Paneth_1931_2003_The Epistemological Status of the Chemical Concept of Element

8/11/2019 Paneth_1931_2003_The Epistemological Status of the Chemical Concept of Element

1/34

F.A. PANETH

THE EPISTEMOLOGICAL STATUS OF THE CHEMICALCONCEPT OF ELEMENT

1. THE NEED FOR EPISTEMOLOGICAL CLARIFICATION OF THEFUNDAMENTAL CONCEPTS OF CHEMISTRY

Discussions of the principal concepts of chemistry are few andsupercial, in striking contrast to the many penetrating investi-gations into the philosophic foundations of physical theories. Inprevious centuries there was an obvious reason for this: manyphysical doctrines were already common knowledge amongst theeducated, whilst chemistry, in the scientic sense, did not exist atall, or was only just beginning. But even today we nd a widerspread of physical than of chemical knowledge and interest amongstphilosophers, whether this be due to tradition in the profession or tothe curricula of our secondary schools. The rst two laws of thermo-dynamics, the questions concerning action at a distance and contactaction, the difculties of the ether hypothesis, and more recently,particularly the principle of relativityand the quantum theory, are all

problems familiar to every philosopher who wishes to take accountof the results of the exact sciences in his system. On the other hand,one rarely nds a discussion of a specically chemical theory inthe writings of a philosopher. The atomic theory is an exception.This, however, belongs just as much to physics as to chemistry, andindeed the physical aspect (divisibility, mutual attraction and repul-sion, and so on) has been discussed much more than the chemical(qualitative characteristics, valency, etc.). As a rule, chemistry is

The text of a public lecture given by Professor F.A. Paneth to theGelehrte Gesellschaft of Knigsbergin 1931. (Germantext published in Schriftender Knigsberger Gelehrten Gesellschaft , Naturwissenschaftliche Klasse, 1931,Heft 4 [Max Niemeyer Verlag].) Translated by H.R. Paneth. Reprinted from the British Journal for the Philosophy of Science, Vol. 13,pp. 114 (Part I) and 144160(Part II), 1962, by permission of Oxford UniversityPress.

Foundations of Chemistry 5: 113145, 2003. 1962 Oxford University Press.


2/34

114 F.A. PANETH

presented by the philosophers as a science which is well on the wayto transforming itself into physics, and to which, therefore, the sameconsiderations will apply in due course.1

Not only philosophers, but many physicists have endeavouredto clarify the epistemological foundations of physics, sometimes insystematic investigations, sometimes merely by way of occasional(but extremely illuminating) comments. I need only recall in thisconnection the names Helmholtz, Kirchhoff, Heinrich Hertz, Mach,Boltzmann, Poincar, Planck, Einstein, Weyl, Jeans, Eddington,Bohr, and Heisenberg. For this reason there has long been anextremely fertile interaction between writings in philosophy and inphysics, while we have hardly anything comparable in the eld of chemistry.

The fact that there has been little thorough philosophical workon the concepts of chemistry is particularly apparent in the caseof the concept most fundamental to the whole science: the conceptof element. Most textbooks of chemistry, when introducing theconcept of element, make a few remarks about its philosophicformulation by Aristotle, and in the Middle Ages, and emphasisethe difference form the present-day scientic idea of element.With stereotyped uniformity the statement is repeated that the fourelements of Aristotle didnot designate substances,but combinationsof the fundamental properties cold, warm, dry, wet; that Boyle wasthe rst to mean by a chemical element a substances that could

not be decomposed further, etc. In most expositions of the historyof chemistry, too, we nd similar statements.2 There is not doubt,however, that it is incorrect to characterise the elements of Aristotleas properties rather than substances, and that Boyles innovationlies in a different direction from the assumption that the elementsare, by their nature, substances and that they cannot be decom-posed. The fact that there is nevertheless general agreement withthese statements amongst the majority of chemists writing on thissubject only shows how little importance they attach, in general, tothe investigation of the concepts used in chemical teaching.

The study of the philosophic-historical development of the

concept of element, in particular, is interesting and important fora further reason. The same epistemological questions which wereraised in connection with it in classical times and were even in


3/34

CHEMICAL CONCEPT OF ELEMENT 115

part answered, still play an important part in present-day chem-istry, I am primarily referring here to the question In what sensemay one assume that the elements persist in compounds? When

one compares the obscure remarks of some modern authors on thissubject with the insight already gained by philosophers of antiquity,one becomes convinced that in chemistry, just as in physics, a philo-sophic clarication of the fundamental concepts should promotescientic understanding as well. There is the further point that wemay, by such an investigation, test whether or not it is true thatchemistry should and will dissolve into physics.

For these reasons we shall try in this essay to present theepistemological status of the concept of element somewhat morethoroughly than is customary in chemical textbooks. To makecommunication possible it is necessary to use certain technicalphilosophical terms; their choice will, of course, vary with onesbasic philosophical conceptions. Since, however, we are here onlyconcerned with the discussion of a particular problem pertaining toscience which must retain its meaning for every epistemologicalposition it is not of decisive importance which terminology weuse. For reasons of expediency we shall follow in the main theterminology chosen by Eduard von Hartmann, since, amongst thetheories put forward anddeveloped in detail by professional philoso-phers, his epistemology possibly comes closest to an understandingof the essential nature of science.3 We can ignore here the question

whether Hartmanns transcendental realism in some points evendepended too much on the science of his time, just as we can, forour purpose, completely ignore his metaphysical system.

2. THE CONCEPT OF SUBSTANCE IN CHEMISTRY

Chemistry, like every natural science, started from the naive-realisticworld-view, and gradually found itself compelled to apply correc-tions to this. It is, however, characteristic of chemistry that it has notadvanced as far in the application of these corrections as some other

sciences; indeed it is of the essence of its fundamental concepts thatthey have retained quite an appreciable naive-realistic residue.Let us consider rst the concept of substance in this respect.

After all, the whole doctrinal structure of chemistry rests on this.


4/34

116 F.A. PANETH

Moreover, it is fundamental to the concept of element, with whichwe are concerned here, in particular. It is well known that it haslong been the aim of physics to reduce sensory qualities to quanti-

tative determinations.4

Nature is assumed to be without qualities,and the properties of colour, sound, taste, smell, etc., are applic-able only to our sense representations. The objective prerequisitesof the last two qualities are still almost totally obscure; but sinceeven today, as formerly, these qualities play an important part inchemistry, in characterising substances, the chemist has no choicebut to assign the properties of taste and smell to the substancesthemselves, in completely naive-realistic fashion. Nobody objects,in fact, to speaking of the salty taste of sodium chloride, or theunpleasant smell of hydrogen sulphide. Here we nd ourselves, as issurely unnecessary to elaborate, still standing with both feet on theground of naive realism, without, it should be emphasized, beingaware that this lack of philosophic clarity entails any disadvantage.Indeed, even in the case of that property of substances which canbe reduced to quantitative determination most easily colour weusually refrain, for the purpose of chemical characterisation, fromso reducing it: cinnabar is red, gold is lustrous. While the intro-duction of the various constants would make a more exact numericalstatement possible in such cases, this would usually be too cumber-some. After all, we nd something similar in physics, where we donot mind speaking of red and green light as rough data; quality is

retained here as useful mental short-hand.5

When one attacks the problem of substance more closely withthe weapons of epistemology, the fact is revealed (as has alreadybeen pointed out by many authors) that not only may none of itssensory properties be taken to exist in nature, but substance disap-pears altogether. The old view, taken by the Greek atomists, thatonly size, shape and motion are real, while colour, smell, etc., aremere appearances (a view revived later by Locke in his distinctionof primary and secondary qualities), marks but an arbitrary stoppingpoint in this process of dissolution, though (as we shall discussbelow) a very useful one for the sciences. The subjective nature

of the primary qualities, still pertaining, according to this view, tothe things themselves, has been demonstrated in the well-knownidealistic lines of thought (Berkeley, Hume, Kant). Exponents of


5/34


transcendental realism, too, who are convinced of the existenceof an external nature which is knowable to a certain extent, ndthemselves forced, in pursuing the logical consequences of atomism

to the end, to declare matter to be not something real, but merely anappearance which is produced by the joint effect of closelyadjacentforces in and for the conscious mind.6

Now it is particularly interesting that, independently of thesephilosophic considerations, progress in theoretical physics, also,has led necessarily to the complete dissolution of the concept of substance. The era of the classical kinetic theory of matter is longpast; molecules and atoms are no longer solid spheres; the spaceassigned to them is lled only by electric forces originating in thepositive protons and negative electrons, which are vanishingly smallcompared with the atomic radius. Now one might regard protons andelectrons as last remnants of the old concept of substance, if modernquantum mechanics did not require the assumption that the motionsof these particles do not take place in the three-dimensions. Thuswe have lost the last possibility of retaining, in any sense at all, thefamiliar notion of substance; we renounce all possibility of visua-lizing, while raising calculability to a level never before attained.Here physics is harvesting the fruit of a tree planted centuries ago.Starting with Galileo, more and ever more weight was attachedto mathematical formulation; the successes of this methodologyprovided encouragement for further and further advances along the

same route, and we must not be surprised when present-day physi-cists, rather like the ancient Pythagoreans, see the true nature of theworld in mathematical relations.7

Chemistry did not follow this course of development. Indeed,already in the seventeenth century, at the time of the rst begin-nings of scientic chemistry, we nd one man who, with the greatestpossible clarity, demanded for it principles different from the math-ematising ones of physics: the Englishman, Robert Boyle. Boyleis one of the few scientists with a mental range wide enough tomaster the procedure adopted in physics, and appreciate it on itsmerits, without thereby according it the status of the only true path

to salvation in natural science. On occasion he states quite denitelythat the secondary qualities are no less real than the primary ones,8and that it is not necessary to give up the former and calculate only


6/34

118 F.A. PANETH

with the latter to obtain scientic results, and nd thereby methodsfor the practical mastery of nature.9

Nevertheless, he himself always considered mathematical formu-

lation to be a higher level of theoretical insight into nature.10

Hissuccessors amongst the chemists, while following his example as apractical chemist, abandoned his penetrating Weltanschauung . Theycompletely neglected the classical, pure, quantitative particle theory,which he had particularly fostered; instead, they put forward withoutreservation purely qualitative hypotheses, which would have beenintolerable from, say, the more severe point of view of Newton.These hypotheses were often very useful for the development of chemistry. We need only remember the introduction of the idea of phlogiston11 at the beginning of the eighteenth, or of catalyticforce12 at the beginning of the nineteenth century.

On the other hand, the Newtonians inevitably low opinion of such tendencies emerges particularly form the oft quoted statementof Kant that in any particular discipline of the study of nature onecan nd only as much actual science as there is mathematics. If one accepts this denition, Kant is perfectly right not to includechemistry amongst the sciences, since chemistry is essentially non-mathematical; he saw this more clearly than later philosophers, whoexpected chemistry to pass into physics. He did, however, recognisechemistry as a systematic art.13 Chemists may be satised withthis title for their subject, if, indeed, they must remain excluded from

the realm of genuine science. However,Kants denition results inan extremely narrow, and therefore probably inappropriate, concep-tion of science. A closer consideration of chemistry not to mentionthe biological and historical sciences sufces to show this. Someof its elds, particularly so-called physical chemistry, admittedlynow include so much mathematics that they would presumablysatisfy Kants denition; but there also exist analytic and syntheticchemistry, which have remained completely unmathematical to thisday. Nevertheless, on the basis of their successes, we can claim therank of science for them too. We need only remind ourselves thatit was the analytic chemist, unhampered by mathematics or indeed

almost any theory, who discovered the majority of all chemicalelements on the basis of the most primitive concept of substance!Much later, when a theory of chemical elements existed, and physi-


7/34


cists realized that the elements must follow each other as the ordinalnumbers, and that their number is therefore strictly limited, theywere astonished to nd that the analysts had already discovered

almost all possible elements. A systematic art that can achievethis may surely be called a science as well. Moreover, already inthe seventies of the last century, the elements had been arrangedby the chemists into a scheme, the so-called natural system of the elements. Present-day workers on the theory of the atom havecontributed enormously to the deeper understanding of this naturalsystem; but the arrangement itself has not undergone any essen-tial change in the last sixty years. Perhaps even more imposing isthe achievement of the organic chemists, who have been able tobring hundreds of thousands of different chemical substances intoa clear system on the basis of a very few hypotheses concerningthe combination of atoms. Here, too, the methods of research usedin physics are already being applied and much splendid progresshas been made in this way; but much more astonishing than indi-vidual corrections, achieved with the most modern aids, is the quiteoverwhelming number of cases where the organic chemists simpleconcept of valency, with its obvious weaknesses, had made possiblecompletely correct statements about the mutual linkages of atoms.Even if the character of chemistry should change essentially inthe future owing to penetration by mathematico-physical methods,its history during the nineteenth century, in which it achieved

such successes without mathematics, must never be ignored in itsphilosophic evaluation.In view of all this, we declare chemistry, too, to be a true science,

even in those branches where it contains little or no mathematics.Indeed, perhaps it is just there that we are dealing with chemistry inthe strictest sense of the word. Emil Fischer, in his autobiography,tells of a colleague at Wrzburg who had the charming denition:Chemistry is bangs and stinks.14 We can essentially agree withthis, and interpret it as follows: chemistry is a science in whichinterest is directed towards the secondary qualities of substances.

Taking this philosophically primitive standpoint chemists have

usually managed very well in the whole vast range of the subjects of analysis and synthesis. For this reason they have remained unawarethat, after all, somewhat deeper epistemological consideration is


8/34

120 F.A. PANETH

necessary for the complete understanding of the reactions they havecarried out and the theorems they have put forward. This neces-sity will, however, at once become apparent when we turn to the

concept of chemical element. Here we are bound to get involved inirresolvable contradictions if we retain the chemists naive notion of substance.

3. THE EPISTEMOLOGICAL STANDPOINT OF THE ANCIENTATOMISTS

It is probably useful to preface the discussion of the modern conceptof element with a few remarks about its predecessors in classicaland medieval times. We do this, not with the intention of presentingan exhaustive history of its development, but because we nd thevarious views very well represented; instead of abstract argumentswe can frequently make the intuitively comprehensible theorems of the ancient philosophers the basis of our discussion.15

How could hair have its origin in non-hair, or esh in non-esh? is the text of one of the surviving fragments of the writingsof Anaxagoras.16 Already in those days, according to Aristotlestestimony, all physicists were agreed that matter is permanent.17Anaxagoras went further than some of his contemporaries andpredecessors only in considering qualitative, as well as quantitativeinvariance a matter of course. In his question the basic problemof chemistry and its conceptual difculty are pointedly presented.When the raised this question there were already some typicalattempts to solve the problem. These we will consider briey, sincethey lead almost directly to the espistemological problem in whichwe are interested.

The conviction that the truly existent could not perish had ledthe Eleatics to declare all changes in the world of the senses to bemere appearance. Conversely, Heraclitus was led, by contemplationof the continual change in nature, to deny any permanence. Theformer took up a position so much in contradiction to everyday

experience that they were exempted, from the start, form the taskof explaining nature. For this reason Aristotle refers to them asthe opposite of scientists (, o ).18 But Heraclitus view doesnot agree well with experience either, for according to his doctrine


9/34


one would expect that every substance could change into everyother, which even in those days must have been seen to be untrue.An intermediate position is taken up by Empedocles. His doctrine

of the four elements constitutes an enormous advance. Indeed,one may justly argue that the assumption of a nite number of permanent substances which, by being mixed in different propor-tions, produce the multitude of objects we perceive, is an essentialpart of present-day chemical doctrines.19 No answer, however, isgiven to Anaxagoras question; it seems that it is not even raisedby Empedocles. Yet the creation of a new substance by mixing twoknown ones is at rst incomprehensible precisely on the assumptionthat the substancesare permanent. Thus we must rate very highly thelogical rigour of Anaxagoras, who rejected this inconsistent idea,and assumed smallest particles of all homogeneous substances to bedistributed everywhere in the universe. Indeed, it is worthnoting thatrather similarobjections were still raised much later by philosophersand chemists. Thus John Stuart Mill asserts for this reason that thelaws of chemistry own their existence to a breach of the principle of Composition of Causes;20 while Berthelot considers the differencebetween the properties of common salt and those of its elements,sodium and chlorine, so incomprehensibly great that he feels boundto refute at length the notion that common salt contains a thirdelement, with less divergent properties, as well.21 Fundamentallythis difculty could be overcome only by adopting a more profound

epistemological standpoint, an achievementwhich we must considerone of the greatest merits of the atomists.The atomists were the rst to realise that one may indeed retain

the demand of the logos for a permanent substratum, but not withreference to our worldof appearances; or, to put the same idea differ-ently, that laws of nature may be constructed not for our subjectiveperceptions but only for an objective world quite different in kind,which lies behind those perceptions. Thus they were not forcedeither to deny the changes in the world of appearances, as did theEleatics, or to deny permanent substance, as did Heraclitus. Whatlooks to us like coming into being and ceasing to be is actually only

re-arrangement of the permanent atoms. The secondary qualitiesexist only in our perceptions, the primary ones pertain to the atomsthemselves; in their realm only size, shape, and motion exist. Thus,


10/34

122 F.A. PANETH

the conceptual view of the world demanded by the Eleatics, wasbrought into harmony with appearances.22

The question how can hair come form non-hair, and esh form

non-esh? is now answered to this effect: the same particles whichin one particular arrangement give rise to our impression of greenmeadow plants, serve in another arrangement for the constructionof the animal body fed on these plants, and appear to us to be hairand esh. In this way the demand for permanence of substance issatised with no necessity for Anaxagoras assumption, viz. thatparticles of esh and hair already exist in the plant. Empedocleshappy idea of replacing generation and decay by mixing and separa-tion also plays its part for the atomists; but while with Empedoclesit is still impossible to imagine how the mixture of indestructibleelements could take place, atomism provides a picture which isboth useful and intuitively very acceptable, by assuming smallestparticles which juxtapose themselves in different arrangements.23It may not be superuous to show that on this basis, which wasconstructed by Leucippus and Democritus, of an atomic worldwithout qualities, it was also possible to arrive without difcultyat a more detailed idea of the nature of a chemical compound.

An important step in this direction was taken by Epicurus.24Aristotle had raised the objection that no substance with new proper-ties could ever result from the juxtaposition of smallest particles,since sharp eyes would surely still recognise the known substances

in the mixture. Epicurus rightly sought to overcome this difcultyby shifting to the realm devoid of qualities, although in devel-oping this idea he did not quite arrive at our present standpoint.In contrast to the founders of the atomistic doctrine, he attributedobjectively real sensible qualities to substances: he denied themonly to atoms. Now if the properties of substances only arise as aconsequence of the coming together of many atoms of the samekind, it becomes understandable why in the case of dispersioninto atoms, as effected by mixing, the properties of the previouslyexisting substances disappear and new ones arise from the jointaction of the mingled atoms. From this we learn how valuable an

instrument the atomic theory was, long before Dalton, in giving anintuitive explanation of chemical processes. The later developmentby Boyle and Lemery of more denite ideas about the cohesion of


11/34


smallest particles on entering into chemical combination, and theattempt by present-day atomic theories to comprehend the changein electronic conguration of atoms when coming together, are only

more detailed developments of the old lines of thought.25

In principle Epicurus would have already reached the presentstandpoint, had he not considered the qualities of substances tobe objectively real. Epistemologically this constitutes a step back-wards from Democritus. But this was of no practical consequence,since Epicurus did not, after all, ascribe qualities to the atoms, butonly to the substances formed by them; in other words, he onlymade the small epistemological simplication (retainingsome of thenaive-realistic convictions) which we mentioned initially as beingthoroughly appropriate and justied in chemistry. Since he was notguilty of this error in his atomism, he was able to explain changesof properties occasioned by chemical combination, in the way wehave sketched, without having to assume changes in qualities aspart of the nature of things. Aristotle, the opponent of atomism, didnot shy away from this consequence (a change in qualities) whichcould be well explained in terms of his philosophic principles. Weshall discuss presently the contradictions which arise if one followshim, as many modern chemists do, in not distinguishing the worldof the transcendental from the world of appearances, but speaksnevertheless of a persistence of the elements in compounds.

Epicurus denial (in common with the old atomists) of qualities

to atoms, and his admission of only one homogeneous primodialmatter does not, in our view, constitute an essential differencefrom the so-called qualitative atomism of present-day chemistry.An approach favoured particularly, but not exclusively, in popularexpositions is exemplied by the formulation that the atoms of goldare gold still, while the properties of a compound are retained inits individual molecules, but lost on decomposition into atoms. Thisversion, which is designed to make the difference between elementsand compounds intuitivelyunderstandable, is, after all, no more thana very inaccurate formulation. Epicurus would rightly have objectedthat only the coming together of many atoms generates the qualities,

and, in fact, no chemist would seriously maintain nowadays thatany property of gold could be discerned in the individual atom.Not only today when the electronic theory has already success-


12/34

124 F.A. PANETH

fully begun to elucidate the structure of atoms, does the notionof a single primary substance seem quite modern; for philosophicreasons qualitative atomism, too, could never ascribe qualities

in the strict sense to individual atoms. We must, therefore, regardthe conception of a primary substance as a necessary consequenceof their correct epistemological standpoint also in the case of theatomists,26 in contradistinction to the preceding thinkers. In the caseof the latter it was presumably only a rash generalisation or, if oneagrees with Joels opinion, shows the genesis of nature philosophyin the spirit of mysticism.27

As a result of these observations we afrm that some Greekthinkers had already realised that it is the aim of the natural sciencesto nd the laws of a world that is objectively real, whose changesare indicated in our consciousness by processes quite different inkind; and that to understand the change of properties of substanceswe require transcendental hypotheses. As the Greeks expressed theantithesis: the o has to be explained by means of the,

o .28After these preparatory historical remarks we will attempt to

determine the philosophic character of the modern concept of element.

4. THE EPISTEMOLOGICAL POSITION OF THE CONCEPT OFELEMENT INTRODUCED BY LAVOISIER

In the important passage in Lavoisiers textbook in which following Boyle, whom he does not name he introduces his newdenition of chemical element, he attacks the prejudice, for whichwe originally have to thank the Greek philosophers, that all bodiesin nature should be composed of only three or four elements andthe metaphysical investigations of the atomists.29 He wishes touse the name element for all substances which we cannot decom-pose further, because only in this way can we in fact recognise them.I take it to be superuous to discuss in greater detail how extraordi-

narily successful this denition has turned out to be. The whole of modern chemistry is, after all, based on it. Here I only wish to showthat it is impossible to overcome metaphysics as Lavoisier thoughthe had done by this denition. This apparent defeat of metaphysics,


13/34


which is still ascribed to him and to Boyle by numerous writersin chemistry even today, resulted in a recurrence of the misunder-standing already discussed: that the properties of simple substances

must necessarily persist in compounds.In the linguistic usage of atomists, which Lavisier seems alsoto have had in mine, the term elements designated the smallestparts of matter, the quality-less atoms.30 But this system had verymuch decreased in importance in scientic circles in the MiddleAges. Under the inuence of Aristotle, the four substances of Empe-docles predominated as elements; later, the three principles of the Spagirists sulphur, mercury, and salt competed with themfor recognition. While the former were, to put it briey, the carriersof physical properties, viz. the different states of aggregation, thelatter were already much closer to the chemical requirements of basic substances. They were supposed to confer the properties of combustibility, metallic nature, and resistance to re. For this reasonwe see, for instance, in the Chymista Scepticus of Boyle a noticeablygreater inclination to allow them as chemical elements. Presum-ably no one would doubt that we are dealing, in the case of theelements of the Aristotelians, as well as in the case of the prin-ciples of the trimaterialists, with the metaphysical concept of asubstance which is itself completely removed from our perception,but produces by its effects all the phenomena which make up theconnections between external experiences.31

As we mentioned at the beginning, the elements of Aristotle werenot properties in contradistinction to the substances of Empedocles.But, since Aristotle considered sensory appearances decisive, hewas forced to assume a transmutation of elements into each other.For this he could nd without difculty an explanation in agree-ment with his philosophic principles: matter, common to all thesub-lunary world, completely indeterminate, and present as merepotentiality, only comes into existence as one of the elements bythe form-giving addition of the properties, cold or warm and dryor moist; and as these properties changes, the elements changeinto each other in a perpetual cycle.32 In some passages Aristotle

expresses himselfsomewhat differently about the relation of the fourelements to the primary matter which latter already appears onoccasion as substance devoid of qualities, as with the Stoics33 but


14/34

126 F.A. PANETH

the polemical writings leave no doubt that the elements were alwaysunderstood as substances by his followers amongst the chemists.What Joachim Jung, for example, fought against was the actual-

potentialdoctrine of the Peripatetics just this change of propertiesof the elementary substances ;34 and what Boyle demanded asproof of their doctrines was the possibility of preparing the foursubstances . (This, incidentally, they believed they were able to showby experiments, to satisfy those that are not capable of a noblerconviction.)35 Only in the demand that elements should be prepar-able, and not, as often stated, in the view that elements must beconsidered non-decomposable substances, lies the advance of Jung,Boyle, and Lavoisier beyond the standpoint of the Aristotelians.Thus, J.F. Gmelin, for instance, denes:36 simple bodies in thechemical sense are those which can be no further decomposed intounlike particles by chemical artices; they are called by anothername elements ; and he explains on the basis of this denition completely modern in appearance, but whose prototype is to befound in Aristotle himself 37 that we have very good reason toallow the name of elements to re, air, water and earth, if we imaginethem in their greatest purity. In the same way the Encyclopaediaof Diderot and DAlembert characterised the chemical elements assimplest, indestructible bodies persisting in their compounds, butnevertheless recognised essentially the Peripatetic elements.38

Lavoisier obtained general recognition for the principle that one

should not make a priori statements about the number of elements,but should determine a posteriori how many substances we actuallynd in nature which cannot be further decomposed, and regard theseas the constituents of the composite ones. In this way an experi-mentally determinable criterion was introduced into the denition of element, and the interminable and obscure discussions about the trueelements brought to an end. Nevertheless I cannot agree with thosewho believe that anything has thereby been changed concerningthe metaphysical nature of the concept of element. The essentialpoint, after all, is the assumption, which is retained, that the simplesubstance is present in some latent form (i.e., in such form that the

properties which it has in the pure state are not recognisable) inthe composite ones, and that it explains their behaviour. Lavoisiersdemand, that it must be possible to isolate any substance assumed


15/34


to be a chemical element, i.e., to prepare it as simple substancewith denite properties, has proved very apt, since it is possible toprepare most elements in the pure state without too much difculty.

For a proper appreciation of this principle from the philosophic pointof view, however, it is very instructive to consider the exceptions.In Lavoisiers rst table of elements we nd light-substance

and heat-substance. These had not been obtained as simplesubstances, and certainly Lavoisier did not believe even in thetheoretical possibility of preparing them in the pure state. Theywere principles for the explanation of certain properties, and itwould be difcult to state any essential way in which they differedfrom the principles of the alchemists or the old element re. Soon,advances in physics caused light and heat to be deleted from the listof substances, and thus also from the chemists table of elements.But even after the chemists had conned themselves to ponder-able substances, the decision as to which should be included in thetable required more than just a schematic application of the newdenition, if the value of the concept of element was not to beconsiderably reduced. Thus Lavoisier allowed a place amongst thesimple bodies to the chemical compounds lime, magnesia, barytaand alumina, which had not been decomposed at the time, althoughhe considered that they were probably oxides; but he omitted potashand soda, because he was already rmly convinced that they werecomposed of basic constituents still to be discovered. This was

almost twenty years before Davy succeeded in obtaining the newmetals potassium and sodium from them. Here, then, Lavoisierhimself assumed hypothetical elements; these had no more beenproduced than, say, the purest earth of the Aristotelians, or thespecial sulphur of the Spagirists (which is not the same as commonsulphur). His translator, Hermbstdt, wanted to avoid the uncertaintywhich lay in the indirect inference by analogy to elements, contraryto the principle of pure empiricism just put forward: he put potashand soda back into the table.39 However, it was not his pedanticlogical consistency, but Lavoisiers happy instinct, that proved to beright.

The inferring of elements which cannot be prepared must notbe regarded as in any way merely a necessary imperfection of the initial stage of chemistry. Later, too, elements were acknowl-


16/34

128 F.A. PANETH

edged with complete certainty, though their preparation as simplesubstances had not been achieved. Thus nobody doubted the exist-ence of uorine even yet before the famous work of Moissan, since

one could not explain the special properties of hydrouoric acidand its salts with the known elements alone, and thus needed a newprinciple. Indeed, after the discovery of chlorine, arguments fromanalogy led to fairly clear ideas about the nature of this radical of hydrouoric acid which is already named amongst the simplesubstances in Lavoisiers rst table. If one wishes to cite a veryrecent example of this procedure: the preparation of radium asa metal was a great experimental success of Madame Curie, butwithout fundamental signicance; radium had already obtained itsassured place in the system of elements, because the assumptionof a new basic substance was necessary to explain the radio-activeproperties of its salts.

If we accept the legitimacy of this procedure, leading, in theinstances quoted above, to the postulation of the existence of parti-cular basic substances, we must not raise any fundamental objectionto the inferences which were used to justify the conception of themedieval elements. For example, the statement by which Stahl justi-ed taking aether, water and earth as principia prima also appliesexactly to the cases cited here: Simplicia corpora et elementa,quamvis non pura et separata dentur, realiter tamen distincta amente possunt concipi .40 Sodium, uorine, and radium could not

be isolated in the pure state either at the time when they were intro-duced into science as basic substances, but could only be rationallyinferred. However, the vast chemical experience gained in the mean-time, and the guidance of analogy, made inferences incomparablymore certain in these cases.

Thus we consider the great advance due to Lavoisier to consistin this: he did not limit the number of chemical basic substances a priori , but he demanded that they could in principle be prepared asnon-decomposable substances. Further and this is the point wherehe went beyond Jung and Boyle he gave a means, the balance, of proving a substance to be non-decomposable. However, as is shown

particularly clearly by those cases in which the demand for prepar-ability was dropped, the metaphysics of the basic philosophic ideaof the concept of element is not touched by this.


17/34


5. THE DOUBLE MEANING OF THE CHEMICAL CONCEPT OFELEMENT: BASIC SUBSTANCE AND SIMPLE SUBSTANCE

Admittedly it seemed as if Lavoisier had changed the philosophiccharacter of the concept as well. His elements seemed as onecan read in numerous books no longer some sort of metaphysicalcauses of the behaviour of bodies, as with the alchemists, but visibleand tangible substances with quite denite observable properties.41Indeed, Pattison Muir, the editor of a new edition of the ScepticalChymist , even wishes to read an analogous view into Boyle, who,as a declared adherent of atomism, should have been safe fromsuch misinterpretations. Muir, the considers that Boyle was wellon the way to the pragmatic point of view that an exact sciencemust not assume an invariant substratum devoid of qualities, butmust regard a substance only as the sum of its properties which areperceptible to the senses.42 To the question: what is the elementsulphur?,Gebers answer might have been: the carrier of mutabilityby re. The present-day chemist describes the element sulphur as asubstance of pale yellow colour, without taste or smell, insoluble inwater, etc. All these are its properties. How is it that it does not retainthem in its compounds? What sense at all is there in saying that theelement sulphur is preserved unchanged in its compounds, such asthe gaseous, colourless, pungently smelling sulphur dioxide?

What has been said in the previous section should make ourmeaning clear immediately, when we say that the element sulphurrepresents today, no less than in the old days, a transcendentalprinciple underlying the phenomena; and that the mistake in thestatements cited above lies in the naive-realistic assumption that theproperties of sulphur are just those properties which it exhibits to usin the special case when it is not combined with any other basicsubstance. Admittedly, as has been explained above, we may, inconsidering a particular substance, speak just as well of its proper-ties as of the properties we perceive in it. But Lavoisiers denitionmust not mislead us into applying this simplication to the conceptof element, which is used in two different senses . I have already

attempted to keep apart these different meanings in the abovediscussion by using the two terms basic substance and simplesubstance.43 I have referred throughout to basic substance whenever the indestructible substance present in compounds and


18/34

130 F.A. PANETH

simple substances was to be denoted , and to simple substance whenever that form of occurrence was meant in which an isolated basic substance uncombined with any other appears to our senses .

We cannot ascribe specic properties to an element in the senseof basic substance, since the latter contributes to the generation of the innitely diverse properties which it exhibits, by itself and incombination with other basic materials; as simple substance it canbe characterised by the statement of its properties without preju-dice to scientic exactness, as we have earlier shown in general forany substance. In the case of the concept of simple substance wemay remain naive realists; but in the case of basic substance, if we are not to get involved at once in contradictions, we must notoverlook the fact that it belongs to a transcendental world devoid of qualities.44

Even to this day the customary denition follows Lavoisier: anelement is a substance that cannot be decomposed into simplerones by any chemical procedure.45 Thus, to be sure, the assumedcriterion for an element is that it may be obtained as a simplesubstance which cannot be decomposed further. But the greatsignicance of elements for the whole body of chemical theorylies in the assumption that the substances which produce thephenomenon of simple substances serve in the quality-less,objectively real sphere of nature as basic substances for theconstruction not only of simple but also of composite substances.

The fundamental principle of chemistry that elements persist intheir compounds refers to the quality-less basic substances only .In any other sense the statement is unintelligible. Ostwalds inter-pretation of the principle to the effect that there is the possibilityof regaining the element in unchanged amount from every one of its compounds46 is no way out, for, if that is all that is meant,one should not speak of persistence in compounds. One is forcedinto such a re-interpretation, and so to an abdication of any under-standing of the phenomenon, if one assumes that any substance,and hence any element, is characterised by the sum of its proper-ties. Probably no other example shows more strikingly that the

simplest and most familiar scientic laws are only valid under theimplied or explicit assumption of a transcendental world.


19/34


If it is maintained that the elements are not present in theircompounds in actuality (as basic substances), but only potentially(as simple substances), then the fact that the properties of their

compounds can form a basis for the most profound systematics of the chemical elements becomes inexplicable. As is well known, thegroups of the natural system of elements indicate the similarityof the properties of compounds to a greater extent than that of theproperties of simple substances we need only recall how littleresemblance there is between gaseous nitrogen and pure antimony.Thus, in terms of the distinction introduced here, we may refer onlyto a natural system of basic substances, not of simple substanbces.47Perhaps it is the many relationships between the forms of combi-nation of chemical elements, expressed by the natural system,that most clearly reveal that, behind the simple substances andcompounds endowed with constant properties, there stand thetranscendental basic substances, producing, Proteus-like, innu-merable forms, yet nevertheless bound by rigid rules; and that it is just the task of chemistry to discover the laws of this transcendentalworld.48

Our present day chemical language does not, in general, reectthe difference between the meaning of basic substance andsimple substance in the nomenclature of the elements. Only ina single case, that of carbon, are different names actually usedin German, depending upon whether we are referring to the basic

substance or the simple substance; and even here the distinction isnot carried through strictly. The term carbon [Kohlenstoff ] usuallydenotes the basic substance only, while the simple substance isreferred to as charcoal [Kohle ]. We speak of the carbon- (notcharcoal-) content of organic compounds, but of the properties of charcoal (not of carbon). Admittedly this distinction is not alwaysobserved. Thus the shorter name Kohlendioxyd is frequently usedinstead of Kohlenstoffdioxyd , and, on the other hand, we speak of amorphous carbon instead of amorphous charcoal. At any rate, wend here a distinct beginning of a linguistic separation of the twomeanings, favoured in this case by the fact that the free element

occurs in different allotropic forms (charcoal, diamond,graphite); inthis way the difference between the basic substance, common to allof them, and its forms of occurrence as a simple substance, is made


20/34

132 F.A. PANETH

particularly apparent. For this reason a new articial term [Kohlen-stoff ] was introduced to denote the basic substance. This name was,to be sure, derived from the name of the most common form [Kohle ]

of the free element, but not made synonymous with it. However, inanother case where two different forms of the free element exist, thename of the more abundant form is used for the basic substance also;oxygen [Sauerstoff ] is the name of the basic substance of both thegas called oxygen and of ozone. It would certainly be helpful tohave a special name for the gas oxygen, as well as for ozone.49,50However, I do not, of course, wish to suggest by this a linguisticreform impossible to carry out so late in the day, but only to pointout what would seem to be indicated in the interest of logic. Muchconfusion could thus be avoided. Take, for instance, the followingsentences, correct according to present chemical usage: Copperoxide consists of copper and oxygen, Copper is a metal; oxygenis a gas. Therefore, one must conclude, copper oxide consistsof a metal and a gas, which is obviously nonsense. The fallacyhere is based on the use of copper and oxygen in the sense of basic substances in the rst sentence and in the sense of simplesubstances in the second. Thus we have a quaternio terminorum made possible by linguistic identity.

To sum up: using the two terms, we may say that the importantadvance in Lavoisiers denition of element lay in inferring theexistence of a basic substance from each simple substance found

experimentally.51

An accompanying disadvantage was that, in theeyes of many, the meaning of basic substance was completelyhidden by that of simple substance; the close relationship of the modern concept of element, also, to the alchemists principleswhich underlie appearances, was obscured, and the principle of explaining the on the basis of the , , already foundnecessary by the Greek philosophers, was abandoned as unscienticmetaphysical speculation.

Of the two senses in which the expression chemical elementmay be used, philosophical writers of today usually consider onlyone: basic substance. Accordingly we nd the concept of element

subordinated to the concept of substance, and its metaphysicalcharacter emphasized. In chemical textbooks, however, this abstractmeaning has been lost, precisely because of Lavoisiers inu-


21/34


ence, and elements are understood to be not devoid of qualitiesbut substances with certain denite properties. With reference tothe elements the chemists, as Wundt justly says, have in many

cases not got beyond the concept of thing derived from commonexperience.52 I have found only two instances in which the twodifferent meanings of the term chemical element are pointed out.In the article Elements by C. Hell, in the Neue Handwrterbuchder Chemie ,53 it is emphasised that this may mean the simplesubstances as well as the material constituents of the simple andcomposite substances themelves. According to the second deni-tion, the article continues, the concept of element coincides withthat of atom, and serves mainly to designate and individualise thelatter more closely; the atoms are the true elements of bodies, astatement which is not, in my opinion, correct. The atomic theorycan, it is true, contribute enormously to indeed, may be neces-sary for visualising how the basic substances persist in simplesubstances and compounds;54 but the concept of basic substance assuch does not in itself contain any idea of atomism. It was, afterall, while explicitly rejecting atomism that Lavoisier carried thisconcept to victory; and also in more recent times, there were, andare, chemists who avoid the atomic theory but retain the elements,including, of course, elements in the sense of basic substances.55

The distinction is also found in Mendelejeffs textbook of chemistry. Mendelejeff emphasises very forcefully the difference

between the concept of simple body as a single homogeneoussubstance, and as the material constituent, not perceptible to thesenses, of a composite body.56 Only in the latter sense does he calla simple body an element. I can think of only one reason why hisvery justiable distinction, and his other passages based on it, whichI have quoted earlier, have not produced any echo in later chemicaltextbooks;57 namely, his inapt terminology. Element and simplebody were introduced by Lavoisier explicitly as synonyms (cf. alsothe expression elementary sulphur, customary today, by which ismeant sulphur as simple substance), and it is thus hardly possible touse them to designate opposites. For this reason I have preferred to

speak of basic substance and simple substance as different aspectsof the chemical concept of element .


22/34

134 F.A. PANETH

6. THE DOUBLE MEANING OF OTHER CHEMICAL CONCEPTS

Our discussion would be incomplete if we did not, in conclusion, at

least indicate that the distinction between basic substances devoidof qualities and simple substances perceptible to the senses, whichwe have applied in the case of the concept of element, is, in avery similar form, of importance in other problems of chemistry.It applies quite generally whenever we speak of the persistence of asubstance under chemical change.

The radicals of organic chemistry, for example, exist almostwithout exception in the world of the transcendental alone.58 Theopponents of the theory of radicals have always demanded, beforethey could believe in them, that radicals be prepared as observ-able substances. But this demand we must consider unjustied, forthe same reasons adduced above to show that the existence of achemical element as basic substance may be completely provedwithout its having been isolated as simple substance.59 Radicalsare, in fact, an even more convincing example of the transcendentalcharacter of the laws of conservation in chemistry; for in sharpcontrast with the case of the elements, the majority of which canbe prepared easily as simple substances the old proofs of theindependent existence of free radicals, perceptible by the senses,were all false and were soon refuted; nevertheless the assumptionof radicals was completely justied. Once the attempt to prepare

radicals had been given up, their use as chemical principles devoidof qualities could be developed in full measure, without danger of epistemological obscurities and misunderstandings.

The situation is different when we consider the statement that theso-called compounds of rst order, e.g., sulphur trioxide (SO3)and water (H2O), combine into compounds of higher order suchas sulphuric acid (SO3.H2O), for here we know the compoundsof rst order as perceptible substances. After what has been saidabove it should be clear, however, that is is again only possible toassume their continued existence in the compounds of higher orderif we transpose them, divested of their qualities, into a transcen-

dental world. This example is, however, less clear-cut, for, from thechemical point of view, it is also possible to hold that the compoundsof rst order are not preserved as such, even in a transcendentally


23/34


real sense, but rather that only the atoms contained in them arepreserved, in a new arrangement. Sulphuric acid, say, would thenbe formed from SO3 and H2O, but would no longer consist

of them (formulation: H2SO4). It is clear that examples are theless suited for discussion of logical principles, the less certain thescientic situation is with regard to them. For this reason we haveconned ourselves in our discussion mainly to the undisputed lawof persistence of the chemical elements.

Nevertheless we will consider one other, somewhat more com-plicated, chemical law of conservation, viz. the well-known law of chemistry that a salt is formed from acid and base with productionof water. We will discuss this theory, which constituted a funda-mental scientic advance, because violent objections were raised byphilosophers when it rst appeared. This led to an interesting contro-versy between one of the greatest chemists, Boyle, whose name hasalready appeared repeatedly in our discussion, and one of the mostfamous philosophers,Spinoza. This scientic polemic may be foundin Spiozas correspondence60 of the years 16611663.

Boyle had recently demonstrated that nitre may be synthe-sized from an acid and a base; he therefore considered these twosubstances, endowed with completely different properties, to be thecomponents of nitre. The secretary of the Royal Society, Olden-burg, had sent the paper in question to Spinoza, asking for hiscriticism. In two long letters to Oldenburg, Spinoza thereupon put

forward his objections to the scientic standpoint of Boyle whichcorresponds completely to the view taken in chemistry even today.Spinoza is a Cartesian and admits only one single primary matter,whose corpuscles differ from each other only in shape, position andmotion.61 He censures Boyle for putting on the same level thoseconcepts which show nature, not as she is in herself, but in her rela-tion to the human senses, and those which explain nature as she is inherself. As the latter he only admits motion, rest and their laws.He considered it so improbable that homogeneous nitre should becomposed of two compounds so different in kind, as Boyle claimedto have proved, that he performed experiments especially to refute

him; and indeed he believed he had found that it was unnecessaryto accept any difference between spirit of nitre [now called nitricacid] and nitre, other than this, that the particles of the latter are in a


24/34

136 F.A. PANETH

state of rest, whilst those of the former are in violent mutual motion.Nitre and spirit of nitre are related as ice is to water. The solidbasic residue, the calx of nitre he declared to be not an essential

component of nitre, but an impurity.We need not dwell on Spinozas experiments. With the limitedfacilities at his disposal and his utter lack of practice in this eld,he was unable to obtain useful results. It was only owing to certainerrors which Boyle, with his great experience, was able to pointout immediately that Spinoza believed that he had also been able toprove his views by experiments. Of interest to us is his rm convic-tion that Boyles experiments and their interpretation must be false,because they do not correspond with his idea of the essential natureof scientic explanation. Here, at the beginning of scientic chem-istry, we encounter, in the person of Spinoza, the rst but by nomeans the last philosopher who will only admit the same conceptsto be valid in chemistry as in physics. We must admit that formallyhe is in the right, in that Boyles statement that nitre consistsof volatile nitric acid and a solid residue shows just that oscilla-tion between the naive-realistic and the transcendental meaningsof the terms which we exhibited above as particularly character-istic of the concept of element. Nevertheless, it is surprising thatSpinoza did not understand Boyles epistemological point of viewat all; for, though the latter nowhere explicitly points out this doublemeaning, he nevertheless discusses at some length the application of

corpuscular theory in the explanation of chemical processes, evenin the paper considered by Spinoza. These deliberations shouldhave shown Spinoza how Boyle really pictured the persistence of chemical substances: he meant it in the sense of constancy, notof their qualities, but only of the structure of their particles in atranscendental world in which acids are not sour, bases are notalkaline, and salts not salty. In the deepest convictions of his philo-sophy of nature he was no less a Cartesian than Spinoza. It is tobe regretted that he never entered into direct exchange of ideaswith him (probably his strong religious sensibility prevented him);62for we may well believe that he would have been able to dispel

Spinozas objects. He conned himself rather to giving Oldenburghis answers in the course of conversation:


25/34


He had not intended to give a truly philosophic and thorough explanation by hisanalysis of nitre. He had never undertaken to discover the nature of nitre, norto reject what anybodycould teach concerning the homogeneityof matter and thedifference of bodies arising only from motion, shape, etc. This only he wantedto show: that the various structures of bodies produce their various differences,and that from these proceed very different effects. Philosophers and others rightlyinferred from this a certain heterogeneity, as long as reduction to one primarymatter had not yet been achieved.

Thus, Spinoza only admits into scientic research mathemati-cally denable data about the particles of primary matter, andexcludes all sensible qualities. Boyle, on the other hand, as empha-sized at the beginning, recognises that this aim is unattainablefor chemistry, and that an intermediate standpoint, which assumessubstances of several kinds, may also be genuine science.63 AsOldenburg wrote to Spinoza: Boyle belonged to those who do notrely so completely on their thought that they believe themselves freeto ignore whether or not it agrees with the phenomena of reality.

In the development of the concept of matter, science and philo-sophy have rarely been in agreement, Wundt64 says at one point,and here, if anywhere, their quarrel is understandable. Philosophylooks above all to the strict logic of its concepts, science to agree-ment with experience. The way in which Boyle and Spinozarespectively interpret the phenomena in the case of nitre is typicalof this difference in mental attitude.

The very thing Spinoza censures, the mixing up of thoseconcepts which show nature in her relation to the human senses,with those concepts which explain nature as she is in herself hasremained the well-founded procedure of chemistry to the presentday. Thus, even today, to repeat Boyles words also, it is still legi-timate to infer a certain heterogeneity from the different effectsof substances, as long as reduction to one primary matter has notyet been achieved. Indeed, we may add that even when one daythis problem is solved, chemistry, as the science of the changes of substances, will still be justied in going no further than the reduc-tion of the phenomena to the chemically indestructible substances,

the elements, and thus in retaining qualitative differences in itsfundamental concepts. The object of the present study has been themore precise denition and elucidation of the peculiar dual episte-


26/34

138 F.A. PANETH

mological status of the concept of element, which is a consequenceof this. As has been shown, this concept must be taken in the naive-realistic sense when meaning simple substance, but has to be

understood as transcendental when meaning basic substance. Wehave also tried to emphasise the close relationship of the concept of basic substance to the metaphysical principles of the Aristoteliansand the alchemists.

NOTES

1. Exceptions to this rule are the writings of H. Rickert (particularly Die Grenzen der naturwissenschaftlichen Begriffsbidung . Tbingen, 1921);mile Meyerson (particularly Identit et Ralit . Paris, 1908, and DelExplication dans les Sciences . Paris, 1921); and E.A. Burtt. The Meta- physical Foundations of Modern Physical Science . London, 1925.

2. To name but a few of the best-known works: E.v. Meyer. Geschichteder Chemie . Leipzig, 1905, p. 7 (English translation, 3rd edition, 1906);W. Ostwald. Der Werdegang einer Wissenschaft . Leipzig, 1908, pp. 5, 9;W. Ramsay. Essays Biographical and Chemical , 1908, pp. 7, 25; S. Arrhe-nius. Theorien der Chemie . Leipzig, 1909, pp. 11, 14; F. Soddy. Matter and Energy . London (no date given: circa 1912), p. 39; W. Herz. Grundzgeder Geschichte der Chemie . Stuttgart, 1916, pp. 9, 35; F. Dannemann.Grundzge einer Geschichte der Naturwissenschaften . Leipzig, 1903, II,p. 188; A. Sieverts in the Handwrterbuch der Naturwissenschaften . Jena,1912, II, p. 438.This questionablepresentationof Aristotles doctrinescan probably be tracedback to H. Kopps. Geschichte der Chemie (Braunschweig, 1843, I, p. 30).We must add, however, that Kopp himself later presented a perfectly correctformulation in the Beitrge zur Geschichte der Chemie ( Ansichten ber die Aufgabe der Chemie , usw., Braunschweig, 1875, p. 5). Unfortunately littlenotice has been taken of this latter version.

3. Cf. E. v. Hartmann. Kritische Grundlegung des transzendentalen Realismus ,1875; Das Grundproblem der Erkenntnistheorie , 1889; Grundriss der Erken-ntnislehre . Bad Sachsa, 1907; Grundriss der Naturphilosophie . Bad Sachsa,1907.I wish to emphasise particularly that, like Hartmann, I am using the wordtranscendental in its epistemological sense only, i.e., meaning beyond thesphere of consciousness.

4. Even epistemologists who deny the actual existence of a world devoidof qualities allow this as an epistemological principle of physics; see,for instance, H. Rickert. Psychophysische Causalitt und psychophysischerParallelismus (in Sigwart-Festschrift , Tbingen, 1900), p. 75.


27/34


M. Schlick ( Allgemeine Erkenntnislehre . Berlin, 1918, pp. 214, 243), in parti-cular, has emphasized that the exclusion of qualities can only be completewith respect to the sensible qualities.

5. Cf. E. v. Hartmann.Grundrissder Naturphilosophie . Bad Sachsa, 1907,p. 23.6. E. v. Hartmann. Grundriss der Erkenntnislehre . Bad Sachsa, 1907, p. 73.

Similarly Grundriss der Naturphilosophie . Bad Sachsa, 1907, p. 105. Cf.also W. Wundt. Logik , Stuttgart, 1883, II, p. 364: Everywhere the concept of matter reduces to moving centers of force.

7. See, for instance, Sir James Jeans. The Mysteriour Universe . Cambridge,1930. For a survey of the changes in the concept of matter in physics seethe essay by H. Weyl. Naturwissenschaften 12: 561, 585, 604, 1924.

8. Robert Boyle. The Origin of Forms and Qualities. The Works of the Honour-able Robert Boyle , Birch Edition, III: 22 sqq., London, 1772.

9. Certain Physiological Essays. A Proemial Essay. The Works, etc . I: 310.10. Ibid. p. 308. I consider then, that generally speaking, to render a reason

of an effect or phaenomenon, is to deduce it from something else in naturemore known than itself; and that consequently there may be divers kinds of degrees of explication of the same thing. For although such explications bethe most satisfactory to the understanding, wherein it is shewn, how the effectis produced by their more primitive and catholick affections of matter, bulk,shape and motion; yet are not these explications to be despised, whereinparticular effects are deduced from the more obvious and familiar qualitiesor states of bodies, such as heat, cold, weight, uidity, hardness, fermen-tation, &c. though these themselves do probably depend upon those threeuniversal ones formerly named. For in the search after natural causes, everynew measure of discovery does both instruct and gratify the understanding;though I readily confess, that the nearer the discovered causes are to those,that are highest in the scale or series of causes, the more is the intellect bothgratied and instructed.

11. See K.A. Hofmann. Sitz.-Ber. Preuss. Akad. der Wiss ., 1931, pp. 56, 64.12. See, e.g., G.M. Schwab. Die Katalyse vom Standpunkt der chemischen

Kinetik . Berlin, 1931, p. 3.13. I. Kant. Metaphysische Anfangsgrnde der Naturwissenschaft . Vorrede,

1786.14. Emil Fischer. Aus meinem Leben . Berlin, 1922, p.118. Cf. H. Dingle ( Nature ,

1931, 127, 526): The schoolboys name for chemistry is stinks, notbalances, and a very appropriate name it is.

15. We wholly disregard here the deliberations of Wald and Ostwald, who deviatefrom the usual notion of element and follow their own anti-atomistic lines of thought. Similarly we shall disregard all remarks (for instance, those of FritzMauthner in his Wrterbuch der Philosophie , Munich, 1910) that remain onthe philological level and fail to grasp the scientic content of the concept.

16. H. Diels. Die Fragmente der Vorsokratiker . Berlin, 1906, I, p. 317, Fragment10.


28/34

140 F.A. PANETH

17. Cf. for instance, Fragment 8 of Empedocles (H. Diels, op. cit. p. 175): Thereis no birth of any mortal things, nor any end in accursed death. Rather thereis only mingling and exchange of things mingled. Birth is only a name givento this by man.

18. Sextus Empiricus. Adv. Mathem . X, 46 (see H. Diels, op. cit. p. 109).19. Th. Gomperz. Griechische Denker . Leipzig, 1911, p. 185. On the other

hand, I do not agree with the view of W. Windelband ( Lehrbuch der Geschichte der Philosophie , Tbingen, 1907, p. 34) who sees in the homoio-meries of Anaxagoras the precursors of the chemical elements. Preciselywhat Anaxagoras denies the generation of completely homogeneous newsubstances from others different in kind is characteristic of chemicaldoctrines.

20. J.S. Mill. A System of Logic . Book III, Chapter VI.21. Cf. E. Meyerson. Identit et Ralit . Paris, 1908; cf. also p. 13, n. 1.22. C. Bumker. Das Problem der Materie in der griechischen Philosophie .

Mnster, 1890, p. 82.23. Aristotle already noted that Empedocles doctrines, worked out to their

logical conclusion, necessarily lead to atomistic ideas (see C. Bumker, op.cit. p. 71).

24. Cf. P. Natorp. Forschungen zur Geschichte des Erkenntnisproblems im Altertum . Berlin, 1884, pp. 209 sqq.

25. In more recent daysLemerys extreme ideas of atoms with points which breakoff and remain stuck in other atoms (see, for instance, his explanation of theaction of aqua regia in his Cours de Chymie ) have been censured sharplyby opponents of atomism. This criticism appears less reasonable when weconsider the striking analogy with the most modern theories, in which wespeak of electrons traveling in far, outer orbits of strongly reactive atoms,and assume that these electrons become ineffective by entering gaps in theelectron rings of other atoms when a chemical compound is formed (cf.W. Kossel. Ann. d. Phys . 1916, 49, 229).Boyles view on the cause of the change in properties on formation of achemical compound is expressed in the second part of the Sceptical Chymist (The Works etc ., I, pp. 458, 506 sqq.).

26. Cf. the passage from Theophrastus in Diels, op. cit. I, 371, 42 (cf. Bumker,op. cit. p. 91). It is remarkable that many present-day chemists have lackedthis epistemological sense. They have seriously maintained that the proper-ties of the elements should be preserved completely even in the atoms.Thus the disappearance of the original properties of an atom, when itundergoes chemical combination, has seemed to them to contradict the basicassumptions of atomism (I.W. Hinrichsen. ber den gegenwrtigen Stand derValenzlehre. Ahrens Sammlung chemischer Vortrge 7: 189, 277, 1922).

27. See Karl Joels essay Ursprung der Naturphilosophie aus dem Geist der Mystik (Jena, 1906) for an attempted justication of this idea.

28. Cf. C. Bumker, op. cit. p. 313.


29/34


29. A.L. Lavoisier. Trait lmentaire de chimie 2nd edn., 1: XV and XVII. Paris,1793.

30. See H. Diels. Elementum, Eine Vorarbeit zum griechischen und lateinischenThesaurus . Leipzig, 1899, pp. 5 sqq.

31. W. Wundt, loc. cit. I, p. 485. Using E. v. Hartmanns terminology we shouldput transcendental in place of metaphysical, pseudosubstance in placeof substance (see, for instance, Grundriss der Erkenntnislehre , p. 221).

32. Cf. principally De generatione et corruptione , Section B. That it is quitewrong to put the Aristotelian elements on the level of properties is shownin concise form for instance by the passage B I, 329 a 32:

o `

, , , o , , o o ` , o

`

` ` o . ` ` ,

. . . o, o .[Therefore,rstly, the potentiallyperceptiblebody, secondly, the contrarieties

(e.g., heat and cold), and thirdly, Fire and Water and the like are sources.For the bodies in this third class change into one another . . . whereasthe contrarieties do not change. (Translation by E.S. Forster. Aristotle, OnComing-To-Be and Passing-Away . Heinemann, London, 1955.)]

33. On this point see C. Bumker, loc. cit., p. 260.34. See E. Wohlwill. Joachim Jungius und die Erneuerung atomistischer Lehren

im 17. Jahrhundert. Abjandl. a. d. Gebiet der Naturwissenschaften 10.Hamburg, 1887.

35. R. Boyle. The Sceptical Chymist ; from the speech of Themistius in Physiolo-gical Considerations. The Works etc . I: 469.

36. J.F. Gmelin. Einleitung in die Chemie . Nrnberg, 1780, p. 36.37. Aristotle. De coelo , C 3, 302 a 15.

38. SeeH. Kopp. Die Entwicklung der Chemie in der neueren Zeit . Munich,1873,p. 111, footnote.39. A.L. Lavoisier. Trait lmentaire de chimie . German translation by S.F.

Hermbstdt. Berlin, 1803, p. 247, footnote.40. G.E. Stahl. Fundamenta Chymiae dogmatico-rationalis et experimentalis .

Nrnberg, 1732, p. 1.41. To quote only one of the more recent authors: Vergleicht man nun die

Elemente Boyles mit denen des Aristoteles und den Prinzipien, so erkenntman leicht, dass sie begrifich vllig von ihnen verschieden sind. Unter denalten Elementen versteht man verkrperte Eigenschaften, unter den neuenin der Natur vorkommende Stoffe. (A. Benrath. Chemische Grundbegriffe .Berlin, 1920, p. 26.) [If one compares Boyles elements with those of Aris-totle and the Principles, one soon realises that the former are conceptuallyquite distinct from the latter. The old elements are to be understood asembodied properties, the new elements are substances occurring in nature.][Note by translator: cf. also J.R. Partington. A Short History of Chemistry ,2nd edition. 1948, p. 67: . . . he [Boyle] gave a clear denition of an element


30/34

142 F.A. PANETH

and showed by experiment that the four elements of Aristotle and the threeprinciples of the alchemists (mercury, sulphur, and salt) did not deserve tobe called elements or principles at all, since none of them could be extractedfrom bodies, e.g., metals.]

42. The Sceptical Chymist by the Hon. Robert Boyle, Everymans LibraryEdition, p. xviii. Similarly, it will be clear from the above argument that wereject completely the view expressed by M.M. Pattison Muir elsewhere (The Alchemical Essence and the Chemical Elements . London, 1894) that there isan unbridgeable opposition between alchemy and chemistry.

43. Many textbooks do not use the term Grundstoff [basic substance] at all,others dene it at one point as synonymous with chemical element, but use itno further. It has, at any rate, not been needed to specify concepts until now,and may therefore be used with a new signicance.[Translators note: throughout this article, we have used the termsbasic substance and simple sunstance as translations of the GermanGrundstoff and einfacher Stoff respectively.]

44. It goes without saying that, in the case of simple substance also, any closerconsideration is bound to lead us to the well-known contradictions of naiverealism. We need only think, for instance, of the change in properties of the simple substance sulphur on transition from the rhombic into the mono-clinic form, and even more on transition into the gaseous state; or again,of the impossibility of explicating potential qualities, such as solubility orcombustibility, in this view of the world.In general, it is true, when we consider the simple substances of chemistry,these contradictions need not obtrude into our consciousness, any more thanthey do in the case of objects of daily life; they are in fact ignored in the usuallaboratory practice.

45. This restriction to chemical procedures in the denition of chemical element

is appropriate today after the discovery that nitrogen, aluminium, are severalother elements, may be disintegrated not indeed by chemical means, butby the impact of alpha-rays and that most elements are mixtures of twoor more chemically inseparable isotopes. Another modern denition to bepreferred for several reasons runs: A chemical element is a substance all of whose atoms have the same nuclear charge. This denition differs from theconventional one in its more theoretical tone. But here, too, the possibility of preparing thesimple substance is demandedas a necessary condition. For thisreason we need not here consider more closely the advantages and disadvant-ages of this denition. (On this point, see Zs. f. phys. Chem. 91: 171, 1916; Naturwissenschaften 6: 646, 1918; 8: 839 1920.)[Note by translator: In these early publications, the author advocated thatthe term chemical element should continue to denote such substances aslead and chlorine, even though these substances were then known to consistof several isotopes of different atomic weight, and although some of theseisotopes could be separated. Against the view of a number of chemists, heheld that, from the chemical point of view, it was better to drop Daltons prin-


31/34


ciple that there were as manyelements as atomic species, rather than abandonthe chemical classication of elements in the periodic system. His argu-ment involved a distinction between physical and chemical (includingelectro-chemical) methods of analysis, and is rooted in the experience of early experiments establishing the practical impossibility of distinguishingbetween isotopes by chemical means. These experiments led to the methodof radioactive tracers.]It is only in the formulation: A chemical element is the class of all atomsof equal nuclear charge that the chemical element would be dened withcomplete generality as basic substance, and no longer as simple substance.

46. W. Ostwald. Vorlesungen ber Naturphilosophie . Leipzig, 1902, p. 287.Similarly W. Ramsay, loc. cit. p. 148.

47. Cf. D. Mendelejeff. Grundlagen der Chemie . German translation byL. Jawein and A. Thillot, Petersburg, 1891, p. 688. [Cf. Principles of Chemistry , 3rd English edition. London, 1905.]

48. The atomic theory of Rutherford and Bohr enables us to visualize parti-cularly vividly how we are to understand the persistence of an element inits compounds; namely, as the unchanged presence of all atomic nuclei of the simple substance. The simple substance with its optical, chemical, etc.properties has disappeared because the electronic shells of its atoms haveundergone changes. (Only two of its properties, weight and where present radioactivity, are unchanged, because they reside in the atomic nucleus.) Thesimplesubstance can,however, always be recovered,with all its properties, aslong as the atomic nuclei are not destroyed, because in this case the electronscan be brought back into their original positions, i.e., the positions whichthey occupy in the absence of alien atomic nuclei. (Signicantly, claricationis obtained here through the notions of physics, a science whose concepts donot have the naive-realistic coloration of the chemical concept of element.)

49. Cf. D. Mendelejeff, loc. cit. English edition, 1: 207, footnote 17.50. Note by translator: I have omitted here a remark to the effect that it wouldhave been desirable to reserve the end-syllable -stoff in German for basicsubstances.

51. For the sake of simplicity we ignore here the possibility that a simplesubstance may occur in allotropic forms.

52. W. Wundt, loc. cit. I, p. 484. Compare with this thestandpoint already reachedby Basil Valentine (circa 1600), who distinguished between salt, sulphur,and mercury as preparable substances and as seeds. This latter term coin-cides with what we refer to as metaphysical principles; for instance, in thepassage: Er ist der sichtige Mercurius metallum, ein Corpus vor sich selbst,aus denen sieben Corporibus, kann derhalben kein Saamen seyn (H. Kopp. Beitrge zur Geschichte der Chemie , III. Braunschweig, 1875, p. 126) [Thevisible metallic mercury is a body in itself, out of those seven bodies, andtherefore cannot be a seed. For a discussion of the complex theory found inthe writings of Basil Valentine (of questionable identity) see H. Kopp, ibid.pp. 124127].


32/34

144 F.A. PANETH

53. Neues Handwrterbuch der Chemie 3: 1. Braunschweig, 1878.54. Cf. Aristotles view, quoted above (note 2, p. 145), concerning the rela-

tion of the Greek atomists to Empedocles. Jung, who wanted to replace theactual-potential doctrine (see p. 146 above) by that of syn-dia-crisis,i.e., atomism, and Boyle saw the inevitability of the process of thoughtleading from the assumption of constant elements to atomism as clearly asdid Aristotle. Lavoisier, on the other hand, felt himself to be an opponent of metaphysical atomism when introducingthe very same conceptof element.

55. Besides theGermanworks written or inuencedby Ostwald,we maymentionhere the deliberate exclusion of all atomistic ideas, presumably a consciouscarrying on of the Lavoisier tradition, in one of the best-known French text-books. (Passage quoted from Habers preface to the German translation of LeChatelier. Vom Kohlenstoff . Halle, 1913.)

56. D. Mendelejeff.Grundlagen der Chemie .1891,p.27(cf.Principles of Chem-istry , 1905, p. 23). Also, previously, Ann. d. Chemie u. Pharm ., 1872, 8,Supplement, 133.

57. Only Meyerson approaches the two views just quoted, in a passage wherehe speaks of the atomic element ( Identit et ralit ). [English translation, Identity and Reality . Library of Philosophy,London, 1930, p. 338.] However,his reference to atoms in this connection seems to me as unfortunate asin the passage by Hell discussed above, and to miss the essential point.For, on the one hand, we may take also groups of atoms and molecules inthe transcendental sense (cf. our comments on radicals below), and on theother hand, there are simple substances too, which are built up of atoms andnot of molecules, as, for instance, the inert gases, metal vapours or activehydrogen. Incidentally, Mendelejeff, too, is inclined to identify element withatom (Principles of Chemistry 1: 207, n. 17).

58. We may ignore here the relatively very rare cases in which so-called freeradicals have recently been prepared.[Note by translator: It might be mentioned here that short-lived aliphatic freeradicals were rst prepared and detected by the author.]

59. Liebig poined out this analogy long ago in a polemic against Laurent. Ann. d.Pharm . 25: 1, 1838.

60. Note by translator: The author mentions two German translations:J. Bluwstein, Leipzig, Insel-Verlag, 1916, which he considers to be veryabbreviated and unintelligible in parts, and J. Stern, Leipzig, Reclam, whichhe recommends. Passages probably corresponding to some of the quotationsin the text may be found in A.Wolf. The Correspondence of Spinoza . London,1928, pp. 110, 112, 130, 136137. The word nitre is used as a synonym of saltpetre.

61. A passage from a letter quoted by Moissan (Trait de Chimie Minrale I: 1.Paris, 1904) shows that Descartes thought of the Peripatetic as well as thealchemical elements as being built up of like particles of primary matter.

62. See A. Wolf, loc. cit.; review: Nature : 124, 787, 1929.


33/34


63. It may be pointed out that Spinozas conviction thatonecould not separate outtwo heterogeneous substances from nitre is not connected by necessity withhis assumptionof a basic matteras eternallypersistingsubstance (against this,cf. the note by J. Bluwstein, loc. cit. p. 327). We must think of the differentsubstances as arising out of this basic matter by different arrangement of theparticles. Thus it is not the change of substances, but on the contrary theconstancy of elements, that is difcult to explain if we assume a primarymatter. Boyle, incidentally, accomplished this in a way that is still acceptedin principle today. (On this point, see Zs. f. phys. Chem 91: 171, 174, 1916; Ergebnisse der exacten Naturwissenschaften 1: 362, 367. Berlin, 1922.)

64. W. Wundt. Essays . Leipzig, 1885, p. 59.


34/34

Documents

Paneth_1931_2003_The Epistemological Status of the Chemical Concept of Element