CORNELL UNIVERSITY LIBRARY BOUGHT WITH THE INCOME OF THE SAGE ENDOWMENT FUND GIVEN IN 1891 BY HENRY WILLIAMS SAGE Q 1 75 H5 1°'"^" ""'^'^'^^'•y '-""■a''y 3 1924 012 262 030 Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/cletails/cu31924012262030 THE ORDER OF NATURE AN ESSAY BY LAWRENCE J. HENDERSON ASSISTANT PKOFES30R OF BIOLOGICAL CHEMISTRY IN HARVARD UNIVERSITY CAMBRroCE HARVARD UNIVERSITY PRESS LONDON: HXnWPHKET MILFORD OXFOBD UN1VEH8ITT raE88 1917 COPTBIGHT, 1917 HAHVAHD TJNIVEHSITT PRESS PREFACE " La nature est un ordre . . . dont Vensemhle constitue une puissance inalterable dans son essence, assujettie dans tous ses actes, et constamment agissant sur toutes les parties de Vunivers." . . . "un ordre . . . capable de donner successivement Vexistence a tant d'etres divers " . . . " cette puissance qui fait tant de choses, et qui cependant est constamment bomie a nefaire que celles-ld." Lamarck, Histoire naturelle des animaux sans vertebres. The study of adaptation, of which Lamarck is the great originator, has not yet won for itself a secure scientific foundation or led to clear and unequivocal interpretations of nature. Although the facts which this study presents are both univer- sal and important, biologists have neither agreed upon their place in the theory of evolution nor discovered any principle by which they may be even unified. This failure of our modem science is not hard to understand, and may fairly be attributed, in part at least, to the lack of a systematic study of adapt- ability; which at bottom is a physical and chemi- cal problem, uncomplicated by the riddle of life. THE ORDER OF NATURE INTRODUCTION Many of the characteristics of inorganic nature, like the stability of the solar system and the en- during movements of the waters of the earth, are the very condition of existence for life as we know it and the source of diversity in organic evolution. This is perhaps one of the oldest interpretations of nature. But since Darwin's time the fitness of the environment has only occasionally aroused passing comment without ever entering the main current of scientific thought. And yet, whatever may be the final judgment of natural science upon either organic or inorganic harmonies, biological fitness is manifestly a mutual relationship. For, however present order may have developed out of past con- fusion, the organism and the environment each fits and is fitted by the other. In a recent book ^ I have tried to recall attention to the many interesting peculiarities of the en- vironment and to state the facts concerning the fitness of the inorganic world for life. This has turned out to be more notable and extensive than biologists had supposed, and more important in 1 The Fitness of the Environment. New York, The Macmillan Co., 1913. 4 NATURE determining the universal characteristics of living organisms. The very nature of the cosmic process and of the physical and chemical phenomena of matter and energy bring about not only stability of the solar system, but very great stability of land and sea. Thus the temperature of the earth is more equable than it could be if the composition of the surface of the earth were other than it is. Thus the alkalinity of the ocean possesses a constancy which is nearly perfect, and this depends upon certain unique properties of carbonic acid. Thus the currents of the atmosphere and of the ocean, the fall of rain and the flow of streams are almost ideally regular, and are so only because water is different from any other substance. Secondly, the properties of water cause a mobili- zation all over the earth of most of the chemical elements in very large quantities, and no other substance could so effectively accomplish this result. Once mobilized, these elements penetrate everywhere, borne by water, and the penetrating qualities of water are imique. In this manner the whole earth has become habitable. Even more significant appear what the chemist calls the properties of the three elements, hydrogen, oxygen, and carbon, from which water and car- bonic acid are formed. These are the most active of all elements (if we take account of both intensity and variety of activity), their compounds are the INTRODUCTION 5 most numerous, the molecular structures which they form are incomparably the most complex and elaborate which have been brought to light. More- over the energy which they yield in their mutual chemical transformations is more than other ele- ments can provide, yet, because of their manifold reactions, more easy to regulate, to store, and to release. In short the primary constituents of the environ- ment, water and carbonic acid, the very substances which are placed upon a planet's surface by the' blind forces of cosmic evolution, serve with maxi- mum efficiency to make stable, durable, and com- plex, both the living thing itself and the world around it. With otherwise unattainable effective- ness they provide both matter and energy in many forms and in great abimdance for growth and for repair, and in the ensemble of characteristics upon which these results depend they are unique. Noth- ing else could replace them in such respects, for their utiHty depends upon a coincidence of many peculiar and unequaled properties which they alone possess. It is therefore certain that in abstract physical and chemical characteristics the actual environment is the fittest possible abode of life as we know it, so far as the elements of the periodic system are concerned. In truth fitness of the en- vironment is quite as constant a component of a particular case of biological fitness as is fitness of 6 NATURE the organism, and fitness is quite as constantly manifest in all the properties of water and carbonic acid as in all the characteristics of living things. Such a conclusion, however, only touches the surface of the problem. For this relationship, although mutual, is not symmetrical: it is some- thing more than adaptation for it involves great adaptability. In every case the particular char- acteristics of the organism fit a special environ- ment, while the general physical and chemical properties of water and carbonic acid fit the general characteristics of life. But it may be shown that stability, mobility, durability, complexity, and availability of matter and energy are favorable not merely to life as we know it; they are favorable to any mechanism, to any possible kind of life in this universe. For it is not by chance that life needs to be stable, that it needs food, that it needs to be complex if it is to evolve. Accordingly it is not for any special or peculiar form of life, whether life as we know it or another form, that this environment is the fittest. Just because life must exist in the universe, just because the living thing must be made of matter in space and actuated by energy in time, it is con- ditioned. In so far as this is a physical and chemical world, life must manifest itself through more or less complicated, more or less durable physico-chemical systems. INTRODUCTION 7 Accordingly it is possible to assert and it will pres- ently be demonstrated that the primary constitu- ents of the environment are the fittest for those general characteristics of the organism which are* imposed upon the organism by the general char- acteristics of the world itself; by the very nature of matter and energy, space and time. I feel sure that this conclusion is but a precise statement of a view that has long been vaguely held by many chemists. The facts upon which this conclusion rests prove, I believe, that a hitherto unrecognized order exists among the properties of matter. For the peculiari- ties that make things what they are have been found not evenly distributed among the com- poimds of all the elements, nor in such manner as the laws of chance can explain, nor altogether in such manner as the periodic system of the elements describes. If the extreme values and unique prop- erties be considered, very many are seen to belong to the three elements hydrogen, oxygen, and car- bon in an arrangement that brings about stabihty of physical and chemical conditions, and diversity of phenomena, and, further, the possibility of the greatest complexity, durability, and activity of physico-chemical systems on the surface of a planet. This order is masked when the properties of matter are considered statically. It becomes evi- 8 NATURE dent only when time is taken into account, for this is the order that determines the later course of cosmic evolution. At present it can be only im- perfectly described, but there is reason to hope that a clearer description is attainable, and if an explanation seems to be beyond our grasp, the recognition of the order may yet serve a useful purpose by helping to define a little more clearly one of the riddles of existence. Proceeding from the results of this earlier in- quiry, I have, in the following pages, endeavored in a more rigorous manner to discuss the import- ance of the three elements for the process of cosmic evolution and by eliminating all biological theories and principles to rest the conclusions exclusively upon the secure foundation of abstract physical science. Such is the principal aim of the present essay. But it has also seemed desirable at least to raise another question. For the fact cannot be escaped that these considerations have a philosophical as well as a scientific bearing. I have, therefore, after much hesitation, ventured to sketch the develop- ment of thought upon the problem of teleology, and at length to confront the scientific conclusions with the results of philosophical thought, in order finally to attempt a reconciliation. I fear that this task has been accomplished with feeble strokes. It was not undertaken confidently. INTRODUCTION 9 but in the sincere belief that when such questions are involved men of science can no longer shirk the responsibility of philosophical thought. Only thus can they hope to escape from many errors, like those that weaken both sides of the vitalistic and mechanistic controversy, and that do really retard the advancement of science. But in thus mingling philosophy with science a danger is incurred. I would, therefore, beg the reader still to remember after he has turned this page that, when all has been said, the scientific conclusions are inde- pendent of the philosophical problem of teleology. And — I wish to say it as clearly as possible — the present essay professes to demonstrate nothing but the existence of a new order among the prop- erties of matter, and only to examine the teleologi- cal character of this order. II ARISTOTLE The teleological appearance of nature and the forms of life is a universal fact of human experience. Hence it has been quite impossible for natural science or philosophy permanently to ignore the problem of teleology. Merely to explain away the order of nature is no more satisfactory than to ex- plain away matter itself. We may argue against such ideas ever so ingeniously, but the experiences of daily life steadily oppose the arguments, and gradually overwhelm them. Thus men must always hiquire into the cause and significance of the teleo- logical appearance of things. Efforts to solve such questions are to be found in every system of thought; they are greatly involved in the earliest of all systems. The peculiar philosophical standpoint of Aris- totle, that position which enabled him to unify his doctrine of the philosophy of nature, can only be understood as the result of many different circum- stances. And yet one feels that the teleological parts of his thought were no mere accidents of time and place. Aristotle's historical derivation from Socrates and Plato, and, on the other hdnd, from Greek physicians is important as a determinant of 10 ARISTOTLE 11 his thought. Not less so are remote influences; for his work is continuous with the evolution of early science and primitive speculation. In short, both the sources of his system in personal experience, and its historical connections are to the point; and we may fairly think that they are often quite decisive in respect of his completed systematic views. But it is, above all other causes, his temperament, his native bent as a philosophical naturalist, to which the general character of his elementary teleological conceptions seems to be due. His system is only intelligible as an histor- ical product, but on this subject his opinions are just his own. No long reading of Aristotle's works is necessary to reveal in his general attitude toward final causes the very essence of his own mental disposition, the reflection of the world of life and thought as neces- sarily perceived through his eyes. Thus and not otherwise Aristotle was destined to see nature, if he was ever to see her as a whole, clearly. Nevertheless, it was an historical accident that the systematic examination of the teleological concepts was first undertaken by Aristotle, and carried through under pecuhar conditions and diffi- culties. The problem arose, to be sure, from his equal interest in the philosophy of forms and in natural history. Moreover, this conflicting in- terest in the end prevented a genuine resolution of 12 NATURE the difficulties. But it was the accident of his time that compelled him to deal with the ideas of both Democritus and Plato. For the same reason he lacked a clear conception of mechanical or even of " efficient " causation, the one essential foundation of a clear theory of final causes. And, in turn, it was this defect which led to his peculiar idea of development, founded upon metaphysical con- cepts of matter and form. Hence his most highly elaborated considerations upon this point are least satisfactory. In this view of development, which occupies the central position in the Metaphysics, one seeks in vain, I believe, for those elementary teleological conclusions which are so important in Aristotle's thought. They have disappeared in a subtle proc- ess of synthesis. It is true, contrary to the opin- ion of Gomperz,* that the modem idea of evolution, hence at least one question of objective science, is not entirely foreign to this analysis of development. The study of the Politics clearly establishes the point. But not in this concept of development, that concerns primarily the logical aspects of the idea of formation, nor in any similar speculations, does the riddle of the teleology of nature reveal itself simply and clearly. This can be grasped only in the scientific investigation of nature itself. Even so admirable a stroke as Aristotle's compari- ' Greek Thinkers, IV, 154. London, 1912. ARISTOTLE 13 son of natural formation with the work of the artist, whereby both analogies and differences are revealed, and the curious unimportance of con- scious design is made evident, only confuses the real questions. In the biological works, however, the teleologicalf problems appear in their simplest forms. Even' here there are difficulties and inconsistencies enough, and too many unlucky errors which count' heavily in the later development of science. But] the ideas are established; they arise as they are to recur for all later generations; and they initiate one of the great currents of human thought. Whatever blimders Aristotle may have made, hd has here avoided fallacies that have proved dan- gerous to his successors, even in our own times The reason for this success is that he adopts as his starting point those ideas which sooner or later must come home to every genuine naturalist. However they may be interpreted, these ideas will forever persist. They are the basis of all later speculations. Aristotle was not equipped with the philosoph- ical and scientific methods that have been found indispensable for a genuinely critical examina- tion even of this simpler problem, and that are absolutely necessary for a formulation of the ultimate concepts. But his efforts, restricted in method though they had to be, were carried out 14 NATURE systematically and with great subtlety of dialectic; they extended throughout the vast field of his scientific knowledge; they especially involved the whole of his favorite science of zoology; and the result, accordingly, was not the least important of his contributions to the understanding of nature. Aristotle's discussion of causation arises from the consideration that we are obliged to assign to nature several different kinds of causation, two of which are especially important for the philosophy of science. He says himself: (* The causes con- cerned in the generation of the works of nature are . . . more than one. There is the final cause and there is the motor cause. Now we must decide which of these two causes comes first, which second. Plainly, however, that cause is the first which we call the final one. For this is the Reason, and the Reason forms the starting point, alike in the works of art and in the works of nature." ^ It should be observed that the employment of but two kinds of causation in the explanation of nature, in place of the four which are to be found in his more philosophical works, is characteristic of Aristotle as a naturahst. The relation between the two forms of causation is not to be judged from the priority of the reason alone for " it is plain . . . that both of these must, 1 De partibus animalium (The Works of Aristotle translated into English), I, 1, 639^ 10-15. Oxford, 1911. ARISTOTLE 15 SO far as possible, be taken into account in explain- ing the works of nature, or that at any rate an attempt must be made to include them both; and that those who fail in this tell us in reality nothing about nature." ^ Democritus, however, neglecting the final cause, reduces to necessity all the opera- tions of nature; but though necessary, they are for a final cause, and for the sake of what is best in each case. Thus it is also possible to understand, according to Aristotle, the failure of Empedocles, who was content to limit his reflections to mechanical causation. On the other hand, in the time of Soc- rates, men gave up inquiring into the works of nature, so that mechanical causation did not receive its due regard. The true method to be employed is illustrated by Aristotle as follows^, " In dealing with respiration we must show that it takes place for such or such a final object; and we must also show that this and that part of the process is necessitated by this and that other stage of it." ^ The study of physiology from this point of view is likely to lead to a vitalistic theory, but it hardly involves the general philosophy of nature. For Aristotle, however, the same considerations apply 1 De partibus animalium (The Works of Aristotle translated into English, I, 1, 642», 15. Oxford, 1911. * Ibid., 1, 642s 30. 16 NATURE to the whole of nature. " Absence of haphazard /and conduciveness of everything to an end are to be found in Nature's works in the highest degree, and the resultant end of her generations and com- binations is a form of the beautiful." ^ Teleology therefore appears to be a universal principle. In the science of life, however, a more subtle consideration arises, and this leads Aristotle to the concept of organization. " As every instru- ment and every bodily member subserves some par- tial end, that is to say, some special action, so the whole body must be destined to minister to some plenary sphere of action." ^ " And the animal organism must be conceived after the similitude of a well-governed common- wealth. When order is once established in it there is no more need of a separate monarch to preside over each several task. The individuals each play their assigned part as it is ordered, and one thing follows another in its accustomed order. So in animals there is the same orderliness — nature taking the place of custom — and each part naturally doing his own work as nature has com- posed them. There is no need then of a soul in each part, but she resides in a kind of central governing place of the body, and the remaining parts live by 1 De partibus animalium (The Works of Aristotle translated into English), I, 5, 645% 20. Oxford, 1911. 2 Ibid., 645b, 10-15. ARISTOTLE 17 continuity of natural structure, and play the parts Nature would have them play." ^ The idea of organization leads at once to a science of physiology based exclusively upon the ; concept of function. But once more a qualification arises; — this will not do as the complete science, for it deals only with final causes, accordingly " We have ... to inquire whether necess ity may not also have a share in the matter; and it must be admitted that these mutual relations could not from the very beginning have possibly been other than they are." " Thus Aristotle arrives, no doubt less clearly than we may be inclined to think, at the conception of mechanism and teleology as complementary aspects of nature, which are always associated in its mani- festations. And he is therefore led to a further question: " Let us now consider the character of the material nature whose necessary results have been made available by rational nature for a final cause." ' This no doubt is an important quest, which con- templates nothing less than the problem whose solution must put teleology in its proper place, or else eliminate it altogether. But the mquiry not unnaturally leads Aristotle from his general prin- 1 De motu animalium, II, 703^, 30-35. Oxford, 1912. ^ De fartihus animalium, II, 1, 646'', 25-30. ' lUd., Ill, 2, 663^ 20. 18 NATURE ciples to the infinite complexity of phenomena; he loses his clear vision of the principles, and stumbles into pitfalls. For dysteleology is hardly less obvious m nature than teleology, and the search for a final cause of everything is a hopeless task. Thus betrayed he concludes that: "... we must not in all cases expect to find ... a final cause; for granted the existence in the body of this or that constituent, with such and such properties, many results must ensue merely as necessary conse- quences of these properties." ^ And he even dares to be specific in his statement of this idea, that: *' Whenever things are not the product of Nature working upon the animal kingdom as a whole, nor yet characteristic of each separate kind, then none of these things is such as it is or is so developed for any final cause. The eye for instance exists for a final cause, but it is not blue for a final cause unless this condition be characteristic of the kind of animal." ^ Such ideas are damaging to the logical consist- ency of Aristotle's views. But they are harmless to his science. Not so the complementary incon- sistency. The explanation of natural phenomena by final causes alone is at once incompatible with his principles and destructive of all sound science. Yet it is only too common in his scientific treatises. 1 De partitnts animalium, IV, 2, 677", 15. ^ De generatione animalium, V, 1, 778*, 30. Oxford 1910 ARISTOTLE 19 We need make no difficulty of understanding Aristotle's perplexities on this point, for they depend upon his ignorance of the true manner in which mechanical processes are to be conceived. Thus for example the phenomena of the heavens were scientifically a complete riddle to him, and/ teleological explanations his only escape from perfect bewilderment. It is singular that within a hundred years Archi- medes and others shovdd have been able to avoid these difficulties and, quite in the modem spirit, investigate the problems of mechanics. But for Aristotle this field of research was closed, and accordingly the use of final causes as a sufficient principle for the explanation of nature became usual in his works and the fatal defect of his natural philosophy. It is unprofitable to continue the analysis of Aristotle's failures as a physical scientist. In certain departments these are too well known and too apparent to call for comment, but the blame for their perpetuation belongs to his successors. And it should never be forgotten that his great merits as a zoologist more than offset his errors. We need not further consider his remarks upon teleology. For though other examples in great profusion could readily be cited, they add little to the essential considerations. 20 NATURE Accordingly, if the preceding analysis be not at fault, the foundations of Aristotle's teleology may be stated as follows: In the study of the living organism the mechanical cause and the reason of everything must both be sought. This is an abso- lute rule, although there is ground for the belief that sometimes the one explanation, sometimes the other, cannot be discovered. The whole of nature is also subject to these two forms of causation. But the diflSculties of the in- vestigator are here multiplied. For on the one hand matter is a refractory medium; it does not lend itself quite perfectly to the working out of the ends of nature. For this reason results may some- times arise which are due to necessity and not properly to final causes at all. On the other hand the wider our experience and knowledge of nature, the more often do we lose track of the chain of necessary causation and discover only the final causes. It is only reasonable to proceed one step further in the elimination of the unessential from Aris- totle's views. We then reach the heart of his doctrine: Teleology and mechanism are in all phenomena, for they are complementary aspects of all things and all changes. Every qualification of this view is evidently due to Aristotle's perplexities as a naturalist or physicist. Therefore, it is the task of the investigator to " consider the character ARISTOTLE 21 of the material nature whose necessary results have been made available by rational nature for a final cause." A more pregnant statement was never uttered. One special view remains to be noted: the con- ception of the living thing as an autonomous unit in which every part is functionally related to every other and exists as the servant of the whole. No external end or purpose guides this being. Here, as in the state, the teleological principle is within. Every activity is subject to the regulative control of the soul. But, in the biological treatises, the soul is nothing more than a name for the principle of autonomy. Kant and the modern physiologists have expanded this view without improving it. It is the complete formulation of the biological prin- ciple of organization. Ill THE SEVENTEENTH CENTURY During the next two thousand years the history of teleology more than of other things is a record of the stagnation and decay of thought. Although the great achievements of Archimedes and the Alexandrians did almost at once afford an example of how unnecessary is a regard for teleology in the development of physical science, the lesson re- mained unheeded, and in the course of time the system of Aristotle won the commanding position in all domains of thought. The followers of Aristotle, Mohammedans and Christians alike, were able in most respects only to degrade his doctrine, for they had lost his spirit of independence; only rarely could they comprehend the precision of his abstractions and generaliza- tions; and above all they were too far from nature. Perhaps, of all his works, the teleological portions suffered most at their hands. The rare independent spirits who from time to time arose — Roger' ^acon, Nicholas of Cusa, Leonardo da Vinci — failed or did not try to shake the authority of the schools, and meanwhile the practice of explaining the phenomena of nature by their supposed final causes, alone increased and developed. Under the 82 THE SEVENTEENTH CENTURY 23 influence of logical pedantry or the belief in author- ity, now arrayed in the service of theology, Aris- totle's worst faults were perpetuated, and when the currents of modem thought began to flow, the abuse of ideas which were admirable in their in- ception had long been complete. i The actual situation is revealed in the works of Francis Bacon. When only thirteen years old his mind had revolted against the accepted doctrines and the mature philosopher was not slow to detect and define the sources of error. These are first stated in a famous passage of " The Advancement of Learning." " The second part of Metaphysic is the inquiry of final causes, which I am moved to report not as omitted, but as misplaced. And yet if it were but a fault in order, I would not speak of it; for order is matter of illustration, but pertaineth not to the substance of sciences: but this misplacing hath caused a deficience, or at least a great inproficience in the sciences themselves. For the handling of final causes mixed with the rest in physical in- quiries, hath intercepted the severe and diligent inquiry of all real and physical causes, and given men the occasion to stay upon these satisfactory and specious causes, to the great arrest and prej- udice of further discovery. For this I find done not only by Plato, who ever anchoreth upon that shore, but by Aristotle, Galen, and others, which 24 NATURE do usually likewise fall upon these flats of discours- ing causes. For to say that the hairs of the eyelids are for a quickset and fence about the sight; or that the firmness of the skins and hides of living creatures is to defend them from the extremities of heat or cold; or that the bones are for the columns or beams, where- upon the frames of the bodies of living creatures are built; or that the leaves of trees are for protecting of the fruit; or that the clouds are for watering of the earth; or that the solidness of the earth is for the station and mansion of living creatures, and the like, is well enquired and collected in Metaphysic; but in Physic they are impertinent. Nay, they are indeed but remoras and hindrances to stay and slug the ship from further sailing, and have brought this to pass, that the search of the Physical Causes hath been neglected and passed in silence. And therefore the natural philosophy of Democritus and some others, who did not suppose a mind or reason in the frame of things, but attributed the form thereof able to maintain itself to infinite essays or proofs of nature, which they term fortune, seemeth to me (as far as I can judge by the recital and fragments which remain unto us) in particu- larities of physical causes more real and better enquired than that of Aristotle and Plato; whereof both intermingled final causes, the one as a part of theology, and the other as a part of logic, which were the favourite studies respectively of both THE SEVENTEENTH CENTURY 25 those persons. Not because those final causes are not true, and worthy to be enquired, being kept within their own province; but because their ex- cursions into the limits of physical causes hath bred a vastness and solitude in that track. For other- wise keeping their precincts and borders, men are extremely deceived if they think there is an enmity or repugnancy at all between them. For the cause rendered, that the hairs about the eye-lids are for the safeguard of the sight, doth not impugn the cause rendered, that pilosity is incident to orifices of moisture; Muscosi fontes [the mossy springs], etc. Nor the cause rendered, that the firmness of hides is for the armour of the body again against extremi- ties of heat and cold, doth not impugn the cause rendered, that contraction of pores is incident to the outwardest parts, in regard of their adjacence to foreign or unlike bodies; and so of the rest: both causes being true and compatible, the one declaring an intention, the other a consequence only. Neither doth this call in question or derogate from divine providence, but highly confirm and exalt it. For as in civil actions he is the greater and deeper poli- tique, that can make other men the instruments of his will and ends and yet never acquaint them with his purpose, so as they shall do it and yet not know what they do, than he that imparteth his meaning to those he employ eth; so is the wisdom of God more admirable, when nature intendeth one thing 26 NATURE and providence draweth forth another, than if he had communicated to particular creatures and motions the characters and impressions of his providence." ^ It is apparent that Bacon does not differ radically from Aristotle. Had he been able to distinguish the original elements of Aristotle's thought from the master's blunders and the school's vagaries, he must have dealt with the problem quite differ- ently. But neither the age nor the cast of Bacon's own mind was favorable to historical criticism. Perhaps for this reason Bacon's one genuine contribution to the teleological problem is to be found in his discussion of the method of science. Admitting the Aristotelian principle that mecha- nism and teleology appear to be two complemen- tary aspects of things, he showed that experience demands their separation in scientific research. Thus he discovered the peculiar feature of physical science that it must proceed as if final causes did not exist, even though he fully agreed that they may be conceived as real. In other words physical science can recognize only one kind of causation, which is physical causation. This is a return to Democritus and Empedocles. Bacon's criticism is quite sound, but it misses that important feature of Aristotle's thought, the ' The Philosophical Works of Francis Bacon, pp. 96, 97. London Routledge. 1905. THE SEVENTEENTH CENTURY 27 concept of organization. And yet this concept had undergone important developments at the hands of Aquinas. Perhaps the defect is fortunate; for certain it is that physiology needs sound physical investigations quite as much as physics itself. But the defect remains, and it is significant of Aris- totle's superiority as a naturalist. In another respect Bacon and Aristotle suffer from a like disabiUty. Neither is able to conceive just how one should go about a physical research. A modern philosopher, to be sure, is far better in- formed upon this subject, for history affords him many more examples, and Bacon's misconceptions now seem almost inexcusable. Nevertheless he is notoriously wide of the mark in his illustrations of scientific method in the Novum Organum, and he failed to see the point of much contemporary re- search. Perhaps one direct influence of his thought concerning the subject of causation was to en- lighten his unmediate successors, if indeed the more clear-sighted were ui need of enlightenment, but it is certain that he had no valid notion of mechanical causation. This first arises in the in- vestigations of Galileo and receives its first critical treatment at the hands of Descartes.^ The ancients, however, did not wholly lack an idea of mechanical causation, although it finds no ' For an excellent and learned discussion of the development of the concept of causation see E. Meyerson, IdentitS et RealiM. Paris, 1908. 28 NATURE place in the thought of Aristotle. Very imper- fectly the old atomic theory served the purpose, and the speculations of Lucretius provided a foun- dation for the conception of the indestructibility of matter,! ^j. e^gn perhaps of the conservation of mass.^ The same idea was destined to find a place in Newton's inquiries,^ and we may note in passing that, for scholastic philosophy, causes had likewise been things or substances, rather than forces or conditions. In accordance with this atomic view of things, that which results from a change is such as it is because it has been formed without gain or loss of substance from that which has disappeared. These ideas, however, were always vague. There can be no doubt that, if they proved themselves impotent in the chemistry of the eighteenth century until Lavoisier introduced them into his experimental researches, they must have been without perma- nent influence of a sufficiently definite nature upon the thought of earlier times. In fact the development of dynamics rather than of chemistry equipped the scientific investigator with his earliest representation of the true charac- ter of mechanical necessity. This concept arises directly from the principle of inertia, and therefore 1 De natura rerum, 1, I, 160, 486-487, 500, 652-565, 584-598. " lUd., 361-363. ' Opticks, 3d ed., p. 375. THE SEVENTEENTH CENTURY 29 from Galileo's experiments on falling bodies.* Long before its definite formulation by Galileo it was grasped by Descartes, perhaps independently,^ and incorporated in his philosophical system. Thereby the defect in Bacon's formulation of the method of science was temporarily repaired. For Descartes, struggling with his new philo- sophical system, the principle of inertia leads directly to the law of conservation of movement, his " memorable error " as it was called by Leib- niz.^ Accordingly Descartes proceeds from the hypothesis that the product of mass by velocity in all natural phenomena is constant, to the formula- tion of a principle of universal necessary causation. From his reflections upon the principles of dynamics he was led to the belief that " God never changes his manner of acting — and in order to maintain things with the same action and the same laws, which he has caused to obtain in their creation, it is necessary that he should now conserve in them all the movement which he then introduced into them, together with the property which he has given to this movement that it shall not forever remain attached to the same portions of matter, but in their encounters shall pass from one portion ^ On the idea of inertia in antiquity, see P. Tannery, Revue gSn- Srale des sciences, xii, 1901, 333. ' Meyerson, loc. cit., p. 104. ' Mathematische Sckriften, ed. Gerhardt, VI, 117. See also below, pp. 33, 36. 30 NATURE to another." ' This is Descartes's first contribu- tion to the definition of the teleological problem. In spite of the unsound foundation, it marks a very great advance in thought. The argument, however, leads him much farther, and in his progress he becomes the founder of the systematic view, long since developed as the basis of dynamics and, in a manner, of all natural science, that every phenomenon is ultimately reducible to matter and motion. He arrives at a consideration of the functions of the living organism. And at one stroke, aided by Harvey's researches upon the circulation, he founds the mechanistic theory of the vital processes. Here too it would seem that mechanical causation must be supreme; but this conclusion is incompatible with the theological view of voluntary action. Thus arises, for the first time clearly defined, the ever-perplexing problem of vitalism. The chains of mechanical causation may be rigorously deter- mined, but where living things are involved they can never seem so. Yet Descartes himself was not inextricably entangled in this difficulty, from which no later thinker has been able to escape. The very errors of his dynamics, which reduce his conception of causation to a mere preliminary sketch of the true principle, left him a way out. Such is the origin of his mistaken notion that the ' Principia, Part II, ch. 42. THE SEVENTEENTH CENTURY 31 will might operate by changing not the quantity but merely the direction of motion. The theory is ingenious and undoubtedly fills a real want, as many analogous speculations of a later date attest. But it could not withstand the most superficial analysis and it soon fell before the rapid advance of theoretical mechanics. All the other contributions of Descartes to the teleological problem may be summed up in one statement. He, more clearly and systematically than any of his predecessors, perceived and eluci- dated the complementary relation of mechanism and teleology, thereby reinforcing Aristotle's position. For him all things are teleological at all stages of their necessary development because they originally possessed a teleological character which is itself necessarily conserved. This view, his valuable though imperfect discussion of causa- tion, his mechanistic theory of physiological activ- ity, and his definition of the vitalistic hypothesis in the discussion of freedom, determine Descartes's position. I have given no account of his admir- able analysis of the problems and his clear exposi- tion of the concepts. In his case, apart from theological interests and such considerations as involve the properties of matter, these may justly be taken for granted. The philosophical system of Descartes moves chiefly in another world from that of these simple 32 NATURE ideas. For that very reason we need not here investigate the subject. Even his strange meta- physical conceptions of matter and extension are without importance for his theory of mechanical causation, because derivative from this and from his other scientific views. In the thought of Descartes, as in that of Aristotle in an earlier day, and of Leibniz a little later, philosophical princi- ples are not to be regarded as consistent with scien- tific principles. The effort is to be scientific, yet ia all three systems the philosophical principles are derived from a number of incompatible sources, of which the scientific is probably most important. But physical science and its mathematical foun- dations seem to have been Descartes's central interests, and the source of a great part of what is truly original in his philosophy. ^ No long succession of centuries was to pass before the fallacies of the Cartesian theory of causation were revealed, for they had been pub- lished to a world seething as never before with scientific thought, and busy with the experimental and mathematical investigation of dynamics. Huygens, Newton, and Leibniz immediately took up the task which Galileo and Descartes laid down. And very soon, with the aid of many lesser workers and thinkers, they had completed the definition of ' Cf. Gilson, La doctrine cartedenne de la liberti et la thSologie. Aican, Paris, 1913. THE SEVENTEENTH CENTURY 33 the scientific principles and improved the idea of mechanical causation. Leibniz's attack upon the Cartesian dynamics has been often discussed, but not always with a clear understanding of the true nature of the under- lying scientific principles. ^ There is not much of originality in the criticism itself, which is largely founded upon the results of Huygens and others. The one point which concerns the teleological problem is a demonstration that the direction of motion cannot be altered by the action of the will, j since the conservative principle in respect of vis viva must apply, if at all, to the sum of moving force in any direction. With this demonstration the peculiar vitalism of Descartes, which had already received its death blow at the hands of his disciple Geulincx ^ passes out of the history of thought. The great interest of Leibniz's position is due to the fact that he first faced the problem which in- evitably arises from a philosophical examination of the completed principles of classical dynamics. His treatment of the subject reveals all the capac- ity of his admirable intellect, and the result is a synthesis of the thought of his century. In grasp of all departments of knowledge and speculation, in skill of dialectic, and in sheer intellectual power it is one of the great examples of philosophical > See Mach, Die Mechanik, 3d ed., p. 274. '' Windelband, Geschichte der Philosophie, 3d ed., p. 341. 34 NATURE analysis. But it suffers from the fatal defect of a constant concern for the interests of theology. It tries to serve two masters. Leibniz could not but believe that mechanical necessity is a principle without exceptions, and accordingly that God " foresaw and arranged everything once for all." ' The doctrine is developed as a result of his prin- ciple of the conservation of vis viva (the conserva- tion of S m v^) though there is reason to believe that it really arises, as in an imperfect form it had occurred to Descartes, more or less directly out of the idea of inertia. For, historically speaking, tne idea of absolute universal determinism seems to be almost necessarily imposed upon the student of dynamics. He may think to derive it from the principle of the conservation of motion, of vis viva, or of energy, or from the two principles of ther- modynamics together; the psycho-physical riddle may lead him to all sorts of ingenious subtle quah- fications; but the idea always stands in his mind as the generalization of his concept of causation or as a self-evident a priori principle from which the notion of cause is itself derived. No doubt, however, the principle is not strictly an a priori judgment, in that it cannot be formed with the necessary precision in the absence of extensive scientific knowledge. Therefore the simplest view is to regard it as a necessary corollary of the 1 Ed. Gerhardt, III, p. 400. THE SEVENTEENTH CENTURY 35 concept of inertia,' upon the assumption that all phenomena may be reduced to matter and motion. In this connection we may recall Newton's first law of motion; " Every body continues in its state of rest or of uniform motion in a straight line unless compelled to change that state by forces impressed upon it." Leibniz is certainly the real author of the convic- tion IbEat every phenomenon, without any excep- tion whatever, is the result of mechanical causation and is therefore rigorously and unequivocally de- termined. The view is almost as old as thought itself, and was very widely held in his day. But Leibniz founded it upon a careful analysis of thej known laws of nature, and thus made it directly/ accessible to the understanding and imagination. His analysis has been criticized, extended, and thereby gradually modified. But throughout the later development of scientific thought it has never for a moment lost the support of the greater number of qualified judges, and today it consti- tutes the first article of the orthodox scientific creed. The fact is historically interesting that Leibniz conceived visible motion of masses as somehow losing itself in that of the imperceptible constituent particles, when motion apparently ceases. The identity of this excellent guess with our modem theories is obvious. ^ But see Meyerson, loc. cit. 36 NATURE The position of Leibniz is made quite clear by many passages in his works, of which one may suf- fice as an example. In the Monadology he declares that " Descartes saw that souls cannot at all im- part force to bodies, because there is always the same quantity of force in matter. Yet, he thought that the soul could change the direction of bodies. This was, however, because at that time the law of nature, which affirms also the conservation of the same total direction in the motion of matter, was not known. If he had known that law, he would have fallen upon my system of Pre-established Harmony. " According to this system bodies act as if (to suppose the impossible) there were no souls at all." 1 . . . Leibniz might better have treated the problem which thus arises in the simple Aristotelian manner. Two considerations, apparently, led him far beyond the purely destructive criticism of the Cartesian position to his theory of Pre-established Harmony. These are the perplexing riddle of voluntary action and the interests of theology. It has become customary, at least in scientific circles, to look upon the monad, which is at the very foundation of Leibniz's Pre-estabhshed Harmony, as a vague creation of fancy, incoherent, unintelli- 1 Monadology, 80, 81, in Leibniz trans, by Montgomery. Chicago, 1908, p. 269. THE SEVENTEENTH CENTURY 37 gible, even preposterous, and quite unworthy of the mind of one of the inventors of the calculus. It is not so. Mr. Bertrand Russell has a right to be heard on this point, and he declares that: "This seemingly fantastic system could be deduced from a few simple premises, which, but for the conclu- sions which Leibniz had drawn from them, many, if not most, philosophers would have been willing to admit." ^ There can be no doubt, however, that this unique example of metaphysical speculation according to the pattern of mathematics does not concern us. It has no importance for the history of the forma- tion of ciu'rent teleological concepts, in so far as I find them relevant to our problem.^ The Monad- ology, to be sure, is nothing if not teleological, and it was designed to be so. But the permanently significant teleological elements are not involved in the fate of the whole structure. This may be over- thrown without including them in the destruction. Only one of these concerns the present subject, and this is present in all of Leibniz's specula- tions. For him, as for Aristotle, the universe is no less teleological than mechanical. But Leibniz, even more than Descartes, is compelled to put the 1 A Critical Exposition of the Philosophy of Ldhniz, p. viii. Cam- bridge, 1900. * Cf. however Ward, The Realm of Ends: Pluralism and Theism. Cambridge, 1911. Lecture III, etc. 38 NATURE origin of the teleological back of all mechanism at the origin of things, and thus in the creation itself. There can never be any genuine teleological novel- ties of whatever origin, for all is order. Even miracles are in accordance with the order of nature, just as Babbage later crudely explained. Details of Leibniz's speculations may appear to contradict or gravely to qualify his fundamental position, but such vagaries are not more impor- tant than the bewilderment of Aristotle under similar conditions. For Leibniz resembled both Descartes and Aristotle especially in this, that his scientific views are fundamental, his philosophical opinions largely secondary. ^ Especially that which is original and valuable in the philosophy of Leib- niz is of scientific origin, for the results of science and mathematics were the only novel sources of his speculations. The necessity of assimilating the idea of abso- lute mechanical determinism to his metaphysical position had one other important result in Leib- niz's thought. It led to an examination of the problem of organization. The investigations of the seventeenth century biologists had gradually revived interest in this question, and at length it had become possible vaguely to perceive how dif- ferent is the problem of organization itself from ' Cf. B. Russell, op. cit; Couturat, La logigue de Leibniz. Paris, Alcan, 1901 ; Cassirer, Leibniz's System, Berlin, 1902. THE SEVENTEENTH CENTURY 39 that of simple purposeful behavior. Not only the vitalism of Descartes, but the far wider view of Stahl, which extends the animistic principle from the operation of consciousness to the organism as a whole, appear to Leibniz as radically false. For him nothing in the living being is heterogeneous with mechanism, and everything has its mechani- cal cause or explanation. Otherwise, as the modem mechanists still contend, it must be quite unintel- ligible. But the organism is a mechanism of exquisite perfection where everything takes place as if the materialistic philosophy of Epicurus and Hobbes were true. Yet this absolute scientific truth possesses only relative philosophical validity. The mechanical character of the organism, like that of nature itself, is just the means by which we attain to the eternal truths. And in the organism we can readily see the philosophical limitations of the mechanical description as an ultimate philo- sophical position. " Therefore, every organic body of a living being is a kind of divine machine, or natural automaton, infinitely surpassing aU artificial automatons. Because a machine con- structed by man's skill is not a machine in each of its parts; for instance, the teeth of a brass wheel have parts or bits which to us are not artificial products, and contain nothing in themselves to show the use to which the wheel was destined in the machine. The machines of nature, however, 40 NATURE that is to say, living bodies, are still machines in their smallest parts ad infinitum. Such is the diflference between nature and art, that is to say, between divine art and ours." * Leibniz's con- ception of organization is thus inferior to Aris- totle's.2 In spite of this he has established the important principle that organization is compati- ble with mechanism. So far as it concerns the development of the theories of teleology the result of the first period of modern science and philosophy is now apparent. Very important is the strengthening of Aristotle's original position. Mechanism and teleology are still to be regarded as complementary aspects of all things, whether physical or biological. The pecul- iar character of the living being as an organism is still recognized, but, without the help of Aristotle's profound insight and in the absence of advanced biological thought, vaguely conceived. Meanwhile nature itself has put on more and more the appear- ance of the organism. On the other hand, the mechanistic has been logically disentangled from the teleological. All is as if mechanism were the only ultimate reality, or at least all phenomena of matter, and motion are so. And all phenomena are reducible to matter and motion. Newton, no less than Leibniz, seems to take this for granted. ^ Monadology, 64, op. cit., pp. 265, 266. ^ Above, p. 16. THE SEVENTEENTH CENTURY 41 Moreover all phenomena of matter and motion are necessarily determined according to that concept of conservation which is the direct consequence of the idea of inertia. Hence teleological principles are involved only in the interpretation of phenomena, especially in the interpretation of nature as a whole and of the! organism. This interpretation, however, involves a form of description, which, though quite inde- pendent of ordinary physical description, is once more correlative with it. Final causes, therefore, remain in high favor. In physical science they have lost their power to do mischief. But in wider fields of thought they are quite as dangerous as ever. IV THE EIGHTEENTH CENTURY It is a dismal period that followed the deaths of Leibniz and Newton. Taking the lead which they had given, physical science went on its way un- troubled by further metaphysical problems save those of its own creation. But meanwhile the philosophy of nature degenerated into eighteenth century Theism, and so departed altogether from the road of progress. And yet, in spite of appear- ances, the field was not quite free for childish play with final causes. Those who were so engaged might suppose that they could find their full justi- fication in the philosophy of Leibniz and in the evolution of Theology. Gradually, however, the ideas of Locke, most independent of seventeenth century philosophers, were working in another direction, and at length Hume arose. By this time the teleological question had come to be regarded as identical with the problem of design; a view for which little can be said except that it reveals the development of a vague conception of organic unity in nature. In every other respect it is a sign of decadence. Hume's historical position no less than his natural temperament was ideally favorable to the 42 THE EIGHTEENTH CENTURY 43 task of examining the problem of design. In the history of the subject this is a rare combination, but necessary for unbiased judgment. After/ Aristotle he was perhaps the first notable thinker again to approach teleology in a thoroughly dis- passionate and impartial spirit. Not that he wag indifferent, as so many have believed, to the in- fluence of his thought on morals, but he evidently possessed the true philosopher's conviction of the supreme value of thought itself. And whatever concern he may have felt for his destructive in- fluence upon religion in general must have been offset by the desire to overthrow the abhorred theological system of his day. Design was the great and primary question of the time in Hume's England, from which he could not have escaped. It is very evident, however, that he would not if he could, for he has himself reported with favor the view that natural theology should come as the mature climax of the other philosophi- cal studies.! It is probable that Hume's Dialogues concerning Natural Religion may justly be regarded as his last word in philosophy. Chronologically the latest of his philosophical writings, their style marks them as the product of much labor and careful recon- • " That students of philosophy ought first to learn Logics, then Ethics, next Physics, last of all, of the nature of the Gods. Chrysip- jnts apud Plut. de repug. Stoicorum." Didoguea concerning Nalural Religion, Edinburgh and London, 1907, pp. 6, 7. 44 NATURE sideration. We know too that, during the long period when they remained unpubUshed, he sought the criticisms of his friends; ^ and he reserved the pubhcation until after his death. In forming a true opinion of the Dialogues two other considerations are important. On the one hand, Hume lived in a day when the echoes of the seventeenth century scientific revolution had, especially in England, almost died away. A period of quiet and continuous progress had ensued, which, except for Lavoisier's great innovation, was to continue almost undisturbed for many decades. For this reason, no doubt, science and mathematics have but a small place in Hume's thought, so that his famous discussion of causation is according to the scientific view one-sided and sterile. On the other hand, Hume was almost equally isolated in the history of philosophy. He is so far removed from the philosophers of the seventeenth century that his works are on the whole, notably in respect of the teleological problems, discontinuous with theirs. And he did not live to see that great result of his own labors; Kant " wakened from his dog- matic slumbers." There are difficulties in discovering Hume's rear opinions beneath the uncertain conclusions of the Dialogues concerning Natural Religion. At the very outset, in an interesting apology for this literary ' Loe. cit., p. ix. THE EIGHTEENTH CENTURY 45 form,' he has himself indicated that his views are not quite established on many of the questions at issue, indeed the Dialogues may fairly be taken as a demonstration that these problems surpass human power, and it is known that he was dissatisfied with the results of his analysis. ^ But the general tend- ency of the argument is not doubtful. For Hume there can be but one foundation of a belief in design. This is the recognition of natural order, which, directly apprehended, is enough. There is, in the last analysis, no need of argument on the point, for the conviction of design arises without process of logic. As he makes Cleanthes say: " The order and arrangement of nature, the curious adjustment of final causes, the plain use and intention of every part and organ; all these bespeak in the clearest language an intelligent cause or author. The heavens and the earth join in the same testimony: The whole chorus of Nature raises one hymn to the praises of its creator : You alone, or almost alone, disturb this general har- mony. You start abstruse doubts, cavils, and objections : You ask me, what is the cause of this cause.'' 1 know not; I care not; that concerns not me. I have found a Deity; and here I stop my enquiry. Let those go farther, who are wiser or more enterprising. 1 Loc. cit, p. 2. ^ Ibid., p. xi. 46 NATUIffi " I pretend to be neither, replied Philo: and for that very reason, I should never perhaps have attempted to go so far; especially when I am sen- sible, that I must at last be contented to sit down with the same answer, which, without farther trouble, might have satisfied me from the be- ginning." 1 "A purpose, an intention, a design strikes every- where the most careless, the most stupid thinker; and no man can be so hardened in absurd systems, as at all times to reject it." ^ But this is Hume's only concession to natural theology. Unlike the bitterly conscientious But- ler, he can find no place for evil in this field of thought, and thus expresses himself: "And is it possible, Cleanthes, said Philo, that after all these reflections, and infinitely more, which might be suggested, you can still persevere in your Anthro- pomorphism, and assert the moral attributes of the Deity, his justice, benevolence, mercy, and rectitude, to be of the same nature with these virtues in human creatures ? His power we allow infinite: whatever he wills is executed: but neither man nor any other animal is happy : there- fore he does not will their happiness. His wisdom is infinite: he is never mistaken in choosing the means to any end: but the course of nature tends not to human or animal felicity: therefore it is not 1 Loc. cU., pp. 70, 71. « Ibid., p. 165. THE EIGHTEENTH CENTURY 47 established for that purpose. Through the whole compass of human knowledge, there are no in- ferences more certain and infallible than these. In what respect, then, do his benevolence and mercy resemble the benevolence and mercy of men ? " Epicurus's old questions are yet unanswered. " Is he willing to prevent evil, but not able ? then is he impotent. Is he able, but not willing ? then is he malevolent. Is he both able and willing ? whence then is evil ? " ^ This is by no means a fair treatment of the problem of evil, but it is sufficient as a reply to the theology of the day. For Himae, such considerations destroy the whole fabric of natural theology,^ and leave noth- ing but that impression which the recognition of the order of nature produces on the mind. The analysis reveals his firm conviction that the himian reason is entirely unqualified for thought upon the subject. In like manner the pre-established har- mony of Leibniz loses its whole foundation because it, too, denies himaan misery, or at least loses sight of it altogether.^ After many other discussions a final conclusion upon design is undeniably reached, and very care- fully formulated as follows: > Loc. cit.. pp. 133, 134. « Other departments of theological thought are not here in ques- tion. 3 Ibid., p. 125. 48 NATURE "If the whole of Natural Theology, as some people seem to maintain, resolves itseH into one simple, though somewhat ambiguous, at least undefined proposition, That the cause or causes of order in the universe probably bear some remote analogy to human intelligence: If this proposition be not capable of extension, variation, or more particular explication: If it afford no inference that affects human life, or can be the source of any action or forbearance: And if the analogy, imper- fect as it is, can be carried no farther than to the human intelligence; and cannot be transferred, with any appearance of probability, to the other qualities of the mind: If this really be the case, what can the most inquisitive, contemplative, and religious man do more than give a plain, philo- sophical assent to the proposition, as often as it occurs; and believe that the arguments, on which it is established, exceed the objections, which lie against it ? " ^ It is but a step from this position to Kant's theory that all teleological conclusions are mere reflective judgments of the human mind. But the historical importance of Hume's criticism is chiefly due to his destruction of all claims of natural theology to a scientific or philosophical standing. The natural theologians, indeed, went on for nearly another century, still working over 1 Loc. cit, pp. 189, 190. THE EIGHTEENTH CENTURY 49 the genuine results of science into the semblance of a theological form. But they did so only by ignoring all that Hume had told them, and with rare exceptions, their labors have nothing more to do with the real thought of the race. Yet the teleological problem remains. Theism had quite unwarrantably injected an anthropo- morphic element into teleology, but it had not altered the appearance of order. And nothing was farther from Hume's mind than to deny the exist- ence of this. He does, however, analyze it, and in the analysis are to be found his positive contribu- tions to the problem. It is apparent to Hume that the mind can con- ceive certain states of a blind chaotic system, aris- ing by chance in the course of time, which must maintain themselves for a longer or shorter period and must, therefore, present the appearance of order. He states his thought as follows: " Is there' a system, an order, an oeconomy of things, by which matter can preserve that perpetual agita- tion, which seems essential to it, and yet maintain a constancy in the forms, which it produces ? There certainly is such an oeconomy : for this is actually the case with the present world. The continual mo- tion of matter, therefore, in less than infinite trans- positions, must produce this oeconomy or order; and by its very nature, that order, when once established, supports itself, for many ages, if not 50 NATURE to eternity. But wherever matter is so poised, arranged, and adjusted as to continue in perpetual motion, and yet preserve a constancy in the forms, its situation must, of necessity, have all the same appearance of art and contrivance, which we ob- serve at present. All the parts of each form must have a relation to each other, and to the whole: and the whole itself must have a relation to the other parts of the universe. . . ." ^ Such an origin and development of the universe might well account, even when we take the living being itself into consideration, for the appearance of order and our resulting impression of design. In fact, says Hume, " It is in vain, ... to insist upon the uses of the parts in animals or vegetables and their curious adjustment to each other. I would fain know how an animal could subsist, unless its parts were so adjusted ? " ^ This, if not the principle of the survival of the fit- test, is at least the principle of the survival of the fit. Properly speaking it is even closer to Darwin's thought, for the positive fact in both ideas is the elimination of the unfit. Another aspect of Dar- win's conception of evolution is clearly developed by Hume in an analysis of the conditions which have governed the evolution of the ship, by a process of trial and error in which human skill and foresight had but small part. In truth the underlying idea » Loc. dt., pp. 105, 106. 2 Ibid., p. 109. THE EIGHTEENTH CENTURY 51 in all these considerations leads to a general state- ment of that tendency toward dynamic equilib- rium which is one of the principles of modem physics anH biology alike. Science has hardly suc- ceeded in formulating the view as broadly as Hume states it. But Hume's position is none the less well founded. This theory, though not entirely original with Hume, marks an important advance in the devel- opment of thought.^ It clearly demonstrates the] manner in which we are to conceive a mechanistic! imiverse at work upon the production of some- thing very like organic unity. I am of the opinion that it also shows Kant's view that teleology is exclusively a function of the reflective judgment to be untenable, or at least unimportant for natural science. * Cf. " But in what ways yon concom«e of matter founded earth and heaven and the deeps of the sea, the courses of sun and moon, I will next In order describe. For verUy not by design did the first-be- ginnings of things station themstives each in its right place by keen- sighted intelligence, nor did they bargain sooth to say what motions each should assume, but because the first-beginnings of things many in number in many ways impelled by blows for infinite ages back and kept in motion by their own weights have been wont to be carried along and to miite in all manner of ways and thoroughly to test every kind of production possible by their mutual combinations, therefore it is that spread abroad through great time after trying unions and motions of every kind they at length meet together in those masses which suddenly brought together become often the rudiments of great things, of earth, sea and heaven and the race of living things." Lucretius: " On the Nature of Things," translated by H. A. J. Munro, London (no date). Soutledge & Sons, p. 163. 52 NATURE There is another consideration, developed at an earlier stage of Hume's discussion, to which these conclusions lead back. The sceptic Philo is made to say that, " For aught we can know a priori, matter may contain the source or spring of order originally, within itself, as well as mind does. , . ." ^ Thus the original Aristotelian position once more reappears. But in this later form it is so developed and refined as to become almost a scien- tific problem. For we may fairly ask what is the nature of this original source or spring of order. The problem might even be solved if we could but construct Laplace's world formula, since the vari- ous quantities therein contained and the functional relations which must obtain between them would provide all the data that could be necessary. In other words, the form of the equation would pro- vide a complete description of the source of order in the world. The question would then arise whether there is any teleological significance in the original properties and arrangement of matter or if perchance the tendency to equilibrium alone is teleological, as seems to be the view of extreme scientific materialism. I think there can be no doubt that Hume's conclusions are hostile to the latter opinion, Hume's discussion of teleology is in many re- spects decisive. It still remains, for the man of I Loe. cit., p. 36. THE EIGHTEENTH CENTURY 53 science, on the whole the best treatment of the subject, for it is clear, specific, and single-minded. Once for all it eliminates dogmatic theology, and in several particulars it provides Kant with the material for his reexamination of the problem, according to his peculiar critical method. All this is due not so much to the novelty of Hume's thought as to an illumination which proceeds from his lucidity, his thoroughness, and, above all, his perfect honesty. Descartes had been lucid and Leibniz thorough, each had brought to the task a better mathematical and scientific equipment than Hmne's, but neither had attained to that spiritual freedom which permits the single-minded search for truth. It is important that until the last half of the eighteenth century there is no effort to separate the philosophical interpretations from the scientific results in investigating the problem of natural teleology. Science had long since won its inde- pendence, which had been demanded by Bacon and declared by Newton, himself a thorough tele- ologist. But there was no suspicion that the teleo- logical problem might perhaps be regarded as exclusively philosophical. This is true in spite of the fact that mechanics was well understood to be concerned in its investigations with mechanical causation alone. The only way to get rid of the teleological in science had been to deny the exist-.; 54 NATURE ence of teleology. To this condition Kant under- took to put an end. The Kritik der UrteilsJcraft contains Kant's ripest and most complete critical examination of teleol- ogy. In his view the belief that teleological forms and combinations exist in nature may be regarded as an aid, when the principle of mechanical causa- tion is insuflBcient, in reducing phenomena to rules. Such a standpoint, however, is an affair of the re- flective judgment, and does not concern physical science. Now this manner of studying phe- nomena, which according to Kant cannot be re- garded as the business of physics at all, has had very important results in our attitude toward nature. " For in the very necessity of that which is purposive, and is constituted just as if it were designedly intended for our use, — but at the same time seems to belong originally to the being of things without any reference to our use, — lies the ground of our great admiration of nature. . . .^ " Yet such ideas can justify a behef in the external purposiveness of nature only if we believe that which they serve, for instance mankind, to be itself a purpose of nature. Kant, therefore, concludes: " Since that can never be completely determined by mere contemplation of nature, it follows that relative purposiveness, although it hypothetically gives indications of natural purposes, yet justifies ' Kant's Krilik of Judgment, trans. Bernard, London, 1892, p. 264. THE EIGHTEENTH CENTUEY 55 no absolute teleological judgment." ' Thus the ox needs grass, and man the ox. But we do not see why the existence of man is necessary. Aristotle's conception of organization presents a different case, for here the teleology is internal, or as Kant prefers to put it, a living being is " both cause and effect of itself." ^ " Every part not only exists by means of the other parts, but is thought of as existing for the sake of the others and the whole." ' Moreover, unlike all machines, the organism possesses formative power as well as mechanism. In Kant's opinion, if the organic products of nature are only imperfectly analogous to the prod- ucts of art, on the one hand, they are, on the other hand, hardly more analogous to the organization of nature as a whole. Therefore the teleology of the living organism and of nature as a whole are both unique, and the conclusion is inevitable that organic beings alone can be regarded as absolute purposes of nature. For all other apparent pur- poses in nature are merely relative. Therefore it is only through the organism that the concept of teleology is necessarily forced upon us. This con- cept, however, naturally leads to the view of col- lective nature as a teleological system. In this system mechanism is regarded as the servant of reason and nothing is worthless or in vain. 1 Loc. cit.. p. 271. 2 Ibid., p. 274. » Ibid., p. 277. 56 NATUKE " It is plain that this is not a principle for the determinant but only for the reflective judgment; that it is regulative and not constitutive; and that we derive from it a clue by which we consider natural things in reference to an already given ground of determination according to a new law- abiding order; and extend our natural science according to a different principle, viz., that of final causes, but yet without prejudice to the principle of mechanical causality. Furthermore, it is in no wise thus decided, whether anything of which we judge by this principle, is a designed purpose of nature. . . ^" " We venture to judge that things belong to a system of purposes, which yet do not (either in themselves or in their purposive relations) necessitate our seeking for any principle of their possibility beyond the mechanism of causes work- ing bhndly. For the first Idea, as concerns its ground, already brings us beyond the world of sense; since the unity of the supersensible prin- ciple must be regarded as valid in this way not merely for certain species of natural beings, but for the whole of nature as a system." 2 " Natural characteristics which demonstrate themselves a -priori, and consequently admit of insight into their possibility from universal principles without any admixture of experience, although they carry with them a technical purposiveness, yet cannot, » hoc. ciL, p. 285. 2 Ibid., p. 287. THE EIGHTEENTH CENTURY 57 because they are absolutely necessary, be referred to the Teleology of nature, as to a method belong- ing to Physic for solving its problems. . . . Even if they deserve to be brought into consideration in the universal theory of the purposiveness of things of nature, yet they belong to another [science], i. e. Metaphysic, and constitute no internal principle of natural science; as with the empirical laws of natu- ral purposes in organized beings, it is not only permissible but unavoidable to use the teleological mode of judging as a principle of the doctrine of nature in regard to a particular class of its ob- jects." 1 After these characteristic critical delimitations of his subject Kant at length reaches the position that we speak of the teleology of nature as ^ it were designed7l5ul wiCBl the understanding that no design, in the proper meaning of the word, is in- volved. Thus ends the first division of the Kritik of Teleological Judgment. He now turns to the Dialectic of teleological judgment which opens with the two maxims of judgment: " All production of material things and^ their forms must be judged to be possible according to merely mechanical laws " and " Some products of material nature cannot be judged to be possible according to merely mechanical laws. (To judge them requires quite a different law of causality, 1 Loc. eit., p. 289. 58 NATURE namely, that of final causes)."^ Turned into objective principles these two maxims are in his opinion clearly contradictory, and one must be false. But as mere maxims of judgment he declares that they involve no contradiction in fact. Regarded as an objective principle the final cause can have no place in Kant's philosophy for the reason that a thing as a natural purpose is objectively inexplicable. " That it is not suscep- tible of proof is clear from the fact that as concept of a natural product it embraces in itself neces- sity and at the same time a contingency of the form of the Object (in reference to the mere laws of nature) in the selfsame thing regarded as purpose. Hence, if there is to be no contradiction here it must contain a ground for the possibility of the thing in nature, and also a ground of the possibility of this nature itself and of its reference to some- thing which, not being empirically cognizable nature (supersensible), is therefore for us not cognizable at all." ^ Or, more concretely stated, " How can I number among the products of nature things which are definitely accounted products of divme art, when it is just the incapacity of nature to produce such things according to its own laws that made it necessary to invoke a cause different from it ? " 3 ' Loc. cit, p. 294. 2 Ihid., pp. 307. 308. 3 /j^^ p ^^ THE EIGHTEENTH CENTURY 59 According to Kant, therefore, the most complete teleology could prove at best nothing more than that the human understanding cannot conceive a world such as ours otherwise than as the product of a supreme cause operating designedly. But he believes that we are driven into this same position even by our restricted teleological conclusions. This view, identical with Hume's, is finally formu- lated as follows: " We cannot otherwise think and make comprehensible the purposiveness which must lie at the bottom of our cognition of the in- ternal possibiHty of many natural things, than by representing it and the world in general as a prod- uct of an intelligent cause." * To the vulgar such a conclusion may seem bar- ren enough, not so to Kant, who with unwonted enthusiasm passes on to one of his most famous remarks: " It is indeed quite certain that we can- not adequately cognize, much less explain, or- ganized beings and their internal possibihty, according to mere mechanical principles of nature; and we can say boldly it is alike certain that it is absurd for men to make any such attempt or to hope that another Newton will arise in the future, who shall make comprehensible by us the produc- tion of a blade of grass according to natural laws which no design has ordered." ^ Those who have mistaken Darwin for the Newton of the blade of 1 hoc. cit., p. 312. 2 Ibid., pp. 312, 313. 60 NATURE grass strangely misconceive Kant's meaning, though not, I think, in all respects. The general conclusion of the Dialectic may perhaps be summed up as follows: Reason must not for a moment overlook the mechanistic prin- ciple. For our knowledge of nature is not at all advanced by explanation according to final causes, in that we can never know their teleological mode of action. But on the other hand we must never lose sight of the teleological. For this would make i reason fantastic, just as a merely teleological view I makes it visionary. Yet the two principles of /mechanical and teleological causation cannot be united. Though complementary, and in no proper sense contradictory, they are independent in such a manner that one method of explanation excludes the other. In short, they are heterogeneous; their assimilation in one principle can occur only in the supersensible; and regarding this we can form no determinate conception, for the principle is trans- cendent. The organism presents peculiar difficulties. Human understanding is quite powerless to con- ceive the existence of the phenomena of organiza- tion as a result of mere mechanical causation. Yet we must not on that account even here attempt to decide against the mechanical principle. For upon the assumption that such things are teleologically determined, we are still compelled to conceive / THE EIGHTEENTH CENTURY 61 them as mechanically produced. And, when we doj so conceive them, we necessarily put aside the teleological explanation altogether. This argu- ment, which has lost none of its force, is the true mechanistic reply to all vitalistic theories. Finally it is to be observed that we can never hope to determine how much mechanism does for the development of the teleological in nature, but rather must seek the mechanical and the teleologi- cal in all things. And this is the reply to blind mechanism. Unquestionably these results, important as they are in the philosophical controversies of the bio- logical theorists, may be regarded as unimportant for natural science itself. Kant has accorded to mechanical causation all that the most thorough- going man of science can ask for it — imless the scientist should turn philosopher. But science had already usurped such rights. At the most Kant merely legalized an accomplished fact. The real purpose of the discussion seems to be to secure a like independence for the teleological principle. This Kant tries to accomplish by re-i moving teleology altogether from the field of the determinant to that of the reflective judgment. Here he hopes to estabhsh it in perfect security. But here it must remain in isolation, fenced oflE from natural science by the critical philosophy itself. Yet even metaphysical fences often suffer 62 NATURE from the tooth of time. And it is very doubtful if at this point Kant's undertaking has been philo- sophically successful. For who shall say that the periodic system of the elements or the second law of thermodynamics is the concern of the determi- nant rather than of the reflective judgment ? Indeed the reverse is clearly the case. Yet no argument can ever remove such principles from natural science. Further, the fully developed hypothesis of natural selection appears to involve mechanism and teleology in a new entanglement, unforeseen by Kant's analysis and inconsistent with it. Thus he declares that nothing is gained for the theory of nature or the mechanical explanation of its phenomena by means of its effective causes, by considering them as connected according to the relation of purposes. And yet without such con- sideration the idea of natural selection could never have arisen, for only the teleological judgment tells us that such things as adaptations exist and we can only prove them to be mechanical products after we have become aware of their existence. According to Kant's view the development of existing organic forms into new shapes must always be judged to result from the purpose that is withm them and conformably to it. In such a view there is no place for the natural selection of random variations. THE EIGHTEENTH CENTURY 63 A vague conception of mechanistic evolution is, however, not unknown to Kant's penetrating mind,^ and he meets the difficulty by pushing back the origin of the teleological to an earlier period, quite after the manner of Leibniz and many others. Thus he arrives at pre-established harmony, as applied to organic evolution, and, as a means of representing evolution, at Blumenbach's views on the theory of epigenesis. But just as there is no room in his view for the development of adapta- tions through the operation of unaided chance, so a fortiori spontaneous generation is out of place: " That crude matter should have originally formed itself according to mechanical laws, that hf e should have sprung from the nature of what is lifeless, that matter should have been able to dispose itself into the form of a self-maintaining purposiveness — this he [Blmnenbach] rightly declares to be contradictory to Reason." ^ Throughout the whole course of the essay Kant struggles hard but in vain to convert his concept of universal teleology into a more synthetic view which shall somehow represent the whole of nature under the guise of an organism. At last he aban- dons the attempt and turns, where we need not follow him, to man and human affairs. Such, so far as I understand it, is the substance of this famous work. I have set it forth, not be- 1 hoc. cit., p. 338. ^ Ibid., pp. 345, 346. 64 NATURE cause it seems to be important for the present pur- pose, but because I wanted a justification for pass- ing it by. Kant has indeed revealed to us the human understanding at work upon the problems of mechanism and teleology. He has enabled us to be self-conscious in this difficult task. But in every other respect he has left the question just where Hume left it. From the problem of the teleology of nature design is eliminated, and with it theology. Organization remains. Universal teleol- ogy remains. And both have a tendency, as science advances, to recede to the very origin of things. Science has never paid much attention to Kant's speculations concerning the judgment. It disre- gards the metaphysical distinction between the determinant and the reflective judgment, just as it disregards the metaphysical distinction between space and time, on the one hand, and matter and energy, on the other. Admitting that these may be vaUd discriminations concerning the operations of the reason, the scientist takes his reason, just as he takes his own sensory apparatus, for granted. The conclusion at which Kant arrives is nothing but the metaphysical formulation, according to the peculiar principles of his critical philosophy, of a distinction between mechanical and teleological explanations which Aristotle, Bacon, and Kant's other great predecessors had with varying success sought to grasp. I have no doubt that it may be THE EIGHTEENTH CENTURY 65 regarded as a genuine metaphysical discovery. Yet from time immemorial a vague realization of this distinction has governed actual scientific re- search. Since the seventeenth century there has never been a question of mingling the teleological judgment with the researches of physical science. A sign of this development of scientific method may be found in the pious Newton's rigorous exclusion of all reference to final causes from the Principia. The concluding schoHum, " On the eternal deity by and through whom the universe exists," which was added to the second edition, reinforces this point. Further, in his first rule of reasoning in philosophy, Newton declares that we are not to assume more causes than are sufficient and neces- sary for the explanation of the observed facts.^ This rule was, I think, not formulated in order to exclude final causes, but it could have been thus stated only by one who disregarded them as a matter of course. Thus, even in Kant's day genuine physical science, as distinguished from the spurious natural theology, could have no interest in such speculations, for it had long since forgotten final causes quite as completely as had mathe- matics itself. In biology there is hardly an enduring trace of Kant's influence. The concept of function, 1 Principia, Thomson & Blackburn's edition. Glasgow, 1871, p. 387. 66 NATURE naively held as a guide in research, has weathered all storms. The physiologist in his laboratory neither thmks nor cares whether the idea is an affair of the reflective judgment. As a rule func- tions raise in his mind no more philosophical doubts than energy or the electric current. He simply goes on investigating them. Thus, to be sure, he is applying Kant's precepts, for all his investigations are rigorously physical and chemical, but he is quite indifferent to any such considerations. He stands with Harvey; his determinism, almost without exception, is that of Leibniz; and for him judgment is so httle suggestive of metaphysics that, in English, it represents sagacity rather than one of the elementary constituents of human reason. Here, as in physical science, the earher vague dis- tinction, not the refined product of Kant's meta- physics, determines the course of events. More important than these well known charac- teristics of scientific research, as a means to the appraisal of Kant's influence, is another fact. It will be remembered that when Leibniz asserted it, the principle of absolute mechanical causation became estabUshed. In truth no assertion of the principle was really necessary to its establishment, for it was already implied in the principles of dynam- ics, and it is undoubtedly an inevitable tendency of the mind.i But though mechanical causation ' Meyerson, loc. dt. THE EIGHTEENTH CENTURY 67 can readily be abstracted from all teleological views, teleological forms, as Kant thoroughly explained, always involve chains of mechanical causation. For this reason Kant has been power- less to do what Leibniz did. Though we can readily separate the mechanical from the teleologi- cal in nature, we can on no account separate the teleological from the mechanical, if we are to think about it scientifically. So, in spite of Kant, when scientific research employs teleological concepts such as function, adaptation, fitness, or natural selection, it is obliged to regard them as cognate with mechanism. And I believe that organization has finally become a category which stands beside those of matter and energy. If this be so Kant's whole position has been judged and condemned. The fact is that for science the idea of organization, like that of energy, be- comes established through a process of induction. It is today a component part of the theoretical de- scription of natm-e, unique indeed, yet thoroughly homogeneous with the other elements of that description. In short, at the one point where tele- ological concepts have always been inevitable, modem natural science holds fast the view which! it owes to Aristotle in his role of biologist, undis-f turbed by the criticisms of Kant. This destructive task of science is completed by the establishment of principles such as those of Camot and Mendeleeff 68 NATURE above mentioned, which are at once the work of the reflective judgment, according to Kant's analysis, and by common consent integral parts of physical science. BIOLOGY Historically the most striking result of Kant's labors was the rapid separation of the thinkers of his own nation and, though less completely, of the' world, into two parties; — the philosophers and the scientists. No doubt this was an inevitable stage in the movement of thought. But it could hardly have come to pass so definitely without the influ- ence of the systems of the philosophers of nature and of Hegel, both of which are Kantian in origin. The consequent reaction of science against meta- physics was extreme, and has of coiu'se had certain unfortunate effects. But even though somewhat uncritical, it was on the whole well justified and certainly most beneficial to the progress of natural science in the middle period of the nineteenth cen- tury. For the time the work of philosophy in discovering concepts and assigning to them their role as regulative principles in scientific research was done. Accordingly the history of science in the nineteenth century bears few traces, and those chiefly of an early date, which point to philosophi- cal influence. Especially the problem of teleology had received a philosophical treatment quite adequate to the 69 70 NATURE needs of the contemporary scientific investigator. He was licensed to postulate absolute mechanical determinism throughout nature. Oh that basis he was quite free to study all things from the stand- point of physical science. But yet, he was told, organization is a concept from which the biologist has no escape, and this can be thought only with the help of ideas which are teleological and not me- chanical. That which is organized, the structure, the process, is indeed exclusively mechanical, but like the idea of beauty, the idea of organization is in no sense a mechanical concept. Finally there is that in the universe, we know not what, which leads all men, the most devout and the most materialistic alike, to speak of nature. This too can on no ac- count interfere with the business of science. And yet it points to a remote possibility that we may some day find in the universe a clue which shall lead us to think of nature as a whole somewhat as we now think of the organism. I have put these ideas in a modem form so that they may be more clearly intelligible. But they were all accessible to the men of a hundred years ago. There was also philosophical ground, though no doubt quite unnecessary and uncalled for, to jus- tify the scientific study of the development of that which we regard as teleological. Kant had allowed this only with important reservations, but Hume had admitted none upon the subject. BIOLOGY 71 Under the circumstances, we may proceed to the examination of the fate of these ideas in the scien- tific history of the nineteenth century without regard to the independent history of philosophy, especially of German philosophy. In France and England no such complete hostility was ever estab- lished between the scientists and the philosophers, for the scientists of these coimtries had no bitter memories of bondage under the rule of meta- physics. Accordingly English and French phi- losophers are more closely related to nineteenth century science. Of course this depends very largely upon the fact that they did not as a rule, especially early in the century, adopt the views and methods of German ideahsm. In the end a few of the results of Hegel's speculation and other products of the German school must be considered. They have finally foimd a place in our present analysis of the teieological problem, as it presents itself to the natural scientist. But these need not be treated in their historical development. Until Lotze's time there is no further association of science with German philosophy. Lotze re- turned to a position which is nearly identical with that of Leibniz, though the estabUshment of the principles of thermodynamics had modified his manner of conceiving mechanical determination. He first established friendly relations with science by founding a criticism of the despised philosophy 72 NATURE of nature upon the basis of broad and profound scientific learning, especially in the medical sciences. He then turned to the teleological con- cepts and reexplained them to a generation of in- vestigators who had not themselves taken the trouble to analyze problems from which they could not escape. But Lotze was a somewhat isolated figure of his day, and though he influenced a num- ber of superior minds, he established no easily recognized current of thought. Indeed, so far as the elements of the teleological problem are con- cerned, he said nothing very new. A more important influence than the German philosophers was Goethe, who stands far above any of Kant's successors in wisdom and an almost in- stinctive recognition of the truth. As a philo- sophical poet of nature he was long, though not permanently, to hold the German people to a view of nature as teleological, a view which has in it something of the true spirit of the Renaissance. His scientific activities were also important. For in judgment he surpassed his scientific contem- poraries like Hmnboldt ahnost as much as in philosophical intuition he surpassed Schelling. Thus m Germany at least, where the philosophical influences have failed, Goethe's unsystematic views have had a wide influence. This has been much needed, for among Goethe's coimtrymen the fan- BIOLOGY 73 tastic excesses of materialism have quite equalled, the visionary excesses of ideaUsm. On the whole neither Goethe nor Lotze, nor indeed Mill, Spencer, or Comte, seriously modified the development of scientific thought, which now becomes our principal concern. Needless to say what is chiefly involved is the progress of biology. In the nineteenth century the concept of organization appears for the first time as an exphcit postulate of scientific research. There has never been a period when the idea of function was absent from physiological investiga- tion. And it would be an almost hopeless task to trace the transformation of this idea, with widening experience, into the larger one of organization. Provisionally, it may therefore suffice to note the conscious and dehberate use of the latter idea in- " Cuvier's Law." According to this hypothesis it is possible after a careful study of any one part of an animal, for example a tooth, to reconstruct the whole. Nothing could correspond more perfectly with Aristotle's original position concerning the organic relation between the parts and the whole. This hypothesis is to be regarded as an induc- tion from the necessary method of paleontology. Although untenable as a strict and imiversal principle, it is well groimded in Cuvier's own pale- ontological studies and in the vast fabric of his comparative anatomy. Cuvier, however, was a 74 NATURE life-long opponent of the theory of evolution. He failed, therefore, to represent his idea in its true Hght as a necessary implication of the historical continuity of the forms of life. In this later aspect it has done good service in one of the principal fields of research developed by the Darwinian theories, and thus it has never ceased to be a con- stant preoccupation of biology. In this special form the idea of organization dominates all modem biology. Physiology was more deliberate in setting up the principle, because organic activity is harder to define and to describe. At least as early as the time of Johannes Miiller the idea was clearly grasrted.i But not until the estabHshment of experimental morphology did it become overtly a guidinig principle of physiological research. One very important influence toward this result is to be found in the speculations of von Baer. He was, however, less concerned with the larger questions than with an examination of the simple concept of teleology as a necessary component of scientific thought. For him the teleophobia of his scientific contemporaries is a mere vulgar prejudice which rests upon the old mistaken notion that somehow the te^eological view interferes with the mechanical explf/nation. He is convinced that " All necessity and compulsion in nature lead to ends, and all 1 Du Bois-Reymond, Reden, u, p. 217. BIOLOGY 75 tendency to ends is accomplished solely throughi necessity and compulsion." ^ The fundamental difficulty of his contemporaries is therefore thatl they do not understand the thought of their pred- ecessors, Aristotle, Bacon, Descartes, and LeibnizJ This, in its simplest form, he adopts. But a further difficulty is involved in the ter- minology of the German language. The word zweckmdssig says too much, for it puts into teleologi- cal the idea of purpose or conscious design. And this von Baer as a man of science cannot allow. He, therefore, coins the term zielslrebig. For the recog- nition of ends in nature is no less certain to him, even as a man of science, than is that of conscious purpose illusory. This position, radically anti- Kantian, is formulated as follows: " I cannot help expressing the conviction that scientffic research, in that it establishes the existence of harmony among the different forces of nature, does and should lead us to the recognition of a general and ultimate principle (Urgrund) thereof." ^ It is interesting to observe that Hegel had long since reached a similar view.^ In some respects these opinions of von Baer have not received the general approval of men of science. But they have greatly supported the teleological 1 V. Baer, Reden, ii, p. 234. 2 IbU., ii, p. 77. ' Cf. J. S. Haldane: Mechanism, Life and Personality. London, 1913. 76 NATURE 'point of view in physiology, for his name is a great one in the foundation of modern biology. These positions being established, von Baer turns to the problem of organization, and fully expounds the necessary concepts. It is curious that he should seem to attach less importance to this idea than to that of Zielstrebigkeit. Perhaps he did not see how useful the larger concept could be in research. Nevertheless the truly Aristotelian idea of internal teleology of the organism is at the bottom of his biological philosophy,* He and Bichat are the first of the organicists.^ Their suc- cessor is Claude Bernard. This great man, whose purely mechanistic researches stand at the foun- dation of many departments of physiology, steadily exerted all his influence in favor of the idea of organization. He recognized a directive and organizing idea in the animal, and again and again insisted upon it.' Yet his analysis of the problem, like that of von Baer, was not complete. Though he, like all other physiologists, employed the idea of fimctional activity as a guide in research, though he was fully aware of Cuvier's method in paleon- tology, his just concern for the integrity of physio- ' Loc. cit., ii, p. 188. 2 It is hardly correct to derive them, as M. Delage has done, from Descartes. {L'HSreditS, p. 408.) Their descent is clearly from Aris- totle; their modification of his view the necessary result of modern physical theories. ' Iniroduclion a la medicine expSrimentale, p. 162. BIOLOGY 77 logical method beguiled him into declaring that " The metaphysical evolutive force by which we may characterize life is useless in science, because, existing apart from physical forces, it can exercise no influence upon them." * This, strange to say, is Kant's own error. It is as if one should declare that the idea of the periodic system of the elements is useless to science, be- cause, existing apart from the physical forces, it can exercise no influence upon them. What Claude Bernard well knew, but failed here to point out, is that organization, like the second law of thermo- dynamics, is a condition of those physico-chemical phenomena which were the subject of his investiga- tions. At times, however, he stated the case more correctly. During the later years of von Baer and Claude Bernard, the ideas of Darwin were accomphshing a revolution in general biology. Not the least im- portant residt was at least temporarily to establish adaptations as the most positive of realities. Yet an adaEtatiDiL. is only to be defined in terms of organization. In the orthodox Darwinian view it is thatj^bi^_con^ibu*es4»-*he.4u:^^ whole. There is nothing in its merely physical character which enables us to recognize it as an adaptation. Only its function reveals its true nature. 1 La science expirimentale, 3^me ed., p. 211. 78 NATURE In the course of time some of Darwin's original positions have been weakened and the more ex- treme views of his followers overthrown. As a result this manner of thinking about adaptation is somewhat out of fashion. But it endured quite long enough to leave its mark upon several depart- ments of the science. And it is very doubtful if any one will be bold enough ever again to put aside the idea of function itself or to deny its necessary implications. Meanwhile a nmnber of independent lines of investigation have developed from Darwin's re- searches. One of the most interesting of these is the study of experimental morphology initiated by Professor Wilhelm Roux. This subject appears to have developed, partly at least, as the reaHzation of a program of research founded upon Roux's quasi- philosophical analysis of the characteristics of life.^ Such a process is a genuine curiosity in the his- toty of science. According to Roux th e livia g ^being^ Qiay-.be'ldeBned as a natural object which possesses the following nine characteristic autono- mous activities : Autonomous change, autonomous excretion, autonomous ingestion, autonomous as- similation, autonomous growth, autonomous move- ment, autonomous multiplication, autonomous transmission of hereditary characteristics, and au- tonomous development. This conception, as Roux * Der Kampf der Telle im Organismus. Leipzig, 1881. BIOLOGY 79 admits, is closely related to Herbert Spencer's famous conception of life as " The continuous adjustment of internal relations to external rela- tions." ^ Roux's discussion of the subject was independent of Spencer's influence and, in its specification of conditions, his analysis possesses certain advantages over the English philosopher's more abstract statement. But, from the stand- point of physical science, it is gravely deficient in method and has never been regarded as more than a preliminary statement of the several physiologi- cal aspects of the fact of organization. What has given Roux's investigation its value and influence is that there is thus presented a pro- visional discrimination of organic activities as a basis for the experimental physiological study of organization itself. Thus regarded, Roux's service to biology may be seen to be both permanent and important. With the foundation of experimental morphology the problem of organization assimies its proper place in physiological research. The experimental results of the new science clearly prove that the place is secure. This department of science has developed inde- pendently, and only in recent years can its influ- ence upon the older science of physiology be detected. The physiologists, in their more abstract and more analytical researches, have usually dealt 1 Principles of Biology, revised edition, 1909, p. 123. 80 NATURE exclusively with physical and chemical phenomena. Unlike Roux's followers, they have been concerned with those things which are organized in the living being, rather than with the organization of them. Their very method of research, which proceeds from a preliminary analysis of the factors of or- ganization, has obscured the larger biological problem. At length Pavlov's researches on the glands of digestion, the study of internal secretions and hor- mones, Sherrington's investigation of the integ- rative action of the nervous system, ^ Cannon's study of the emotions,^ and many other independ- ent lines of investigation have cleared the ground, and at the present moment the physico-chemical treatment of the problem of organization is widely if somewhat vaguely recognized as the ultimate goal of physiological research.* j In the study of metabolism, which has also had an independent development, the idea of organiza- tion has long dominated research. This is due to the fact that here the concept of equilibrimn can- liot be avoided. At an early period in the history of the science it was discovered that a normal or- ganism is in a state^ nitrogen eguilibrium. That is to say, the composition, in respectof compounds 1 New York, 1906. 2 Ibid., 1915. * Cf. Mechanism, Life and Personality, J. S. Haldane, London, 1913, and my review. Science, N. 8. xlii, 378. BIOLOGY 81 of nitrogen, is steadily preserved, through the regulation of a long chain of most intricate chemi- cal processes. Day by day the ingestion of nitro- gen is approximately equal to the excretion. A modification of the diet may cause a temporary disturbance of the condition, but this is soon restored. The phenomena of growth and disease are found to involve more enduring changes. Here- upon, by a process of reasoning patterned upon that of physical science, growth is declared to involve nothing more than other phenomena superimposed upon the underlying conditions, thereby modifying the observed facts in such manner that the funda- mental state is partly obscured. And disease is after all, in its very essence, a disturbance of or- ganization; in short, diseases of metaboUsm involve by definition disturbances of equilibria, which may or may not be compensated. Further research reveals similar equilibria con- cerning carbon, sulphur, phosphorus and the other elements. The results are extended to defi- nite chemical compounds such as water, salt, sodium bicarbonate, glucose, and the like. It is perceived that the equilibria of temperature, of osmotic pressure, of alkalinity, which involve physico-chemical states rather than chemical sub- stances, are truly analogous phenomena. Meanwhile it has always been clear that within certain limits the existence of these equiUbria is 82 NATURE essential to the preservation of life itself, and that they might have been taken for granted. The real question has been to define the normal and pathological fluctuations, their duration, their limits, and their relations to other phenomena. In short, so far as these problems are concerned, the study of metabolism has consisted in an attempt to describe as thoroughly as may be, and if possible to explain, the fluctuations of the approximately constant physical and chemical conditions of the body. In other words, the task of the investigator has been to make known the facts concerning the regulation of the ultimate physical and chemical constitution of the organism. In this undertaking he has always kept in mind the idea that the organ- ism exists in a state of dynamic equiHbrium, just as it was long ago conceived by Cuvier. Now this idea of regulation, so familiar in the investigations of the temperature of the body, and in many other general problems of metabolism, is the very concept to which all the other independent investigations of organization as a physiological problem also lead. Thus Roux has long since declared, and recently reasserted * the belief, that the capacity of autonomous regulation of all nine of his elementary characteristics is quite the most important of all the pecuharities of life. For example, this is what makes possible the direct ' Die SelbstregulcUion, Halle, 1914. BIOLOGY 83 adaptation to the environment, or, in other words, the acquiring of characteristics. In like manner the action of hormones, the integrating function of the nervous system, and the phenomena of emo- tional excitement investigated by Cannon are all regulatory. It is now possible to see that Herbert Spencer's conception of life as " the continuous adjustmep.t of internal relations to external relations, "though doubtless far from satisfactory as a characteriza- tion of life itself, is really a true statement of the phenomena of organization. Vague though it may be, it is confirmed by the results of experimental morphology, of physiology, and of the science of metabolism. I should be sorry to produce the impression that this idea of regulation is well defined in general physiology. The fact is that current views upon the subject are very generally loose and perhaps in part contradictory. But it is certainly rigorously defined in some departments of the science. And it is unquestionably everywhere in use. Perhaps the most convenient definition of regu- lation is Driesch's: " We shall imderstand by regulation any"occurrence or group of occurrences in a living organism which takes place after any disturbance of its organization or normal functional state, and which leads to a reappearance of this organization or this state, or at least to an approach 84 NATURE thereto." ' This statement bears the mark of having been formulated for the purposes of ex- perimental morphology, and accordingly lacks the quantitative character that one finds in the ia- vestigations of physico-chemical regulations, such as temperature. It will, however, suJBfice to estab- lish the point that the concept of regulation is governed by that of organization. Thus, at length, Aristotle's original view of the internal teleology of the living thing, which is nothing more than self-regulation, has completely established itself in physiology. Meantime still another development of biologi- cal research, the study of animal behavior, which is an offshoot of psychology, has undertaken the task of systematically examining the phenomena of organization as they appear in the integrated activities of the individual. Among the investigations which have contrib- uted to the establishment of the principles of regulation and organization as subjects of physio- logical investigation, several have had an unex- pected result. Out of them a new doctrine of vitali^ has arisen, to trouble the even progress of biological thought. Not many years ago such a development would have been considered simply impossible. But in reality there is a more diflficult » This concept is developed in Die organischen Regulationen, Leipzig, 1901. BIOLOGY 85 riddle hidden in the fact of regulation than at first appears. And gradually, with the transition from the older purely analytical investigation to the modern more synthetic researches, this riddle has been revealed. The principal figure in the new development is Professor Hans Driesch. His thought originates in the idea, whose foundation we have just traced, that the phenomena of regulation disclose " Teleol- ogy as an irreducible peculiarity of the phenomena of life." This standpoint is quite unexceptionable. But while it has led the greater number of investi- gators merely to redouble their efforts toward a mechanistic description of the various organic regulations, Driesch has sought to utihze such descriptions as a means to overthrow the mechanis- tic theory itself. In his opinion no machine could possibly produce many of the phenomena of regulation. It may be at once conceded that nobody has ever given a mechanistic explanation of a single organic regulatory process. In spite of this, however, many simple mechanical analogues like the ther- mostat and the gyroscope are well known. As an explanation of the present state of the question we may consider the regulation of the temperature of the htmaan body. There exists an admirable de- scription of the various methods by which the tendency to a rise or fall of temperature is counter- 86 NATURE acted. Conduction, convection, and radiation of heat and the evaporation of water are all involved. These are varied according to the condition of the organism and its environment. The process is con- trolled in a most intricate manner through the operation of physiological activities such as the distribution of the blood in the different parts of the body and the intervention of the sweat glands. Beneath these conditions in turn is the regulation of the production of heat, which at need rises and falls so as to preserve the equiUbrium. Under ordi- nary circumstances the oxidation of carbohy- drate alone may be concerned in this fluctuation, but, if the supply of carbohydrate fails, other sub- stances will be employed in its stead. Moreover, all of these processes are involved in and modified by other physiological activities such as the per- formance of mechanical work. Clearly the physiologist is here confronted by an incomplete, but yet partially successful mechanis- tic account of one element of the functional organization of the hxmian body. But how is the process governed ? By what chain of mechanical causation does a fall in the temperature cause a rise in the rate of oxidation .? This question, in turn, is not altogether beyond the scope of our present physiological investigations. But sooner or later, when the problem is studied, we come upon the fact that a certain organ or group of cells accom- BIOLOGY 87 plishes that which is requisite to the preservation of the equilibrium, varying the internal conditions according to the variation of the external condi- tions, in a manner which we can on no account at present explain. The same difficulty is encountered in the analysis of every other organic regulation, of whatever sort. There is no physiological phenom- enon of regulation the autonomy of which we can today understand. This is Haldane's ground for the rejection of mechanism. The distinction here involved is by no means easy to grasp. In the qualitative researches of experimental morphology, from which Driesch's speculations arise, the difficulty of understanding how our knowledge is restricted becomes magni- fied. But it remains even in the quantitative in- vestigations of physico-chemical regulation. One- thing is evident, we possess no good device for imagining a cell at work, and until we can do this we shall never know just what to think about any regulation. This, at least, is a fact on which mech- anists and vitaUsts can agree. It may, perhaps, be said that in performing such functions as adjusting the regulatory processes, ceDs seem to act as if they were controlled by something which remotely resembles intelhgence, but which is really far superior in efficiency, in that it operates necessarily, according to the needs of the moment, without the guide of previous expe- 88 NATURE rience, and without those errors of judgment which are all too common in voluntary action. Such is the entelechy of Driesch. Driesch's effort to prove the existence of entelechy in the organism culmi- nates in what he regards as a demonstration that mechanism is necessarily unable to determine some of the phenomena of organic regulation. In the absence of any clear understanding of the operation of cell mechanisms, such an effort is, 1 think, clearly in vain. It may carry conviction to those who are already predisposed in its favor, but no one else can accept the argument, and an opponent will always regard it as worthless. The opponents of Driesch are in certain respects better equipped than he is for the controversy, because they are able to appeal to the authority of general principles of science. There is, for instance, the theory of natural selection. That has already had a large share in overwhelming the older vitahs- tic speculations. For it undertakes to reveal the development of all the most complex forms of life out of the simplest forms, as the result of a purely mechanical process. And whoever accepts it may be disposed to regard the original organism itself as trapped in a stable form, according to that prin- ciple of the survival of a dynamic equilibrium which Hume had recognized. ^ ' Above, p. 49. BIOLOGY 89 From such a beginning Darwin's evolution might proceed. This appears to be the view of Roux. Hume's idea is, indeed, just as appHcable to any other material system as, in a refined form, it has proved to be to the organism. But this criti-' cism of Driesch is not decisive. In the first place it begs the question, for the nature of regulations is at issue in the problem of evolution. Further it is now evident that we are not justified in regarding our existing mechanistic theories as suflScient com- pletely to explain the evolutionary process. And finally, any theory about the origin of life is nothing but an unfounded guess. But there is a far more formidable objection to vitaHsm. More explicitly than ever before the modem principles of physical science seem to com- pel us to recognize absolute mechanical necessity in all things. We may not imderstand organic regu- lations, or organic evolution, or the origin of life; in fact we are still imable with the necessary clear- ness to represent to ourselves the structure of a cell; yet these are at least phenomena. As phe- nomena they are subject to the laws governing all phenomena, that is to say to the two laws of ther- modynamics. For the laws of conservation and degradation of energy have long since supplanted Leibniz's rudimentary idea of the conservation of vis viva, as the ground of our conception of neces- sary causation. 90 NATURE Now the laws of thermodynamics contemplate all forms of activity, not merely mechanical force, but heat, light, electricity, and the rest, in all kinds of material systems, not merely among moving masses, but in gases, electrical machines, steam engines, and all others. We have a good deal of experimental evidence that they hold for the living organism. And everywhere they are beUeved to involve that same absolutely necessary mechanical determination which Leibniz had postulated upon a more slender foundation. It is important to understand the foundation of the belief in mechanical determination, which seems to be as follows: The world of physical science consists of matter and energy existing in space and time; in any particular case these four things are always to be represented by mathemati- cal terms which are functionally related together in the equations expressing the laws of physical science. This functional relationship, although often unknown, is believed to be rigorously and unequivocally determined by the laws. There- fore, in accordance with the laws of conservation and the second law of thennodynamics, the non- mechanical, i. e. any factor which is non-material, non-energetic, non-spatial, and non-temporal, can- not enter into or modify any physical or chemical process. Thus " vital " processes can no more modify mechanical determination than mechanical BIOLOGY 91 processes can modify geometrical determination, and mechanism is conceived to be no less absolutely a condition of life than geometry of mechanics. ! Driesch is not immindful of this difficulty. As a way out he suggests that entelechy may, perhaps, j operate by suspending, as occasion requires, the/ operation of the second law of thermodjoiamics. This theory is ingenious, but I believe untenable. In fact it involves a reduction to the sphere of molecules of the old fallacy of Descartes. For to suspend the operation of the second law of ther- modynamics would be precisely equivalent to an alteration, without the expenditure of energy, of the direction of motion of the particles of a mate- rial body. Under these conditions an object which had fallen to the ground might, by cooling itself, rise again into the air. Nothing could be more radically inconsistent with the fundamental prin- ciples of physical science, as now generally ad- mitted, than this assumption or the theory which it is designed to support. Driesch's discussions also extend to voluntary action, which is the most familiar means to the foundation of vitaUstic theories. But, except in one important respect, the ground for a conclusion favorable to the vitaUstic hypothesis is here identi- cal with that in the case of the organic regulation. The difference consists in the fact that we know the mind, but not the entelechy. In fact every- 92 NATUEE body is conscious of his ability at will to change the course of mechanical events in the world around him. No conviction of the truth of the principles of physical science, however firmly grounded in scientific and philosophical criticism, can ever eradicate this belief. As a practical guide in daily life it is perfectly indispensable. Yet there seems to be a confiict between the behef and the prin- ciples of thermodynamics. It is a strange irony that the principles of science should seem to deny the necessary conviction of common sense. For was it not a similar denial of the external world that led men of science to their most contemptuous rejection of metaphysics ? And if Berkeley's idealism is to be rejected in so far as it denies the real existence of our tables and chairs, then the principles of thermodynamics must be rejected in so far as they deny the justice of our common-sense idea of voluntary action. But another kind of consideration further com- plicates this problem, for the common-sense notion of voluntary action is badly in need of analysis. The most elementary of psychological considera- tions leads us to the perception that our choice of action, even though it be conceived as the result of no mechanical process, is yet far from free. Not only does our whole past experience provide us with the material for choice, but it rigorously limits the choice. In spite of ourselves we do always act in BIOLOGY 93 character, and the suspicion cannot be escaped that all mental processes may be unequivocally determined. If this be admitted we have once more arrived at the conception of a world in which all is in some sense absolute necessity. Yet even then the psycho-physical riddle per- sists. In truth it is undiminished, for how does the psychical determination get its effect in the physi- cal world ? How does an idea change the course of events ? Is the relationship a mere illusion ? Is there merely a pre-established harmony between the two absolutely independent worlds of mind and matter ? That common sense cannot tolerate such a view is proved by the experience of more than two centuries. But physical science seems to deny the possibility of any other theory, unless we admit that mind is a mere epiphenomenon upon a mechanical process in the nervous system. The latter process would then become a part of the chain of physical causation and the difficulty would be removed from physical science. But what becomes of the biological function of con- sciousness, upon this assumption ? Consciousness was never produced in the process of evolution merely as an impotent accompaniment of reflex action. Obviously there is a necessary postulate of biology which declares that in its simplest form the function of consciousness is a regulation of the reflex processes in such a way as to modify in a 94 NATURE peculiar manner the chain of mechanical causation. The question is to discover how, if at all, this can be itself a mechanistic process. The vitalistic reply is that a mechanistic explanation of the phenomena of the most trivial voluntary action is impossible. Driesch's view of this argument in favor of vitalism is as follows: "Any real action is an individual answer to an individual stimulus — founded upon the historical basis. " And this individual correspondence, occurring upon an historically created basis, cannot be under- stood as a case of mechanical causality. For there is not a ' sum ' on the side of the stimulus that corresponds to a ' sum ' on the side of the reaction, and, further, not even the possibilities of acting are in any way ' preformed.' " From this point of view, the brain and the nervous system appear as nothing but a necessary means for putting the ' acting ' factor in connexion with material nature, but they are not themselves the acting factor." * This result appears to be substantially identical with the conclusion of Hobhouse's careful analysis of the same problem: " In a simple purposive action — where I re- quire a book which I remember to have left in a particular place and go to fetch it, my memory, which, mechanically interpreted, must be some 1 The History and Theory of Vitalism, London, 1914, p. 213. BIOLOGY 95 deposit of the effect of my previous dealing with the book in my brain, is so combined with my need and my physical surroundings as to discharge in succession the actions appropriate to fetching the book. This deposit — complex enough in that it must have its exact point to point correspondences with the several physical relations of the rooms of the house, etc. — is only one among the millions of deposits that my experience has formed. Yet provision must be made for selecting it out of them, and bringing it, and none other, to bear upon the physical tension, which may be supposed to correspond to my felt need, and thereby to effect the successive discharge of a complex series of actions. If we try to formulate a general plan for effecting such selection and correlation, we find ourselves speaking of a state of want, picking out from experience whatever is relevant to its satis- faction, and guiding action accordingly. But though we might find terms other than these which would avoid all reference to feeling or conscious- ness, the explanation would imply that there exists a something determined in its actions by their rela- tion to their results, i. e. something purposive. Abstract the notion of the relevancy of the means to end, and the bottom of the whole proceeding! timibles out. In short, in the activity which we claim as purposive, we find repeatedly that one factor of our life (e.g. an experience) may bej 96 NATURE brought to bear upon another (e. g. a want) in a manner that varies indefinitely from case to ease. The only principle uniting the otherwise unique combinations is that of the relevance of the com- bination to the end. Admit this principle, and we recognize a structure determined by purpose. Deny it, and we have no general plan to explain the unique combinations. Either horn of the dilemma excludes mechanism. " The denial of purposive causation, therefore, is not suggested but repelled by general experience, and owes its existence only to the theory that everything must act by mechanical laws. But this theory is a pure assumption, which derives its apparent cogency from confusion with the quite different principle that everything must act in accordance with some law. The leading mechani- cal principles I take to be adequately proved for mechanism, and, therefore, for any structure which is purely mechanical. Now the organism is a physical structure, but to assume that all its actions conform to mechanical laws is to assume that it is a physical structure only. Consciousness directly informs us that it is more than this — that it is ... a psycho-physical whole." ^ At this point I cannot help thinking that Hob- house has momentarily lost his grip on the argu- ' Hobhouse, Development and Purpose. London, 1913, pp. 325, 326. BIOLOGY 97 ment. Consciousness does indeed inform us that the organism is more than a physical structure; no doubt it is a psycho-physical whole. Accordingly some of its actions do not, strictly speaking, con- form to mechanical laws.^ An instance of this is choice or any other psychical activity. But, even so, it involves a further assumption to assert that the 'physical activities of the organism even when parts of psycho-physical activities can ever be explained as not in conformity with mechanical laws,2 Countless attempts so to represent them have been made and they have all failed. What Hobhouse seems to mean is that, if we are to understand the true nature of organic activity, we must not consider the physical structure at aU. In short we must adopt Haldane's view that the concept of organization somehow excludes or elimi- nates that of mechanism and is inconsistent with it.' This seems to be imphed in a later statement: " In this account the living being is regarded as a system of what must be called forces, in which mechanical relations are qualified by teleological relations. When these two sets of relations are hypostatised as Mind and Body they become two substances, and in place of a system whose mode of action as a whole departs from that of mechani- cal systems in virtue of its specific quality, we have ' But see below, p. 113. ^ Cf. the relation of geometry to mechanics: mechanics is more than geometry, but never ungeometrical. 98 NATURE the problem of interaction between two distinct and separate systems, each with laws of its own. If interaction is admitted, we have the conception of body as a purely mechanical system, whose operations at a certain point come plmnply to an end, while at another point they as plmnply begin, the intervening stage being filled by actions within the other system. Body is thus a purely mechani- cal system which does not conform to laws which, it is not denied, are adequately proved for mechani- cal systems. To escape this conclusion it must be admitted that Mind exerts force and is acted on by force. But Mind was precisely the concentrated essence of that which is opposed to force. Thus the contradiction of a purely mechanical system which does not act mechanically is balanced by the con- tradiction of a non-mechanical system which does act mechanically. To escape from this dilemma the Parallelistic scheme is propoimded, according to which the mental and the bodily riui on side by side in point to point correspondence, but without interaction. This scheme, however, in effect renders the mental element superfluous. A complication of mechanism is all that is required to explain the actions of livmg beings. On the other hand, the rise of the psychical stream in coincidence with a certain point of the physical, and its disappearance at another point, are left unexplained." ^ ' Hobhouse, Development and Purpose, p. 329, note. BIOLOGY 99 So far as I can understand the problem, Hob- house's argiunents and those of Driesch before him against the mechanistic character of the mental processes involved in voluntary action are at present unanswerable. But, in view of our present ignorance of the underlying phenomena, they may I fairly be regarded as inconclusive. On this subject we simply do not know what we are talking about. No doubt these processes have their physical basis, but the fact remains that science, like philosophy, cannot regard thoughts as the activities of material systems. All attempts that have been made in this direction are unworthy of the slightest considera- tion. Nevertheless biology is obliged to assert that ideas, whatever the philosopher may think of them, at least have a function, and that function, physiologically considered, can only be to regulate action. Thus we come to the conclusion that ideas, which are nevertheless non-material and non- mechanical, do change the course of mechanical processes. We may hope that in time this difficulty will somehow be circumvented. Meanwhile I think it is true that Hobhouse's own resolution of the difficulty, like that of Lotze or of Leibniz, is unacceptable to science. K this be so we are confronted by a genuine paradox. Looking at physico-chemical phenomena on the one hand, we declare that the principles of thermodynamics fully prove absolute mechanical 100 NATURE determinism in all material systems of whatever nature. Turning then to the phenomena of volun- tary action, we can see no escape from the view that, if determinism be imiversal, it is not at least always in the strict sense mechanical. The only principle of determination in the sequence of events seems to be teleological. Who indeed can so far forget common sense as to deny this in the case of any plan ? Here is a stark contradiction. This psycho-physical paradox is one of the most tormenting that the human mind has ever con- structed, and countless efforts have been made to escape from it. One of the most curious of these is the tychism of Charles Peirce. According to this idea the laws of nature possess not an absolute but only a statistical character. ' Not even the laws of conservation are absolutely true, but they are only approximations. In this approximate character there is found the possibility of a belief that the psychical may impinge upon the physical; that mind may move matter. A somewhat similar theory, but more m accord with the ideas of most men of science, has been propounded by the emi- nent mathematical physicist, Boussinesq.^ This is developed from the theory of singular integrals in differential equations. And the conclusion is 1 This, of course, is Maxwell's view of the second law of thermo- dynamics. 2 Conciliation du veritable dStermination mScanique avee Vexistence de la vie et de la libertS m&rale. Paris, 1878. BIOLOGY 101 reached that mechanical processes are conceivable which arrive at situations where a further progress in one or another direction might be determined without the expenditure of energy. Thus the mind, even though it has no energy to expend, might determine the outcome of such a process. This idea may be regarded as a further develop- ment from the theory of Descartes on the basis of Leibniz's criticism. The principal conclusion is thus stated: " The equations of motion of the organ of thought admit of singular integrals; and for geometry these integrals are the expression of the influence of morals upon physics. In this mysterious field two coexistent orders which are perceived as quite distinct — on the one hand the geometrical or material order extended in space, on the other the psychological or moral order com- prising that rich web of sentiments, thoughts, and volitions whose interconnections and successions constitute the marvelous spectacle of our inner life — correspond and touch each other. It is in this field, the only one where it may set foot without ceasing to be free, that the mind, deprived of all material force, succeeds in ruling the world of material things. Here it directs and conquers the blind forces which struggle for dominion, by setting them against one another. Here it modifies the geometrical order of objects, without being obliged to find in their actual state the principle of its 102 NATURE determinations, but rather guiding itself by a pre- vision of a future which exists only for the mind, and realizing plans ideally conceived in view of a desired end." ^ This theory rests upon a mathematical formu- lation of that view of contingency which from Cournot to M. Boutroux has been so prominent a feature of French philosophical thought. There can be no doubt, if the mathematical analysis is soimd, if singular integrals are indeed possible for the unknown differential equations which mathe- matically express any of the phenomena of the central nervous system, that a possible escape from the psycho-physical diflficulty becomes con- ceivable. But it must be remarked that the peculiarities of an equation cannot help us to imagine mind operat- ing upon matter. And it may be asked if the theory does not prove too much. Is it not de- structive, ideally regarded, of all mechanical deter- minations ? These, however, are not questions which now concern us, for they are not involved in the history of the teleological problem. The ideas of Boussinesq, like those of Peirce, have not yet exerted an appreciable influence upon thought. It ' Boussinesq, op. cit., p. 60. See also for two less consistent theories Cournot, TTaiti de T enchainemeni des idies fondamentales dans les sciences et dans I'kistoire, Paris, 1861, i, ch. 4, and Saint- Venant, Comptes Hindus, Ixxxiv, 419. In the work of Cournot there may be also found an excellent statement of the concept of organization. BIOLOGY 103 is even hardly possible to guess whether the neglect of them may be due to some radical defect which makes them incompatible with scientific thought, or whether they may have fallen into obHvion because those whom they should have interested were imable to understand them.^ Thus we come back to two counter propositions, which were rejected by Kant because contradic- tory, as an expression of the conclusion of two lines of scientific thought in the nineteenth century. " All production of material things is possible according to merely mechanical laws." " Some production of material things is not possible according to merely mechanical laws." ^ These propositions are, indeed, contradictory. But I think there can be no hope of an immediate generally acceptable decision between the two. It is well estabhshed that a study of physical science nearly always leads to the first, and that few men can escape the second when, like the historian, they study human actions. At present there seems to be no way open to science of further investigating the question. Conceivably the ideas of Boussinesq and Peirce, foreign as they are to orthodox scienti- 1 Clerk Maxwell, whose qualifications for the task were the very highest, has discussed this question of freedom in a little essay which may be found in the Appendix. It is to be observed that, though he reaches no conclusion, his discussion revolves about the concepts of singularity and statistics. ^ Kritik of Judgment, pp. 294, 295. 104 NATURE fic thought, may some day lead to a novel develop- ment, and of course no one can foresee the new thought of the future. But the counsel of discretion is to leave the question as it stands, and to turn to other matters. This is the more suitable since, as we have seen, the study of psycho-physical phenomena leads, not to indetenninism, but to a new determinism in which voluntary action is thought to be no less subject to law than inorganic phenomena them- selves. Whatever our metaphysical views, this we are obliged to admit as a necessary postulate of scientific research. For we can on no account think about the phenomena except on the assiunption that even the most casual of human actions would again necessarily recur if all the conditions which preceded it could be perfectly reestabhshed. The mind simply cannot escape the necessity of operat- ing in this manner. Even supporters of free inter- vention admit so much.i I shall not however, seek systematically to estabhsh the proposition. It will suffice to note that the whole tendency of psychol- ogy is in this direction and that it is generally allowed by the vitahsts. Moreover, I think it is evident that the operations of Driesch's entele- chies no less than the laws of gravitation would suffer from purely chance occurrences, and this seems to be their author's own view. ' Ward, The Realm of Ends. Cambridge, 1911, chap. xiv. BIOLOGY 105 The question which remains is, therefore, the old problem of the teleology of nature as a whole. Each advance of the scientific description, Newton's Principia, Camot's Reflections, The Origin of Species, or the concept of organization, refers some aspect of things as they are to the earliest conceiv- able state of the imi verse. Certain things still seem to have originated quite inexplicably during the course of evolution. Such are life and consciousness, to say nothing of histori- cal events. But the whole tendency of science is either to destroy the novel character of the prod- ucts of nature by discovering how they did really originate through necessary processes, or else to regard them as contemporaneous and coexistent with the imiverse itself. We cannot doubt that this process wUl continue. It is not restricted by the doubts which we have just reviewed, and it does not directly touch certain ethical and philo- sophical problems which cannot be avoided in the vitaUstic controversy. It does not depend upon any particular way of looking at natm-al phenom- ena. For it is nothing more than an expression of that general principle of continuity, which from Galileo's discovery of inertia till today has gov- erned all scientific thought. In the course of this movement of thought Driesch's "dynamic" teleol- ogy of vitalism loses itself in the larger problem of the " static " teleology of nature, and Bergson's 106 NATURE Slan vital, if it be admitted, becomes a question of detail. I camiot think this altogether a misfor- tune, for as Professor Bosanquet says: " Purpose only means, prima fade, that, using consciousness in the widest sense, some creature consciously wants something. But . . . does the something lose its value when it is attained ? " . . . " Things are not teleological because they are purposed but are purposed because they are teleological." ^ ' The Principle of Individualiti/ and Value. London, 191S, pp. 136, 137. VI NATURE Infinitely curious and varied is man's attitude toward nature. The savage, the craftsman, the sailor, the artist, the philosopher and the scientist each contributes to it; yet all seem to advance toward a common understanding. There is no such agreement upon any other great subject in the whole domain of thought, in all the manifold forms of hmnan expression. Coumot has com- posed a rhapsody upon this theme: " Men early felt the need of a term to designate that hidden power which maintains the cycle of life; to repre- sent it in possession of those attributes which vital phenomena reveal to us, but without a mingling of other ideas suggested by phenomena of another order, such as consciousness of our moral personal- ity, of our reasoned conclusions, of a moral law which governs them, of good and evil. The term which they employ for this ptirpose is nature taken actively {Natura naturans, as the schoolmen had it) : an indispensable term, which corresponds to an idea so well determined, yet so hard to define, that we see all the world making use of it, the believer and the skeptic, the philosophers of all sects and the learned of all schools, those who profess the 107 108 NATURE grossest materialism and those who enshroud themselves in the cloudiest mysticism alike. There must indeed be a reason for such an agreement, and this reason is the need of distinguishing and mark- ing off that which equally impresses every one, that iwhich every one feels obliged to recognize, to whatever philosophical or religious system his reason or his faith may attach him. It is as if there were a territory whose neutralization had been prescribed by a common interest, in order to carry elsewhere the ardors of war. Whether we believe in a supernatural providence which in its goodness and justice rewards and punishes, which yields to prayers and repentance, or reject this consolatory dogma, still must we admit that in the visible world, save for hmnanity, the action of the supreme cause only manifests itself deprived of such moral attributes, as it suffices for a world where morality has no place. " The idea of nature is the idea of inexpressible divine power and divine art, beyond comparison or measure with man's powers and industry, impress- ing on its works an intrinsic character of majesty and grace, yet operating under the sway of neces- sary conditions, tending fatally and inexorably to an end which surpasses us, yet in such manner that the mysterious chain of finality, whose origin and term we cannot scientifically demonstrate, appears to us as a guiding thread, with whose help order NATURE 109 introduces itself into the observed facts, and we find the trace of the subject of our investigations." ^ Is it indeed vain to seek an explanation of the order of nature beyond the laws of nature's uni- formity ? So it would appear to one who regards but the surface of things. Only the poetic philoso- pher like Cournot or the philosophical poet like Goethe seems to find something more. Yet posi- tive thought can never rest in the face of such a question, and I think that it has fotuid a clue. But if we are to grasp this clue we must ascend to a region of colder thought. Lacheher is one of the most notable successors of Cournot in France. He is known for his brief essay on induction, but has produced little beside to in- dicate the originality of his mind. His essay con-; sists in a metaphysical examination of the problem why nature is such that the inductive process lays bare our scientific laws. And he reaches a novel conclusion. In his opinion the fundamental axiom of induc- tion is that in living beings as in all material objects the conditions of the existence of phenomena are absolutely determined. Accepting this view it is easy to see how we can pass from the fact to the law. For the conditions of any case must then be identical with those of every case of a phenomenon. 1 Traits de V enchainemml des idSes fondammtales dans les sciences et dans Vhistoire, Paris, 186L, vol. 1, pp. 497, 498. 110 NATURE But this is not all, for in addition to the laws which we thus recognize there is also the law of 2^^' iza.tion. And, as Lachelier believes, a similar principle of order is to be seen in the inorganic world. " The conception of laws of nature seems, therefore, to be founded on two distinct principles: one by virtue of which phenomena constitute series, in which the existence of the preceding de- termines that of the succeeding; the other by virtue of which these series in turn constitute systems in which the idea of the whole determines the existence of the parts." (In LacheHer's opinion this is especially to be seen in chemistry.) " In a word, we may say that the possibility of induction depends on the double principle of efficient causes and final causes." ^ This idea is also stated on the ground of a dis- crimination between the existence of serial unity or causal enchainment and unity of system or har- monious unity in nature. ^ If I rightly understand the idea, Kepler's first and second laws, considered with regard to one planet alone, would be an illus- tration of serial unity, Kepler's third law might be, and the periodic classification of the elements certainly would be an illustration of systematic unity. ^ Lachelier, Dufondement de Vinduction, Paris, 1871, p. 16. 2 Ibid., p. 83. NATURE 111 Again the idea is put forward as follows: " While the mechanism of nature fills up, by a continuous evolution, the infinity of time and space, the final- ity of this same nature, on the contrary, concen- trates itself in a multitude of systems, which are indeed distinct, but yet analogous to one another."' But, moreover, every phenomenon is in fact mechanically determined, not merely by those phenomena which precede it in time but also, as Lachelier points out, by all those which accompany it in space. Without following Lachelier in his more strictly metaphysical discussions, we may note a final ob- servation, that, if finahty is in all phenomena the hidden spring of mechanism, there is nothing in the formation of an organism which exceeds the ordi- nary powers of nature, and which requires the interference of a special principle.^ The essential idea of Lacheher's essay seems to be that the chains of causation in nature weave themselves into an inteUigible pattern, and that this pattern, quite as much as the chains of causa- tion, is the subject of om- scientific investigations. Through this, as Coumot has said, " We find the trace of the subjects of our investigations." Now there can be no doubt that this is true, at least in part. For Hume's idea of the survival of ^ Loc. cit., p. 90. * Cf . The Fitness of the EnviTonment, p. 300. 112 NATURE dynamic equilibria is but an example of such a process, and it meets all the conditions of Lache- lier's analysis. The ideas of this essay on induction are some- what similar to the central points in Lotze's phil- osophical system. They appear to be even closer to certain opinions now held by Professor Bosan- quet.i The English metaphysician is concerned to reveal the error of those who " rest the case of teleology within the universe exclusively on the capacity of finite consciousness for guidance and selection." ^ This of course, though contemplating mere psycho-physical action, is to found a discus- sion exclusively on the logic of the extreme vitahs- tic position. In Bosanquet's opinion such a philosophy " is going near to destroy the idea of the reign of law, and to enthrone the finite subject as the guide and master of nature and history." ^ But " it is vain to look to the bare fact of conscious purpose for the essence or significance of teleol- ogy." ^ In truth not conscious purpose but uni- versal determination is essential to the existence of a plan in nature. For " plan involves determinate- ness, and determinateness continuity, and that in all directions. Everything must be followed by something — must be continued by something on 1 " The Meaning of Teleology," Proceedings of the British Acad- emy, ii, p. 235. April 30, 1906. 2 Ibid., p. 235. 3 Ibid., p. 235. < Ibid., p. 236. NATURE 113 every side, and between any two somethings within a unity there must be a determinate inter- connexion, prescribed by the content of that unity." 1 " Mind and individuahty, so far as , finite, find their fullest expression as aspects of very complex and precisely determined mechanical systems. This is the law, I believe wholly without] exception, for every higher product of hmnan souli and intelligence, and also of cosmic evolution. The mechanical appearance must be granted to be imi- versal and imbroken." ^ Yet from this point of view " we can freely suppose the world-plan to be immanent in the whole, including finite mind and also mechanical nature." ^ " It is impossible . . . to treat part of the world as primary and part as a secondary superstructure. We must interpret the nature of nature as much by the flower as by the law of gravitation. If we come to that, there are appearances, which we cannot on any sound prin- ciple refuse to call teleological, in the most direct and simple reactions of mechanism." * "... the foundations of teleology in the imiverse are far too deeply laid to be accounted for by, still less re- stricted to, the intervention of finite consciousness. 1 Loc. cit., p. Si38. '^ Ibid., p. 240. This is of course not meant as an assertion that the operations of mind are to be regarded as physico-chemical proc- esses, indeed the term mechanical is here used in rather too general a sense for the purpose of scientific analysis. 3 Ibid., p. 240. * Ibid., p. 241. 114 NATURE Everything goes to show that such consciousness should not be regarded as the source of teleology, but as itself a manifestation, falHng within wider manifestations, of the immanent individuality of the real." i " The contrast, then, of mechanism with teleol- ogy, is not to be treated as if elucidated at one blow by the antithesis of purposive consciousness, and the reactions of part on part. It is rooted in the very nature of totality, which it regards from two complementary points of view, as an individ- ual whole, and as constituted of interreacting members. Of the two points of view, it is impos- sible for either to be entirely absent. Assuming this impossibility to be possible, a total failure of mechanical intelligibility would reduce the spir- itual to the miraculous, and destroy teleology, as a total failure of teleological intelligibility would reduce individuality to incoherence, and annihilate mechanism." ^ As philosophical doctrine of the present day I can see no escape from Bosanquet's conclusions. At the very least it is a necessary postulate of science " that the mechanical [i. e., naturally deter- mined] appearance must be granted to be universal and unbroken." However interesting may be the organism as such, however alluring the vitalist's conception of the world, these, without determin- » Loc. cit., p. 242. 2 IMd., p. 244. NATURE 115 ism, are no secure foundation for a philosopliy. Rather are they greatly involved in the rapid movement of scientific thought. There is indeed in the concept of organization that which has defied time and change, and endured from Aristotle to our own day. But the organism is now imder in- vestigation. Year by year we see more clearly, in accordance with elementary physical concepts and quantitative measurements, what is the nature of this harmonious imity. The advance of science has assuredly not made the origin of life easier to imagine, or even to think about. On the contrary I am fully persuaded that it has made the task far more difficult. Least of all does it lead us unduly to prize those analogies between organic and inorganic phenomena that have been so much discussed. The growth of a crystal and of a living body are less similar than the growth of a bank accovmt and of a great com- mercial " organization." The dynamic equilibria of life and of a whirl-pool are entirely unequal in complexity and in the very essence of the physical and chemical processes by which they are adjusted and controlled. Yet it is quite impossible to escape from the idea of living things as natural products, for science involves determinism and determinism imposes this very concept. With the increase of our knowl- edge of organization we see ever more clearly the 116 NATURE interdependence of all living things and the har- mony between the organism and its environment. This leads us to a conception of the organism as intrinsically a part of nature and so to the idea of nature as a whole. The essential feature of Cour- not's position regarding the necessity of thus hypostatizing nature is today better than ever before justified by science. And thus the problem of the teleological form and behavior of the or- ganism merges in the larger question of the order of nature. Nothing can oppose the tendency toward this idea; it is the modern echo of Aris- totle's thought, which made him seek " the charac- ter of the material nature whose necessary results have been made available by rational nature for a final cause." ^ Thus we arrive at a clear philosophical conclu- sion. But science can never accept this result until it has been founded upon the scientific evidence by a process of scientific reasoning. And it is only too apparent that progress in this direction has been scarcely perceptible. We do indeed scientifically recognize the truth of Hume's concept of the tendency of dynamic equilibria to survive. The living thing itself is one example and Newton's Principia gives a full account of another. But though Newton himself and many others have not failed to form teleological inferences from such 1 Above, p. 17. NATURE 117 facts, these have never been generally accepted as scientifically valid. However teleological may be the appearance of the products of nature, the teleology of nature itself cannot be scientifically established unless some kind of connection, con- ceivable only as teleological, can be shown to exist among nature's laws. Lacheher has imagined such a relation and em- ployed it as the foundation of his philosophical thought. But it is very doubtful if science can ever thoroughly establish such a proposition. The exhaustive examination of all the laws of nature from this or any other point of view is quite incon- ceivable, if for no other reason, because we shall never know them all. And perhaps science can never decide whether the organic and the inorganic are ultimately to be philosophically conceived as a single order, for the task of scientific synthesis will never be completed. Nevertheless we may now see that the whole movement of scientific and philosophical thought upon this subject does lead to a more modest scien- tific problem. For if it be quite inconceivable that science should ever completely solve the riddle of the order of nature, it is clear that nothing but the inherent diflSculty of scientific research is to hinder an inquiry, step by step, into the problem. In biology this question has long been recognized and efforts to imderstand the origin of life, as well as to 118 NATURE account for the process of organic evolution, have followed. But at this point of attack the difficul- ties are almost insurmountable. Therefore, in spite of Darwin's great labors, we remain largely in ignorance. Apart from the imperfect generaliza- tion of natural selection and the rudimentary beginnings of a science of heredity, we still have but the vaguest ideas concerning the development of living things as products of nature. And regarding their origin we have no ideas at all. The simpler and more general problem of the teleology of nature as a whole has been neither recognized nor investigated by science. Yet the problem is now clear enough. All men admit in the teleological appearance of the world something that is real. There is order, stability, and a richly varied collocation of material objects at the basis of it. When we think of the solar system, the meteorological cycle and the organic cycle we distinguish that which quite inevitably and directly impresses us as harmonious. Now, as we have seen, it is no longer permissible to doubt that this impression of harmony corresponds to an order in the universe. No doubt science must put aside the philosophical problems which thus arise, and phi- losophy must deny to all men the right to f oimd a system of natural theology upon the fact. But it is a false and discredited metaphysical hypothesis which leads to the denial of the order of nature as a NATURE 119 subject of scientific research. How then is the production of this order to be scientifically ex- plained ? What is the mechanistic origin of the present order of nature ? Only if we turn to the facts concerning the evolu- tion of our solar system and of the earth can we investigate the problem. But in following this as a special case of the whole cosmic process we are in danger of bewilderment. The natural history of the earth involves a mass of particular facts which are not yet well coordinated and can seldom be referred to the laws which govern them. If, however, we seek a more general and abstract point of view, we find a clearer issue. This process of the evolution of our world, however manifold in its details, is at least governed and directed by the general laws of physical science. It cannot be doubted that others among them beside the tend- ency to formation and survival of stable systems, as formulated by Newton for dynamics, by Darwin for biology, and by LeChateHer for physical chemistry, are intelUgibly concerned in the pro- duction of the order of nature. In like manner the properties of matter and energy are concerned. It is clear, therefore, that the real scientific problem may be approximately solved by discovering, step by step, how the general laws of physical science work together upon the properties of matter and energy so as to produce that order. Thus, and thus 120 NATURE only, can we iinderstand it. Speculation a priori on such a question is in vain; only the scientific in- vestigation can reach a result; only this scientific result can determine the importance of the question for philosophy. Somewhat vaguely, from the biological point of view, I have already discussed one aspect of this problem. 1 In the following pages the question will be investigated more rigorously and systematically according to the principles of physical science. ' The Fitness of the Environment, New York, 1913. " The Func- tions of an Environment,'' Science, n.s., xxxix, p. 524. VII EVOLUTION In the history of thought there is, beside the es- tablishment of the second law of thermodynamics, one systematic effort to discover a general law of nature governing the whole process of evolution. This is to be found in Herbert Spencer's First Principles, where it serves as the foxmdation for his Synthetic Philosophy. This law of evolution, as its author called it, is developed from a rather vague conception not unlike Lachelier's later ideas. Spencer perceived that we can know a complex phenomenon only when we imderstand both its elements and how these elements cooperate in order to produce it. " That which alone can unify knowledge must be the law of cooperation of all the factors — a law expressing simultaneously the complex antecedents and the complex conse- quents which any phenomenon as a whole pre- sents." 1 Such a law, Spencer declares, must be regarded as quite generally valid, for: " If the law of operation of each factor holds true throughout the cosmos, so, too, must the law of their coopera- tion." 2 1 First Principles, New York, reprinted from the fifth London edi- tion, p. 468. ^ Ibid' P- 468. 121 122 NATURE In Spencer's opinion this law is apparent in all the phenomena of evolution; it governs all pro- duction and dissipation, and necessarily concerns matter and motion alike. Such is " the law of the entire cycle of changes passed through by every existence — loss of motion and consequent inte- gration, eventually followed by gain of motion and consequent disintegration. Besides applying to the whole history of each existence, it applies to each detail of the history. Both processes are going on at every instant; but always there is a differential result in favor of the first or the second. And every change, even though it be only a trans- position of parts, inevitably advances the one process or the other." ^ " There is habitually a passage from homogeneity to heterogeneity along with the passage from diffusion to concentration. \\'hile the matter composing the solar system has been assuming a denser form, it has changed from unity to variety of distribution. Solidification of the earth has been accompanied by a progress from comparative uniformity to extreme multiformity. In the course of its advance from a germ to a mass of relatively great bulk, every plant and animal also advances from simplicity to complexity. The increase of a society in numbers and consolidation has for its concomitant an increased heterogeneity both of its political and its industrial organization. 1 Loc. cit., p. 469. EVOLUTION 123 And the like holds of all super-organic products — language, science, art and literature." ^ " In all evolutions, inorganic, organic and super- organic, this change in the arrangement of matter is accompanied by a parallel change in the ar- rangement of motion; every increase in structural complexity involving a corresponding increase in functional complexity." ^ All this depends, in the first place, upon the fact that " Any finite homogeneous aggregate must inevitably lose its homogeneity, through the unequal exposure of its parts to incident forces." ' In Spencer's opinion this instability of the homogeneous is a perfectly universal phenom- enon; it holds for the parts as well as for the complete system. As a result there is a progressive tendency for the less heterogeneous to become more heterogeneous. This tendency even advances, according to his quaintly simple mathematical view, in a geometrical progression as the effects multiply. Such a process can end only in equilibrium. " That continual division and subdivision of forces which changes the uniform into the multi- form and the multiform into the more multiform, is a process by which forces are perpetually dissi- pated, and dissipation of them continuing as long 1 Loc. cit., p. 471. * Ibid., p. 473. 2 Ibid., p. 471. 124 NATURE as there remain any forces unbalanced by opposing forces must end in rest." ^ " This general principle of equilibration . . . was traced throughout all forms of evolution — astronomic, geologic, biologic, mental and social. And our concluding inference was that the penulti- mate stage of equilibration, in which the extremest multiformity and most complex moving equilib- rium are established, must be one implying the highest conceivable state of hiunanity." ^ From this exposition it is apparent that, what- ever philosophical use he may have made of it, Spencer believed his law of evolution to be a well- founded induction, and, therefore, a law of nature. In this he was probably mistaken. There is indeed a measure of truth in the so-called " law." And his generahzations, regarded as provisional and tentative hypotheses, possess genuine importance. But Spencer seems to have had no idea how ardu- ous would be the task of estabhshing such a prin- ciple even in physical science. He had Hterally no conception of the nature of the problem which he was raising, for rigorous mathematical proof was foreign to his nature. Under the circumstances it is not surprising to find his views meeting the open hostility of mathematical physicists like Lord Kelvin, Clerk Maxwell, and Tait.' ' Loc- cii-. P- 475. 2 Ibid., p. 475. » Cf. Knott's L!fe of Tail, pp. 281-288, Cambridge, 1911. EVOLUTION 125 At this very time, however, Willard Gibbs was attacking the problem of heterogeneous equiHb- rium in a rigorous manner, with a full mathe- matical equipment, a clear understanding of the principles of thermodynamics, and a power in the formation of abstract concepts hardly rivaled in our time. Maxwell at once perceived the connec- tion with Spencer's theories, and wrote to Tait: *' Have you (read) Willard Gibbs on Equilibrium of Heterogeneous Substances ? Refreshing after H. Spencer on the Instability of the Homogene- ous." ^ This investigation indeed leads to that very " blank form of a universe " which, accordiug to Tait,^ is the outcome of Spencer's speculation. One great result of Willard Gibbs's thermody- namic researches was to estabhsh the concept of a system as a genuine abstraction. Until the results of his labors were published the mathematical physicists possessed rigorous definitions of time, space, and mass. These, with the aid of their various quantitative measurements, enabled them, after the example of dynamics, to treat many prob- lems quite rigorously and exhaustively. But, wherever chemical composition or constitution was involved they were powerless. If we may judge by the published works on this subject, even Newton had contented himself with the demonstration that mass is independent of chemical composition. * Loc. cit, p. 284. ' Nature, November 25, 1880. 126 NATURE The significance of the new ideas is readily apparent. Thus the concept of a line is a pure abstraction, for geometrical lines do not exist in natiu-e. Nevertheless it is necessary for the very existence of geometry. The concept of mass, independent of all other forces than gravitation, is a similar fiction; for electrical, magnetic, and other forces, are never quite absent; but it is indispen- sable to the development of dynamics. In like manner the concept of an independent system is a pure creation of the imagination. For no material system is or can ever be perfectly isolated from the rest of the world. Nevertheless it completes the mathematician's " blank form of a universe " without which his investigations are impossible. It enables him to introduce into his geometrical space, not only masses and configurations, but also physical structure and chemical composition. Just as Newton first conclusively showed that this is a world of masses, so Willard Gibbs first re- vealed it as a world of systems. In this way physical chemistry has learned what manner of world is the subject of its investigations. It is a world made up of systems and nothing else. This conception of the universe, like that of classi- cal dynamics, which perceives only masses, is both exhaustive and rigorous, though purely imaginary and abstract. It is true that the gravest practical difficulties are sometimes involved, and that these EVOLUTION 127 difficulties have led to certain widespread miscon- ceptions. Thus, the modem chemist hardly real- izes the necessity of taking account of electrical and various other forces in his definition of sys- tems. Yet such forces are much more generally involved in the phenomena of heterogeneous equi- librium than in those with which dynamics is con- cerned, and they were discussed in Gibbs's original publication. But such fallacies regard practice and not the principle itself. The characteristics of a system are revealed in Gibbs's development of his mathematical analysis. They do not appear as entirely novel concepts, but like those of line and mass, as the results of old familiar ideas transformed by critical analysis. The proximate subordinate parts of a system or isolated aggregate of matter are the phases. A phase is, first of all, a homogeneous body. " We may call such bodies as differ in composition or state different phases of the matter concerned, regarding all bodies which differ only in quantity and form as different examples of the same phase."^ Accordingly a phase may be solid, liquid, or gaseous. Its only essential characteristic as such is physical and chemical homogeneity within the limits of our analysis. It is simply the sum of all the parts of a system that possess one perfectly definite and absolutely uniform structure and com- 1 Wilkrd Gibbs, Collected Papers, i, p. 96. 128 NATURE position. A system, therefore, includes as many phases as it contains physically distinct varieties of homogeneous aggregation. Beneath the phases are the components or primary constituents of the system. These were originally defined by Gibbs in a less elegant manner than the phases.^ But by a sUght modification in the mathematical development, which involves no change in the principles, components may be re- garded as the several species of chemical sub- stances, in so far as they are not decomposed, which are to be foimd in the system as a whole. ^ Every distinct variety of molecule, regardless of its physical state or states in the system, and re- gardless of the manner of its distribution through- out the system, provided only it is not liable to decomposition in this system, is a component. Such is the generalized material composition of the system. It is characterized by two types of aggregation; the physical and the chemical. Each of these is to be logically analyzed into its several uniform constituent parts. These niay be dis- tributed or put together in the simplest or most complex manner. But there is never any theoreti- cal difficulty iQ recognizing and distinguishing them. * Willard Gibbs, Collected Papers, i, pp. 63 ff. ^ Cf. Richards, Journal of the American Chemical Society, May, 1916. EVOLUTION 129 More puzzling is the treatment of the system's energy or activity. It was chiefly in order to cir- cumvent the diflBculties here involved that Gibbs introduced the idea of isolation. His own pre- liminary statement best illustrates the nature of the case: "We will examine the conditions of equilibrium of a mass of matter of various kinds enclosed in a rigid and fixed envelop, which is impermeable to and unalterable by any of the substances enclosed, and perfectly non-conducting to heat. We will suppose that the case is not com- plicated by the action of gravity, or by any elec- trical influences, and that in the solid portions of the mass the pressure is the same in every direc- tion. We will farther simplify the problem by supposing that the variations of the parts of the energy and entropy which depend upon the sur- faces separating heterogeneous masses are so small in comparison with the variations of the parts of the energy and entropy which depend upon the quantities of these masses, that the former may be neglected by the side of the latter; in other words, we will exclude the considerations which belong to the theory of capillarity. " It will be observed that the supposition of a rigid and non-conducting envelop enclosing the mass under discussion involves no real loss of gen- eraUty, for if any mass of matter is m equiUbrium, it would also be so, if the whole or any part of it 130 NATURE were enclosed in an envelop as supposed; therefore the conditions of equilibrimn for a mass thus en- closed are the general conditions which must al- ways be satisfied in case of equilibrivun. As for the other suppositions which have been made, all the circumstances and considerations which are here excluded will afterward be made the subject of special discussion." ^ We need not follow such special discussions. It will suffice to note that all forms of energy and ac- tivity are involved in the definition of systems, but that temperatvire and pressure are of very general importance. Yet gravitation, which can never be screened, as well as electrical, magnetical, and optical phenomena, and all other activities may often be involved. Moreover if phases are finely divided as in colloidal systems, there will be a great increase in surface area, and capillary phenom- ena must ensue. Another fundamental characteristic of a system is the magnitude of the concentration of each com- ponent in each phase. The recognition of this is perfectly essential to the description. But it is the last of the characteristics which must be taken into account in a system that has reached a state of equilibrium. It is well, however, to go beyond Gibbs's discus- sion as presented in his formulation of the Phase * Loc. cit., p. 62. EVOLUTION 131 Rule, and to note that so long as the condition of equilibrium has not been attained it is also neces- sary to take account of volume and configuration in the phases. With these discriminations the task is completed. Every physico-chemical aggregation as such, that is to say disregarding the functional relations of its parts as in a machine, the struc- tural configuration as in a crystal, and the infra- molecular characteristics such as the nature of molecular structure or the phenomena of radio- active transformations, may thus be ideally de- scribed. Often, as in the living organism, the actual task presents insurmountable difficulties; but these difficulties are practical, rather than conceptual or ideal. And no one, not even the vitalist, doubts that the organism is a Gibbs system. The difficulties involved in the use of this instru- ment of thought were recognized by no one more clearly than by Gibbs himseK. They led him to his last and, as some of his pupils think, his most novel contribution to science, the work on Statis- tical Mechanics. 1 This book, written after long years of meditation, but, as it seems, almost with- out notes to aid in the task, and completed in a period of less than a year, is perhaps the greatest example of sustained thought in the history of America. Gibbs's motive in turning his attention 1 New York, 1902. 132 NATURE in this direction is made clear in the preface, where, after pointing out discrepancies between thermodynamical theory and fact in the study of individual systems, he says: "Difficulties of this kind have deterred the author from attempting to explain the mysteries of nature, and have forced him to be contented with the more modest aim of deducing some of the more obvious propositions relating to the statistical branch of mechanics. Here, there can be no mistake in regard to the agreement of the hj^otheses with the facts of nature, for nothing is assumed in that respect. The only error into which one can fall, is the want of agreement between the premises and the conclusions, and this, with care, one may hope, ia the main, to avoid." ^ The specific object of the inquiry, so original and daring as to be almost inconceivable to those who have not the advantage of Gibbs's insight into mathematics, is thus stated: "We may imagine a great munber of systems of the same nature, but differing in the configurations and velocities which they have at a given instant, and differing not merely infinitesimally , but it may be so ks to embrace every conceivable combination of configurations and velocities. And here we may set the problem, not to follow a particular system through its succes- sion of configurations, but to determine how the ^ Loe. cit, p. X. EVOLUTION 133 whole number of systems will be distributed among the various conceivable configurations and veloc- ities at any required time, when the distribution has been given for some one time. The fimdamen- tal equation for this inquiry is that which gives the rate of change of the number of systems which fall within any infinitesimal limits of configuration and velocity." * The undertaking seems to have been notably suc- cessful, for Gibbs goes on to say : " The laws of sta- tistical mechanics apply to conservative systems of any niunber of degrees of freedom and are ex- act." ^ "The laws of thermodynamics may be easily obtained from the principles of statistical mechanics, of which they are the incomplete ex- pression." ' "We may therefore confidently believe that nothing will more conduce to the clear apprehen- sion of the relation of thermodynamics to rational mechanics, and to the interpretation of observed phenomena with reference to their evidence re- specting the molecular constitution of bodies, than the study of the fiindamental notions and principles of that department of mechanics to which thermo- dynamics is especially related." * It is apparent, therefore, that Gibbs has pro- vided physical science with a rigorous mathemati- 1 Loc. cit, p. vii. ' Ibid., pp. viii, ix. * Ibid., p. ix. * Ibid., p. ix. 134 NATURE cal analysis of the conditions of equilibrium in any system and also in any ensemble of similar systems. I cannot pretend to understand more than a little of Gibbs's analysis, and regarding the interpreta- tion of his statistical inquiry I am obliged to rely upon help from the mathematicians. Nor should I wish to be understood as venturing to accept or in any way to pass judgment on all of the results. The presumptions are solidly in their favor, but time alone can test the productions of even so great a man. Yet this is clearly the best that we now pos- sess as a means to the general and abstract physico- chemical characterization of cosmic evolution, for it involves our general concepts of matter, energy, space, and time, it includes the one known law of evolution,* the second law of thermodynamics, as an implication of its own more general results, and it is, so far as we can now see, rigorous, exhaustive, and exact. The results of Gibbs's thermodynamical studies clearly prove that Spencer's generalization bears no simple and intelhgible relation to the laws of equihbrimn. The Phase Rule may serve as an il- lustration of this fact. According to this rule the niunber of degrees of freedom, other things being equal, increases or diminishes as the number of phases diminishes or increases. In other words, roughly speaking, the greater the number of phases, • Cf. Perrin, TraiU de chimie physique, Paris, 1903, ch. 5. EVOLUTION 135 the smaller is the number of kinds of variation which can occur in the system. This may be illustrated by the case of pure water. For example, in a sys- tem which consists of ice, water, and steam the composition of each phase, the temperature and the pressure are all absolutely fixed. And thus, compression or the addition or subtraction of heat from the outside can only change the quantities of the several phases until at length one of them may cease to exist, meanwhile leaving temperature, pressure [and composition of the phases unchanged. But in the system of ice and water alone the appli- cation of pressure will at once produce a change in the pressure of the system, which will be accom- panied by a change in the temperature. On the other hand, if either temperature or pressure be fixed in such a system, then the condition of the system is fully determined and the other factor — pressure or temperature as the case may be — can- not vary. A similar statement applies to the sys- tems steam-water and steam-ice. Finally, if a system consists of the steam phase alone, it will not suffice to fix the temperature in order to fix the pressure, or vice versa. In order to fix the tempera- ture it will be necessary to fix the pressure and the composition, i. e. the concentration or volume. And in like manner both temperature and pres- sure must be fixed in order to fix volimae, both tem- perature and volume in order to fix pressure. The 136 NATURE conditions for more complex systems are perfectly analogous. Thus, at first sight, it appears that Spencer's idea of the greater stability of the multiform is justified. But a closer examination shows that his conception of multiformity involves not merely heterogeneity, or multiformity in phases, but also diversity in chemical composition and in the activities due to energy. At this point his views are radically con- tradicted by the Phase Rule. For the number of degrees of freedom increases by the same number as the number of components or diflferent forms of energy which are involved in the system. Thus, if to the system steam-water-ice a little alcohol be added, there will result a system as variable as the simpler water-ice system. And the same thing is true if gravitation is appreciably involved in the original system. In short "the instability of the homogeneous" tends to reappear in the heteroge- neous of Spencer though not in that of Gibbs. It is now important that two facts should be well understood: (I) Other things being equal the stability of a system increases with the number of phases and also with the number of restrictions upon the mtensities of energy, e. g. temperature, and upon the concentrations. Thus a system of three phases is more stable than a similar system of two phases; a system of constant temperature is more stable than a similar system in which the EVOLUTION 137 temperature is variable; and a system in which the tension of carbon dioxide is constant is more stable than one in which this is a variable quantity. (II) Other things being equal the stability of a sys-j tern diminishes with increase of the number of its' undecomposed constituent molecular species, and of the nimiber of different forms of energy, e. g. heat, pressure, electrical potential, surface tension, which are involved in its activities. Thus we must conclude that Spencer's view, though not contradicted, is also not supported by the Phase Rule. For while certain kinds of multi- formity tend toward stability others tend toward instabihty. Apparently the resultant of these two tendencies in the phenomena of nature as a whole can only be estimated by an objective considera- tion of the properties of matter. On this point we may at once note that there is a general tendency of the more complex molecular structures to in- stability, and that there is some reason to suppose such a tendency to exist in the elements of greater atomic weight. Moreover, complexity in the struc- ture of a phase, whether through irregularity of its configuration, or through its dispersion into sepa- rate fragments, is as a rule accompanied by a de- crease of stabiUty. But it is very doubtful if the problem as a whole, in the present state of knowl- edge, can be solved. The Phase Rule indicates a tendency toward the greater stability of a certain 138 NATURE kind of multiformity, which is exactly defined by Gibbs's term heterogeneity. But equally it proves other forms of multiformity to be unstable. I cannot find any further support for Spencer's hypothesis in the results of Gibbs's Statistical Me- chanics. And I think that we may admit, there- fore, that the physicists' hostility to Spencer's theories has been well founded. Many of his views are indeed clearly in error, though, as a rule, from excessive generalization rather than from radical inconsistency with the elementary principles of science. Spencer's belief in the tendency toward dynamic equilibrium in all things is of course fully justified. Its foundation may be discovered in the Phase Rule, and especially in the theorem of Le Cha- telier.i But as formulated by Spencer this is nothing more than a return to Hume, and taken by itself this principle could never have served the purpose as a foundation for the Synthetic Philos- ophy. 1 An interesting essay on the wide significance of this principle may be found in Bancroft's paper, Journal of the American Chemical Society, 1911, xxxiii, p. 92. VIII THE PROBLEM Herbert Spencer's law of evolution undoubtedly fails as a complete and rigorous principle, and it seems imlikely that any such law is for the present to be discovered. Yet, as we have seen, it is not altogether fallacious; nor is it without very sub- stantial foimdation in fact. As a general principle the instability of the homogeneous may be doubted and the invariable tendency to multiformity, as stated by Spencer, categorically denied. I am not sure, however, that the difficulty on the latter point may not be due to inconsistency between different statements in Spencer's own writings, and that so much of the idea as pervades his whole work, is better founded. In any case, there can be no doubt that he correctly analyzed many of the phenomena of nature. The extent to which he anticipated Darwin proves that. And it is certain that in the course of evolution at many points, perhaps even as a rule, there is a marked tendency toward dif- ferentiation and the production of complexity from that which is relatively simple. This is accom- panied, moreover, by a uniform tendency toward equilibrium. In short, Spencer's "law " is a reason- ably correct description of the evolutionary process. ISO 140 NATURE The history of the earth clearly illustrates this. There was a time, according to an ancient theory now often disputed, when the earth was a molten mass, approximately homogeneous except for the continuous variations in the concentrations of the different elements which make it up, or possibly consisting of a small number of distinct phases. Provisionally adopting this theory, we perceive that this phase or these phases were enveloped in an atmosphere, likewise approximately homogeneous except for the influence of gravity upon the con- centrations of the several constituents, which there- fore made up a single gaseous phase. If the earth was once molten there is no other conclusion in- volved in remote geological and astronomical his- tory so nearly certain as this, that the earth was once a two or three phase system. The few con- stituent phases were all, as phases go, of very unusual complexity ; first, because of the large munber of components of the system, and secondly, because of the magnitude of the continuous varia- tions in density and in concentrations throughout the phases. Such variations had been produced by gravitation. This condition may be taken as the origin of terrestrial evolution. Of course this is a purely ar- bitrary proceeding, but any analysis of the nature of the evolutionary process must begin somewhere, and it had better begin not too far off in time and THE PROBLEM 141 space, and with that condition which can be most probably made out. The argument in favor of the behef that the earth was once in a thoroughly molten condition is moreover, from the physico-chemical standpoint, very strong, and certainly far stronger than that in favor of any one of the numerous complete cos- mogonical hypotheses. This is perhaps best illus- trated by the fact that none of the recognized theories of the origin of the earth appears to be radically inconsistent with such a view, while nearly all clearly involve this stage in the evolu- tion of the earth,' On any other assumption it is hard to see how the most general characteristics of the earth's crust, or the nature of the igneous rocks can be explained.^ Further evidence is afforded by the molten condition of the sun and the stars, by the phenomena of vulcanism, and by many other consid- erations bearing on the present state of the interior of the earth, especially its great density. Finally, it is nothing short of fantastic to assume that the processes by which the earth was heated were just sufficient simultaneously to melt the whole mass, except a very thin crust. Such a coincidence can never be allowed without the support of arguments far more cogent than those which depend upon 1 Cf. Poincare, Legons sur les hypotheses connogoniques, Paris, 1911. 2 Cf. Daly, Igneous Rocks, chap. 8, New York, 1914. 142 NATURE any cosmogonical theory. Yet, without fusion how could the difference in density of surface and in- ,terior become estabhshed ? In truth, no hypothe- sis concerning the unknown past appears to be better founded than this hypothesis of a molten earth. It is established quite as securely as most of the scientific theories which are constantly em- ployed without question. We are, however, only concerned with it provisionally as an hypothesis. The conclusion that under these circumstances the earth must have consisted of a small number of phases rests upon experience. For the labors of the physical chemists conclusively prove the co- existence of a large number of liquid phases or of more than one gas phase to be impossible. The only restriction upon this statement is foimd in the case where mixing is incomplete. But though the size of the earth greatly restricts mixing, the long periods of time involved in geological processes must have gone far to neutrahze such a tendency. It is to be noted that the only current hypothe- sis that denies a molten epoch to the earth as such, assigns to the earth an origin by disruption from the molten sun. But to this early sun the above considerations equally apply. Thus we reach a definite position: the earth, as such or as a part of the sun, was probably once in a molten state. Under these circumstances it is most conveniently regarded as a single system. This THE PROBLEM 143 system consisted, for a long time, of a small nvmiber of phases of very great volume. These phases pos- sessed one peculiar characteristic, for the force of gravity, even though opposed by diffusion, must have accomplished a differentiation with respect to the concentrations of the components at differ- ent levels throughout each phase. The components of this system were very numerous, in that they, included at the very least all of the chemical ele- ments. Conceivably certain chemical compounds which may happen to be quite stable at the tem- perature of the system may also have been involved as components. I think, however, that the existence of compounds which are imdissociable at such tem-| peratures may be regarded as unlikely. Possibly a few were present in the atmosphere. It is clear that we have thus imagined a condition of relative instabihty in the relatively homogeneous. Indeed, taking everything into consideration, the niunber of degrees of freedom must have been about one hundred. This may be compared with our labora- tory systems where the number of degrees of free- dom is rarely as great as ten. The instabihty of such a condition depends upon the fact that the nimiber of components is large, while the number of phases is small. Out of this condition the almost infinite variety of the present world has been evolved. We see about us countless systems — not Ladeed strictly 144 NATURE independent, but nevertheless best conceived as such — which present the greatest diversity in re- spect of both phases and components. Such are the geological strata, the rocks, the sands, the soil, the lakes and streams, the ocean itself, and the at- mosphere; and such is every Uving organism. We see, moreover, orderly relations between these sys- tems. , It is evident that in the course of the evolution ^ the earth systems have evolved in great profu- sion, in inconceivable variety, with almost infinite diversity in phases, components, concentrations, and activities, and always ia coordination. This indeed, abstractly stated, is the very essence of the evolutionary process. This is what evolution is. And we may now see that Herbert Spencer was not far wrong about it. Whatever may be the other peculiarities of the evolutionary process, relative stabiUty in relative diversity has certainly suc- ceeded relative instability in relative uniformity. And so it had to be if anything interesting (to in- troduce the teleological impUcation) was to happen. Apart from all theories regarding the formation of the crust, we shall soon see that this conclusion is established upon the foundation of geological fact. It is now apparent, however, that the general laws of science do not sufficiently account for the evolution of the globe. The Phase Rule, the second law of thermodynamics, the principles of statis- THE PROBLEM 145 tical mechanics and the fact of the stability of dynamic equilibria are all, like the laws of conser- vation and of gravitation, conditions of the process. But the process itself is the evolution of the original matter and the original energy of the globe. It is the properties of this matter and of this energy which chiefly bring to pass the manifold events in the history of the earth, or at least which make it possible that they should be manifold. Perhaps it may be said that the above mentioned laws organ- ize the historical events and the systems which are the sole actors. But that which permits their diversity, as thus organized, is the nature of the matter and the energy themselves. Or, to put it in another way, the characteristics of matter and energy condition that to which the laws apply. Spencer failed clearly to imderstand this, and there- fore to establish his conclusions. There is, perhaps, danger of making too much of the antithesis between the laws of phenomena and the characteristics of the various forms of matter and manifestations of energy. Yet there is un- doubtedly a certain logical priority in the law of conservation of energy as compared with the phe- nomenon of an electric charge, or in Newton's law of inverse squares as compared with the properties of copper. This difference finds its expression in the tendency of many thinkers to regard the laws of conservation as necessary a priori truths, or to 146 NATURE consider Newton's law as a necessary consequence of the principles of geometry. A similar a 'priori character might easily be assigned to the second law of thermodynamics on account of its statistical foundation and, for similar reasons, to the tend- ency toward dynamic equilibrimn. I am not here concerned to justify or to criticize this view, but to point out that no one is likely to take a simi- lar position regarding the specific characteristics of things. Possibly the second law of thermodynam- ics, in one or another of its forms, might have been worked out by a mathematician in perfect igno- rance of how energy should be conceived. This would hardly be a more remarkable achievement than the creation of non-Euclidian geometry ; and in a way, it is not a bad description of Camot's actual performance or of some of Gibbs's more finished productions. But no one can imagine the origin of the concept of an electrical charge imtil the phe- nomena of electricity had been investigated. In other words, the prediction of electrical phenomena by one ignorant of all such phenomena seems to be quite impossible. Admitting that there is sufficient ground for a distinction, established solely for convenience and without philosophical impHcations, between the laws and the specific properties of things, we may now take another step. In the first place we note that the production of diversity would be impos- THE PROBLEM 147 sible if matter and energy were uniform. It is be- cause there are not far from a hundred elements, for the most part capable of entering into a great variety of chemical reactions, and because beside mechanical forces, there are many other ways in which energy manifests itself, that the world can become diversified. But this is far from sufficient as an analysis. For, secondly, we may note that if there were no tendency when solids are deposited from a molten mass to the separate formation of individual compomids, the process of evolution would hardly be more varied than the freezing of a huge mass of water. Thus we may vaguely per- ceive how the general and individual properties of matter and energy are alike concerned in the pro- duction of the manifold forms of nature. So we have at length reached the real problem of the order of nature. Admitting that evolution consists in the evolution of systems, because sys- tems make up the whole world of physical chem- istry, we have to inquire in what manner the properties of matter and energy make possible that orderly diversity which is so conspicuous a result of the evolutionary process. The other character- istic result, the stability of the products of nature, we can more clearly see to be largely effeg^§^rby the operation of natural laws. But we must not forget to consider this too. The question now arises: how far did the properties of matter and 148 NATURE energy admit of freedom — using the word in its recognized scientific meaning — in the evolution of systems ? To what extent, considering only these properties, are mere number, variety, and durability of systems possible ? Or, in short, what are the properties of matter and energy which must be taken into consideration when we regard them as material for the construction of systems and of ensembles of systems of every kind, i. e. of any kind ? I hope that a knowledge of the character of Willard Gibbs's researches may make this ques- tion seem not quite illusory. In this inquiry it will be necessary to take ac- count of number, of diversity, and of durability in systems as a whole, in their phases, their compo- nents, their concentrations, and in all the forms of their activity. Further, in that the concept of isolation is a fiction, it will be necessary to consider relations between systems, thus introducing the ideas of pattern and organization. But nothing else need be considered. For these are the primary qualities of the world, established at length by the analysis of modern science after centuries of vain philosophizing. It must not be supposed that the problem of the cooperation of the laws of nature has disappeared in the course of our analysis, nor that it has been solved. This is still a genuine problem and an open question. Yet I think it is less promising than THE PROBLEM 149 that of the cooperation of the properties of matter and energy. At any rate there is nothing to hinder us from now disregarding it. For we have arrived at a clear issue which is open to an independent examination. Nearly all the phenomena of the evolution of the earth have taken place upon the surface during the existence of the crust. This fact is a necessary consequence of the principles of physical science just examined. For the evolution of systems could only begin on a large scale with the intervention of solid phases. There was, however, one great proc- ess that involved not the crust only, but the whole earth, and lead to a partial separation of the chemi- cal elements iinder the action of the force of gravitation. It is not unlikely that this separation de- pended especially upon the existence of just two hquid phases in the molten earth; — a central me- talhc core and an outer slag. Such conditions would correspond to the state of affairs in a blast furnace. The structure of meteorites, moreover, seems to be consistent with such a two-phase ori- gin, and I know not what other physico-chemical explanation can be suggested.' Thus, or otherwise, the lighter elements have come to the surface in relatively great quantities. 1 These considerations were suggested to me by Professor T. W. Richards. I cannot think that a physically and chemically intelli- gible alternative hypothesis has yet been made out. 150 NATURE Moreover, such diflferentiation into phases as had always existed since the earth somehow became a large dense aggregate involved the existence of an atmosphere. In this atmosphere great quantities of certain elements, all of them also relatively light, which could exist free or in combination as stable gases imder the existing conditions, were present. Thus it has come about that only a few of the elements, and those as a rule of low atomic weights, have had a large part in the evolutionary processes. Such are hydrogen, carbon, nitrogen, oxygen, so- dium, magnesium, aluminium, silicon, chlorine, calcium, and iron. We may at once note the fact that the elements of low atomic weight are gener- ally more intense and more diverse in their chemi- cal activity. In this manner the very earhest stages of differentiation have led to an increase in the possibilities of chemical changes during the course of the evolutionary process. It must not be supposed, however, that any elements have thus been excluded from the crust. The effect of these processes has been only to alter the distribution of the elements in the maimer indicated, and to pro- duce a relatively light exterior and a relatively dense interior of the lithosphere. While the formation of the earth's crust was pro- ceeding and thereafter until the present time the atmosphere has persisted. Meanwhile it has im- THE PROBLEM 151 dergone great changes in composition so that its early history is little known. But in its composi-/ tion the lightest of the important elements — hyf drogen, carbon, nitrogen, and oxygen — have never failed in recent times. For a long period, certainly since an early stage of organic evolution, these elements have existed in the forms of chemi- cal combination, if not in the proportions, in which they are now present in the air, viz., as water, car- bon dioxide, nitrogen, and oxygen. Still earlier, for a very long period, at least water, carbon di- oxide, and nitrogen were present. As the cooling of the earth progressed a temper- ature was finally reached at which water began to condense out of the atmosphere. Before that time the differentiation of systems upon the earth's sur- face had been in steady progress. Igneous rocks had been formed. They had probably been torn and twisted in theu- structure and variously segre- gated by volcanic upheavals and other great changes. But such processes are as nothing to those which were to follow. For water is the most powerful and most universal agent in moulding the surface of the earth. The meteorological cycle resulted from the precipitation of water, and has continued, presumably without interruption, until the present time. Any objection which may be felt to the above provisional account of early geological processes 152 NATUKE may now be readily obviated. For it is possible to look back with certainty at least to the earlier epochs of the meteorological cycle. At that time the crust of the earth, however differentiated into systems as compared with a molten sphere, was still almost perfectly homogeneous or at least dis- organized if contrasted with its present condition. For all finer comminution and all the organic phe- nomena are later in epoch, and there were then no intricate orderly relations between the different systems. From this point, no less than from the earlier hypothetical single system, the process of the evolution of systems has steadily gone on in the general manner which has been above remarked, or, in other words, more or less according to the requirements of Spencer's " law." In the course of the meteorological cycle the movements of water became canalized. Streams, lakes, and the ocean assmned a somewhat definite form, water began to penetrate the debris resulting from its own action, and from that of dissolved carbonic acid, to set this in motion, and thus in certain localities to form deposits. Some of these have become strata, others, with the help of fur- ther agencies, earth and soil. And at length nearly everything that meets the eye, except life and the products of life, has been moulded into its form by the action of water and carbonic acid. THE PROBLEM 153 The only other great event in the history of the earth — but of this we have no knowledge — is the beginning of the process of organic evolution. Yet' if the meteorological processes have multiplied a thousand fold the evolution of systems, organic evolution has again multipUed these in a like pro- portion. The elements here chiefly involved are, once more, hydrogen, carbon, nitrogen, and oxy- gen. Thus what is known with certainty about the history of the earth enables us to see that a few ele- ments, and especially the four organic ones, are the chief factors. Among these nitrogen plays a somewhat subordinate role, especially in the min- eral kingdom, while hydrogen, carbon, and oxygen, notably as constituents of water and carbon di- oxide, are almost everywhere of equal importance. This conclusion admits of another advance in the analysis and a final formulation of the problem. We have noted, step by step, that Gibbs's ab- stract concept of system is a means to the exhaus- tive characterization of the world as contemplated by physical chemistry: that systems are made up of phases and components: that these are charac- terized by the concentrations of the components in the phases, and by the various manifestations of energetic activity: and that in certain cases vol- ume and configuration must also be regarded. We have also seen that the process of evolution of the 154 NATURE earth appears, when examined in the light of this concept, as a continuous production of many sys- tems related together in an orderly manner from few original systems, and that these systems are not only very numerous but also very diverse and often very stable. Further, we have seen that there is ground for the belief that the more impor- tant conditions which make possible this evolu- tionary process are the specific characteristics of matter and energy as they cooperate in the proc- ess, rather than the most general laws of physical science. And at length we have discovered that the elements which, by the combination of their characteristic properties and activities, chiefly make possible the greater part of the results of the evolutionary process are but three — hydrogen, carbon, and oxygen. Now we may ask what is the relation of the prop- erties and activities of hydrogen, carbon, and oxygen as causes to the evolution of numerous, diverse, stable systems as effects ? How is it that, on account of the peculiarities of these three ele- ments, there are so many degrees of freedom left open in the evolutionary process ? This is the question which will be discussed in the following pages. I have already examined it in a less sys- tematic manner in The Fitness of the Environ- ment. I hope now to simplify and to generalize the analysis. IX THE THREE ELEMENTS Components Of all the chemical elements, hydrogen, carbon, and oxygen possess the greatest number of com- pounds and enter into the greatest variety of reac- tions. The known compounds of carbon, which very often contain all three elements, are num- bered by ten thousands, while the possible carbon compounds are almost innumerable. The com- poimds of inorganic chemistry in a very large number of cases also contain oxygen or hydrogen. Thus these elements afford by far the greatest number of components for the constitution of systems. This unique combining power of the three ele- ments with other elements, but especially among themselves, depends of course upon their very nature, upon those characteristics which are pecul- iar to them and mark them off from all other elements. These properties, moreover, produce characteristics in the compounds which distin- guish these from the compounds of other elements. Thus the compounds of oxygen are commonly very reactive, the compounds of carbon with hydrogen, 155 156 NATURE taking account of the size of the molecule, very stable. But, in a somewhat less conspicuous man- ner, numerous other elements present similar phe- nomena, and the most striking chemical property of the three elements is therefore the variety of their combinations. The necessary condition for the mmiber and diversity of carbon compounds is the abiHty of the single atom of carbon to combine with several other atoms, actually with four, and thus, accord- ing to the atomic theory, to make possible the for- mation of chains, forked chains, and rings of atoms in the molecule. But this fact of the quadriva- lence of carbon is in itself by no means unique, nor is it evident that a combining power of four is necessarily better than five. What is remarkable is the ability of chains and rings of carbon atoms to hold together, especially when hydrogen is the principal other element of the molecular structure. The history of chemistry indicates that this is a unique phenomenon, for in our experience it is not paralleled. Compounds of hydrogen and carbon, free from all other elements, and therefore known as hydro- carbons, exist in the most bewildering profusion. One such compound probably possesses a straight chain of sixty carbon atoms, and there is no reason to suppose that chemists have approached the limit of length of such chains. Moreover, these may THE THREE ELEMENTS 157 apparently be forked at any point, they may be joined into a great number of ring systems, and rings and chains of all kinds may be combined within a single molecular structure. Finally, more than one valence of each of the carbon atoms may, as it seems, be involved in the union between two such atoms, A single example of molecular struc- ture expressed according to current theories may serve to illustrate these considerations. H I H— C— H I C ^ \ H H-C C<„ H I I g H>y