at places where the ends of a broken bone, failing to unite, remain moveable one upon the other. According to the character of the habitual motions there results a rudely formed hinge-joint or a ball-and-socket joint, either having the various constituent parts—periosteum, fibrous tissue, capsule, ligaments. Now Mr. Darwin's hypothesis, contemplating only normal structures, fails to account for this formation of an abnormal structure. Neither can we ascribe this local development to determinants: there were no appropriate ones in the germ-plasm, since no such structure was provided for. Nor does the hypothesis of physiological units, as presented in preceding chapters, yield an interpretation. These could have no other tendency than to restore the normal form of the limb, and might be expected to oppose the genesis of these new parts.

Thus we have to seek, if not another hypothesis, then some such qualification of an existing hypothesis as will harmonize it with various exceptional phenomena.

§ 97d. In Part II of the Principles of Sociology, published in 1876, will be found elaborated in detail that analogy between individual organization and social organization which was briefly sketched out in an essay on "The Social Organism" published in 1860. In §§ 241-3 a parallel is drawn between the developments of the sustaining systems of the two; and it is pointed out how, in the one case as in the other, the components—here organic units and there citizens—have their activities and arrangements mainly settled by local conditions. One leading example is that the parts constituting the alimentary canal, while jointly fitted to the nature of the food, are severally adapted to the successive stages at which the food arrives in its progress; and that in an analogous way the industries carried on by peoples forming different parts of a society, are primarily determined by the natures of things around—agriculture, pastoral and arable, special manufactures and minings, ship-building and fishing: the respective groups falling into fit combinations and becoming partially modified to suit their work. The implication is that while the organization of a society as a whole depends on the characters of its units, in such way that by some types of men despotisms are always evolved while by other types there are evolved forms of government partially free—forms which repeat themselves in colonies—there is, on the other hand, in every case a local power of developing appropriate structures. And it might have been pointed out that similarly in types of creatures not showing much consolidation, as the Annelida, many of the component divisions, largely independent in their vitalities, are but little affected in their structures by the entire aggregate.

My purpose at that time being the elucidation of sociological truths, it did not concern me to carry further the biological half of this comparison. Otherwise there might have been named the case in which a supernumerary finger, beginning to bud out, completes itself as a local organ with bones, muscles, skin, nail, etc., in defiance of central control: even repeating itself when cut off. There might also have been instanced the above-named formation of a false joint with its appurtenances. For the implication in both cases is that a local group of units, determined by circumstances towards a certain structure, coerces its individual units into that structure.

Now let us contemplate the essential fact in the analogy. The men in an Australian mining-camp, as M. Pierre Leroy Beaulieu points out, fall into Anglo-Saxon usages different from those which would characterize a French mining-camp. Emigrants to a far West settlement in America quickly establish post-office, bank, hotel, newspaper, and other urban institutions. We are thus shown that along with certain traits leading to a general type of social organization, there go traits which independently produce fit local organizations. Individuals are led into occupations and official posts, often quite new to them, by the wants of those around—are now influenced and now coerced into social arrangements which, as shown perhaps by gambling saloons, by shootings at sight, and by lynchings, are scarcely at all affected by the central government. Now the physiological units in each species appear to have a similar combination of capacities. Besides their general proclivity towards the specific organization, they show us abilities to organize themselves locally; and these abilities are in some cases displayed in defiance of the general control, as in the supernumerary finger or the false joint. Apparently each physiological unit, while having in a manner the whole organism as the structure which, along with the rest, it tends to form, has also an aptitude to take part in forming any local structure, and to assume its place in that structure under the influence of adjacent physiological units.

A familiar fact supports this conclusion. Everyone has at hand, not figuratively but literally, an illustration. Let him compare the veins on the backs of his two hands, either with one another or with the veins on another person's hands, and he will see that the branchings and inosculations do not correspond: there is no fixed pattern. But on progressing inwards from the extremities, the distribution of the veins becomes settled—there is a pattern-arrangement common to all persons. These facts imply a predominating control by adjacent parts where control by the aggregate is less easy. A constant combination of forces which, towards the centre, produces a typical structure, fails to do this at the periphery where, during development, the play of forces is less settled. This peripheral vascular structure, not having become fixed because one arrangement is as good as another, is in each determined by the immediately surrounding influences.

§ 97e. And now let us contemplate the verifications which recent experiments have furnished—experiments made by Prof. G. Born of Breslau, confirming results earlier reached by Vulpian and adding more striking results of kindred nature. They leave no longer doubtful the large share taken by local organizing power as distinguished from central organizing power.

The independent vitality shown by separated portions of ventral skin from frog-larvæ may be named as the first illustration. With their attached yolk-cells these lived for days, and underwent such transformations as proved some structural proclivity, though of course the product was amorphous. Detached portions of tails of larvæ went on developing their component parts in much the same ways as they would have done if remaining attached. More striking still was the evidence furnished by experiments in grafting. These proved that the undifferentiated rudiment of an organ will, when cut off and joined to a non-homologous place in another individual, develop itself as it would have done if left in its original place. In brief, then, we may say that each part is in chief measure autogenous.

These strange facts presented by small aggregates of organic matter, which are the seats of extremely complex forces, will seem less incomprehensible if we observe what has taken place in a vast aggregate of inorganic matter which is the seat of very simple forces—the Solar System. Transcendently different as this is in all other respects, it is analogous in the respect that, as factors of local structures, local influences predominate over the influences of the aggregate. For while the members of the Solar System, considered as a whole, are subordinate to the totality of its forces, the arrangements in each part of it are produced almost wholly by the play of forces in that part. Though the Sun affects the motions of the Moon, and though during the evolution of the Earth-and-Moon system the Sun exercised an influence, yet the relations of our world and its satellite in respect of masses and motions were in the main locally determined. Still more clearly was it thus with Jupiter and his satellites or Saturn with his rings and satellites. Remembering that the ultimate units of matter of which the Solar System is composed are of the same kinds, and that they act on one another in conformity with the same laws, we see that, remote as the case is from the one we are considering in all other respects, it is similar in the respect that during organization the energies in each locality work effects which are almost independent of the effects worked by the general energies. In this vast aggregate, as in the minute aggregates now in question, the parts are practically autogenous.

Having thus seen that in a way we have not hitherto recognized the same general principles pervade inorganic and organic evolution, let us revert to the case of super-organic evolution from which a parallel was drawn above. As analogous to the germinal mass of units out of which a new organism is to evolve, let us take an assemblage of colonists not yet socially organized but placed in a fertile region—men derived from a society (or rather a succession of societies) of long-established type, who have in their adapted natures the proclivity towards that type. In passing from its wholly unorganized state to an organized state, what will be the first step? Clearly this assemblage, though it may have within the constitutions of its units the potentialities of a specific structure, will not develop forthwith the details of that structure. The inherited natures of its units will first show themselves by separating into large groups devoted to strongly-distinguished occupations. The great mass, dispersing over promising lands, will make preparations for farming. Another considerable portion, prompted by the general needs, will begin to form a cluster of habitations and a trading centre. Yet a third group, recognizing the demand for wood, alike for agricultural and building purposes, will betake themselves to the adjacent forests. But in no case will the primary assemblage, before these separations, settle the arrangements and actions of each group: it will leave each group to settle them for itself. So, too, after these divisions have arisen. The agricultural division will not as a whole prescribe the doings of its members. Spontaneous segregation will occur: some going to a pastoral region and some to a tract which promises good crops. Nor within each of these bodies will the organization be dictated by the whole. The pastoral group will separate itself into clusters who tend sheep on the hills and clusters who feed cattle on the plains. Meanwhile those who have gravitated towards urban occupations will some of them make bricks or quarry stone, while others fall into classes who build walls, classes who prepare fittings, classes who supply furniture. Then along with completion of the houses will go occupation of them by men who bake bread, who make clothing, who sell liquors, and so on. Thus each great group will go on organizing itself irrespective of the rest; the sub-groups within each will do the same; and so will the sub-sub-groups. Quite independently of the people on the hills and the plains and in the town, those in the forest will divide spontaneously into parties who cut down trees, parties who trim and saw them, parties who carry away the timbers; while every party forms for itself an organization of "butty" or "boss," and those who work under him. Similarly with the ultimate divisions—the separate families: the arrangements and apportionments of duties in each are internally determined. Mark the fact which here chiefly concerns us. This formation of a heterogeneous aggregate with its variously adapted parts, which while influenced by the whole are mainly self-formed, goes on among units of essentially the same natures, inherited from units who belonged to similar societies. And now, carrying this conception with us, we may dimly perceive how, in a developing embryo, there may take place the formation, first of the great divisions—the primary layers—then of the outlines of systems, then of component organs, and so on continually with the minor structures contained in major structures; and how each of these progressively smaller divisions develops its own organization, irrespective of the changes going on throughout the rest of the embryo. So that though all parts are composed of physiological units of the same nature, yet everywhere, in virtue of local conditions and the influence of its neighbours, each unit joins in forming the particular structure appropriate to the place. Thus conceiving the matter, we may in a vague way understand the strange facts of autogenous development disclosed by the above named experiments.

§ 97f. "But how immeasurably complex must be the physiological units which can behave thus!" will be remarked by the reader. "To be able to play all parts, alike as members of the whole and as members of this or that organ, they must have an unimaginable variety of potentialities in their natures. Each must, indeed, be almost a microcosm within a microcosm."

Doubtless this is true. Still we have a consensus of proofs that the component units of organisms have constitutions of extremely involved kinds. Contemplate the facts and their implications. (1) Here is some large division of the animal kingdom—say the Vertebrata. The component units of all its members have certain fundamental traits in common: all of them have proclivities towards formation of a vertebral column. Leaving behind the great assemblage of Fishes with its multitudinous types, each having special units of composition, we pass to the Amphibia, in the units of which there exist certain traits superposed upon the traits they have in common with those of Fishes. Through unknown links we ascend to incipient Mammalian types and then to developed Mammalian types, the units of which must have further superposed traits. Additional traits distinguish the units of each Mammalian order; and, again, those of every genus included in it; while others severally characterize the units of each species. Similarly with the varieties in each species, and the stirps in each variety. Now the ability of any component unit to carry within itself the traits of the sub-kingdom, class, order, genus, species, variety, and at the same time to bear the traits of immediate ancestors, can exist only in a something having multitudinous proximate elements arranged in innumerable ways. (2) Again, these units must be at once in some respects fixed and in other respects plastic. While their fundamental traits, expressing the structure of the type, must be unchangeable, their superficial traits must admit of modification without much difficulty; and the modified traits, expressing variations in the parents and immediate ancestors, though unstable, must be considered as capable of becoming stable in course of time. (3) Once more we have to think of these physiological units (or constitutional units as I would now re-name them) as having such natures that while a minute modification, representing some small change of local structure, is inoperative on the proclivities of the units throughout the rest of the system, it becomes operative in the units which fall into the locality where that change occurs.

But unimaginable as all this is, the facts may nevertheless in some way answer to it. As before remarked, progressing science reveals complexity within complexity—tissues made up of cells, cells containing nuclei and cytoplasm, cytoplasm formed of a protoplasmic matrix containing granules; and if now we conclude that the unit of protoplasm is itself an inconceivably elaborate structure, we do but recognize the complexity as going still deeper. Further, if we must assume that these component units are in every part of the body acting on one another by extremely complicated sets of forces (ethereal undulations emanating from each of the constituent molecules) determining their relative positions and actions, we are warranted by the discoveries which every day disclose more of the marvellous properties of matter. When to such examples as were given in § 36e we add the example yielded by recent experiments, showing that even a piece of bread, after subjection to pressure, exhibits diamagnetic properties unlike those it previously exhibited, we cannot doubt that these complex units composing living bodies are all of them seats of energies diffused around, enabling them to act and re-act so as to modify one another's states and positions. We are shown, too, that whatever be the natures of the complex forces emanating from each, it will, as a matter of course, happen that the power of each will be relatively great in its own neighbourhood and become gradually smaller in parts increasingly remote: making more comprehensible the autogenous character of each local structure.

Whatever be their supposed natures we are compelled to ascribe extreme complexity to these unknown somethings which have the power of organizing themselves into a structure of this or that species. If gemmules be alleged, then the ability of every organ and part of an organ to vary, implies that the gemmules it gives off are severally capable of receiving minute modifications of their ordinary structures: they must have many parts admitting of innumerable relations. Supposing determinants be assumed, then in addition to the complexity which each must have to express in itself the structure of the part evolved from it, it must have the further complexity implied by every superposed modification which causes a variation of that part. And, as we have just seen, the hypothesis of physiological units does not relieve us from the need for kindred suppositions.

One more assumption seems necessary if we are to imagine how changes of structure caused by changes of function can be transmitted. Reverting to § 54d, where an unceasing circulation of protoplasm throughout an organism was inferred, we must conceive that the complex forces of which each constitutional unit is the centre, and by which it acts on other units while it is acted on by them, tend continually to remould each unit into congruity with the structures around: superposing on it modifications answering to the modifications which have arisen in those structures. Whence is to be drawn the corollary that in course of time all the circulating units,—physiological, or constitutional if we prefer so to call them—visiting all parts of the organism, are severally made bearers of traits expressing local modifications; and that those units which are eventually gathered into sperm-cells and germ-cells also bear these superposed traits.

If against all this it be urged that such a combination of structures and forces and processes is inconceivably involved, then the reply is that so astonishing a transformation as that which an unfolding organism displays cannot possibly be effected by simple agencies.

§ 97g. But now let it be confessed that none of these hypotheses serves to render the phenomena really intelligible; and that probably no hypothesis which can be framed will do this. Many problems beyond those which embryology presents have to be solved; and no solution is furnished.

What are we to say of the familiar fact that certain small organs which, with the approach to maturity, become active, entail changes of structure in remote parts—that after the testes have undergone certain final developments, the hairs on the chin grow and the voice deepens? It has been contended that certain concomitant modifications in the fluids throughout the body may produce correlated sexual traits; and there is proof that in many of the lower animals the period of sexual activity is accompanied by a special bodily state—sometimes such that the flesh becomes unwholesome and even poisonous. But a change of this kind can hardly account for a structural change in the vocal organs in Man. No hypothesis of gemmules or determinants or physiological units enables us to understand how removal of the testes prevents those developments of the larynx and vocal cords which take place if they remain.

The inadequacy of our explanations we at once see in presence of a structure like a peacock's tail-feather. Mr. Darwin's hypothesis is that all parts of every organ are continually giving off gemmules, which are consequently everywhere present in their due proportions. But a completed feather is an inanimate product and, once formed, can add to the circulating fluids no gemmules representing all its parts. If we follow Prof. Weismann we are led into an astounding supposition. He admits that every variable part must have a special determinant, and that this results in the assumption of over two hundred thousand for the four wings of a butterfly. Let us ask what must happen in the case of a peacock's feather. On looking at the eye near its end, we see that the minute processes on the edge of each lateral thread must have been in some way exactly adjusted, in colour and position, so as to fall into line with the processes on adjacent threads: otherwise the symmetrical arrangement of coloured rings would be impossible. Each of these processes, then, being an independent variable, must have had its particular determinant. Now there are about 300 threads on the shaft of a large feather, and each of them bears on the average 1,600 processes, making for the whole feather 480,000 of these processes. For one feather alone there must have been 480,000 determinants, and for the whole tail many millions. And these, along with the determinants for the detailed parts of all the other feathers, and for the variable components of all organs forming the body at large, must have been contained in the microscopic head of a spermatozoon! Hardly a credible supposition. Nor is it easy to see how we are helped by the hypothesis of constitutional units. Take the feather in its budding state and ask how the group of such units, alike in structure and perpetually multiplying while the unfolding goes on, can be supposed by their mutual actions so to affect one another as eventually to produce the symmetrically-adjusted processes which constitute the terminal eye. Imagination, whatever licence may be given, utterly fails us.

At last then we are obliged to admit that the actual organizing process transcends conception. It is not enough to say that we cannot know it; we must say that we cannot even conceive it. And this is just the conclusion which might have been drawn before contemplating the facts. For if, as we saw in the chapter on "The Dynamic Element in Life," it is impossible for us to understand the nature of this element—if even the ordinary manifestations of it which a living body yields from moment to moment are at bottom incomprehensible; then, still more incomprehensible must be that astonishing manifestation of it which we have in the initiation and unfolding of a new organism.

Thus all we can do is to find some way of symbolizing the process so as to enable us most conveniently to generalize its phenomena; and the only reason for adopting the hypothesis of physiological units or constitutional units is that it best serves this purpose.