Experiment in early greek philosophy and medicine

whieh deens Manti ie karlism, ak MERE io gen uit hirorians te Waa ru ard the ramone for fis unpravizled at mien wha, Greek agi nod ehe AL, gici Kat the reason, wl Indeatod ;

EXPERIMENT IN EARLY GREEK PHILOSOPH AND MEDICINE EXPERIMENT IN BRANLY ORERK PHILOSOPHY AND MEDIGINE si analyse their own words and notions, So they initiated science but failed to sustain It, d partly because modern science, which resta on expert : and never disengaged ‘physica’ from the general philosophy in whieh it began. There and exper was a fatal neglect of detail most craftsmanship was left to slaves. Thus ment, is in a sense not really continuous with ancient”? e Many other commentators have lald a similar emphasis on the failure ofthe Greeks ure of the role of experiment in Greek science Anyone that attempts to tackle the question on the subject, and what must first come to terms with the extensive modern literat different view on thin question, Burnet, an ie well known, attempted to defend the Greeks against the charge that they made no use of experiment, "The rise of the experimental method dates from the time when the medical schools began to influence to use experiment, or at least on thelr fallure to use itoystematleally,” On the other hand what one may call the ‘positivist’ school af Interpremtton has è a very um of Bacon strikes one first about this literature la that ever since the Novum Organ l Bacon's judgement is well known: in tone. (1620) much of it has been highly polemica ulares ‘Atque ex philosophila hetis Graeeorum, et derivationibus earum per partic vix unum experimentum adduci potest, quod ad scientias, jam per tot annorum apatia speculationibus ac hominum statum levandum et fuvandum apectet, et philosophiae at Aristotle In particular Bacon dogmatibus vere acceptum relerri posait," Again Aristotle's works the development of philosophy, and accordingly we find that the first recorded "Te is ine with experiment of a modern type Is that of Impedoalen applied the shou conceivable’, Burnet went on, ‘that an inquisitive: people experimental method in a single case whheut extending it to other problema’, and elsewhere he argued chat the reason why we have no little information about their valve opinions, According to Bacon, then, Aristotle was more blameworthy than his modern adherents who neglected of experience to fit his pre of twisting the data references to experiments are 10 be found, but neeusen Aristotle adopts a different line of attack, for he concedes that In some of the data of experiobservations and experiments la that “nearly all that we know on this subject comen lers on whom we are mar atelidenda World af the Greeks, In to be understood,” the » 1996), p. a1 With very few exceptions, lit d years made no attempt at ayatematnte experimentation, (Greeks throughout a period of elght hu | Sapientiae, Cambridge, 1992, Ps 4), Spirit ond Method (New York, Asoù pr ® In an article entitled ‘Did the XVII (1949), pp. s84f On the tople aaa w È Beobachtung und Experiment In der antiken W: 1, dus avan (1999), ; In the Tatars V‚ P, Zubov, pe Alf, and v 1919), ppi aar fl. 1, Mdetaveln, ‘Recent Barly Greek Philosophy Cath ed., London, 1940), pi an. + Essays and Addresses (London, 1919), pp, sat (of, Comford'a remarks in Principium trans. M, Dagut (L Compare the rather more cautions judgement of Samburwey, The difference, which it is essential to reoognise If the history of Greek rd Nore. per recently H, D, P, Lee, for example, in hi tendano,o he the Greche They observed logica (1952), p. xxvil, puts it that che experimental method eluded but they did not experiment, and between observation and experiment there la à fundamental ! History of Ancient Geography (Cambridge, 1040) of authorities are oled by Thomson, Ate: * A number remarkable tendency to generalige on the queation of whether or not the Greeks seem implausible or exaggerated, but 10 examine these in further detall would serve no useful purpose, There is, however, one Importi whieh I hope emerges from nd ercommentators have shown a the quotations I have given, and that la Several of the theses which have been maintained.by either side In this controversy dependent’, and a similar line of argument has been used beth by Heidel! and, in more recent years, by BIUh® si ip id à ei nothing because they did not happen to Interest the observation, we are justified In supposing. that there were others of whieh we know themselves interested in sclence, and for readers wha were even leut 10°,4 According to Burnet, then, ‘if we can point to Indubltable examples af the une of experiment and from compilations and manuals composed centuries later, by men who were not tuenda decreta ta rite eonsulult; sed postquam pro arbltelo «uo decrevimet, experlentiam ad ence entirely: ile enim prius deereverat, neque experlentiam ad consti et axioma ndus sit, sua placita rortam elreumduelt et anptlvamı ut hoe etlam nomine magis accusa entiam quam sectatores ejus moderni (scholastteorum philosophorum genus) qui experi s why a seventeenth. omnino deseruerunt’* Now there were, of course, good reason telianism. Bur century sclentist should remet strongly against contemporary Aristo in general, and science similar exaggerations mar much of what was said about Greek entury historians of science, Aristotle's contribution in partioular, by nineteenth-c edition of the 5 H, Lewes! quotes John Playfair's Dissertation, prefixed to the 1842mid-nineteenth the view prevailing in the Encyclopaedia Britannica, us Greeks and the reasons for its century on the subject of the physleal selence of the who, however shortcomings. ‘Extreme eredulity disgraced the speculations of men ided with the ingenious, were little acquainted with the laws of nature, and unprov ed, Though great criterion by whieh the evidence of testimony can alone be examin instituted; and philoobservations were sometimes made, experiments were never , did not seek sophers who were litte attentive 10 the facts which spontaneously offered sions of increase their number by artiflolal combinatione,' In more modern discus were nearly to fundamental distincthis topte, 100, the view has often been expressed chat there in a failed to employ the ts tion between ancient and modern selenge In that the ancien mental method, J, O, Thomson, for example, put kt that’ the Greeks ly, preferring * Op, elt 1, 6: ry of Seienee (London, 1864), Pi 49: things without firet studying and describing them proper experi to collect and sift facts, always too ready to theorize, and had Hiele of the modern will to explain or of the instruments and techniques necessary for this purpose: they tried instead to ' Novum Organum, ty The Europ 1 Aristotle, A Chapter from the Histo of the progress of mathematical and 4 J, Playfair, ‘Dissertation, , ; Bxhibleing a generale’,view Encyclopaedia Britannica (71h ed., 1842), physical science, since the revival of letters in

experimented. Much of the discussion has been carried on as if the important thing was to decide the answer to this globa/ question, and in attempting to settle this question one way or the other it has too often been forgotten that experimentation is of varying, usefulness and relevance in different fields of scientific investigation, or may also be said to vary (in certain respects) at different levels of scientific developeven on different problems within the same field: indeed the role of experimentation ment, that is according to the level of knowledge attained in a particular field at a particular time. To my mind, then, the question that needs to be asked is not the nor even ‘how far did the Greeks recognize global one ‘did the Greeks experiment", the value of the experimental method?’ Rather we should ask the more concrete, and at the same time much more difficult, questions “what experiments were open to the Greeks on different problems or In different branches of science which they did not perform?’ and on the other hand ‘what tests did they actually carry out, in different fields, and with what success?’ To these questione there is, of course, no simple yes or no answer, Rather we must try to assess the Greek performance in each department of science, indeed on each problem, independently, and to do this generally requires a thorough knowledge not merely of the contributions made in a particular field at different stages in antiquity, but also of the history of subsequent developments, that is of how thought developed on the subject in modern, post-Renaissance times. Of course the extent of the subject I have outlined precludes anything like a full, let EXPERIMENT IN RARLY GREEK PHILOSOPHY AND MEDICINE 53 obscure problem," he prefaces his account with this disclaimer: ‘ Perhaps to try to give an explanation about some things, or about everything without excep tion, may seem to indicate a high degree of simple-mindedness or a high degree of seul, But’ he goes , on, ‘this objection in not always justi one should consider the reason for speaking , and also what kind of convie is beingtto aimed n at, whether merely humoran something, stronger, Whenever anyone lights upon more exact proofs, then we mustbe grateful to the discoverer, but for the present we must state what seems plausible." However, my immediate point is not thar in asseusing the work of early Greek astronomers we should welgh against the obvious dogmatie and teleological features their use of empirical data? and certain passages in whieh they draw attention to the tentative nature of some of their theories, but rather this, that to speak ofa ‘failure to conduct experiments’ in this fleld would be quite Inappropriate, One may, perhaps, go further, fin astronomy it was physloally impossible for the Greeks to carry out experiments, the name also applied for them, at least, in a large part of what they called ‘meteorology’, Nowadays, It In true, It la possible to manufacture a lightning-flash artificially, But that the Greeks did nor attempt to do so is hardly surprising (they got no further than notlelng certain examples of statie electricity, such as the attraction of amber when rubbed), Por the Greeks most of the problems of ‘meteorology’, the nature and causes of ligghening and thunder, meteors, winds, earthquakes and so on, were beyond the reach of direct experimental investiga tion, But if the theories put forward in this field hy the Presocraries and later writers are generally highly dogmatic and speculative, this la not to say that no attempt whatsoever was made to adduce empirical support for some such doctrines, 1 have in to the fourth century n.6, I shall take in turn each of the main departments of inquiry alone an exhaustive, treatment here, and 1 shall confine myself to making, certain tentative suggestions concerning some of the problems which occur when we consider the earliest period of Greek science, that is, roughly speaking, from the sixth familiar experience, mind the occasions when the explanation of an obscure natural phenom enon was y year, why the heavens revolve in one direction attention Improvemen tang ps Bunoo some of the heavenly bodies have ‘complex’, others ‘simple et, tI, 3, ta; DK, 12404, 1 5 Mate, 3690, ol, = =p were made of the length of the nolar ) Sambursky, op, et, pp, soll, has drawn ments that took place In 2 Greek es motions, = Cael, er st8 Tas = # similar disclaimer at Ce, B, eh, 1a, 2914 aaf, where Aristotle rather than the other (wo far an I know, the problem has still not been solved), rather than the other, or, as we should put It, why the earth vovolven on bus axle In one direction ' Cael, B, ch, 5, 3874, aaff, The problem In question In what happens when an unlit lamp In placed below another one which Is alight and the one thing to another, he draws on two analogies: either, he suggests, It may be like exhalation'. And when he tackles the question of whether shootingestars are actual projectiles, or whether this phenomenon Is due to a train of fire passing rapidly from by Anaximander concerning the nature of lightning (namely that de ie due to a cloud being split by the force of wind), Aristotle himself, In his necount of thunder, compares it with the crackling of logs in « fire, attributing both phenomena to the ‘dry rofl, seems to have been used by Anaximenes to back a theory originally advanced illustrated or supported by an appeal to an analogy drawn from there is in our extant sources (prinelpally the Presocratics, the Hippocratic Corpus on which the earliest Greek investigators were engaged, and consider what evidence The comparison between the flash of lightning and the flash made by an oar or some such object on striking the water, whieh da mentioned by Aristotle at Mere, 3709, and Aristotle) which will enable us to answer the two questions | have propounded. The first point that may be made Is the obvious one that much of the speculative effort of the earliest Greek natural philosophers was concentrated on a department of nomy. Astronomers can and do attempt to verify, by further observations, the prescience which, strletly speaking, Is not experimental at all: I mean, of course, astrodictions they make on the basis of theorles or hypotheses, but in the nature of things they cannot conduet experiments, that is they cannot vary or govern the conditions of the objects they are observing, Anelent astronomy Is, no doubt, in certain respects less rigorously empirical, more aprloriate, than modern, but how far we can talk of a fundamental difference in method between the two is more difficult to decide. While many of the explanations which Aristotle, for instance, put forward are, we should say, highly arbitrary, It is worth recalling that on more than one occasion he draws on the results obtained from many years of observations by Egyptian and Babylonian astronomers," that a passage In the Mereorologiea suggests that his own observations extended, in some cases, over a period of fifty years," and that in a striking passage in the de Caelo, where he introduces one of his more fanciful explanations of a highly * Cael, 2920, 708. Mate, 9494, off and a8 ff, and cf, also Caad, 2704, 1308, * Mete, vraa, aff, (referring to the rare occurrence of moon rainbows),

GEN LLOYD lower lamp is lit by a flame travelling rapidly downwards from the upper one, or else the shooting star may be a solid body forced downwards under pressure, and here he compares them with such objects as the stones of fruit being shot out from between the fingers, in order to illustrate how things may move, under pressure, in a direction contrary to their natural tendency,' Now many of the illustrations used by Greek writers to suggest or support explanations of meteorological phenomena might strike us as rather far-fetched, though this is not always the case: one might cite, for example, the passage in which Aristotle explains why we see the flash of lightning, before we hear the sound of the thunder by referring to the experience of watching, a ship at sea, where the oars are already going back again by the time the sound of their striking the water reaches us." But the point I should like to suggest is that in ‘meteorology’, where direct experimental investigation was often physically impossible, the appeal to analogies drawn from more familiar experience served the Greeks as a Berrepos 1rAo0s, a next best method, indeed in many cases the only empirical method open to them. Moreover some of the illustrations they use refer not to well-known facts or common experience, but to quite rare phenomena, and in certain instances It seems that the analogy may even have involved the deliberate undertaking of a plece of research, If we take the illustration of the two lamps which Aristotle uses to suggest one of two possible explanations for the phenomena of shootingestars, it is fairly clear that Aristotle himself had observed quite closely the way in which one lamp may be lit piece of research, comes in Aristotle's discussion of the rainbow (Mere. T, ch. 4): from another: ‘the speed with which this happens’, he notes at Mere, 3420, 5ff., ‘is extraordinary, and it resembles the projection of a missile, rather than fire passing from one thing to another’, And a second passage which shows how fine may be the distinction between an illustration drawn from familiar experience, and a deliberate there one of the illustrations which he adduces to support his suggestion: that the rainbow is due to a reflection caused by minute drops of water is what happens when a man sprinkles water in a room which is so placed that it faces the sun and is partly illuminated by it and partly in shadow, ‘Then if one man sprinkles water in the room, another standing outside sees a rainbow at the place where the sun’s rays stop and the shadow begins.’) The invention of prlams was necessary before the spectrum could be fully investigated under strict experimental conditions.4 But it is apparent that at an early stage the Greeks exploited thelr knowledge of rainbows formed under other, artificial conditions, in trying; to explain the meteorological phenomenon. In astronomy, then, and in much of what the Greeks called ‘meteorology’, we are at liberty to suggest that the early theorlsts were often uneritieal, or that they seriously underestimated the complexity of the phenomena; but where experiment is impossible for objective, physical reasons, the eritielam that they failed to use the experimental ' Mete, 1480, all, ef 9440, asf. on cometa, * Mete, 3696, Hl, The expression that sight mperepelv ris diofis might be thought rather odd in view of the fact that elsewhere Aristotle flatly denies (against the opinion of Empedocles) that 3 Mete, 744, 3, Me refers, also, to the rainbows formed under certain conditions when an light travels at all (e.g. Sans, 4464, ag fl, 4, ag). oar ls raised out of the water (3744, 20 (1), 4 De was first achieved by Newton In his Opricks (1704), following on the work of Descartes i, and EXPERIMENT IN BARLY GREEK PHILOSOPHY AND MEDICINI method is, clearly, wide of the mark. But we have now to deal with the more ing and more difficult problems of the role of experiment in other fields, and f in the sciences we know as physics and chemistry, First physics, and 1 may| hydrostatics, in which the achievements of Greek scienc are by e no means ne devoted to the subject, ı med to be a digen Concerning the Greek contribu to the study tion of opties, in particular, « comparatively good information, since there are several treaties extant eithe original or in Latin translations which are special y important being, Kuelid'a Optles, the De Speeulis Catoperies of Hero of Alexandria) and a Latin translation of an Arable ve another Greek treatise on opties whieh there ie no good reason to doubt ie the : Ptolemy.' These works make it abundclear antly not only that the Greeks were ful in their application of geometry to the study of refleetion and refraction, | that some investigators undertook quite extensive experiments to corrobo: principles of optics which they formulated, Some passages from Ptolemy's ©, worth quoting, to illustrate one Investigator's methods, At the beginning of n (ch. 3, 88, off. Lejeune) the writer sets out three elementary principles or | reflection. These are (1) that objects that are seen in mirrors are seen in the di of the visual ray which falls on them when reflected by the mirror; (a) that thir are seen in mirrors are seen on the perpendicular whieh falla from the object surface of the mirror and is produced; and (3) the position of the reflected ray with the perpen the eye to the mirror and from the mirror to the object, ie such that each of parts contains the point of reflection and makes equal: to the mirror at that point." The truth of these prinelples, he goes on, is confirr the phenomena, and he proceeds to cite a serie of simple s experiments to corro his laws. ‘Thus to confirm his first peine he remarks iple (it, eh 4, 89, sfr case of all mirrors (that is plane, convex and concave) we find that if we ma points on the surface through whieh the images are seen, and cover these point the image of the object will certainly no longer be visible, But then when we ur the points one by one and look at the uneovered points, both the paints and the of the object will be seen together on the straight line drawn to the summit visual ray (ie. the eye)” Further experiments follow: one whieh Is parti: remarkable involves the use of three mirrors, one plane, one convex and one col 7 "This is the conclusion of the mont recent editor, A, Lejeune, 4'Oprigue de Claude P (Louvain, 1956), Introduction, pp, 13-40, * With reference to the diagram where AZA In the mirror, 4 the aye, # the object, 1948, p. 269, N. 1.) mirror, these three principles ares (1) 4° lien on AO prod O the point at which the visual ray silos the mirror, and TO wi ap hd ERBE is rin ed ol

in each case, Ge Be LLOYD and confirms very neatly that the angle of incidence is equal to the angle of reflection These are simple, but generally very effective, tests undertaken to corroborate the elementary laws of reflection, but elsewhere the author carries out experiments of a greater complexity, notably in his discussion of refraction in Book v. There (v, ch. 3, 224, 10ff.) he first points out that as in reflection, so too in refraction, the image is seen at the intersection of the line of the visual ray and the perpendicular drawn from the object to the reflecting, or refracting surface, and here he refers to an experiment which goes back at least as far as Archimedes,’ in whieh a coin is put into an opaque vessel in such a position that It is just hidden by the lip of the vessel, but comes into EXPERIMENT IN KARLY GREEK PHILOSOPHY AND MEDICINE 7 angle of refraction, Both the problems in question, and the experiments used to study in them, are, of course, quite elementary, ButI have chosen this example deliberately late antiquityè antiquity, | Hut is this the ease, perhaps, only in quite order to indicate that} where the circumstances were favourable, thatin where the be investigated emp simple and where they could problems are relatively ), we do find evidence that detailed experiments were nome» without too great times carried out in ments of a Ptolemy In opties, or of a Philo or a Hero in Are we to consider the in fact, a radical break with the methods pneumatics, as quite exceptional, as marking, of investigation used in earlier periods? Or is what distinguishes Ptolemy from earlier or thoroughness with which he applied techniques writers on optics merely the success of investigation which had been used, though not fully exploited, before? Without entering into the various probleme raised by the development of each one of the branches of physics in antiquity, I may briefly consider the question in so far an it for poor though our information is, concerns the earliest period of Greek of physica had already begun to there are grounds for believing that some view when water is poured into the vessel, But not content merely to state certain general principles of refraction, he undertakes detailed investigations to measure the amount of refraction which takes place at different angles of incidence, and in different media, First (v, ch, 7, 227, 11.) he describes the setting up of his apparatus. To meawhich he attached different weights to a atring tension and pitch, a third that he filled jars and observed à relationship between varying amounts of water and noted made different notes when struck, another that he did ao as a result of experiments in numerical relations between the musteal intervals of the octave, fifth and fourth, One be investigated with the help of simple tenta In the Presocratic period or in the early fourth century. Some of this evidence In well known, We are all acquainted with the legends that purport to describe the experiments which led Pythagoras to discover the marker is to be placed at different points along the circumference of one or other of circular disk, each quadrant of whieh ia divided into ninety parts (he, like a protractor): this disk Ie set up in a bowl of clean water «o that the water just covers the bottom half of the circle, He then describes how a coloured story has it that he made this discovery by measuring the weights of hammers which sure the angles he uses a the two quadrants which are above the water level, and how sightings are to be taken so that the coloured marker and the centre of the disk are aligned with the eye. The the coloured marker and the centre of the disk, This enables one to determine both the next operation is to move a small, thin rod along the circumference of the opposite quadrant which is under the water until the extremity of the rod appears in line with a relation between the quantity of water and the sounds the Jar made when struck, a or pipe which gave various notes, fourth that he measured the lengths of These accounts, as is well known, contain much that ie pure fantasy! In the majority of cases when the test is carried out in the manner described In our sources, it does not, in fact, reveal the simple relations between the various musical intervals, Nevertheless that the angle of incidence is always greater than the angle of refraction, and that as angle of incidence and the angle of refraction, and when these are measured, we find greater. The results obtained by Ptolemy are given in detail: when the angle of accuracies. For one thing the very faet that such legenda elreulated about this dis- “this is the method by which we have discovered the amount of refraction in the case incidence is 10 degrees, the angle of refraction will be about 8 degrees; when the angle the angle of incidence increases the amount of refraction becomes progressively of incidence is 20 degrees, the angle of refraction will be 154 degrees, and so on for angles of incidence up to 80 degrees. Concluding the passage with the remark that Untersuchung der Wirkliehkekt,' 7 Api O) he carried out his experiments, indeed Pythagoras’, It was no doubt on the * Cf. the judgement of Zubov, ap, ef pi 3941 “Allein die Tarache, daf) eine solche Legende in der Antike aufkam, zeugt von dem Verstündais des Wesons des Experimenten als einer ratlonalen stories contain, Gurhrie coneludes shat if the esp. 223 ff, who gives the Greek sourves and ! See, for example, Guthrie, A History af GreekA Pila valt Ce tension of the string) the relations between the musical intervals are readily determined with reference to the lengths of the string or pipe, We have other evidence, too, pipes, for here, of course, other things being equal (e,g, the tiieknews, material and covery in antiquity suggests some recognition (if only a theoretical recognition) of we should not dismiss these storied an entirely worthless, for all their obvious inof water’, Ptolemy adds the noter ‘we have found no perceptible difference in this respect between waters of different densities or rarities’ (ch, 12, 230, afl). Ie is worth drawing attention to the point that Ptolemy evidently tried the same experiment with the value of the experimental method, for example in determining which are the causative factors governing a particular effect! Secondly, if most of these supposed tests do not, in fact, yield the results reported, exceptions must be made of two of the investigations referred to, those with the monochord and with the columns of air in Opties we find set out in detail the results of his Investigations of the refraction of other different kinds of waters to see whether thls gave different results, and elsewhere in the media, namely from air to glass and from water to laws," In investigating the problems of reflection and refraction Ptolemy carried out extensive experiments to corroborate his general laws and in particular to establish that there is a definite quantitative relationship between the angle of incidence and the * See Lejeune, op, eft, n, 9 to aay, 9 * It should be maid that Ptolemy Lair the ray from the eye, not that from the object, the incident ray. For his results, which are expressed to within half a degree, compare the tables given by Brunet and Mieli, Histoire des selances: Antiquird (Paris, 1935), pp, Has

early Pythagoreans carried out certain empirical investigations in acoustics. There is besides the stories which refer to Pythagoras himself, which tends to confirm that the (ch. 17, L. vit, 498, 174). This involves putting, three different substances (ear And another quite ambitious test is described in the treatise On the Nature of6 EXPERIMENT IN RARLY GREEK PHILOSOPHY AND MEDICINE the interesting, though admittedly obscure, report that Hippasus constructed bronze Hero in pneumatics, for instance, or by Ptolemy In opties. In some fields, ther if these are neither so precise nor so complex as those conducted later by Ph experiments were conducted even as early as the (flh or the fourth century mc and where tests could be carrled out without any great difficulty, certain rudin considered that in some branches of physics, where the problems are fairly elen the Hippocratic Corpus but also from Aristotle," but it is apparent from those v action of like-to-tike, but tostudy the reactions of substances of different gravities in partial suspension in water, and that the author of On the Natur Child attempted to adapt this experiment to suit his own purposes. Further examples of simple physical experiments might be mentioned not on, and lead-filings) into a bladder full of water and agitating them by blowing « through a tube let down Into the vessel, The author refers to this test to sup; theory that the various parts of the body are formed by the action of like coi like, but it seems probable that this test was originally designed not to illust he cited seems to derive from first-hand investigations (e.g, the variations in the pitch disks of varying thicknesses to produce certain harmonies, and in a fragment preserved by Porphyry, Archytas refers to a variety of phenomena in an attempt to establish his theory relating the pitch of a note to its ‘speed’, and some of the evidence of the sound produced when a stick is moved at different speeds, and the notes proprovides evidence of early empirical investigations in acoustics, for in the Republic duced by different lengths of pipe)’ Finally it is worth recalling that Plato too Socrates is made to refer (with disapproval) to those who ‘measure the harmonies and sounds they hear against one another’ (531 A, 1ff,) and who ‘look for numbers in these heard harmonies’ (0, 1 6), Acoustics, then, is one field in which the Greeks both observed and conducted rudimentary empirical tests from a very early period, and the same may be said of pneumatics and of what was later to beeome (with Archimedes) the science of mental methods to the study of certain physical phenomena, and in some ca is a definite (if only a quite slow) progress, in antiquity, in the application of| Greek theorists who Investigated the phenomena of air pressure by observing, the method, while the second was indefatigable in devising and conducting practle of these pales by comparison with the manifest fallure to Investigate the probl motion experimentally. Can we deny that the main reason why Aristotle in pa failed to formulate adequate laws of motion, and why Galileo was ao mucl successful in this field, was simply this, that the fret quite ignored the expert hydrostatics, Beginning with Kmpedocles and Anaxagoras we find a whole series of behaviour of air enclosed In the clepsydra and other vessels, both repeating and modifying the simple experience described by Empedocles himself (fr. 100). The author of the pseudo-Arlstotellan Problemata, for example, objects against Anaxagoras that it is not enough simply to say that air is the cause of the water not entering the bulb of the clepsydra when the tube at the top is blocked, for in some circumstances, viz. when the clepsydra is immersed obliquely in the water, this does not happen (9144, gft). In the Hippoeratie Corpus there are several texts which refer to simple investigations of air and water pressure, as for example the creation of what we should call a partial vacuum by inverting a narrowenecked vessel containing water or oil, and before Aristotle himself there is nothing that can be called dynamies at all in we should first consider the early history of dynamics and the relation b Aristotle's theory of motion and earlier views. It is hardly an exaggeration to 8 to corroborate his theories? To assess how much truth there In in these judge exceptional cases where the Greeks did conduct some simple tests, but the signi by their failure to experiment? It la all very well, I shall be told, to point te earliest Greek scientists, and indeed of Greek nelentists an a whole, were quite + beginnings of this development can be traced already in the Presocratic pet shortly afterwards. But is it not the case that In dynamics, at least, the efforts the release of the vacuum by piercing a hole in the vessel.) But two rather more elaborate tests are worth mentioning particularly, One writer describes a test which demonstrates that water finds its own level: this involves setting up an apparatus of three or more intercommunicating vessels on level ground (this point is stressed), and support his theory of the a ' Senn, op, ein pp, agafl, concludes that the test described was probably in origin science. The Presocratics refer in various context to the attraction of like th one another, but this generalization embraces a very wide range of phenomen the writer describes how the whole system may be filled or emptied by filling or emptying any one of the vessels (though he then uses this piece of information to support a highly speculative theory of the interplay of the humours in the body). mentierungaversuch", which the Hippocratic author applied to * Aachollum on Plato, Phaedo vol p (DK, riba 13) contalna a report of Arlatoxenus which refers to Hippaaus’ construction, and it seems possible that this may have been designed (as Burnet suggested In a note In his ed, of che Phaedo ad foe.) to provide a model to Iustrate the harmony of kovırd with Platt, for dndvera), a im references to the fact that ifa vessel ls heated and then cools it contracts and some of the water In dirawn up Into the vessel, though unlikely that Aristotle himself originated thie an ka wab, aft. it may be Waters, Places, cl, 4, CMG, 1, 1, 64, vat, in whieh a howl of water In left out of doorst and when the water Is thawed it is found on belng remenaured to he low than the original« u la the vent dencribed in ( the ‘breath’ causing like substances to come together In the denumiadanbeye. chy 97, Lo vin, Gia, GI, Con which see Senn, op, cit, pp. a4BfT) and ch. 51, On the Nature of the Child, ch, 18, sas, af, and ch, 23, 532, aaff, (on which see * Perhaps the best known example in the Hippoeratie the spheres, (a test which the writer supports hie contention that freesing causes finest’ part of the water to x dried up and disappe 1 On Diseases, tv, Fr. 1, The significance of thin fragment for our understanding of the role of experiment in Prosooratlo, particularly Pythagorean, philosophy, has been discussed by G, Senn, ‘Ober Herkunft und Stil der Beschreibungen von Experimenten im Corpus Hippocraticum', Arch. f. Gesch, d, ff, Medizin, xxn (1929), pp. 271 488, 17%, Cf also Senn, pp. 245 fl). 4 On Diseases, 1, ch. 39, Le var, 559, ULE, on which Senn commented (op, alt, p. 232): ‘wieder eine anschauliche Beschreibung eines einwandfrei durchgeführten physikalischen Experiments’,

G.E Re LLOYD action of gravituonal forces may be included under this heading, but so 100 may the behaviour of animals; in a fragment quoted by Sextus Empiricus, Democritus refers explicitly to bothinimate and inanimate examples to illustrate how like Is drawn to, (pigeons and crays are mentioned in particular) and on the other to the separation of and known by, li, hat is on the one hand to certain gregarious species of animals seeds according heir kinds in a winnowingebasket, and to that of pebbles according to their different tapes on the sea-shore,* It is Aristotle himself, then, who makes the first serious attenp to formulate general theories in dynamics in which the relevant factors governing he velocity of amoving object are interrelated, In various passages in the Physics and he de Coelo he Wwolates two main factors which govern the speed of a moving objecı, mmely its weight (which in the case of objects of the same kind varies directly wih the size) and the ‘density’ of the medium.’ And he puts forward 6 his discussion that motion Then too tt le an = ik in motion that takes Place in a medium, EXPERIMENT IN BARLY GRERK PHILOSOPHY AND MEDICINE relationship between weight and velocity although the relationship Isnota simple one of direct dense me moti that on th obvious fact of experi en ce must take place through a medium, force, mass and velocity, Aristotle assumes in In the PAyrler and de Caelo In highly The ‘laws’ of motion whieh Arlitotle proposes are not utterly at varlance with dies, experience: rather they are hasty generalizations based on superficial ol His discussion of the problems af theories concerning the relation between these factors and velocity in ‘natural’, and theoretical (it is not often that he oxplielily refers to emplrieal evidence in this connecin ‘forced’, mot, ‘Thus in natural motion, that is In the case of freely falling or freely rising; bodi, velocity ie said 10 be directly proportional to the weight of the tion at all), but even so it in at least arguable that in wasuming, place’, in forced mtion the velocity decreases as the object progresses further away predecessors or contemporaries, are mentioned even mare rarely.* Furthermore, context in the Physics and de Cuelo, experiment, whether his own or those of his the data of experience, Yet this does not alter the fhor thar he quite failed to verify his takes place through a medium, he stayed voo clone, rather than not close enough, to that motion necessarily body and inversely proportional to the density of the medium, while in forced motion theories experimentally, for if empirieal evidence la only seldom referred to in this These, the main heories of Aristotellan dynamics, represent, then, the first attempt velocity is sald tu directly proportional to the force applied and inverely proportional to the may of the body moved! (and Aristotle also suggests that wherens in natural motion i speed of the object increases the nearer it comes to its ‘natural from the propelliy agent). from a considerable height (though ulum and with an inelined refers 10 the more famous nothing precluded the carrying out of certain simple tests whieh would have indicated the falsity of his general theories of natural and forced motion (even If they would not necessarily have suggested a better alternative account), tn the Discorsi, published in experiment of dropping balls of different 1638, Galileo describes in detail his exper iment with sa plane (down which he rolled bronze Wall andy) he to formulate abstuy ‘Jaws’ of motion, to establish the relation between the various factors which dewmine the speed of moving objects, while discounting irrelevant considerations, Antotle assumes, for example, that the medium is completely homogeneous, although; js never so in faety! and while he notes that the speed of an object is influenced by is dupe, he leaves this too out of account when proposing his general several scholars doubt Viviani's report that he carried our this experi ment from the Leaning Tower of Pisa), None of these teste was heyond the range of what was technically possible for the ancient Greeks,’ and any of them might have been used to doctrines of natunland forced motion, But then the next point that should be made is that the theork he suggests correspond to observed phenomena much more closely chan might ar first sight som likely in view of the discrepancy between Aristotelian and Nrvtonlan dynamles, Aristotle has often been taken to task for assuming that the locity of a freely falling body varies directly with It weight, but more when inflated than when empty, This experiment has heen eg, by R data contains a higher proportion of carbon dioxide), (a) Introduction to his ed. of the Physler, pp, aa, on the grounds (+) > upponed experimenta fact is not correct, and (2) that Aristotle's theory In not true to the experimental fret he thought | les pa off, whore he va u Dre ri =ee ' Ph A, ch, 8, esp, 2150, 241.) 2100, vele * Anotable exception Is the passage at Ewel, vind, excopt fire, has weight in ite own ‘natural shebi it is as well to be dar where his mistake Hew, The fact is that in alr heavier bodies do fall more rapidly lun lighter ones of the same shape and size (though this is not true, of course, ins vacuum), Aristotle was correct in assuming that there is some he had at his command, But whether or nota der will wel more when on, among other things, (1) whether ie da inflated with Patni aly ar - and (3) the amount of water vapour in the weer, t results Simplicius At lenet-ehvee different and dough hl Jey, pedati» 1411), ne whieh di je In da inflated under pressure or not ' Sext Adv. math vi, 117 (DK, 080164), ef, also Aut, 19, 19, 19 (DK, 4128). Plato uses a similar model to une Pa of Tike to-like in the Receptacle at 71, 526-134. * He alto notes, eu Cael, 9134 140, that the speed of an object In Influenced by its shape, j A ei ren aaa 2774, fl; Ph, 214 a, ag fl, Con natural motion) and Ph 2496, a7ff, (on bend, | y Arda ile hi motion), were obtained la, perhaps, pe up Een woo Simp. in Cael, 710, 241, and al, alan Anon, Attempts were made to carry out this test in antie oreed * Ph, 2308, aff. al Caal, 1779, 270 At Caed afta, rofl, it da suggested dut In some cases the highest speed is mined ijn ps point from which flight begins nor where It ends but dvà middle of the flight, Itihardly possible to interpret it consistently with Aristotle „known doctrines, of the principle of the watersetogk, and the number of factors which might 3 The Greeks had no precise means af he describes the method he emplayed for ı that of weighing the amounts of water wh a uégov, but if this mew what it appears to mean, namely that the highest speed Is attained in the | See Cohen and Drabkin, op. cir. p. 203, n. te See the notes of Stods in the Oxford Translation and of Guthrie in the Loeb edition of the de Caslo.

demonstrate he inadequacies of Aristotelian dynamics (and in partic ular the doctrine that in naunl motion velocity is directly proportional to weigh t). seems thatve have a clear case in which the earlier Greek investigat Here, then, it ors might have carried out certain tests in conjunction with their theories, but quite failed to do so; but if this judgement is true in the main, two reservations should be added, First we must repeat hat what Galileo understood, but Aristo tle failed to appre and indeed would have denied, was that in studying the conditions of uniformciate or accelerated motion, theeffects of friction and of the resistance of the medium should be discounted. Thus whilei was certainly open to Aristotle to perform the three types iment I have mentioned, yet so long as the necessary corrections were not madeoftoexper offset the effects of friction and air resistance, the actual results obtained from such experiments would cerainly be difficult to interpret and perhaps even positively misle ading . Second, and more important, while it in undoubtedly the case that Aristotle himself failed, in general, to undertake practical tests in connection with his theory of motion, this is much less true of some of his successors, Some of the examples of objects in motion which are discussed in the Problemata appear to derive less from common experience thin from deliberate investigations, and Strato too carried out certain tests in this field, for example to establish the fact that falling bodies undergo acceleration.’ Such empirical investigations as are recorded are all quite rudim entary (Strato apparently observed the impact made by stones dropped from different heights, and the way In which when water falls from a considerable height, its flow is continuous at the top, but broken at the bottom of its fall), but they may be taken indicate that the possibility of carrying out practical tests to illuminate the probltoems of that the dynamics was not wholly ignored by Aristotle's immediate successors. Nor should we fail to note that this method was used with much greater success in later antiquity. Philoponus, for example, challenged the fundamental assumptions of Aristotelian dynamics on both theoretical and empirical grounds. Argui against the doctrine that the speed of a falling body is proportional to its weighng, t, he says ‘but this is completely file. And this may be confirmed more forci bly by actual obser vatio n than by any sort of verbal demonstration’, and he then proceeds to adduce (as a thousand years liter Galileo was to adduce) the evidence of what happens when you drop two diffrent weights from the same height: ‘you will see that the ratio of the times required for the motion does not depend on the ratio of the weights, but difference in time is a very small one,’ The next feld we must discuss In chemistry, or rather what passen for chemistry in antiquity, chat Is, the study of the constituent elements of substances and their "Tig the acount of objects rebounding from a plane, and that of the figures described by certain solide, namely the cylinder and the cone, when these are revolved (Probl, xvi, 4 and $e 9134, GM, and 9154, 18M, and xvi, 5, 9134, 71), A passage in the Machanios (isha, De 0 worth quotingds It Iustrates the difficulty which the writer experienced in tackling the prob = of ‘why abject Which are hurled come to a standatill’, ‘Does it stop when the force which starte it falls, or because the object In drawn In a contrary direction, or is it due to ite downward red Which la stronger thin the force which threw it? Or is It absurd to discuss such questions while the Principle encape in (A derottov rö rar’ dirropetv, dpivra Thv dpyiv;)?' (trans, B.S, Forster), * See Simplicius,in Ph, 916, vof, (Weheli, Seraton von Lampsakos, fr. 73): * Philoponu, in Ph, 683, 164, CINE RXPRRIMENT IN BARLY GREEK PHIL OSOPHY AND MEDI interactions, Here the Preso 6 cratics made several very important ibutions, thougl | nearly all of them relate co the conceptual framework of chemical’contrtheor y, rather ther compo to the discovery of facts, Thus it was one of K that he explained how a variety of dif number of primary elements by another in different in! somet portions’, which states that chemical contain their constituent elements in fixed and invarlable proportions by y alwayswas In fact assumed long before it could be demonstrated experimentally, Aristotle, in turn, succeeded, for example, in distinguishing various of mixture and combination, Including owvBsans (i.e. an aggregation of diflerent cen in what we should term a mechanical mixture, for example a pile of barley and wheat grain s) and wikis (in which the prope rties, for example when tin and copper com bin to form e bron ugh it should be noted that not all of the examples which Aristotle uses to illust rate u(fis to verify or faluify refer to what we should class as chemical compounds)’ Hut then one asks, how far is it the cave that early Greek theories concerning the constituent elements of things were merely a series of conjectures which they never attempted experimental investigations? Now as In the early history of dynamics, by means of #0 too in the afterwards detailed empirical empirical research had been , attempts were made development of theories about the constituent elements of things to formulate systematic general theories before much undertiken, But by the time of Aristotle or investigations had begun to be de Generations si Corruptione in largel y carried out In this field, Avistotle's discussion of the nature and number of the elements in the dogmatic.) but if we turn to the Meteorologlea, the exten t of the knowledge which it displays concerning the reactions of different wubswnces to various simple toate In remarkable! It is true that Aristotle tende to pr hisen result ts in the form of generalizations, for example ‘of things whieh are sol es idiby fie d or cold, those that heat are soluble are dissolved by their opposites, Wor those that are solidified by dry heat du mintu eh, 10, esp. * See frr, 96 and 98 on the formula or compo sitio n of hone, bl and oo ‘othe d r forme of flesh’. * Under ulfis Aristotle includes the «patois af liquida and ‚for here too an ch ae intera ction may take place and the renult ant he 3284, 260, and 4, 3), On Aristotle's theor compound y of different wa h ru pamela, the Loeb edition, pp. xill ff, there are no good reasons iesrehgine as this In some have denied that it in Crans ua fl, and Gott PAUL UT Meworony further (3294, aq ft; fo two palrs of re and à na esp. HH, Joachim, Journal of Philology, XXIX (1904 ), pp HM En dt * Ths in GC, B, che, 1 fl, when he sets out te deter mine the na [ e r of the elements, he argus (1) that coming to be and passi ng away are lm Witho ut perceprible bodies (3284, nf), which in turn cannot exlat apart from contravi ettony fo a w i l y must be either heavy body will be tangible > dry, but these cannot be» and (3) that the tangible conte il) or light either hot or cold (3290, rofl), (a) that the principles contraria, (1294, 71), 3300, ult), * Forthe purpo ses of this paper I shall inclu de the | dutt hingen | the work o A mmodiate an auttintie text of Aristotle himself (eg, Hamm ui schalk, CQ, ns, x1 (1961), pp. Ari, and contr dogica, Hook 1v (Göteborg, a, pp. oll, and Price to the second edition, 1 [2 È for not sccepting the book as a wi

are dissolved by water (which is wet and cold) and those that are sldiied by n more likely, It is commonly assumed,I believe, that if for one, concludes that ‘Aristotle had apparently only pe ser, the vapour being colleeted in an ab method of distillation employedIn a treatise On Breathe refers to the account of how sweat is formed EXPERIMENT IN BARLY GREEK PHILOSOPHY AND MEDICI rer on a good deal of in generalizations, Aristotle evidently collected dissolved by fire’ (Mere, A, 3826, 3311). But much detailed — En these (le. Which sub. concerning the physical properties of a wide varlety of substances stances are ductile, which malleable, which fissile and so on) and asoconcerning; their various + si reactions to fire and to water, to being burned, boiled or dissolved in no doubt common u liquid. Much of what is contained in Meteorologica A was ledge. Some of his more specialized information comes from are th such contemporary industrial processes as Iron-making.‘ But some? is “ty edge clearly derives from deliberate investigations (whether or not it et È meelf held over the boiling liquid.* tut if wine ie orate and the | plate or some such objeet held over the bolling wine,the liquid ı eolleeted evap NE 65 collected on a ba cond seawater, To my mind, aleoholie content which would pass for ou and botany, But why, btey which were parallel to, en though less intensive than, the researches then, one may ask, did these inves [ Hut to any radical advance in doctrine, or at least te. theory of Aristotle? Now the fact In that tu an increase of factual knowledge and to the wy Aristotelian doctrine, Aristotle himself concert pounds according to whether earth or wate problems which he found difficult to resolve,®. ! Flat, ch, 83 CMG, 1, 1, 96, 15/8, 5 à and air or trvedua (More, 309 present difficultien whieh he a that are solidified by cold and melted by fire that are solidified by heat to contin mare | varieties of water and of earth and of thely 5 In general, Aristotle considers mont em 4 Mate, A does, however, owe a good deal to the influence of heat and to solidify least under the tiflvenee ı sol afl. and 3874, off), Ma, woeording mont under ga fl) of, ao 3840, "This method of distillation In desorbed in Diasvoriden, a Med. * It may be noted that elsewhere in Mure, A different sorta af wine are inks to their reactions to being heated or frasen: new wine, for example, iesald te then the work ofTheophrastus inquiries were undertaken in the field we should eal a quite detailed discussion concerning the species of fire and the effects of different forms of heat. It appears, indeed, that In the later part of data concerning, the properties and kinds af mineral aubatances, and On Pre contains vestigators. Theophrastus’ book On Stones, In partloular, le an important collection of the properties of different substances or thely renettona to certain simple tests,* but thereafter Aristotle's work in this field was followed up and extended by other ine The Meteorologica itself Is the frat exta work nt whieh deals in any great detail with on hearsay,’ conclude that he is simply repeating something then, it seems probable enough that thls In what Aristotle Is referring to when he says from ‘water’ as naturally as che liquid colourless, almost flavourless fluid of low who originally undertook these), Thus he says that salt and soda an #0 = a n er liquids, such as water, but not in others (he specifies olive oll, Mete.3 n° = I), ur > the substances which he says freeze solid with cold art not ved that ‘wine and flavoured liquide’ hecome water, and we should not necessarily Among vinegar, whey and lye (xovia, the alkaline solution used as a dewrgent), ri = Ixcop (serum) (3894, off). And he distinguishes between diflret cm = according to their combustibility and thelr readiness to freon (3970, 9Ht.; 3880, > n e d som. or prove vered n discon Sometimes, indeed, Aristotle explicitly claims to havehav e 3300). thing by a practical test, although the experiments w Meteorologica have often been severely criticized by modern schohm’ In one cie to show wishes nt when hera 9688, 34 gaf,ve these criticisms are probably quite justified At 3584 ien. that the saltness of the sea is due to an admixture, he descri the jar is recovered, fresh water will be found to have percolated thy <w a wax jar is let down into the sea with its top securely fastened, ation throug, ore Wik walls. But, as many scholars have pointed out, this does not, in ee 5a have concluded that Aristotle never undertook this experiment himself = the most pecan? er pd repeating the story on hearsay, This is, no doubt, +5 yet it seems just possible that he did carry out the test, and that he > into the ven thee quantity of fresh water inside the jar which was the result of the hr È vapour which the jar had contained before it was let down would not, of course, alter the fact that his theory as to how the water ame to be is quite incorrect). Another passage in which he claims to hive aarried out om experiments occurs in the same chapter at 3584, 16, ‘We have prow apr n (mereipaptvor)’, he says, ‘that salt water when evaporated bev vapour does not form sea water when It condenses again’ he # a me it Aristotle clearly deserves full marks both for the method he ~~ “sep |i In otherinocams obtained, But he then goes on::! ‘and the name In truedense kidagin, becon other xvpol, (flavoured liquids) that evaporate and co Wet water,’ But it is far from being the case, an editors of the en have pointed out, that when wine is evaporated In a still It become of potty ' See Mets, 383.0, Jaflı (with Lee's useful note) and cf, 3830, aaf on dumaheture and 3834, 7ff, on that of millstones. * Tg. H‚ Diels, Hermes, x1 (1905), pp. groff. was apparen tly accepted on tn by Pliny Goxxt, 1 Cf, HA, $90a, 2411, Arlatotle's statement and by Aelian (ix, 64).

EXPERIMENT IN BARLY ONBEK PHILOSOPHY AND MEDICINE 6 _ embryo derives nourishment from sucking the side of the womb,* From the Hippo notes that ‘this is not an operation that the man inthe street can undertake’, The theory the writer believes he has proved is in thet incorrect (If the pig's throat was stained, this was no doubt because the animal choked), but the method he uses to verify his theory is exemplary, Nor is this use of the experimental method an isolated example in this work. Later on he refers to another more complicated experiment which met windpipe as it drinks: ‘you will find It stained with the drink’, he says, although he which he stains some water, given It to a pig to drink, and then cuta open the pig’ into the lungs, This he proves, 10 his own satiuftetion at least, by an experiment in but suggests neverthel that a amall ess part of what we drink penetrates by this route the trachea into the lungs, He remarks that the epiglottis covers the larynx exactly, cratic texts one may mention first a passage in On the SacredDisease where the writer — refers to the evidence of postemortemdissections of goats to establishthat the ‘sacred ‘earth’ (On Stones, chs, 48f., espechlly) which implicitly raised the problem of the basic nature of this ‘simple body”, and then in On Fire (chs. 1ff.) he explicitly raised the question of the nature of fire and drew attention to certain important respects in which describ in On the ed Heart are eapeelally worth nating, In eh, a CL, 1x, No, off) the writer considers thequestion of whether any drink passes from the oesophagus via it differs from the other ‘simple bodies’ (notably in that it always exists in a substrate). If the four-element theory as a wholesurvived these and other challenges in antiquity, this was due to a complex of reasons, partly no doubt because the methods of investigation used were still fairly rudimentary (in particular insufficient use was made of quantitative measurements, even though the Greeks were accustomed to distinguish different types of waters, and indeed also different solids, by their weight‘), but partly also because of the nature of the fourelement theory itself: this provided at once an extremely comprehensive, and a quitelmprecise, theory embracing each of the different states of matter, solid, liquid and gaseous, so that the effect of much of the research that was undertaken was to modify the application of the theory in detail rather than to show that a more adequate general theory was needed, So far I have dealt entirely with the physical sciences, bearing in mind that it is to notices, among other things, that the ventricles and the auricles do not contract with greater success, Having described the general structure of the heart (where he wards, was dissection, including not oily the dissection of dead adult animals, but also these that reference is usually made to substantiate the view that there is a radical difference in method and outlook between ancient and modern science. My remarks on biology and medicine must be ever more compressed than my discussion has been so far. In the biological selences, the key method of research, from Alcmaeon onsimultaneously)? he turns in ch, 10 to tt “hidden membranes’, giving a brief but exact description of the semi-lunar valves whieh lie at the base of the aorta and the pulmonary artery. But he not only describes thelr structure: he investigates their by suggesting that a litle air can and does penetrate Into the heart e Gite | Be known gna: nils MA, sou, ay m— a animals + fa blood into the differs gira ho pf of + Ie is notable chat in the De mom cordis ot wann In antmalikur not only to Galen, but also to Arlitotte and ve Hi ton (men in ch, 19) and among the passager he elten In a vert ‘in ek often 10 apselmens of g aart prevent of the ns valves placed at the openin 5 Aristotle, for Instance, refers were, apparently, deliberately en method of locomotion of different ele of EA, Th a ri through the aorta or pulmonary artery inte the heart, but he qualifies thie in eli 1a, 90, 14 = 93, 1, Ch. 10, 86, 13 = 88, 9, At MD, 6, the welter amerte that nov even ale ean be pumped back Ch, 8, La 14, 86, af, supported cà various arguments, eg. that the mt Und cited from early Greek investigatore in the biological selenees.! Hut while the extent Many other experiments (whether involving the use of dissection or not) might be blood essential (if only a preliminary) step towards thediscovery of the elroulation of the is more effective on the left side of the heart (he; at he base of the aorta) than on the right (at the base of the pulmonary artery)’ tt la perhaps not out of place to note that this demonstration of the irreversibiliey ofthe flow of the blood out of the heart was an function, showing that if on removing the heart you attempt to force water down either the aorta or the pulmonary artery, the samielunar valves will prevent any fluid passing into the ventricles of the heart, and he even notices that the seal formed by the valves (as time went on) the dissection of embryos and vivisection.! Now dissection is always an experimental procedure inthe weak sense in that it involves not simply direct observation, but ‘observation provoquées”, that is, a piece of research deliberately undertaken to discover fact, But sometimes we find dissections used not merely to uncover facts but to proveor disprove suggested theories, At GA, 764a, 3348, Aristotle is able to refute the doctrine which we find expressed in various forms in earlier writers, that the sex of the mbryo is determined by the side of the womb on which it is conceived. ‘Male and female twins are often found together in the same part of the uterus’, he says, ‘ Thi we have observed sufficiently by dissection in all the Vivipara, both land-animals ad fish."* And in another passage (GA, 746a, 19fT.) he again refers to the evidenee of dissections to disprove the notion that the ! The author of On Airs, Waters, Pla (ch, 1, CMG, 1, 1, 96, 71) for example, notes that waters differ a good deal from one anothe both In taste (lv rq) errbgerri) and in weighe (tv ré oradu@), while in On Stones (cha, aa ani 19) Theophrastus refers to differences in ‘density’ Cruxvérens) and weight (P&pos) an method of distinguishing between different types of ‘stones’ der Sektion in der Antike', Qui u. Sud, e. Gesch, der Naturwissenschaften u. der 1 The hisgory of the use of dissection h antiquity han been deserthed by L, Edelatein, ‘Die (including, eg, ores). À {have dealt solely with the fourselemen theory, But it may be noted that ite main rival in the fourth century, the atomistic theory, was equally comprehensive and vague, and equally incapable of practical tenta, of being corroborated or falsified by mean Geschichte Medizin, ut (1933), pp. 100ff. 4 At GA, 764 a, 21 fl, Aristotle refers to ose who held that the sex of the embryo ls determined conclusion, by whether the seed of the male comes fom the right or the left teutiele, and who apparently thought chat If one of the testicles In tled upor excised, the oflupring produced are all of the same vena passage which indicates that some arlier writers were aware of the possibility of putting this theory to the test, even if they evidenly assumed the results of such a test to be a foregone

LLOYD EXPERIMENT IN RARLY GREEK PHILOSOPHY AND MEDICINE Files es, for Ina of the researches which had been carried out in such fields as anatomy, zoology and botany by the end of the fourth century nc. is generally recognized, it is often ae that this is, as it were, the exception that proves the rule—that the fact that men w - inconclusive, The author of On cosmologists on the similar, though more general, problem of the eleme physical substances as a whole, — = Several Hippocratic writers elte practical tests concerning the elements of the body, although 1 were mostly trained in the practical art of medicine successfully applied — various subst ances among which ‘the glutinous | are particularly important, but he suggests (ha methods to the study of biology shows up all the more dramatically the failure o other Greek theorists to do the same in other fields, Now it is well known that some of the most rigorously empirical texts of early Greek science come from the medical writers (the collections of case histories presented almost without interpretative comment in the Zpidemies are the most obvious examples), though equally some of the physiological and pathological treatlses of the Hippocratic Corpus are as si readily cook, while what is ‘fatty’ will). ‘This seems to be atrletly comparable wi type of observations whieh appear in the fourth book of the Afezeoro whe reactions of different substances to being boiled, burned, frozen, ete, are noted « Then the author of On the Nature of Man refers to certain rather more ambitiow carried out with drugs to back his theo that the ry primary constituents of the box and speculative as any of the commologieal writings of the philosophers. But that some medical theorists, at least, believed there were important differences between the four humours. First he says that one may discover that these exist as differen stances in the body by givinga man differ types ofdrugs, ent which will draw out ph their own methods of research and those used In such fields as ‘meteorology’ is clear from On Ancient Medicine, even though It should be pointed out that in practice, in his bile and black bile respectively (eh, $, Ly vi, 48,61), But then he says that the ‘el lives up to aie own physiological and pathological doctrines, this writer hardly that the results of the various tests to | a process of coagulat substances put Intoa bladde whieh were (very roughly) similar, The rosta the intercommunicating vessels, and with different ch, 6, 44, 111,3 46, 4, I Several examples from the trontises On Genera gecon, Cha oge wik - ] N Al bur * forme sh 4 Where he ren Pr * Ch. 7, esp, yo, off, Elsewhere too thin weiter describes the affeta of different drugs " Ch, 4, La vini, 588, 25 = 590, di eure the behay in connection with his theory that the liver ie formed erally simpler, subst: given from On #4 a saerifielal v humours he had observed, or thought he had© served, In a man's vomit, ar mental. Many other ingenious, but quite undanı tenta are found 4 Hippocratic Corpus relating to sueh toples na the | formation of the parts of the or the pathogenesis of various diseases,’ Son simple tests are carried out dir what he was attempting to establish, for he assum what es is the point at iasue, th even if the tests produced the results he mentions, they would not, of course, | this writer refers would correspond at all preeisely with those whieh he describe biackest in aurumn’.? Now it is wal proof” that the humours alternate In the body In a eyele according to the seas Yet notwithstanding the evidence of On Ancient Medicine, the differences between expressed ideal of excluding all unwarranted assumptions from the study of medicine. that ‘if you will give the same man the same drug four times in the year, his vom be most phlegmatic in winter, most liquid In spring, most bilious in summer Greek medicine and Greek philosophy in the matter of the attitude towards, and the use of, empirical verification and experimentation may not be so great as is sometimes made out, On the one hand I have already suggested that we should not underestimate the extent to which simple practical tests were carried out in some branches of the physical sciences: the early Greeks took some, if not all, of the most obvious opportunities for experimentation that presented themselves. Nor, on the other hand, should we overestimate the extent to which the medical theorists were able to devise experiments to illuminate the more difficult problems that faced them.’ We have a that some problems in the biological sciences were investigated experimentally with - But more often attempts were made to derive rt of substances in the body by Meer ben on organic substances from the badyt one example has been and elsewhere the same author deseribes testa on blood take good deal of success at an early period. But there were, of course, whole areas o physiology, embryology and pathology where the Greeks posed major questions which could not be resolved by the use of dissection or the carrying out of simple tests. It is instructive, then, to consider the methods used, and the results obtained, by outside the body under conditions the medical theorists in attempting to determine the constituent elements of > body, for example, and to compare these with the work of men who were primarily it cn : È ' ‘Thus although the writer attacks those who based their pathological theorles onthe hot, È the cold, the wet and the dry, he himself numbers such things as the salty, the eri the giro rn acid and so on among the constituents of the body (eh, 14, € MG, ) di 4, 2 ( vo ey i sf) : gf © x 11,96 zr criticizes his opponents for oversimplifying the causal prinelples © similar eriticiam might also be levelled against hie own pathologieal theories, pe rich ni È , for example, to the diseases which arise from ‘depletion (véveooig) and from ‘repletior WA 16) (eh, 10, 42, 11 fl), : peneen md Hai the Greek biologiste might have carried ente een a ] without great difficulty, but failed to do so, In in investigating hybrida (although coagulate 4 Ch.8, Le van, 594, off, He notes that and of, alaaf tv are discussed by Senn, op, ein pp. atoll, di antiker Naturwissenschaft’, uel u, Studs ti : er. ; ll, 1 (1931), pp. 131 remarked that for all the interest in heredity in post-Renalssance times the simple nen a ch x crossing peasplants which led Mendel to the discovery of the law of the gaten © = = + Bals dad con I he also observes that it does not experimen is formed), that when the ‘skin’ ie removed from afterwards (an observation whieh he uses to wi sup were not DRE mee until the mid-nineteenth century). Equally ige the Ss “ftme a generation was challenged by Redi in the seventeenth century, from the putrefaction of the meat itself, were not technically Impossible in antiquity. show that the worms found in decaying meat derive directly from the droppings

O, Te R LLOYD of water and blown on, have already been mentioned, but the Hippocratic writers used these in the former case to illustrate a theory concerning the passage of the humours from one part of the body to another," and in the latter case to suggest how the parts of the body are formed by the breath causing like substances to come together.’ Similarly we find the author of On the Diseasesaf Women |, forexample, illustrating his theory that the flesh of women is more absorbent, because rarer, than the flesh of men, = EXPERIMENT IN BARLY GREEK PHILOSOPHY AND MEDICINK 7! part of the fourth century n.0, sees the firs collections ncerning different properties of substances and chele ric to nl = collections of factual information were no negligible one Ze. lead to the founding of che mic selence al on a firm basis, AM an a whatever the apparent Aristotle him le | feld (and we cannot doubt the success with shorteomings of ran we een hime mn a ani Ma doctrines of Aristotle in late antiquity), Siege en successi : vere, for ti | ; ackn elect, =a methods in chie feld is ear through the Alexandrlans En A une of experimen and here an increasing, and by referring to a test in which equal quantities by weight of unwoven wool and a made-up garment are left suspended over a howl of water, and the greater absorbency y of the unwoven material is shown when the two are collected and reweighed.? A “i (fb submi in iù ne remem) ar relative ease with which la à pointie which Aristotle stressen trouble euch one of , ’ ri gg can we much ie concerning om gs w et, Having er en emp has some hi opinion may have gained curreney, Here one might vefer 16 4 completely failed to experiment, la false, 1 should “y A final question arises which must, however, suggested that the view prevalent In da a their kinds,'* anyone that is willing to take su the things that perish, that is plants and animals,ads passage in the de Partibus Animali where he contrasta the study ma a fino! animals with astronomy. ‘We have better means aw ledge’ he el en - information can be gained in blology logy, by means of dissections), It will he ignored, is the relative ease with whieh some test in certain branches of the biologlonl sen mental methods to biology: a more obvious polnt, but one that ze, lu not the only > (differences 4 which On Ancient Medicine draws attention), this mind when considering the relative success with whieh the Oreoka pen A the other, while there are certainly Important differences in out een” à, between some medical theorlats and the major of the Men Be mer tional and due to the fact that from the Hippoeratle writers and Arlator inei tests to corroborate their theorles, and indeed trled to do this in practice (even though the accuracy of their observations often leaves much to be desired), In investigating Herophilus, down to Galen! This often been suggested that the use of the structure of the body, for example, they wed empirical techniques such as dissection most successfully, Yet on such intracuble problems as the constituent elenumber of medical writers undoubtedly recognized the desirability of undertaking, ments of the body, the formation of the different organs, the origin of diseases and so practical discipline of medicine, bega pon pen, 3 sn = exaggerated. On the one hand we should nor underestimate a. ge hich += simple experiments were performed in other fields of sclence eend cons to on, their attempts to make une of practical tests met with little success, for the tests to which they refer were at best inconclusive, and at worst quite irrelevant, Yet if they failed to devise crucial experiments on the more complex problema of physiology, biochemistry and pathology, this was evidently not always for lack of trying: rather their failure must be considered to a large extent inevitable, given that the successful experimental investigation of many of the questions they raised had to wait for (among, other things) the development of chemistry. In conclusion 1 should summarize the main points I have tried to make in this very compressed discussion of certain features of the role of experiment in early Greek philosophy and medicine, First It seems to me that the question of whether the Greeks experimented cannot be meaningfully discussed in global terms: what we must try to do is to assess the achievements and failures of the Greeks in each branch of scientific inquiry and at each period independently. Secondly, 1 suggested that even where the failure of the Greeks to experiment la most notorious, that is in the physical sciences, we find that in such fields as acoustles, opties, pneumatics and hydrostatics, quite successful experimental investigations were carrled out in antiquity, and if the most striking, successes are all the products of the Alexandrian period or later, the first attempts to undertake simple testa In those fields can be traced back to the fifth or fourth century n.C. Here, where the problems investigated are relatively elementary, on the passages in the Aepubliet in whieh plato us. + re animal superior oe ia iat wen om en rn Mot È ce wi astronomy, acoustics and so on, and deerles empirleal lavigne ' I may mention especially the remarkable and where tests can often be carried out without great diflieulty, the Greeks were far from ignoring, the experimental method. The early history of mechanics and what we should call chemistry seems to tell a rather different story, but in chemistry the latter 1 On Diseases tv, ch. 39, Le vit, 556, 19 ff, muiggizonte dat as all the Vowels may be filled or emptied » by filling or emptying any one of them no the reservolin (myal) of the humours in the body are vessel without food for a given pertod of | plants and our in which Fraslstrarus ews that there are filled or emptied by the stomach being filled or emptled, ae A, ch, 5, 6444, ee excreta and comparing, tile with tie I Eg, saga, fl, and 4414, Mh that our knowledge af on about them, that en gives un greater ernia} e I wei * On the Nature af the Child, ch, 17, L vir, 496, 1710, argues that growth taken place when the breath, 1rveOua, in the body separates the different mbitances according to thelv kinds (the dense, sand, leadefilings) will be the rare and so on), just as the substances put into the bladder (earth, found to be sorted according to thelr kinds when the bladder is left to dey and opened, » Ch. 1, Le vin, 13, off, It is worth remarking that this test, like chat in On Airs, Waters, Places, ch, 8, mentioned above, p, $9, n. 2, involves the useof quantitative measurements,

Plato's view was no doubt influential in antiquity, it did not act as the deterrent that some commentators have supposed. But more important, there is a tendency in several nineteenth-century works on the history of science to assume a constant, linear development of science, and it might be suggested that it was for this reason (among others) that there is a certain reluctance, in many such works, to allow the claims of Greek science whether in respect of methods or of results. That Whewell, for one, experienced a certain difficulty in giving credit where credit is due, may, I think, be seen from a passage from the History of the Induetive Sciences where he discounts the claim of Aristotle to have formed a systematic classification in zoology: ‘it would be difficult’, Whewell says," ‘to reconcile such an early maturity of zoology with the G. i. Nn LLOYD succeeding each other, are requisite to the formation of any considerable science’.* conviction which we have had impressed upon us by the other parts of our history, that not only labour but time, not only one man of genius but several, and those KING'S COLLEGE, CAMBRIDGE 5.2.6. IX, 63 AGAIN involves a state of - altogethe: _Hoyev, ant time. Whi has brou, priest’, This suf not only u hold offie _reuodkpeve Is a refe Robert (pi . recent deli historical i end of the: was so fam the accom) In the last number of these Proceedings (no, 189, n.8, 9, p. 2) Miss Joyce Reynolds discussed this well-known inseription, whieh is dated in A.b. 2/3 and is incised below a relief of a man reclining on a couch, cup in hand, It was found in the Sanctuary of Apollo at Cyrene in 1925; for a bibliography, see Miss Reynolds's article. The text she prints at the beginning reads as follows: from dang, AvorrréAspov (trous) Ay’ Erri leploas Tauoavia Dilan plan Bè Evpdveus vacat mavoapivav rrévraov Tig dvelas AovKios “Oppros Aouklou muAokAnoThs TÒv fivixa MappapiKod Afigev roAtuoio wuBormds yrjOnoev Bérrrou mroXAk mróAis peporrcov rhuos &v[a] yNinpas xorraxeluevov fBurro{mo)raûvre Aetos elvoBicot Ofike rapd mpo@Upe vAat[Bat] mvAns Biérroov “pon plAen of xs foyav Morvoovier fepfi kaıpsdı mavodpevov The last sentence raises acute difficulties, It has driven Adolf Wilhelm and Louis Robert to the desperate assumption that we have here one of the rare cases of a spondee Miss Reynolds has made an important contribution to the solution of this baffling standing in place of the penultimate daetyl of a pentameter, Wilhelm read Keapous and Robert kipav: both imagined that the words meant the same as diddy travodpevov, The authority of these two eminent scholars should not prevent ua from seeing that the suggestion is ridiculous, problem by reading the final letters as konpòs è rravadugyols. But she has not offered a convincing, interpretation of the sentence as a whole; the words printed above as où ! History of the Inductive Sciences from the Earliest to the Present Times (London, 1847), vol. 3, pe 344: fe of tht À T must express my gratitude to Dr M, B, Hesse, who read and criticized an earl paper, and to Dr H‚ B, F. Dixon and Mr F. H, Sandbach for thelr pre and en particular points,