1 Mechanical evolution, the process by which the mass and energy of the universe have passed from some assumed primeval state to that distribution which they now present. Mr. Herbert Spencer the best accredited exponent of this doctrine.

He regards the universe as a single object, which is alternately evolved and dissolved. But the universe cannot be so regarded; and, if it could, Mr. Spencer's mechanical principles forbid such alternation. He ignores ‘dissipation of energy,’ and confuses energy with work. The thermodynamic zero. A finite universe must have time limits.

But is the universe finite? The Kantian antinomies and their solution. The notion of evolution not applicable to ‘the totality of things.’

2 The doctrine of the dissipation of energy and questions of reversibility. Limitations introduced by Lord Kelvin, Helmholtz and Maxwell. Two alternatives thus appear equally compatible with Mr. Spencer's ‘fundamental truth.’—(a) evolution without guidance, and (b) evolution with guidance. To account for the visible universe according to (a) requires a definite ‘primitive collocation.’ This Mr. Spencer rejects; for him then the cosmos can be but a chance hit among many misses, a mere speck of order in a general chaos. In expecting more from his mechanical principles he is guilty of the fallacy of confounding (a) with (b).

IN resuming our discussion after so long an interval it may be well briefly to restate what it is that we have set out to discuss. Naturalism we have taken to designate the doctrine that separates Nature from God, subordinates Spirit to Matter, and sets up unchangeable law as supreme. It means, to quote again the words of Huxley, “the extension of the province of what we call matter and causation and the concomitant… banishment from all regions of human thought of what we call spirit and spontaneity… [till] the realm of matter and law is coextensive with knowledge, with feeling, with action.”1 This naturalistic philosophy consists in the union of three fundamental theories: (1) the theory that nature is ultimately resolvable into a single vast mechanism; (2) the theory of evolution as the working of this mechanism; and (3) the theory of psychophysical parallelism or conscious automatism, according to which theory mental phenomena occasionally accompany but never determine the movements and interactions of the material world. With the first of these we have already dealt, and we now come to the second, in which it is applied.

Yet evolution, as commonly understood, is as far as possible from suggesting mechanism. By evolution or development was meant primarily the gradual unfolding of a living germ from its embryonic beginning to its final and mature form. This adult form, again, was not regarded as merely the end actually reached through the successive stages of growth, but as the end aimed at and attained through the presence of some archetypal idea, entelechy, or soul, shaping the plastic material and directing the process of growth. Evolution, in short, implied ideal ends controlling physical means; in a word, was teleological. In this sense mechanical evolution or development becomes a contradiction in terms. Nevertheless we shall find that the category of End, equally with the categories of Substance and Cause, is nowadays outside the pale of natural science. The term ‘evolution,’ though retained, is retained merely to denote the process by which the mass and energy of the universe have passed from some assumed primeval state to that distribution which they have at present. Also it is implied that the process will last till some ultimate distribution is reached, whereupon a counter process of dissolution will begin. Let us now turn to Mr. Herbert Spencer, the best accredited exponent of this doctrine, for details.

“An entire history of anything,” Mr. Spencer tells us, “must include its appearance out of the imperceptible and its disappearance into the imperceptible. Be it a single object or the whole universe, any account which begins with it in a concrete form, or leaves off with it in a concrete form, is incomplete.” “The sayings and doings of daily life,” he continues, “imply more or less such knowledge of states which have gone before and of states which will come after.… This general information which all men gain concerning the past and future careers of surrounding things, Science has extended, and continues increasingly to extend. To the biography of the individual man, it adds an intra-uterine biography beginning with him as a microscopic germ; and it follows out his ultimate changes until it finds his body resolved into the gaseous products of decomposition.” So as to the clothes he wears—“not stopping short at the sheep's back and the caterpillar's cocoon, it identifies in wool and silk the nitrogenous matters absorbed by the sheep and the caterpillar from plants.” So also as to “the wood from which furniture is made, [this] it again traces back to the vegetal assimilation of gases from the air and of certain minerals from the soil. And inquiring whence came the stratum of stone that was quarried to build the house, it finds that this was once a loose sediment deposited in an estuary or on the sea-bottom.” In these and such like instances Mr. Spencer sees the formula of evolution and dissolution foreshadowed. To quote again his own words: “In recognising the fact that Science, tracing back the genealogies of various objects, finds their components were once in diffused states, and pursuing their histories forwards, finds diffused states will be again assumed by them, we have recognised the fact that the formula must be one comprehending the two opposite processes of concentration and diffusion.… The change from a diffused, imperceptible state, to a concentrated, perceptible state, is an integration of matter and concomitant dissipation of motion; and the change from a concentrated, perceptible state, to a diffused, imperceptible state, is an absorption of motion and concomitant disintegration of matter.”2

Now, there is one obvious yet serious objection to this theory. It proposes to treat the universe, in fact requires us to treat the universe, as we treat a single object. Every single object is first evolved and then dissolved; it emerges from the imperceptible and into the imperceptible it disappears again. And so of the universe: “Any account which begins with it in a concrete form or leaves off with it in a concrete form,” Mr. Spencer tells us, “is incomplete.” Surely we have here a case of what logicians call the fallacy of composition; what is predicable of the parts severally is predicated of the whole collectively. It reminds us forcibly of Locke's “poor Indian philosopher, who imagined that the earth always wanted something to bear it up.” The stability of everything on the earth was manifestly due to a support, therefore the stability of the solid earth itself seemed explicable in no other manner. So the poor Indian; and similarly Mr. Herbert Spencer. As science deals with any visible, tangible thing, so the “synthetic philosophy” will deal with the totality of things. Let us take as a simple instance of the first, the familiar case suggested by Mr. Spencer himself, that of a cloud appearing when vapour drifts over a cold mountain top, and again disappearing when it moves away into the warmer air. The cloud emerges from the imperceptible as heat is dissipated and the vapour condensed, and the cloud is dissolved again as heat is absorbed and the watery particles evaporate. How shall we apply this conception or anything like it to the universe? The stronghold of Mr. Spencer's argument is the nebular hypothesis. A nebula, no doubt, is an object among other objects, though a most sublime and stupendous one. It presupposes colliding stars or meteoric swarms, whose material constituents are dissipated by the heat which their collision has produced; but then these colliding masses in their turn imply still earlier nebulæ, whose materials concentrated as their heat diffused. So the cloud presupposed vapours that had previously condensed; and the vapour, cloud or water that had previously evaporated. And much as clouds dissolve in one place and form in another, and are to be found at any time in all possible stages of evolution and dissolution; so with sidereal systems and nebulæ. The telescope and spectroscope tell of stars and nebulæ in every phase of advance or decline to be found in every quarter of the heavens. To ask which was first, solid masses or nebulous haze, is much like asking which was first the hen or the egg, and like that famous problem, may lead us to conclude,—neither the one nor the other. Meanwhile, it does not surprise us to learn that, though Mr. Spencer is quite sure that the universe began as imperceptible mist, others, like the late Dr. Croll, who have incomparably more right to an opinion on the question, prefer to think that there was an earlier or prænebular stage of the universe; during which large, cold masses of protyle or primal matter were moving through space in all directions with excessive velocities.3 Such an hypothesis, whether otherwise admissible or not, at least recognises a problem with which Mr. Spencer scarcely attempts to deal—I mean the evolution of the chemical elements. It thus suffices to convict Mr. Spencer's work of a certain incompleteness. For surely to begin with some seventy distinct forms of matter with very various and definite properties is not to begin at the beginning, however much we may imagine them to be diffused. We must return to this question of qualitative diversity presently. But the prior question I am anxious to put as pointedly as possible is this: On what grounds is it assumed that the universe was ever evolved at all? A given man, a given nation, a given continent, a given sidereal system, as particular objects, have their several finite histories of birth and death, upheaval and subsidence, fiery mist and cold, lifeless, consolidation. But growth and decay, rise and decline, elevation and degradation, evolution and dissolution, are everywhere contemporaneous. We have but to extend our range to find a permanent totality made up of transient individuals in every stage of change. But so enlarging our horizon we are not warranted in saying, as Mr. Spencer does, “there is an alternation of Evolution and Dissolution in the totality of things.”4 Of the totality of things we have no experience. But now what we do find, so far as experience and observation will carry us, is that, be it great or small—once an object has disappeared into the imperceptible, once it is dissolved in Mr. Spencer's sense, that object never reappears. We do not find dead men alive again, effete civilisations rejuvenated, denuded continents again restored, or worn-out stars rekindled as of yore. If there were any justification for the phrase “visible universe” and if we could conceivably represent the totality of things as a single concrete object,—both which suppositions I deny,—then by all analogy and experience ‘alternate eras of Evolution and Dissolution’ would be physically impossible. So surely as ‘the appearance out of the imperceptible’ was the beginning, so surely would ‘the disappearance into the imperceptible’ be the end. As, according to Mr. Spencer's own description, the entire history of anything, ‘be it a single object or the whole universe,’ lies completely within such limits, it is a manifest contradiction to turn round and say: After all the end is not the end and the beginning is not the beginning, and what we have called an entire and complete account of the totality of things is only one wave in an endless rhythm. It is true, of course, that the history of many concrete objects is marked by periodic phases; but never by dissolution and reëvolution, i.e. by the disappearance of the concrete individual followed by the reappearance of that individual—in short, by what is tantamount within the scope of such terms as visible, tangible, concrete, and perceptible—to as complete a breach of individuality as we should have in annihilation and recreation. It is also true, as we have already noted, that within a given totality, one individual may succeed another, but so far that totality—the universe of discourse, so to say—remains permanent. “One generation passeth away, and another generation cometh: but the earth abideth for ever.”5 6

Moreover, on the physical assumption from which Mr. Spencer sets out, viz. that the mass of the universe and the energy of the universe are fixed in quantity—which seems to mean are finite in quantity—there can be no such alternations as he supposes. Certainly not if we are to accept the second law of thermodynamics, the law, that is, of the dissipation of energy, along with the first law, that of the conservation of energy. But of this second law, commonly accepted though it is by physicists at the present day, Mr. Spencer seems to take no account. Apparently, too, Mr. Spencer confuses energy or the capacity of doing work with work actually done, and imagines that so long as the quantity of energy persists, it must be manifest in perpetual changes of equivalent amount. But this in any case is not a necessary consequence of the conservation of energy, and if the dissipation of energy be true, it is an impossible consequence. For it is not on the bare persistence of energy, but on the transference and transformation of energy that physical changes depend. But energy, whatever be its form, is only transferable from places of higher ‘intensity’ to places of lower intensity, to use a convenient term. So we find heavy bodies tend to fall, hot bodies to cool, and so forth. Thus the amount of energy available for work of the total energy possessed by two bodies is a function of this difference of level or intensity, and is nil when this difference is nil, whatever the total energy may be. Generally speaking, energy is not transferred without an equivalent transformation into work; but to this rule thermal energy is an exception. And it is here that the so-called waste or dissipation of available energy comes in. Putting it quite popularly, in the partnership of energies, heat is the one squanderer, and may scatter without producing. Whenever energy passes into this form, some of it is always, and all of it is sometimes, lost for purposes of work. As Mach puts it, “heat is only partially transformed into work, but frequently work is wholly transformed into heat. Hence a tendency exists towards a diminution of the mechanical energy and to wards an increase of the thermal energy of the world.”7 In other words, though the energy of the world remains constant, the unavailable energy or thermal level, so to say, tends towards a maximum. There is still a peculiarity of heat to be mentioned that will make the significance of the thermal degradation of energy clearer—I refer to Lord Kelvin's definition of an absolute zero of temperature. If—whatever were the temperature of a body—we could always imagine another body with a temperature still lower, just as whatever be the position of a body we can always imagine another at a distance from it towards which it can gravitate, then, so far as in this way differences of temperature would always be possible, the transformation of heat into work might always be possible. But if there be, as is supposed, a thermodynamic zero, there is an end to such a possibility; beyond that zero temperature cannot fall. And so while all transformations of energy lead directly or indirectly to transformation into heat, from that transformation there is no complete return, and therefore finally no return at all. This then is the conclusion to which Mr. Spencer's premisses lead. Two eminent physicists who accept those premisses may be cited at this point: “It is absolutely certain,” they say, “that life, so far as it is physical, depends essentially upon transformations of energy; it is also absolutely certain that age after age the possibility of such transformations is becoming less and less; and, so far as we yet know, the final state of the present universe must be an aggregation (into one mass) of all the matter it contains, i.e. the potential energy gone, and a practically useless state of kinetic energy, i.e. uniform temperature throughout that mass.… The present visible universe began in time and will in time come to an end.”8

To this conclusion we are surely led from such premisses.9 But again I ask what warrant is there for the premisses? Our experience certainly does not embrace the totality of things, is, in fact, ridiculously far from it. We have no evidence of definite space or time limits; quite the contrary. Every advance of knowledge only opens up new vistas into a remoter past and discloses further depths of immensity teeming with worlds. The physical principles of the conservation of mass and energy are, as I have already urged, essentially formal and regulative; they do but formulate the common postulate of all science—the uniformity and continuity of nature as presupposed in all physical measurements. They do not justify us in assuming, what we certainly cannot prove, that the universe as a whole is measurable and therefore finite. And when we pass to more purely a priori considerations, the case against a universe with fixed and finite limits is equally strong. It is needless to attempt even the most cursory discussion of the antinomies as to the finitude or infinitude of the universe in respect of time, space, divisibility, or mass, that have constituted the chief cosmological problems of philosophy, notably since the time of Kant. They have only justified in the main Kant's own solution. We cannot say that the phenomenal universe is infinite in any of these aspects, but just as little can we say it is finite. Since Kant's day, more cogent arguments both for the theses and for the antitheses of the cosmological problem have been advanced. None of these invalidate the claims of reason to regard the universe as a systematic whole, but they set in a stronger light than ever the impossibility of treating it as an arithmetical sum. “Say that the universe is limited,” says Kant, “and it is too small for your concept; you have a perfect right to ask what determines that limit: but say that it is unlimited, and it is too large for every possible empirical concept.” The reason of this is plain. In the empirical regress, to which the understanding, that is science, is entirely confined, “no experience of an absolute limit, that is, of any condition as such, which empirically is absolutely unconditional, can exist.” On the other hand, this regress from any given phenomenon as conditioned to another as its condition, though not truly infinite, is never suspended yet never completed; in other words, such regress must proceed in indefinitum.

But what Mr. Spencer calls a single object, must surely have an assignable beginning and end in time and assignable bounds in space; it is precisely through such time and space marks that the notion of singleness or identity becomes applicable. Those marks, however, are not given by empty time or space, but by other objects relatively defined in the same fashion. The universe, then, we may safely say, not only is not, but never can be, a single object in this wise; and Mr. Spencer's attempt to treat it after the fashion of an evolving nebula, evinces an unexpected paucity of imagination and is philosophically unsound. Experience provides us with instances of evolution and dissolution on the most varied scales, from the grass of the field or the cedars of Lebanon to the solar system or the Milky Way. But of a single supreme evolution embracing them all we have no title to speak: not even to assume that it is, much less to say what it is; least of all to affirm confidently that it can be embraced in such a meaningless formula as the integration of matter and the dissipation of motion—doubly meaningless unless a partial system, such as a nebula, is concerned, and even then assuming the greater portion of molecular physics without explanation. We have no evidence to shew that what we miscall the ‘visible universe’ is coming to an end, for we have no evidence to shew that it is finite. If we had such evidence, we should probably then and there conclude that we were dealing with but a part of the true universe and not with the totality of things. Again there is no physical evidence to compel the application to this absolute totality of such conceptions as increase and decrease, ebb and flow, development and decay; no warrant for attributing to the universe a destined perfection, that if future must either be attained and past; or approached but never completely attained at all. The former, if, as Mr. Spencer supposes, the mass and energy of the universe are finite and fundamental; the latter, if, being still the fundamental factors, the mass and energy are mathematically infinite in amount. Whether the world be absolutely perfect, or merely the best of all possible worlds, or indeed the worst world possible, as actual, it is—so far as we can judge from its physical constitution—just what it always has been, the permanent theatre of perpetual changes.

At any rate such a conception is less conjectural and more adequate than Mr. Spencer's ridiculous comparison of the universe to a spinning top that begins by ‘wabbling,’ passes into a state of steady motion or equilibrium mobile, and finally comes to rest.10 Referring to this second phase as one of perfect moving equilibrium, he finds in it “a warrant for the belief that evolution can end only in the establishment of the greatest perfection and the most complete happiness.”11 Let us not pause now to ask what sort of perfection and happiness that must be which depends on and necessarily follows from such physical equilibration: let us note only that, whatever it be, it is after all, according to Mr. Spencer, neither final nor established. It is but the “penultimate stage,” as indeed he calls it, and gives place, as he tells us, to “Dissolution, which inevitably, at some time or other, undoes what Evolution has done.”12 And again I say that the absurdity to which Mr. Spencer betakes himself does not suffice to put a better face on his doctrine—the absurdity, I mean, of supposing that, though there cannot be two universes in space, there may be any number in time. Beyond the final quietus of cosmical equilibration the doctrine of energy, in which Mr. Spencer puts his trust, affords no hope of a new evolution. The dead bones, the black ashes, may or may not live and glow again, but if they do it will not be from the mere ‘persistence of force’ that the quickening burst will come. Why, if the thing is so obvious, not to say necessary, is it never elucidated by ‘the familiar example’ of the spinning top? No doubt two consolidated sidereal systems may diffuse again after a collision, but how is one to do this? And what can well be less suggestive of recurring cycles than universal concentration of mass and uniformity of temperature on the one hand and indefinite diffusion of mass and diversity of temperature on the other? It must be allowed, in so far as Mr. Spencer is personally concerned, that the doctrine of the dissipation of energy was scarcely in the air when his First Principles were published. Meanwhile, for us at any rate, that doctrine seems to put an end to the alternate eras of evolution and dissolution which Mr. Spencer has vainly striven to derive from the doctrine of conservation. On the whole then we may for the present reasonably demur to Mr. Spencer's attempt to bring the universe under a simple formula of evolution and dissolution, as if it were a single object emerging out of the imperceptible and dissolving into it again. Before proceeding to discuss his formula in more detail so as to ascertain its adequacy where evolution in some sense is admissible, let me first ask attention for a little longer to consider one or two reflections suggested by our inquiry thus far or by points incidentally raised in the course of it.

Among the last in particular is this doctrine of the dissipation of energy, which excludes such reversibility as Mr. Spencer supposes. Lord Kelvin, who was, I believe, the first to formulate this doctrine, has been frequently commended for the caution which led him to restrict the impossibility to cases in which the agency of inanimate matter is alone concerned. Thus Helmholtz, referring to this reversion in a review of Lord Kelvin's papers, says: “Such a reversion is a postulate beyond the power of human means to fulfil. We have no agency at our disposal by which to regulate the movement of atoms. Whether, however, in the extraordinarily fine structure of organic tissues a mechanism capable of doing it exists or not is a question not yet to be answered, and I deem it very wise on the part of Sir W. Thomson that he has limited all his theses respecting the necessity of increasing dissipation by restricting their validity to ‘inanimate matter.’”13 I Dissipation of energy Lord Kelvin himself tells us, “follows in nature from the fortuitous concourse of atoms. The lost motivity is essentially not restorable otherwise than by an agency dealing with individual atoms; and the mode of dealing with the atoms to restore motivity is essentially a process of assortment, sending this way all of one kind or class, that way all of another kind or class.”14 Many here will remember a fine passage in Mill's Political Economy on the function of labour, in which he shews with impressive detail that in what is called the action of man upon nature it is “the properties of matter that do all the work, when once objects are put into the right position. This one operation of putting things into fit places for being acted upon by their own internal forces, and by those residing in other natural objects, is all that man does, or can do, with matter. He only moves one thing to or from another:” all his vast command over natural forces immeasurably more powerful than himself “is obtained by arranging objects in those mixtures and combinations by which natural forces are generated, as when by putting a lighted match to fuel, and water into a boiler over it, he generates the expansive force of steam, which has been made so largely available for the attainment of human purposes.”15 Here then we have the materials and powers of nature, as they fortuitously occur, incapable of, and unavailable for, results, to which nevertheless they can be guided by intelligent assortment and arrangement. And in a precisely analogous way we can imagine finite intelligences disequalising temperature and undoing the natural diffusion of heat, or assorting atoms and undoing the natural conglomeration of matter, and so reversing the downward trend, and even disturbing the final quiescence, to which the dissipation of energy or ‘cosmic equilibration,’ to use Mr. Spencer's term, inevitably leads. The conception of such an intelligence we have in “the sorting demon of Maxwell,” as Lord Kelvin has called it.

This brilliant idea was devised by Maxwell primarily to illustrate “the limitation of the second law of thermodynamics,” to shew, that is, that this second law, the law of the degradation of energy is not like the first—the law of conservation—a fundamental, dynamical law; that, on the contrary, it is properly a statistical law and confined to our experience of secondary bodies consisting of an immense number of molecules, none of which are individually perceptible. And so he remarks: “This law is undoubtedly true as long as we can deal with bodies only in mass, and have no power of perceiving or handling the separate molecules of which they are made up. But if we conceive a being,”—and here we are introduced to the ‘sorting demon’—“whose faculties are so sharpened that he can follow every molecule in its course, such a being, whose attributes are still as essentially finite as our own, would be able to do what is at present impossible to us.” To most of you I am sure the modus operandi of this possible but imaginary being is perfectly well known; still, to add to the clearness of our discussion, I will venture to quote the rest of Maxwell's paragraph. “For we have seen,” he continues, “that the molecules in a vessel full of air at uniform temperature are moving with velocities by no means uniform, though the mean velocity of any great number of them, arbitrarily selected, is almost exactly uniform. Now let us suppose that such a vessel is divided into two portions, A and B, by a division in which there is a small hole, and that a being, who can see the individual molecules, opens and closes this hole, so as to allow only the swifter molecules to pass from A to B, and only the slower ones to pass from B to A. He will thus, without expenditure of work, raise the temperature of B and lower that of A, in contradiction to the second law of thermodynamics.”16

Now, what I think we may fairly deduce from this piece of physical exposition is that conservation of energy at any rate,—and this is Mr. Spencer's one dynamical principle,—is compatible with either of two alternatives.17 The first is that steady fall in the level of available energy, which finds expression in the second law of thermodynamics, technically given in the statements of Lord Kelvin and Clausius already referred to,18 viz., that, though the energy of the universe remains constant, the entropy of the universe tends towards a maximum.19 The second alternative is the process of assortment and guidance—without expenditure of work—by a selecting and directing intelligence, which process may, to an indefinite extent, reverse and overrule the dissipation of energy, that tendency merely to run down. For, granting the energy of a material system, however large, to remain constant, and granting change of direction without work to be always theoretically possible, we may infer that, until—after a lapse of time indefinitely great—a state of absolute dissipation is reached, it would be possible for intelligent beings, without infringing any dynamical principles, to inaugurate changes, and for an adequate intelligence to start that system anew on a fresh round of evolution. This is forcibly put in the paper of Helmholtz's, from which I have already quoted: “The ascertained laws of dynamics,” he says, “yield the deduction that, if we were able suddenly to reverse the total movements of the total atoms of an isolated mechanical system, the whole system would of necessity retraverse all the states which up to that point of time it had passed through. Therewith, also, would all the heat, generated by friction, collision, conduction of electrical currents, etc., return into other forms of energy, and the energy, which had been dissipated, would be all recovered”20 And I presume that an intelligence that could precisely reverse the directions could alter them as easily in other ways. But the point is that, apart from intelligent guidance and arrangement, no such recovery or alteration would be possible.

It will be quite worth while to compare these alternatives somewhat further. Though both are equally compatible with the persistence of energy, yet Mr. Spencer, as we have seen, admits only one, and ignores the fact that that one entirely precludes such alternations of evolution and dissolution as he assumes. According to that, which we may fairly call the mechanical view, evolution, or rather, as Mr. Spencer ought to say, a given era of evolution, begins at an initial extreme, characterized by him as an imperceptible state of absorption of motion and concomitant disintegration of matter; and ends with a final extreme, equally imperceptible, of integration of matter and concomitant dissipation of motion. In conciser and more intelligible language, the whole process ranges from an extreme with very large potential energy to an extreme in which all available energy is dissipated. The other alternative, which we may perhaps call the teleological view, neither requires an initial stage, such as Mr. Spencer's, in order that evolution may begin, nor is debarred by the dissipation of energy from all possibility of further change. Without postulating the creation of energy it recognises the direction of energy by intelligence. Under what circumstances and by what means such intelligent guidance is effected we need not now inquire; it is allowed to be possible, and for the present that is enough.

And now let us attend to the important difference between these two views,—evolution without guidance, and evolution with guidance. According to the former, the entire course of things is once and for all determined singly and solely by the initial distribution. It is here that the Laplacean calculator comes in, prepared from the mechanical data of any one moment to find the state of the whole world at any other. For there is one, and only one, course that a system of inert matter will pursue without guidance,—the line of least resistance: it will run down, and it will run down by the easiest and shortest way. But the directions that such a system may be led to take under guidance, but still conformably to the law of conservation, may be innumerable. To forecast the actual progress on this view it is useless to know merely what would happen in accordance with mechanical laws, if the system were left to itself: for any forecast in this case a knowledge of the end or meaning of such progress would be indispensable. Let us take one or two familiar instances by way of illustrating the difference. Imagine a derelict ship and a sea-worthy vessel fully manned: if you know enough of the winds, tides, and currents, you can say where the derelict is likely to be after a week's interval, but this information will be but of secondary importance if you should attempt to predict the position a week later of the ship under sailing orders. Take two trains running opposite ways on a single line of rails,—of which there are hundreds in this country every day: if you know their distance apart, their rates, and that they are left to themselves, you can calculate when and where they will collide. Yet the extreme rarity of collisions is secured simply by what is practically “guidance without work,” by ‘pointsmen’ directing energy which in itself is directionless.

But however impressive the difference between these two forms of process, the blindly mechanical and the intelligently guided, and however surely common sense in our ordinary affairs enables us to distinguish between the two, yet in so far as both are compatible with mechanical principles it is obvious that strictly mechanical considerations will not enable us to distinguish between them. A bullet aimed to hit the mark conforms to the law of projectiles as completely as one fired at random. But now, of a thousand bullets so fired haphazard, probably one or more will hit equally truly. This simple instance may serve to characterise two ways in which the teleological aspect of things can be viewed mechanically. The first is by way of primitive collocations. As the marksman's aim determined the initial movement of the bullet with a view to its final impact on the bull's-eye, so the Creator chose that particular configuration of nebulous matter from which the existing cosmos would mechanically ensue. So Whewell, Chalmers, Jevons, and others represent the beginning of evolution. “Out of infinitely infinite choices which were open to the Creator, that one choice must have been made which has yielded the Universe as it now exists,” says Jevons. We may venture, I think, to call this a short-sighted and fatalistic view; but I am quite aware that those who first propounded it had many qualifications in reserve, qualifications, however, which must logically resolve the external Artificer into an immanent Spirit. But at all events this half-way house, whatever it be worth, is closed against Mr. Spencer, if even he were disposed to occupy it. For him there can be no ‘ultimate properties of kinds,’ and no specific collocation of diverse natural agents. Thoroughgoing homogeneity, diffusion, and imperceptibility, are, as we shall see presently, incompatible with such variety in the positions and mechanical endowment of the primitive particles as Jevons, for example, supposes. To Mr. Spencer there is open only the second way of one chance hit out of many misses. We have all of us to admit that facts are by no means wanting that may seem to justify such a view of Nature at least in details, as when finding, for example, “that of fifty seeds, she often brings but one to bear.” For the mechanical theory of evolution, however, this second way is absolute and universal. But it will be best here to cite Mr. Spencer's own words: “We have to contemplate the matter of an evolving aggregate as undergoing not progressive integration simply, but as simultaneously undergoing various secondary redistributions; we have also to contemplate the motion of an evolving aggregate, not only as being gradually dissipated, but as passing through many secondary redistributions on the way towards dissipation.”21 Such is Mr. Spencer's general summary; but it would be useless, I fear, to attempt to quote also any of the numerous instances even of physical phenomena, to say nothing of phenomena of a higher order, which he has gathered together in such impressive and bewildering variety in order to substantiate it. I can put the case best, as I understand it, by taking an illustrative instance of my own. Imagine a single drop of water falling alone over Niagara: it will go with accelerated velocity straight from top to bottom. Such a process may typify simple evolution. Now try to realise what happens when the full volume of the river pitches at once over the Falls. The end is in the main the same as before, but in the course of simple evolution on this larger scale there occur many, and some very striking, instances of compound evolution, in other words, of redistributions, arrests, and reversals of the main process. Individual drops and groups of drops may dash each other into mist, fall, rise, and fall again, eventually joining the stream below only after a long time and by the most devious routes. Imagine the height of the Falls and so the time of falling to be vastly increased, and the secondary results will be more varied still. To this head of compound evolution, then, we are asked to refer all the complexity of structure and movement, all the varieties of form and rhythm, of which the actual world consists. “Hence,” says Mr. Spencer, “other things being equal, in proportion to the quantity of motion which an aggregate contains will be the quantity of secondary change in the arrangement of its parts that accompanies the primary change in their arrangement. Hence, also, other things equal, in proportion to the time during which the internal motion is retained, will be the quantity of this secondary redistribution that accompanies the primary distribution.”22 A little reflection will shew, I think, that on this doctrine what others secure by primitive collocations is secured by taking things on a sufficiently large scale, and trusting to the combinations which haphazard will give. Shuffle an adequate number of fonts of type long enough and a given play of Shakespeare will be among the throws; for it is a possible combination, and in time all possible combinations may be expected. In fact, Mr. Spencer's law of evolution seems to consist essentially in treating the universe as a vast problem in thermodynamics, so to speak.

Apropos of this I cannot do better than quote a striking passage from a letter of Boltzmann's that appeared in Nature about a year ago: “We assume that the whole universe is, and rests forever, in thermal equilibrium. The probability that one (only one) part of the universe is in a certain state, is the smaller the farther this state is from thermal equilibrium; but [on the other hand] this probability is greater, the greater is the universe itself. If we assume the universe great enough, we can make the probability of one relatively small part being in any given state (however far from the state of thermal equilibrium) as great as we please. We can also make the probability great that, though the whole universe is in thermal equilibrium, our world is in its present state. It may be said that the world is so far from thermal equilibrium that we cannot imagine the improbability of such a state. But can we imagine, on the other side, how small a part of the whole universe this world is? Assuming the universe great enough, the probability that such a small part of it as our world should be in its present state, is no longer small. If this assumption were correct, our world would return more and more to thermal equilibrium; but, because the whole universe is so great, it might be probable that at some future time some other world might deviate as far from thermal equilibrium as our world does at present.”23

By ‘world’ I take Boltzmann to mean what is commonly called the ‘visible universe’ or ‘our galaxy.’ The return to thermal equilibrium again corresponds to Mr. Spencer's simple evolution, assuming a like fortuitous initial distribution or absence of specific collocations, and a universe indefinitely great. Of course there is no lack of space and time; even energy too is cheap, when you have only to imagine it. But such a chance oasis of life and order in an illimitable desert of monotonous irregularity is, I need hardly say, not what Mr. Spencer means by evolution. So much the worse, however, for his synthetic philosophy. For while that is the most that his law entitles him to, he assumes not merely that the present throw—to recur to my illustration—is comparable to a play of Shakespeare, but he assumes also that, after the processes of dissolution shall have broken up the type, another play will be thrown next time. In other words, he is guilty of the amazing fallacy of supposing that, because the laws of energy are everywhere present, they are everywhere sufficient to explain what we see; which is much the same as assuming that, because a painter's palette, like his finished canvas, shews us a mixture of colours laid on with a brush, therefore what sufficed to produce the one would equally suffice to produce the other.

But the further exposure of this prime fallacy of Mr. Spencer's synthetic philosophy must be reserved till next lecture.

• 1.

  Cf. Lecture 1.

• 2.

  First Principles, §§ 93, 94, stereo. ed., pp. 279-281; rev. ed., pp. 253-256, materially altered.

• 3.

  Cf. Croll, Stellar Evolution, pp. 3, 109.

• 4.

  First Principles, § 190, stereo. ed., p. 551; rev. ed., p. 506, altered.

• 5.

  Eccl. i. 4.

• 6.

  Note i.—In an article on this book (Fortnightly Review, Dec. 1899) Mr. Spencer states his essential purpose to be that of ‘exemplifying my controversial method,’ and concludes by warning his readers that before accepting my version of his views “it will be prudent to verify them.” But, strange to say, in a revised edition of his First Principles, published in 1900, a large number of the passages on which I have animadverted— passages that had remained unchanged for thirty years—are now silently either suppressed or altered. Only in a brief appendix of some five pages is there any direct reference to this work. There Mr.Spencer begins by saying: “It is half instructive, half amusing to observe what trivial difficulties, and even what imaginary difficulties, are urged by those who seek reasons for rejecting doctrines they dislike.” He then dismisses my criticisms with the remark: “Were I to notice all of them at length, half a volume would be required.… So far as I have observed, he has throughout followed the course which generally characterises controversy—that of setting up men of straw and knocking them down.” His readers are thus left to infer that in general Mr.Spencer has found it unnecessary to pay any attention to my objections, and the numerous alterations or suppressions of passages, to which I have alluded, will therefore strike them as interesting coincidences. I have indicated some of these in the footnotes given in the text—stereo. ed. referring to the stereotyped editions, and rev. ed. to the revised edition.

  In the stereotyped editions Mr.Spencer treated the universe as a single object which is alternately evolved and dissolved, and my first criticism was that the universe cannot be so regarded. Instead of the words “Be it a single object or the whole universe any account which begins with it in a concrete form…is incomplete” (see above), we now find merely “Any account of an object which begins,” etc.—no reference to the universe at all; and in like passages elsewhere all reference to the universe is suppressed. Again, in the earlier editions we find Mr.Spencer saying: “It is obvious that we have not acquired all the information within the grasp of our intelligence until we can, in some way or other, express the whole past and the whole future of each object and the aggregate of objects”; and then concluding: “May it not be inferred that Philosophy has to formulate this passage from the imperceptible into the perceptible, and again from the perceptible into the imperceptible?” (stereo. ed., p. 280). He declares “that a Philosophy stands self-convicted of inadequacy” if it fails of such formulation: for “if it begins its explanations with existences that already have concrete forms, or leaves off while they still retain concrete forms; then, manifestly, they had preceding histories, or will have succeeding histories, or both, of which no account is given. And as such preceding and succeeding histories are subjects of possible knowledge, a Philosophy which says nothing about them falls short of the required unification” (stereo. ed., p. 541 fin.). In the revised edition all these passages are omitted, and Mr.Spencer, with commendable candour, confesses that they imply an unattainable ideal. “Complete accounts of the beginnings and ends [even] of individual objects,” he now allows, “cannot in most cases be reached.… Still more, then, with the totality of things must we conclude that the initial and terminal stages are beyond the reach of our intelligence” (rev. ed., p. 256).

  But now Philosophy, according to Mr.Spencer's definition, is completely-unified knowledge; knowledge partially unified is only Science (§ 37 fin.); his theory of evolution, then, on his own showing can be no more. Further admissions, pointing in the same direction, will appear presently (see below).

  My second criticism was that even regarding the universe as a single object, we are not warranted in saying that “there is an alternation of Evolution and Dissolution in the totality of things.” Prior to the publication of his revised edition, in the article above mentioned Mr.Spencer complained that in so objecting I had treated a tentative opinion as a positive assertion. “He does not,” says Mr.Spencer, “quote the whole clause, which runs thus:—‘For if, as we saw reason to think, there is an alternation of evolution and dissolution in the totality of things, etc.’ Here there are two qualifying expressions which he suppresses” (Fortnightly, p. 902). But the odd thing is (as I pointed out in “A Reply to Mr.Herbert Spencer,” Fortnightly, March 1900, p. 469) that even Mr.Spencer himself does not quote his own words without suppression. Here is the passage in full;— “For if, as we saw reason to think, there is an alternation of Evolution and Dissolution in the totality of things—if, as we are obliged to infer from the Persistence of Force, the arrival of either limit of this vast rhythm brings about the conditions under which a counter-movement commences—if we are hence compelled to entertain the conception of Evolutions that have filled an immeasurable past, and Evolutions that will fill an immeasurable future; we can no longer contemplate the visible creation as having a definite beginning or end” (stereo. ed., p. 551—italics mine). As one out of many possible passages in which Mr.Spencer seemed to have committed himself to a positive assertion, I also quoted this one: “Thus we are led to the conclusion that the entire process of things, as displayed in the aggregate of the visible universe, is analogous to the entire process of things as displayed in the smallest aggregates…now an immeasurable period during which the attractive forces predominating, cause universal concentration, and then an immeasurable period during which the repulsive forces predominating, cause universal diffusion—alternate eras of Evolution and Dissolution (stereo. ed., pp. 536 f.). Of course Mr.Spencer knows best what he meant to say: his readers must judge how far he succeeded in saying it. At any rate in the revised edition he is clearer, for not only are these and other seemingly positive assertions withdrawn, but it is expressly admitted that “the question whether there is an alternation of evolution and dissolution in the totality of things is one which must be left unanswered as beyond the reach of human intelligence,” and even “as passing the bounds of rational speculation (rev. ed., pp. 506, 492). Once again, then, Mr.Spencer's theory of evolution drops from the level of philosophical synthesis based on “the ultimate datum of consciousness” to the level of science, “unable to trace the entire history even of a small aggregate!” (rev. ed., p. 493).

  But, in truth, if the appeal is not to that hopelessly vague conception, Mr.Spencer's Persistence of Force as an ultimate datum of consciousness, but to the conservation of energy as commonly understood—and this is what Mr.Spencer usually has in mind—then the question whether there are alternations of evolution and dissolution in the totality of things is not ‘transcendental’ at all. It is neither to be positively asserted nor to be left in doubt. The energy of the universe is either finite or infinite. In both cases there may be alternations of evolution within the universe, but in the one they will come to an end, in the other they will not: in neither will there be such alternations of the universe as a whole. See next note.

• 7.

  Popular Scientific Lectures, Eng. trans., p. 175.

• 8.

  Unseen Universe, § 115.

• 9.

  Note ii.—In his article in the Fortnightly Review, mentioned in the previous note (p. 901 fin.), Mr.Spencer contends that he had himself anticipated this criticism before I was out of my teens, and then proceeds to quote a paragraph of his First Principles (stereo. ed., pp. 535, 536), in proof. “Unhappily,” as I have already said in reply (Fortnightly, March 1900, p. 470), “the facts are quite otherwise. Not only are Mr.Spencer's reasons not the same as mine, but they are not reasons against the doctrine of the dissipation of energy at all; though they refer to something that sounds rather like it, viz. to what Mr.Spencer is fond of calling ‘the dissipation of motion:’ That dissipated or degraded energy means not energy that is ‘diffused’ or ‘radiated’ but energy that is no longer available for work, is a point that Mr.Spencer has entirely overlooked. In the revised edition (p. 492) he has amended this paragraph: there is now some mention of energy and of heat, but the result only shows still more conclusively Mr.Spencer's ignorance of thermodynamics. In fact his second version is, if anything, more inaccurate than his first, for he seems to think that the dissipation of energy may be counteracted by maintaining the thermal equilibrium of space.

  Of course it is conceivable that the energy dissipated at any time is always a constant fraction of the energy remaining available, so that the process would never end. If we then suppose farther, as Professor Poynting has suggested, that “living beings became capable of using more and more minute differences, life might persist as well.” This very theoretical possibility the authors of the Unseen Universe did not take into account.

• 10.

  Note iii.—Mr.Spencer replies that he has nowhere asserted moving equilibrium of the universe, but that on the contrary he has expressly negatived a moving equilibrium of our sidereal system, thereby implying that he would still more definitely negative such an equilibrium of the universe (Fortnightly, p. 904).

  It is true that the spinning-top is only mentioned to exemplify the nature of mobile stability; but not only is the principle itself an integral part of the Laplacean hypothesis upon which Mr. Spencer's theory of evolution really rests, but his own statements of the principle in the chapter on Equilibration as manifest deductions from the Persistence of Force are made without any reservation whatever. In the following chapter dealing with Dissolution, in order to show “that the structure of our galaxy is undergoing change and must continue to undergo changes;” he refers to its irregular distribution as “being such as to render even a temporary moving equilibrium impossible.” But this, even if true, does not affect the existence within our sidereal system of stellar systems, and some of these systems far more complex than our solar system, which are stable in Laplace's sense: indeed the little we know all points this way. To meet Mr.Spencer's criticism it would be enough to say that on his theory the universe consists of an indefinite number of spinning-tops, and that as time goes on the tops collide, tops ever larger in size and fewer in number being the result.

  His admirers will be depressed to find that in the revised edition Mr.Spencer has withdrawn the “warrant for the belief that evolution can end only in the establishment of the greatest perfection and the most complete happiness,” which he had previously deduced from his equilibrium mobile.

• 11.

  Unseen Universe, stereo. ed., p. 517; rev. ed., omitted.

• 12.

  Unseen Universe, § 190, stereo. ed., p. 550; rev. ed., p. 505, altered.

• 13.

  Wissenschaftliche Abhandlungen, Bd. iii, p. 594.

• 14.

  Properties of Matter, p. 139.

• 15.

  Principles of Political Economy, Bk. i, chap. i, § 2. Mill attributes this observation to his father, but even he was anticipated by Bacon (Novum Organum, vol. i, p. 4), and again by Playfair.

• 16.

  Theory of Heat, 1894, pp. 338 f.

• 17.

  But our difficulties, no doubt, increase when we take into account other dynamical principles which Mr. Spencer neglects. Cf. below, Lecture 12.

• 18.

  Cf. above.

• 19.

  Strictly speaking, we are not warranted in applying metrical motions to the universe. Cf. pp. 87, 171.

• 20.

  Wissenschaftliche Abhandlungen, Bd. iii, p. 594.

• 21.

  First Principles, § 145, stereo. ed., p. 396; rev. ed., p. 367, omitted.

• 22.

  First Principles, § 99, stereo. ed., p. 289; rev. ed., p. 264.

• 23.

  Nature, 1894-1895, vol. li, p. 415.