The use of Eidos and Idea in Hippocrates

Gi res Pe att. C À Kale DI Ci wt \ea V2 THE USE OF Elöos AND ‘Téa IN HIPPOCRATES, AFTER reading carefully the essay which, in his recently-published Varia Socratica, Part I., Prof. A. E. Taylor has written on the use of the words elôos and iôéa in the Greek literature of the Socratic and Platonic periods, I find myself on the one hand in agreement with him as to the importance of such linguistic investigations for the understanding of Plato, and on the other in frequent disagreement with him as to the meaning of the words in the passages he cites, and the inferences he draws from them. Thinking that Prof. Taylor’s analysis of the use of the terms in Hippocrates is very far from final, I offer in this paper a further contribution to their interpretation, confining myself as far as possible to the question, What do elöos and iôéa mean in the Hippocratean writings ? The date of the several works is a problem into which I do not enter, assuming with Prof. Taylor that they belong as a body to the late fifth and early fourth centuries; but I am fully aware that this collective treatment is not altogether satisfactory. My citations of the text are from Kühlewein’s edition (K), so far as published; for the rest of the works from Littré (L): the references are by paragraph, volume and page. I make no attempt to distinguish between eidos and iôéa, which appear to be used almost interchangeably. Prof. Taylor's results may be thus summarized. I am not quite certain about some of the details, because of the unsystematic nature of his exposition, but believe that my summary is substantially correct. Prof. Taylor nowhere formally classifies the meanings of elôos and iôéa, but an arrangement under four main heads may be extracted from his account. The nucleus of meaning is form, or structure. (i.) The first meaning is taken directly from popular usage. The words are here applied specifically to human beings in the sense of bodily shape or form, primarily as it presents itself to the eye, but extended to the structure and constitution of the body as a complete thing, not as merely seen. Physique, body, constitution are the proper renderings of the words in this application. (ii.) The words are also used more widely for the form of things other than the human body. This group of meanings contains, as might be expected, considerable developments. (1) Sometimes the outer visible structure is emphasized, when the words may be rendered by shape, ‘in shape like a trough.’ (2) Or the inner structure may be included : form, nature, like puo. NA TAS CLA. AQHA Nat. UY

The Classical Quarterly, Vol. 6, No. 3. (Jul., 1912), pp. 179-203. Stable URL: http://links.jstor.org/sici?sici=0009-8388%28191207%291%3A6%3A3%3C179%3ATUOEAI%3E2.0.CO%3B2-W The Classical Quarterly is currently published by The Classical Association. Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/classical.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact support@jstor.org. http://www.jstor.org Fri Jan 26 05:20:03 2007

AFTER reading carefully the essay which, in his recently-published Varia Socratica, Part I., Prof. A. E. Taylor has written on the use of the words eldos and idéa in the Greek literature of the Socratic and Platonic periods, I find myself on the one hand in agreement with him as to the importance of such linguistic investigations for the understanding of Plato, and on the other in frequent disagreement with him as to the meaning of the words in the passages he cites, and the inferences he draws from them. Thinking that Prof. Taylor’s analysis of the use of the terms in Hippocrates is very far from final, I offer in this paper a further contribution to their interpretation, confining myself as far as possible to the question, What do eiôos and iöca mean in the Hippocratean writings ? The date of the several works is a problem into which I do not enter, assuming with Prof. Taylor that they belong as a body to the late fifth and early fourth centuries; but I am fully aware that this collective treatment is not altogether satisfactory. My citations of the text are from Kihlewein’s edition (K), so far as published; for the rest of the works from Littré (L): the references are by paragraph, volume and page. I make no attempt to distinguish between eidos and (déa, which appear to be used almost interchangeably. Prof. Taylor's results may be thus summarized. I am not quite certain about some of the details, because of the unsystematic nature of his exposition, but believe that my summary is substantially correct. Prof. Taylor nowhere formally classifies the meanings of eidos and idéa, but an arrangement under four main heads may be extracted from his account. The nucleus of meaning is form, or structure. (i.) The first meaning is taken directly from popular usage. The words are here applied specifically to human beings in the sense of bodily shape or form, primarily as it presents itself to the eye, but extended to the structure and constitution of the body as a complete thing, not as merely seen. Physique, body, constitution are the proper renderings of the words in this application. (ii.) The words are also used more widely for the form of things other than the human body. This group of meanings contains, as might be expected, considerable developments. (1) Sometimes the outer visible structure is emphasized, when the words may be rendered by shape, ‘in shape like a trough.’ (2) Or the inner structure may be included: form, nature, like püous.

(3) They may be extended to objects which are not in themselves bodies, and in this connexion may have a semi-classificatory sense: forms or types of disease, etc. But Prof. Taylor insists that they never mean class or species, as eidos does in Aristotle. (iii.) This use is specifically medical. It emphasizes and extends the appearance-side of the meaning, till the words stand for symptoms of disease, the form, shape, structure, etc., of the observed body or limb as it appears to the observing physician. I gather that Prof. Taylor regards this usage as underlying the employment of idéa by Thucydides. (iv.) In the fourth use the bodily side of the first and popular meaning is emphasized. (1) Eidos is used of the stuff, matter, substance, of a thing, e.g., of the heart and kidneys. And more especially it is Professor Taylor’s object to prove that it is used (2) in a highly technical metaphysical sense to denote the ám)\â ocwpara, the elementary bodies. To say that dip is an elôos is to say that it is one of the ultimate unanalyzable bodily constituents of the «édcpos (p. 228). It is thus nearly equivalent to the later aroıxetov. This meaning he supposes to have come into medicine from the puouxoi; and to explain how the word elöos can advance from the meaning of form or structure to that of primitive substance, he suggests as the connecting link the Pythagorean doctrine that the difference between one kind of matter and another is that of geometrical figure (p. 246). Because the Pythagoreans regarded geometrical figures as the ultimate substances, the word was extended to mean things or substances. I am not quite sure, but I think that Prof. Taylor considers the use of the word as the matter or stuff of an ordinary empirical object to be derived from this. Prof. Taylor distinguishes between ‘popular’ meanings and ‘technical’ meanings, i.e., those belonging to the scientific vocabulary of the day, but it is not easy to determine where he draws the line. Groups (i), (iil), (iv) present no difficulty: (i) body or bodily structure he regards as popular; (iii) symptoms is obviously a use technical in medical clinique; (iv) substance and element are specifically physical conceptions. But in respect of group (ii) it is hard to say whether he considers these meanings to be popular or technical, or some of them popular and some technical. When he infers from the absence of the words from many of the practical treatises that they do not belong to the language of the working medical man, except in a current nontechnical sense (p. 225), he seems to imply that the whole group of meanings, found regularly in such practical treatises as wep! dpOpwv éuBorrs and uoxAurov (see p. 224), is non-technical. But on p. 243, apparently, the only current non-technical sense he recognizes is the first, living body, physique, constitution. His rendering of eldos as geometrical figure on p. 224 and elsewhere, coupled with his account of the development of meaning on pp. 243 sqq., seems to indicate that he regards the second group of meanings as technical, and primarily due to Pythagorean influence. I believe that Prof. Taylor's most important conclusions are based on an

erroneous analysis of the data; in particular, I can find no evidence whatever for either of the meanings under group (iv), substance and element. (i.) With his account of eidos as applied to the form, physique, constitution of the human body I agree in the main. He shows that in nineteen passages of the treatise srepì dépwv vddtwy Torwy, as in many passages of Herodotus, the word means the form or structure of the body as a whole, not the face only. But in one or two points his treatment is open to question. In his rendering of eidos in this connexion as ‘ body,’ he lays too much stress on the ‘ bodily’ factor in the idea of ‘ bodily constitution.’ It is true that in several passages of the above-mentioned treatise elòos is contrasted with #0ea and rpórrot, where the opposition of body and soul seems implied. But elêos is not a synonym for ooua; in wept doréwv II, ix. 182 L, the bones furnish the côga with its elôos or shape. When the opposition between body and soul is explicitly put, capa is used; srepì dpxains intpexfs i. 11 K, wept dépwv i. 67 K, mepi Téxvns vi. 10 L, mrepì dvo@v vi. go L. The word lays stress rather on the structure element of the idea. This is especially the case in certain passages where the word has a technical rather than a popular use, as the constitution of the individual patient. In these passages the writer enumerates the special conditions which the physician must observe in his clinical practice :— mrepì xvuÔv I, v. 476 L. aœvyyerès eldos édos mauxin Opn kr. T.’s list.) mepi pvatos avOpwmov 9, vi. 52 L. (Not in Set tov inrpòv évavrior toraadau rois xabeotedot Kal vovonuaci Kal elÒeot Kal bpnot Kal NAıkimat. Ibid., vi. 54 L. oxeÿräuevor tod avdpwmov Tv piow, Thy Te NAıkinv Kal TO eldos xa) thy dpnv tod Ereos Kal THs vovaov Tov Tpórrov Tv Oeparreiny mouetobas. Prof. Taylor gives ‘ bodily constitution ’ as his rendering of elöos in this passage, but most perversely translates ‘substances’ in the earlier passage of the same treatise. In this sense of particular constitution púots is much more frequent, as, e.g., in srepì yvvareiwv B' 111, viii, 238 L. œxemréov de Tas púÚoras Tv yuvatkav Kal Tas xpods kab tds huxias x.7.A. This is only one among many instances of the close parallelism between gúots and eldos in the corpus. In this usage the idea expressed by eldos is just as distinct from that expressed by cpa as the idea of constitution is from that of body in a modern physician’s thought. Similarly when a popular lady novelist speaks of the hero’s stalwart form, she does not mean the same thing as body. Hence Prof. Taylor’s rendering ‘ body’ in this group of meanings is misleading : if he uses it merely to emphasize his point that elôos refers to the shape and nature of the whole body and not to the face only, it would have conduced to clearness to say so explicitly; if he had in his mind the alleged extension of the word to mean bodily substance, his procedure recalls the «A&yaı introduced by Socrates into

the argument by which he makes Polemarchus admit that the just man is the best thief. (ii) A well-marked meaning of elöos is that of bodily shape as applied to parts of the body or to inanimate objects :— mepi kebaAns Tpwudrwv 5, ii. 7 K. idées rijs pAdotos: the shape of a bruise. mepi dpPpwv éuBorñs 3, ii. 32 K. eldos Tod yerptLouévov: the shape of the limb to be operated on, contrasted with its oyua, the word regularly used in the surgical treatises for posture, position of the limb. Moxey I, ii. 246 K. eldos fasBoudéoraror Tav Egwv : the most crooked shape (of the ribs). Ibid. 1, ii. 245 K. eldos kovövA@des: the bones of the thigh are of knobby shape. Ibid. 6, ii. 251 K. dxpdusov àmoomac0év: To pèv elòos paiveras olóv ep Öpov exreoövros. Shape or visible appearance. mept xapdins 5, ix. 84 L. B60pov To eldos elkeXov öXuw: a trough-shaped depression. mepì doréwv 4, ix. 170 L. eldos Kxapdins oi vedppol Exovaı' Kal odrot KoiNımdees. T. renders ‘substance,’ but the context shows that the writer is entirely taken up with the mutual relations of the organs; the only other use of eidos in this work is in the sense of shape: 11, ix. 182 L. Ta òoréa TÔ omyarı ordow Kal òpdórnra Kal eldos TapéxovTat. In these cases Prof. Taylor renders by ‘shape,’ ‘geometrical figure.’ I submit that the introduction of the adjective ‘geometrical’ is utterly without warrant. In not one of the passages cited is there the slightest suggestion of geometrical ideas. I can understand a knobby shape, but not a knobby geometrical figure. In point of fact, the writer of mepì kapdins, who describes the depression in the heart as in elöos like a trough, does describe the heart as pyramidal in shape, and then he uses the technical geometrical term oxnua (capdin oxnpa pèv oKxoin Tupapis, Xpoinv Sè karakopys powikéa 1, ix. 80 L). This usage is very closely connected with the meaning, bodily form of a man. The derivation of the words shows that the idea of the form of a man is specialized from the more general idea of shape or visible appearance. As compounds with elèos are specially frequent in Hippocrates, either in the sense of x-shaped or the general sense of x-like, this idea of the intervention of geometrical notions in the development of the use of elôos is neither more nor less than absurd. Here the parallel with duo is close: the latter word is constantly used in the specific sense of human nature, individual constitution, constantly in the sense of nature, constitution of anything; there is no need to derive the second meaning from the first, as the first is a technical specialization of the second. I may be wronging Prof. Taylor here; he may include these

uses as non-technical; but if so, he must strike out his ‘ geometrical figure,’ which can only be an intelligent anticipation of his hypothesis. Prof. Taylor acknowledges a use of eidos, easily derivable from that of the nature of a thing, in which it has a classificatory suggestion: e.g., ‘ many other types or forms of fevers’ (7oAAd Sè xa) ddXa Tuperav Erreönumoev eldea. émudmu. y, 12, i. 229 K) (p. 222). But he insists that the word never means class, kind, sort, but always form or type. This meaning is indeed an intermediate link in the extension to the meaning of class, which was only arrived at later, viz., in the Aristotelian developments of Platonism (p. 221). The words used to mean kind in Hippocrates are rpömos (p. 222), and é0vos (p. 236). I agree with Prof. Taylor that the word does not mean logical class in Hippocrates: but it frequently has a classificatory use, just as in English we speak of ‘these types of character,’ ‘a malignant form of the disease.’ Every language contains words which express a logical division and something more besides, as they contain metaphors of various sorts: e.g., in English, form, type, sort, kind, class, etc. In use the metaphorical origin sometimes prevails, sometimes the implied logical idea: in the former case we feel that one of the words is more appropriate than another in the context, in the latter we tend to use our classificatory words interchangeably. Now Prof. Taylor’s procedure is misleading in the last degree. He will not allow elêos to have a purely logical meaning, but apparently assigns this to rpómos and é0vos. The truth is that these words stand on the same footing of elöos: they do not mean class primarily, but have classificatory uses, in which the root-idea is sometimes strong, sometimes has practically disappeared into the division idea. There is no word in Hippocrates for a logical class; when he wishes to classify, or when classification underlies his thought, he uses elòos or rpórros (very rarely and in a special context é6vos) according to the nature of the things classified. And this much may be conceded to Prof. Taylor, that the classification is more properly division; it is not regarded as collecting things and arranging them according to their common qualities, but rather as taking a unity (not explicitly thought as a logical unity, but ultimately analyzable into it, as Plato saw), and dividing it. Yet the classificatory idea is always near, and sometimes predominant: when we read that there are several tpdmot karıkıos, the unity of the idea of fracture is contrasted with its modes or varieties; when it is said that there are many other Tpômoe xatataciwy (genitive plural), this is not far from ‘several other karardotes,’ and the plurality-side of the result is emphasized. Ifthe division deals with events, acts, processes, Tpé7ros is used: e.g., a limb is dislocated in four Tpómot :— erminte dè Téooapas tporovs, N) dvw N käTw ú és TO mAdyıov éExaTépwOer. 3 Id x ! [4 A OM A 4 A 2 x 4 e LA mrepì ÄpOpwv éuBorijs 80, ii. 240 K. unpoî apOpov emimre Kata réooapas Tpôrovs ibid. 51, ii. 188 K. There is more than one rpórros of fracture (iy Sè 1) pls karayÿ, Tpômos pèv oùx els dort Karijkios ibid. 35, ii. 154 K), of distortion (ku\Xwoios yap oùx els

TpoTos. poyxMiòv 32, ii. 263 K). The word is often used for modes of treatment (e.g., moXAoì de Kal äAAor TpdTOL KaTaTaciwy. Tepì dypÔv 13, ii. 68 K); often for modes or varieties of disease (ds oùv roÚrov Tod Tpörmrov oparepod éévros. mpoyv. 22, i. 102 K; To Sè dppotov &b’ éxdotoroi TOY TpÔTY eipnoerat. mepi o&éwv 12, i. 114 K; etc.). Prof. Taylor objects that elòos, if translated by kind, class in Hippocrates, always loses some of its force; the remark applies equally well to rpomos. His statement that in mepè dvoôv the word for class, kind, is throughout &#vos (p. 236), is most misleading; the word occurs three times in one paragraph with the same sense, in which men are said to be one &0vos of £a; surely the original idea of race is dominant over the class-idea. The only other example is in the same paragraph, where the MSS. vary between évea and eidea (ore Sì Stood eldea muperdv. epi buoûv 6, vi. 96 L) ; the Hippocratean usage is strongly in favour of elôea. I may remark in passing that the Greek mode of expression, with its rotodros (or rovovrórporros, frequent in Hippocrates), often gives the idea of kind without its name; we have to say ‘of the same kind.’ That the word elöos has a classificatory sense in Hippocrates over and over again can be easily shown. I maintain that Prof. Taylor's treatment unduly ignores this sense, and that it is just because he wrongly ignores it that he is led to the quite unfounded view that the word comes to mean stuff, matter, substance, primary body. Observe, in the first place, how often it is conjoined with numerals in a context which plainly indicates division of a unity into varieties, kinds, forms, or whatever may best express the idea in the context. three examples :— kar’ inrp#io contains 3, ii 30 K. avyñs pèv Òúo eldea, TO pèv Kowòv, TO de reyvnrov. T. ‘kinds.’ 7, ii. 34 K. &miöeoios Slo eldea, elpyacpévov kaì Epyaböuevov. There are two ways of bandaging, one after, the other during the operation. T. thinks the idea is ‘figures,’ as in a dance; as if the writer visualized the whole process of disposing the bandage. 8, ii. 34 K. dyadds dè dio eidea tod émideouévou" ioxvos pèv 4} miete ij mAndeı 68oviwv. The phrase is awkward; the writer means that you must either bind the bandage very tight or apply a great thickness. T. ‘figures,’ ‘ ways of construction.’ I suggest that Prof. Taylor’s way of getting a meaning by a fully concrete image of the situation is mistaken. A careful but less subtle translator would see that the writer has a formula for dividing into kinds or sorts; the very difficulty of getting a precise meaning shows that the writer’s thought is abstract, that he is thinking of the illumination and the bandaging simply as of two kinds (or some such very general word); i.e., that the classificatory aspect of the word is predominant. And where does Prof. Taylor find eldos used literally of a dance-figure?

mept bÚotos dvOpérov 15, vi. 66 L. of mAeîoror TÔv muperav yivovrar amo xoNÿs" eldea Sè chewy éorì récoapa. T. ‘types,’ ‘sorts’; and he adds ‘meaning, no doubt, originally geometrical figures’! mept guody 6, vi. 96 L. tori de rood eldea muperay: reading doubtful; v. 1. é0vea. mepì yovns, 3, vii. 474 L. eloì Sè récoapes iSéat Tod Úypod, alma Yon Üdwp kaì dréypa. T. ‘figures, i.e., types of moisture, or perhaps ‘substances of the moist <kind>.’ Ibid. 11, vii. 484 L. «al Tod bypod aùr®, ad’ où Td omépua Yiveras, Téocapes ideaı Eodoaı oxooa Ev duce dmp£ar. T. ‘ bodies’ or ‘constituents’ ; but surely the meaning must be the same as in the preceding passage, which is directly referred to. This is one of many instances where the neglect of a simple comparative method leads Prof. Taylor into unnecessary refinements. mept voúcov & (by the same author as the srepì yovjs) 32, vii. 542 L. Ever Kal 7 youn Kal 6 dvip Téooapas idéas bypod Ev TH pari . . . abra dè ai ideas eiot phéyua alpa yor) Kal Ddparp. T. ‘substances’; but the meaning must be the same as in the two preceding passages. I will discuss them fully later. Ibid. 57, vii. 612 L. adraı ai Tpeis iSéar Tv vovenudrwv amd Tod Döpwmos. ep yuvaixetwv a’ 96, viii. 222 L (not in T.). 7ò péXav gdpparov . . . mouéeuv Sio 4 Tpia eldn Tod dappdrov, Tò pèv loyvpórarov TO avOos Tpurnuópiov Ts Aemidos, TO Sè Sebrepov Teraprnuöpıov, TO Sè Tpirov TEUTTNAOPLOV. ep) duodv 2, vi. 92 L. ore Sè pia dmaoewv votowr idén Kai aitin (one form, and not many). mepì rpodns 1, ix. 98 L. tpopy kaì Tpobñs eldos pia kai moddal* pia pèv i yévos Ev, eldos Bè dypormre Kal Enpórnru: Kai &v Tosrorowv idéar Kai mocôv kort kal és tiva kal és toradta. The use of yévos as opposed to elòos suggests a late date for this Heraclitizing work, which is suspected on other grounds. A comparison of these passages, except perhaps the last two, shows a common attitude of thought—that of division and enumeration. It strikes me very forcibly that the numeral and the substantive have become a single formula, expressing numerical division of a unity into forms, sorts, kinds. Suppose we asked the question, What formule do the Hippocratean writers use when they divide objects into groups, which they proceed to enumerate? I may remark that this procedure is infrequent in these works; the writers do not as a rule explicitly classify; they name, describe the symptoms and the treatment of ailments, but do not classify them. Many of the treatises contain no phrase corresponding roughly to ‘there are two, three kinds of so-and-so.’ The author of zepi évros madav uses the simple plural: 10, vii. 188 L. rpeîs Sè eioi pOictes: mpwrn Sevtépn tpitn: NO. III. VOL. VI. 35 vii. 252 L. ixrepor

reooapes: 52, vii. 298 L. reravor tpeîs. Tpomos is occasionally used in the same sort of numerical formula; I have cited the examples above. Once we have púous in böpwrwv Slo ev pores (mepi Suairns o&éwv i. 172 K). The most detailed classification is found in mrepì T@v év redaXfj Tpwudrwv 4-7, ii. 6-9 K): the writer distinguishes five rpómot of injuries to the bone, fracture (xérn£ts), bruising (gAdous), etc.; and each has several iSéas, according as the injury is more or less deep, affects more or less of the bone, etc. Here the idea of form is undoubtedly prominent. These examples, I submit, show that in the Hippocratean usage eldos has advanced much further in the direction of the purely logical meaning of kind or class than Prof. Taylor allows. The writers are not far removed from the attitude of the modern unmetaphysical Englishman, who applies such words as ‘kind,’ ‘sort,’ ‘form,’ ‘type,’ ‘class,’ without very much difference, but still in a vague, unconscious manner chooses the one that he feels to be most appropriate. That is one of the reasons why I think that Prof. Taylor’s frequent rendering as ‘figure ’ is so misleading: ‘form’ would more accurately represent the Greek, because with us figure is not used in a general classificatory sense, whereas form is. The same divisory or classificatory suggestion is found in the numerous examples where eiéos or idéa is conjoined with modus, dXXos, or érepos, Tavroios, or sravrodarrós. I give these examples :— (1) With words: (a) wept apxains inrpuñs 12, i. 13 K. moAXà Sè eidea kar’ intpixny és Toraúrnv äkpiBeav fer Tepì dv eipnoera. medicine. T. ‘branches’ of (b) wept trav Ev rebaNfj Tpwudrwv 4-7, ii. 6-9 K. Ta» Sì tporav éxdotou treioves iSéat yivovrai ToD KaTHYypaToS Ev TH TpwaeL... ideaı Sè tis PAdotos mAeioves ylvovrat . . . éopratar Se TO datéov MOAAAs idéas . . . Ev Sè TH Tpomw érdorp mAcioves iÒéar yivovral . . . toMai idea yivovtar Kal rijs pAdotos Kal Ths por pijs. (c) wept puotos avdpwrov 2, vi. 36 L. modral pèv idéar TÔv vovomparov, TON dè Kal 1) inows adrôv éori. T. ‘substances.’ (d) mrepì évròs madôv 20, vii. 214 L. mepi Sì Tod pXéyparos Tas aùrâs yvópas exw, ds mepi XoAis, idéas avtod Todds elvat. T. ‘figures’ or perhaps ‘symptoms.’ (2) With &XAos, érepos: (e) wept dpxains inrpixfs 15, i. 17 K. où ydp éotw aùroîs, olyaı, éfevpnpévov aùrò Tu ep’ Ewurod Oeppòv 1) Wuxpòv 7) Enpòv 4 dypov under) GrrA@ eldeı xowwvéov. This is one of the passages in which Prof. Taylor finds his meaning of ultimate substances. (f) Ibid. 23, i. 30 K. ei yAuads xupòs doy peraBddrra és dXNo eldos, WA dard avyKpyaotos, GAN’ aùròs éÉrordpevos, motos Tus ay mpÔros

yévorro, minpös 7 adwupds 7) otpupvòs i) o€ds; olpar pèv oËús. T. ‘structure,’ but this is too definite a word: kind seems to me the nearest equivalent because the writer evidently thinks of xupos as having five eldea, yAurús mınpös dApvpòs oTpupvòs 6&ds. Prof. Taylor may be right in supposing that he had no clear conception of a quality; but there is nor need to suppose that he was thinking of the bodily nature of things; he probably had in mind nothing more explicit than the distinctions implied in the everyday use of nouns and qualitative adjectives. (g) wept dépov 11, i. 53 K. TA pev (vocetpata) drropdiver, tà dé Anyet, TA Sè GAra mévra ebiorTara és Erepov eldos nal érépnv kardoraoı. T. ‘phase,’ but this is inaccurate: the writer does not think of the disease as persisting in another shape, but as passing into another disease or another form of disease; see rep) mradôv 8, vi. 216 L. xpiveodaı Bè gorw év riot votcoow, Stay alfwvraı ai voDcoı N) papatvwvrat î) peramimTwa és Erepov vovonpa N) TEMEVTOCIV. (h) sept dpOpwv éuBoXîjs 34, ii. 153 K. AAN’ où BoúXopar amomAavav Tòv Adyov, Ev äAAoıcı yap eldeoi voonparwv TEpl ToUTwY elpnrar. T. ‘figures of disorders,’ but adds that nothing is lost by ‘kinds.’ (i) wept téxvns 6, vi. 10 L. viv 88 8% dalvovraı TÔv inrpdy of udMiota èmawedpevor Kai Startypacw iduevor nal dXXowok ye eldeow, à our dv tis aim, um Ste intpos, add’ ovde turns dverioryumv axovoas, pi) où Ths Téyvns elva. T. ‘things,’ ‘substances.’ Why not ‘other kinds of treatment’? (j) mepl iepñs voúoov 13, vi. 386 L. óróoa Te Ev olxjnpact repdpia Kata yas éoti peotà olvov 7) dXXov TIvös Üypod, mAvta Tara aicOdverat xa) SiarrAdooe Tv nopbnv és Erepov eldos. T. wop py = elôos = pois = substance. Elos certainly means nature, but does the writer think of this so definitely as substance? Cf. (f) above: the xepdua éÉioraras Ex rod eldeos. (3) With mavroîos: (k) wep trav Ev kepaXñ Tpwpdrov 4, ii. 6 K. idéa. Sè poyuéwv mavrolaı yivovrat: xa) yap AewTOTeEpai Te Kal AeTTal mavv, dore où karadavées yivovrat tori al TÔv poyyewv. (Not in T.) (1) wept Gialrns òEéwov 43, i. 130 K. Staywoonev . . . Soa Te muéwr ú pÜous Kal 1 eis éxdorovow éxrexvoi méôea Kai eldea mavroîa. T. ‘symptoms.’ (4) With road «al dMa : (m) wept apxains inrpichs 23, i. 29 K. moAAd de Kal dMa Kai gow xa) Ew rod owuaros elÒea oxnuarwv. This collocation troubles Prof. Taylor greatly; he thinks the expression is pleonastic,

both elôea and oymudTowr meaning ‘configurations.’ He seems to overlook the fact that the writer has explicitly defined oxfjpa as a technical term; the oxnuara are dca &veoriv ev TÔ avOpare (p. 26, 22), ie, the head, bladder, uterus, etc. (p. 27, 16); nearly equal to organs, the writer considering that their shape or configuration is their essence. He explains that diseases arise Sua Tijv piow To oxúpmaros (p. 28, 11): thus wide, open organs like the stomach and lungs produce a noise when wind enters them; and there are many other eldea oxnudrwv, forms of organs which have influence on health, the size of the head, the proportions of the neck, etc. The expression is not therefore pleonastic, and its collocation with a word of originally similar meaning shows that it has become very general in signification. (n) émidmuiov y' 12, i. 229 K. moAAà Sè Kal dAXa muper@v èrrediunoev eidea, Tpitaiwy Terapralwv K.T.A. ” # # (0) wept kepaññs Tpwudrov 7, ii. 9 K. «al moral arrau idéa Tod TOLOÜTOU Tpórov. (p) wept Tpodfjs 33, ix. IIO L. ydAa rpodn oloı yara tpopy Kata pÜoiv, äNXouor Sè oùyi, GAXovot Sè olvos Tpopy Kal AAA OIL ovyi, Kal odpres Kai adda idéar Tpopîjs moAAal Kal Kata Kdpnv Kai ediouov. T. ‘bodies.’ (5) With mroXAà kai mravrodarrd : (q) mepi Siairns a’ 4, vi. 474 L. oùrw Sè Toúrwv éyovTwy Todds Kai mavrodamas idéas drrorpivovrau am’ AM AO Kal orreppáTov Kai Ewwv, ovdev ömovov (ouoiwv, Kühn) aAANAcıcıv oÙTe THY öyıv oùre Tv Siva. T. ‘bodies’ on the strength of dyww and Òúvauuv; but the passage is clearly a reminiscence of Anaxagoras, fr. 4, FVS? 315. roúrwv Sì oùtws Exövrwv xp Soxeiv évelvat moXAd TE Kal Tavtoîa Ev maou Tois ouykpuvouévous Kai omépuarTa mrdvrov xpnuatov Kal idéas mavroias éxovta Kai xposas kai ndovas. Here idéa = vous, nature or constitution, the xpovat and oval being Suvdyes. Prof. Taylor gives ‘shapes’ in the passage from Anaxagoras. (6) With dra Kai mavrota : (7) wept apyains inrpwns 19, i. 23 K. petaBddrew . . . és yvpòv eldos Sv dXNXewv eidewv rad mavrotwv. T. ‘structures,’ apparently: why not ‘ body,’ for the sake of consistency with his rendering of similar passages, I do not see. (7) I add some passages in which the classificatory notion seems more or less important :— (s) ériômuiov a’ 20, i. 197 K. dteoglovro mavtes ods ndryw olda, olow ai Úmroorpopaì (recurrences) Scà tod eldeos roÚrov yevoiato,

oùdè rev Òravoonodvrwv Sa TovToV TOD Tporov oddevì olda UTOotpopyy yevopévnv radu. T. ‘symptoms,’ but the similar phrase dia ToÚrov tod Tpómrov seems to show that the writer is not thinking of appearances, but of forms or fashions of disease. (t) éridmuiov y 3, i. 225 K. Td pèv eridnunoavta vovenpata Tadra : éxdotou Sì TÔv Úrroryeypaupévov eidéwv Hoav of xdpvovres Kal éOvnoxov moAAoi. T. again ‘symptoms’; but in view of eldea voonuatwy as kinds or types of disease in srepì dpOpwr éuBorijs 34, ii. 153 K I see no reason why this should not be the meaning here. (u) kar’ imrpiuor 7, ii. 34 K. Ta dè eidea (of bandages) azAobv okémapvov ouudv 6pbaruòs Kal poußos Kal muiTovor. dpuólwv TO eldos TO elder xa) TO mädeı Tod émideomévou. This is immediately after émidécv0s duo eldea and immediately before ayabas dio idea Tod émideouévou (p. 184, above). (v) Ibid. 19, ii. 42 K. 1% dvddm bus, 1) Odors, 1 érrideous ds ev TÔ aur® oxnnarı 7, dapuhdooev. repdrara oxnudtwv ea bücıes ékäoTou TOV peréwv: Ta Sè eldea ex Tod Tpexeıv ödormopeiv éotavat karareiodaı Ex Tod epyou, ex Tov üdelodaı. T. ‘ positions.’ But cyîjua, as here, is the technical term in surgery for the posture or position of a limb; surely the writer means that in general the determinants of such oy#uara are nature and habit, and that the different types or kinds of oymuara come from different ways of using the limbs. (w) mrepì yupev 13, v. 494 L. Örav oùv Kai fp obtws dyayn (sc. Wigea ex Oarméwv ra) Oddros Ex Yrüxeos, the characteristic of P@womwpos, which gives rise to xrepos, jaundice), Kal pos yivovras ixtepos* éyyurdrw yap aurn 1 Kivnow TH @py Kata Toûro Td eldos éoruw. Apparently=#v oörws aydyy. This affection, when it takes this form, is most akin to the season. Not in T. (x) mep? iepíjs vóoov I, vi. 360 L. ÄvÔpwrrou Biov Öeópevor moXkà Kal mavrola Texvéovrar Kal Touxidhdovow & Te TAAAa mdvra Kal és nv vodcov Taúrnv, érdorp elder Tod madeos Beò Tv airinv mpooribévres. The eldea are Hv Bpvxavraı, alya wiu@vraı, tà defià amÔvrau, OEUTEpov Kal eùrovorepov POéyyntat, KTA. T. ‘symptoms’; but here again analogy with other uses suggests that the writer is thinking of these not as appearances but rather as forms which the disease assumes. (y) mepi Giairns B' 66, vi. 582 L. mepi de komwy TÔv Ev Toîor copact Eeyywouevav abe exer of pèv aytpvacto. TÔv dvOpwrev amd mavròs Kotor Tovou . . . Ta Où yeyvuvaopéva TOY cwpdrwv Úmò TÔv dveblorwv Tovey kom: tà Sè imo TÔv cuvjdov yupvaciwy komi&, UmTepBorn Xpmodueva. Ta pèv oùv eldea TÔv

konwv Totaûrd éorw, ú de Sivapus aitéwy de Eyes. Kinds, types. T. unnecessarily speaks of a suggestion of causality in the idea: it is a classification by causes. When the Hippocratean writers speak of the nature of a thing in a context which brings out the causal aspect of the idea, they use pÜous rather than eidos. (z) wept diaitns y' 69, vi. 606 L. The author claims as a discovery of his own that the secret of good clinical practice is mrpoÖudryvwors before disease and Sidyvwors during disease, determining whether there is equilibrium between the nutriment and the work done: èmi radra è) tà eldea errekeim Kai SciEw oxoia yiverart Toloıv dvOpdrrovow ünıaiveıv Soxéover Kai écbiovoiv ndéws x... Types or forms of situation, the equilibrium or dominance of rpopú or mövos. (aa) mpoppnudrov B' 1, ix. 6 L. xal TAMA Tovovrórporra mpoppipara réyeras Ev TovoÚTp TÔ elder. ‘Of this sort’; this writer seems to use cYfua and rpömos in the same sense. The passages I have cited in this section, forming as they do a large proportion of the total number of passages in which the words eldos and idéa occur in the corpus, seem to me to establish conclusively that the words have acquired distinctly classificatory associations. (iii.) Of Prof. Taylor’s contention that the words are used in the sense of symptoms, I will say little. ‘Sometimes, at least,’ he writes, ‘they mean the collective variety of the symptoms presented by the disease, as distinguished from the one hidden cause of the mischief’ (p. 245). In other words, the notion of appearance is present. I doubt myself whether the words ever bear this technical meaning; in many of the passages in which Prof. Taylor renders by symptoms I have suggested another translation. I think that in his cases the writer has symptoms as his context, but does not mean symptoms by the word. In any case, the medical writers have their technical terms for symptoms, onwijia and often madea. But as this point is not of importance for the main issue, I pass on to the alleged use of the words to mean ‘stuff,’ ‘ matter,’ and especially ‘ primary substance.’ (iv.) Prof. Taylor’s most important result is that in Hippocrates elôos and iöea have acquired the meaning of ‘real things,’ not merely the real nature of things, but the things themselves, ‘substances.’ And in close connexion with this, they often mean the stuff, matter, substance of a bodily organ. His case rests so exclusively on the interpretation of two or three treatises, all of which belong to the class of those in which the influence of of puouxoi is manifest, that it will be most convenient to collect the passages from each work, even at the cost of some repetition, and examine them together.

I begin with ep yovns, with which I join repì guotos sraudtov and the fourth treatise repi vovowv, as they are all by one author, and really form three parts of a continuous work. (1) repli yovîjs 3, vii. 474 L. eiol Sì réoaapes idéar Tod Üypod, alua yor) Üdwp Kal DAéyua. Tocavras yap ideas exer ovupuéas 6 dvOpwros év éwuT@ Kai ard Tovtéwy ai vóoor yivovrat Ÿ ai Ex vóowv Svaxpictes. T. idéa (probably) =‘ figures,’ ‘types’; i0éas =‘ substances.’ (2) Ibid. rr, vii. 484 L. emmv dé Ti of vóonpa mpoomeon Kal tod üypod aùrò, ad’ où TO omepua yiverat, Téooapes idéar Eodaaı óróoar Ev pioe borfjpEav, Thy yovijv oùK OAmv Tapéywow, dodevéorepov Sè TO Kal TO memnpwpévov, où Oadpud por dorées cal mrnpwOijvar Borrep Kal 0 rokeüs. (3) mept piatos maudiov 18, vii. 502 L. iv Sè un üyialvn 7 yuvn nde perry Uiaivev, Kwpéer de 1 xdlapois ENdoowv Kal eldos mrovnporépn, Kat où TAXÙ TMHYyvuUTaL. (4) Ibid. 19; vii. 506 L. órórav Sè Gap0pwôm To matdlov, Ta cidea Toy peréwv, ad£auévov aùrod, TA Te dotéa emicxAnpoTepa yivetat, Kal Kothaiverat. (5) Ibid. 21, vii. 512 L. okórav adixntar ès tas umrpas, idémv laxeı Tod ydhakTos, Kal TO mardtov am’ adtod émaupioxetat òXiyov. | (6) mept vovowv È' 32, vii. 542 L. Ever dè 1 your) Kai 6 dvijp Técoapas idéas bypod év TO omparı, ad’ bv ai voboos yivovraı, oKooa um amd Bins vovonpata yiverar» adrar Sè ai idéar eloì préypa alpa yor Upon, Kal dd ToÚrwv é TO oméppa oùk EAduxiorov oùdè dodevéorarov auvépxetat, Kai érrevd) TO C@ov éyévero, KaTà TOÙS Tokÿas Tocavras ideas bypod tyunpod Te Kal vooepoù Éye Ev éwuTe. drropavéw dè Sxws Ev éxdoty Tovrewv TÔv ler kat mreiw kal éddcow Ev TÔ chart yiveras k.T.À. (7) Ibid. 57, vii. 612 L. avraı ai tpeîs id€a Tv vovonudrwv ard Tod Üòpwrros. In (4) rà eiöca rôv peréov evidently means the shape of the limbs: in (5) idémv tod yddaxtos probably means nature rather than appearance (T. renders ‘ substance’): in (3) eidos moymporépn means inferior in character or nature : Prof. Taylor gives ‘inferior in quality,’ but adds that the writer had probably no distinct notion of quality, and means rather ‘and is of inferior stuff” On which I would remark that the rendering of eldos as=qious, a vague word susceptible of many shades of meaning, a sense very close to the original meaning of bodily form, seems quite satisfactory; the remote derivative meaning of stuff is not required. There is no need to suppose that a Greek physician of this period was less capable of abstract ideas than the modern journeyman mason or gardener, who will tell you that the mortar or the manure has lost its ‘nature.’ In (7) ai rpeis iôéa TÔv vovanpáruv is obviously classificatory: three forms of diseases. In all the other cases Prof. Taylor renders ‘substances’ ‘ bodies,’ except that in (r) he thinks that

ai Téooapes idéa means ‘figures,’ ‘types’ of moisture, though possibly it may mean ‘substances of the moist (kind). Can Prof. Taylor seriously ask us to believe that a writer could say that there are four iSéæ rod úrypo and in the same breath apply rocaúras idéas to the same objects in a new sense? Now all the other uses of iöen in these works are precisely the same: (2) Tod dypoù auto Téooapes idéai, (6) réooapas ideas dypod: the meaning must be governed by the first that occurs. Prof. Taylor himself gives this as ‘types,’ ‘ forms.’ That this is right seems proved by (7) adraı ai rpeîs idéau Tv vovonudrwv, which Prof. Taylor has not thought fit to render by ‘ substances of diseases,’ as he does the sroXAal idea trav vovonuarwv of the treatise ep) pvavos dvÔporrov (p. 229). The passages in which e&ôea sruperôv, vovonuarwv occur as forms of diseases confirm this interpretation. Observe that in all these cases there is a genitive following iôéa except in the second occurrence of the word in (1), where it is easily supplied from the context. I may add that when the writer wishes to describe these four substances, without reference to the úypóv, of which they are ‘forms’ according to Prof. Taylor himself (‘any one of the four forms of rù dypov,’ p. 240), but as substances in themselves, he does not use éôéa at all, but such phrases as erw ToUTwv TL Tv reoadpov TOY Yırouevav TÔ compari erérOn mAkov Te un moAAD (45, vii. 568 L), réocapa pèv dj vypà éovra amedcıfa Tov avOpwrov ouveoueva (41, vii. 562 L). When he thinks of them as substances with a moist quality, he uses the simple neuter and expresses the quality by the natural use of an adjective. The main result of this examination is that /8é is only applied to the four constituents when they are explicitly connected with ro üypóv; that the sense of ‘four forms of the moist’ (the moist of course thought of as a substance), a sense which has abundant parallels in the Hippocratean corpus, fits the passages admirably; and that, on the contrary, the sense of body or substance which rather excludes than requires a genitive, is quite inappropriate. Prof. Taylor seems to infer erroneously throughout that because the writer is talking about substances he must use the term iö&n to express the notion of substance. I turn to the passages from the mrepì picvos avOpwrov : (1) 2, vi. 34 L. Some physicians assert that man, like 6A púous, is one; & ydp rı elvai paci, ri Exactos aùr@v Bovrerat òvoudoas, kaì Toûro êv éòv peraX\Xdocew Tv idénv Kad Tv Sivapuv, ävayrasopevov ÙT6 Te Tov Oeppoù Kal Tod uxpod, cal yiverdai kal yAvKY Kat mukpòv Kal AevKoy Kal médav Kal mavroiov TL Ado. (2) Ibid. vi. 36 L. The author denies this; woAAd ydp éoTiw &v TÈ odpart éveovra &, omörav bm’ dddjAwy Tapa picw Oeppaivntal Te kal Wúynrau, cal Enpaivnraı kal bypaivntat, vovcous Tikre Gore moAXal pev idéar vovonudtwr, Tord) Sè Kal ij mous aùréwv écrir. GEO Sè &ymrye Tòv péorovra alpa elvaı poûvov Tòv dvOpwmov Kal dXXo pnôèv Serevivar abròv um pmeraNddooovra Tv idénv pndè yiveodaı mavrolov, adr’ i) Spnv twd tov Eviavrod 4 THs Hrcclys TIS TOD avOpwrrov, ev 7} alua évedv haiverau poùvov ev TH dvOpwrryw.

(3) 5, vi. 40 L. «al rovrav mp@rov pèv katà vouov TA ovvopata Èiwplodaı brul Kai ovdevt aitéwy TWUTO oùvoua elvat, Éreura KaTa giow tas iddas Kexwpicbat, Kal ote TO préypa oddèv Eoıneva TO alyarı, obte TO alpa TH YoAR ovTE THY xXoAyv TÔ HrCypaTL. TAS yap dv éouxôrTa ein Tara AAAMAoıcı, dv ote TA Ypwpara dpoia paiverar Tpooopoueva, ote TH Veupl yrasovri öpora Sonée elvai; oùre yap Oeppà opoiws éoriv obte yuypà ote Enpà ovte ürypd. dvdrykn Toivuv, te Toaodrov SunrAXaxTat GAANAwY THY idénv TE Kal Tv Sivapw, un ev abrà elvai, elmep un wip Te Kal Üdwp & Te Kal TaùTóv Eoriv. (4) (5) already quoted on p. 181: in both eî$os means individual constitution. (6) 15, vi. 66 L. of mAcioTo TÔv muperdv yivovrar amo KoNijs, eldea Ôè ogéwv elot técoapa . . . dvopata È aùréouotv tori oÚvogos dubnuepiwvòs TpıTalos Terapraîos. Thus this writer uses elòos in the well-attested senses of bodily constitution and of form or type (classificatory). In (2) roANal ev idéar vovonuarov seems on all fours with e/Sea muperôv réooapa : for un neraAAaocovra TH idénv we may compare the passage of wepi a&pwv (rr, i. 53 K., quoted on p. 187), where a disease is said uebloraoôai és Erepov eldos, ‘another form,’ while the contrast with py and quxin brings it into close connexion with (4) and (5), where elöos means human constitution, i.e., the specific bodily nature of man or the nature of an individual man. Prof. Taylor’s ‘substances’ is quite unnecessary in both cases. In the other two examples iôém is conjoined with Sivapus. As the collocation of dvcis and Òúvauus is frequent in Hippocrates (my notes show at least ten instances), the meaning of iö&n here must be nearly that of vous, the nature of a thing in itself, contrasted with its Sdvayss or power to affect other things, nearly equivalent to properties. There is no need to press the bodily implications of iöen or búous: both words are freely used when body has nothing to do with it: iöen al Sdvapis is virtually equivalent to the modern ‘nature and properties.’ Now in Prof. Taylor’s remarks on these passages he takes iö&n to mean primitive elementary substance. Thus he writes: ‘When we come to the further proof “that man is many,” karà gvow, we see that this is proved by the argument that the various sensible properties of bXéypa yoy alua are all different. Ergo, they are distinct idéau. Thus the éSé means that which is not directly perceptible itself, but reveals its nature to us through its sensible properties, a “ substance,” “ monad,” or “ thing-in-itself,” and vous or Nature at large is simply the aggregate of such elòn’ (pp. 230-1). These words seem to me to give an unwarranted metaphysical cast to the thought of the writer, which need be no more than physical; however, be that as it may, I utterly fail to see what evidence there is for the allegation that the writer called his elementary substances iddaı or eldea. It is not to be found in (1); this passage criticizes the doctrine that

in man there is one substance (observe the simple neuter, without any elòos or iö&n), which changes its iôén and Ôvvamus; these are both aspects or possessions of the one substance, which the substance has, and not substances. In (2), where the author propounds his pluralistic theory, he does not say that there are moXAà elòn ev TÔ compari, but moAAd, again the simple neuter, and follows immediately with his inference that there are moAAal idéar vovonudrwv, a phrase which no one, except in support of a theory, would seek to force into the impossible meaning of ‘substances of diseases,’ i.e., substances in which diseases arise (p. 229). In (3), where it is said «ara pow ras idéas xeywpioôau, followed by the statement that their sensible properties differ (the Òúvagus side of the púots or eldos), ‘natures’ will do as well as ‘ substances’; roúrov evidently depends on both oùvómara and iôéas. The final proof is that they cannot be one (i.e., one substance), because their iöeaı are different; the writer cannot be arguing that they are not one substance because they are several substances, as he must if Prof. Taylor’s rendering is correct ; he must mean that they are not one substance because they have different natures. I cannot see the smallest ground for supposing that (dé) here has any ‘ element’ meaning; it is indeed applied to the simple constituents of complex organisms, but not one of the meanings varies from that with which the word is applied to so complex a being as man himself in the same treatise. The writer is speaking of simples, but the word iôéy no more connotes their simplicity than it connotes the complexity of the other objects to which it is applied. The work srepl äpxains inrpurijs contains the following examples :— (1) s.7,i. 8 K. Ti dì) roûro ékeivou Siadéper GAN % To y’ eldos Kal dre mrouki\drepov Kal melovos mpnyparins, apx) Sè Exeivn 1 mpórepov yevopévn. T. ‘appearance.’ (2) s. 12,i. 13 K. yaremdv è rovaúrns ampıßeins éovons mrepì Tv Téyvmv Tuyxaveıv ale TOD drpekeordrov. TOAAA Sè eldea kar’ imrpinnv és Tooaúrnv àxpiBevav ijker, mrepì dv eipnoerar. T. ‘ branches,’ ‘ departments’ of the art. (3) s. 15, i. 17 K. dmopéw 8’ &ywye, of tov Aöyov éxeivov Aéyoyres Kal äyovres éx Taurns Ths òdod él Ümobeow tiv Téyvmv Tiva Tore Tpómov Oepamrevover Toùs avOparrous, &omep Úrroridevrar. où yap éotw abrois, oluai, éEmupmuévor aùró Tu éd’ Ewvroù Beppòv 7 Yuypòv À Enpòv 4 Oypòv umdevi GrAr@ elder kouvwvéov. GAN’ olouar Eywye radrà Bpopara kul ropata adroïouv bmapxeıv olor mdvres xpwueda. (4) s. 19, i. 23 K. meooeodaı Sè Kai peraBdMewv Kai Mertiveodai Te Kal maxiveodaı ès KvuÔv eldos È GAdwv eidéwv ra) mavroiwv . . . mavrwv È) ToÚTwv fora mpoonKer Oepud D WvxpÒ mäaxeıv. T. ‘patterns.’ (5) s. 23, i. 29 K. odd dè wal àNAa Kal Eow Kal ew Tod awuaros idea oxnudrov à peydha adrjrov Siadépes mpòs TA mabmuara Kal voceovri xa) byıalvovrı. See p. 187, above.

(6) s. 23, 1.30 K. ei yAvrds xunös ddv peraBddro és äXAo eldos, un amd ovykpijovos AAN adròs éÉsorduevos, motos Tus Av TPOTOS yévorro, mukpòs aXpupôs 7) os; olpar per, öfüs. See p. 186, above. In five cases the meaning form is evident: in (1) I think ‘appearance’ is incorrect; the word elòos is explained by TowiAwrepor, i.e., its form is more complicated: in (2) the divisory notion is prominent: in (4) form will do; Prof. Taylor’s reference to geometrical figures is gratuitous: with (5) and (6) I have already dealt. The remaining passage, (3), forms the text of a characteristic excursus by Prof. Taylor. He renders by ‘they have not discovered anything which, by itself, is hot or cold, moist or dry, and shares in no other elöos’ (p. 214). Eidos he leaves untranslated, as being a technical term for a simple element: ‘as the example shows, Oeppòv, yvypov and the rest of the “opposites” are eiön, but also, each of these “opposites ” is looked on as a substantial thing’: ‘hence eldos here means at once an ultimately simple body and an ultimately simple sense-quality, and is, as for Plato, an exact equivalent of duos’: he represents the writer as arguing thus against the Empedoclean physicians, “You have to prescribe just one or more of the articles with which we are all familiar, and none of these is a pure eidos’ (p. 216). Now we look for some proof that the word is used in this highly specialized sense. Prof. Taylor himself translates all the other examples of the word in this treatise quite differently. He infers from this passage that ‘the terms eldos, aùrò éd’ éwvroù (Plato's aùrò rab’ auto), xowwvia already had a known and definite meaning in the science of the fifth century’ (p. 215). So wide an inference from one passage is surely rash. Thus, the words ‘known and definite meaning’ must signify that aùrò éd’ éwvroù was a technical phrase regularly used to express a particular scientific doctrine, viz.: the independent existence of a simple substance in the Empedoclean sense. The phrase does not seem to be used more than once in this connexion by Hippocrates, and then it is used by an opponent. It and its analogous forms aùrò rab’ aùró, aùrò ad’ éwvroù, etc, are common phrases in Greek, and particularly common in Hippocrates: writers who regularly used such phrases in medical prescriptions as ‘this may be taken aùrò kaf’ aÿrd or mixed with water' (see epi mra0ôv 51, vi. 260 L; 52, vi. 264 L; 61, vi. 270 L; srepì vovowv y' 17, vii. 158 L; rep.) yvvaureins bücios 32, vii. 358 L; mepl yuvairelwv 192, viii. 372 L; 205, viii. 394 L) might easily criticize the Empedoclean doctrine in some such terms. This being so, a single occurrence of the phrase in this context does not establish so special a meaning. In point of fact Anaxagoras does employ éd’ éwvroù several times to denote the separate existence of ultimates. It occurs once in connexion with the poîpa. of things: dre TobAdxıorov un totw elvar, our Av Sivarro ywpio0ivai, odd’ Av ed’ éwvTod yevécOat, AAN ömwaorep dpxijv elvaı Kal viv mévra opod. fr. 6, p. 316 FVS? It is used three times of ‘Nous’: voîs . . . pépeuxras obdevi Xprparı, dNAÀ móvos

aùròs éh’ Ewvrod torw. el un yap &b’ Ewvrod Tv, GdAd Tewr ÉpéuexTo GAL KT... povoy &övra Eb’ Ewvrod. fr. 12, p. 318 FVS? where the use seems borrowed from medical clinique. In Anaxagoras this phrase is not associated with elSos or kouvwvia, so that it is not legitimate to infer much from its collocation with them in the Hippocratean passage. Prof. Taylor’s procedure in regard to eldos is equally dogmatic: he simply says that the word is used in this technical sense of ‘simple body.’ But this is not enough; it may be that if this were the meaning, the sentence by itself would be quite intelligible; but we must surely inquire what is the usage of the same writer in other places. Now the second passage conclusively proves that the writer attaches no such meaning at all to the word. When he says that the cold and the hot least of all are liable to be changed by processes of expansion and compression into the elòos of yuuoi through all sorts of elôea, he means by elòos the form of a thing, whether it is simple or compound, original or derived. The same meaning applies admirably to the passage in question; ‘they have not discovered anything which by itself is hot, and shares in no other form or nature, viz.: cold, dry, moist.’ It is the hatchet argument of Anaxagoras. Prof. Taylor calls attention to the point, remarked by Prof. Burnet, that in the time of Anaxagoras the distinction between body and quality was not distinctly felt (p. 216); that may be so, but Prof. Taylor himself holds that elòos is constantly used in the sense of nature, structure: why then is it necessary to suppose that here the writer had mainly o@ua and ÿAn ideas in his mind ? Thus there is no evidence for eldos as simple nature in this work: we have seen a similar absence of evidence in the first group of treatises which we examined. Lastly, I take the mepi réyvns: the passages are :— (1) s. 3, vi. 4 L. yırwareraı Toivvv on Òederypévov TÔv Texvéwv, Kai oùdegta Eariv ij ye È Tivos eldeos oùy öparaı. olpai È’ &ymye Kal tà > + 2 x Cd » M > € A aM \ Övópara aùràs (so Littré: aùríjs Kühn) d&4 Ta eldea NafBetvGroyov yap ard Tv Òvopdrwv TA eldea myeiodaı Bractavew, Kal dòúvarov: Ta pèv yap dvopata dio.os vouodernuara éariv, Ta be eldea où vouodernuara àNAQ BracTHpaTA. (2) s. 4, vi. 6 L. &meura de kad was olóv TE Eart rois dycacbeiou dio Tu aitınoacdaı ú Tyv réxvny, elmep Ypwpevor adTÿ Kal dToupryéovTes iyidcOncav; Tò pèv yap Ths rüxns eldos yriXòv ode HBRovAnOncav dencacdaı, ev d rn Téxvn Emerpeyrav opâs adroús, dore THs pev és miv TÚXnv avahophs amndrAaypevot eloi, THs pévToL ès THY TEXumv ove amndrXaypevor’ Ev © yap Emerpeyav Kal èriotevoav adTÿ opâs aùroús, èv roÚrp aurns Kal Td eldos éoréyravro Kai Tyv Sdvapuy, mrepav0évros Tod &pyov, &yvocav. (3) s. 6, vi. 10 L. viv Sì 89 dalvovras TÔv intpdv oi pdMiota èrarvedpevot Kat diartmuaciv idpevot kal AAAoıci re eldeaiv à où dv Tus > A > \ > 14 \ La \ 3 >, e LA I # 2 | LA 4 LI \ W- A LA > \ # > » >

gain, un Ste intpds adr’ oùdè Dors Avemornumv axovaas, un où 4 AY e > x THs Téxvys elvaı. > , ION 2 LA > la 3 # AN > Omou oùv oùdèv oùre ev rois &yaboîor TAY Inrpav ovte Ev TH intpixh adTÿ axpelov &orıv, Add’ év Toloı mAeioToiot TOV Te puouévwy Kal TÔv Trovevpévov éveotì Ta elöca TÔv Oeparmevûv Kat M > a? A > A > at LI > I 3 A , A TÔv bapudrwv, oùk Earıv Erı oùdevì THY dvev intrpdv byralopévav TO aùróparov aitıncacdaı op0@ Ado. In (3) Prof. Taylor gives for the first case ‘ things,’ ‘substances’: this is far too materialistic: at this stage of his analysis he seems obsessed by the idea that an early Greek physician could think of nothing but bodies: ‘curing with articles of diet and other substances’ is an absurd rendering: something much wider is required, ‘curing by modes of diet (see L and S) and other kinds (of treatment), e.g., by the use of the ddpyaxa mentioned below. Lower down the translation ‘substances of treatments and of drugs’ offends by the very different meanings of the genitive implied: eôta surely means the specific nature or virtue (púoves) of the different modes of treatment and drugs. Hence Prof. Taylor’s description of eldea as ‘specifics,’ ‘the healing substances contained in plants, minerals, etc,’ seems to have no basis in the text. In (2) the writer uses eidos for the specific nature of ruyn and rexvn; he can use it therefore where no idea of body is required at all. On (1) Prof. Taylor writes: ‘It is clear here that idea means simply the real things or bodies’ (italics mine) ‘ which are the objects studied by a science. . . + We shall catch his meaning if we say that, taking ¢vovs as a collective name for bodily reality (the only kind of reality known to the early men of science before the rise of atomism) the eiSea are the individual constituents of which gvous is the aggregate’ (pp. 225-6). I hardly know how to criticize these remarks as comments on the text. What the writer is trying to prove is that medicine as an art is real; he argues that the name presupposes an eldos; now to establish that medicine is an art through the conception of an eidos, he cannot possibly mean that there is an individual body or cpa, a kind of philosopher’s stone, on the existence of which the art depends. Let elôos mean ‘essence’; but it can only mean the essence of an art, not a bodily essence. We have only to render eiöca by ‘bodies studied by a science’ to make nonsense of the passage: ‘I think that the arts’ (including the art of medicine) ‘have obtained their names on account of the bodies studied by them.’ Eidos is used here in the familiar sense of the nature of a thing; the thing need not be a body, but may be any determinate object of thought, e.g., an art. Eidos and övoua are regarded as possessions of the thing, which the thing has; the writer argues that if it has the one, it has also the other. Every existent art has an empirical existence (öpäraı), is currently recognized as something; therefore, the writer urges, if it is something, it has a determinate nature (eldos). Eldos is not a ‘real thing,’ but the determinate búous which every real thing must have. The meaning is easily derivable from the popular one, though it is here applied in an unusual context. That this is the

meaning is proved beyond doubt by the general argument of the treatise, which seeks to show that the art of medicine has an elôos, or, as it is also expressed, an obota, and not merely a name. He defines medicine as removing the sufferings of the sick, reducing the violence of diseases, and letting alone those who are overcome by them. This, then, is the essence or nature or elSos of the art. To the objection that success is due to rúxn, he replies that those patients who have submitted themselves to the physician have experience of its eldos and recognize its Svvayus. Eidos here= püous, and must be used in the same sense as in the earlier passage. Medicine is not a re without a ‘what,’ as is shown by its effects (Öúvagus). The eldos is the system of rules by which the art of medicine actually does heal. The author does not establish the reality of medicine by proving the existence of a determinate body studied by it, but by insisting that medicine does something. I may add that Prof. Taylor’s canon, that we must not attribute to the early men of science any but a bodily conception of reality, is not to be applied to this treatise without question, because there is internal evidence to show that its date is comparatively late. It is a rhetorical apology for medicine. The word ovcia, used several times with a meaning not far removed from that of elôos, does not belong to the usual vocabulary of the medical writers (it occurs three times in this work, and elsewhere only in rapayyaia 9, ix. 264 L). Both elòos and oùoia are used with meanings which suggest that the work belongs to a period not far removed from that of the Platonic Cratylus. If I am right in my analysis, there is no instance of elòos or idéa in Hippocrates with the meaning of substance. We are now in a position to examine Prof. Taylor’s case as a whole. I give it in his own words. ‘We may thus, I think, take it as established that eldos and iö&a, wherever they appear as technical terms, alike in rhetoric, in medicine, and in metaphysics, have acquired their technical character under Pythagorean influence’ (p. 257). With the rhetorical use I do not propose to deal. What ‘technical terms’ in medicine are intended to include, whether the application to simple natures only or the common meaning of type as well, is not quite clear from Prof. Taylor’s summary. How does he substantiate this statement? ‘From the popular sense of body (especially used of the living, and still more especially of the living human organism) come alike the applications of the word to supposed simple bodies, such as the four of Empedocles, to the figures and tropes of rhetoric, and to the hyperphysical “ monads ”—the word is Plato’s own—of the Socratic-Platonic philosophy’ (ibid.). The very first words contain a suggestio falsi which colours the whole account. By his own showing the popular sense is not body, but bodily shape, structure, form. When the medical writers use it of the shape of an organ, when Herodotus uses it of the shape of a toy, the idea of visible structure is patent. In the

technical medical sense of individual constitution we have the idea of inner structure developed. In the generalized sense of form, type, etc., with a classificatory meaning, we have a natural extension from the shape meaning, which requires no intermediary of Pythagorean mathematics. The idea of shape was in existence long before the idea of a mathematical figure. If therefore elöos in Plato means type, form, and the like, the Hippocratean evidence is in favour of his having derived it from a perfectly natural extension of a common word. Here, however, I am in some difficulty as to Prof. Taylor’s distinction between popular and technical meanings. He lays great stress on the difference of usage in the practical manuals and in the works which bring in cosmological topics (p. 225). The words are freely used in two exclusively practical manuals, zrepì Tôv Ev xepary Tpwudrwv and kar’ inrpniov, with the general sense of form, type, shape, which are apparently, therefore, current non-technical uses: on p. 257 the only popular sense is body. He points out the total absence of the words from many of the manuals of practical medicine. In point of fact, the whole of this argument from the distinction of practical manuals and other works is thoroughly misleading. The comparative absence of the words from the practical treatises as a whole is capable of another, and a simpler, explanation. For one thing, as I have remarked before, these treatises do not classify; if they do, as the manual on wounds in the head, they use the words freely in the classificatory sense. If they wish to express the popular idea of shape, they use elôos, as three times in woydtxdv. When the writers of more scientific or philosophic treatises have occasion to use a word for form, type in a sense applicable to the simple bodies of physics, they use elôos, just as they use it for the form of anything else. The medical use suggests rather that the rise of scientific thought, whether medical or more general, brought with it the need for a word to express the idea of form in all sorts of contexts, and that the development was common to medicine and philosophy. The frequency of compound adjectives of the Oupoedys type at all periods of the language is enough of itself to show that the general idea of form or nature attaches to elôos in popular thought; this makes the body factor so much emphasized by Prof. Taylor quite irrelevant and disposes of all necessity for invoking mathematical conceptions in order to explain the developments of meaning. Prof. Taylor’s application of the Joint Method of Agreement and Difference to prove that the words must have come into philosophical use from the Pythagoreans, is utterly inconclusive. He tries to show that it is absent from the remains of philosophical writers not under Pythagorean influence, but found in the remains of those who were under it. Unfortunately, the line he draws is also a line of date, as he includes among the thinkers influenced by Pythagoreanism all physicists subsequent to Parmenides, viz.: Empedocles, Anaxagoras, Melissus, Diogenes of Apollonia, Democritus (pp. 246 sqq.). Hence his results do not exclude the explanation that the words were extended naturally and directly from a common use— in which I should certainly include the shape of a common object—to meet

the requirements of a scientific vocabulary, unless he can show that the use of the words in these writers is so remote from the popular use as to require a special explanation. Let us look at the facts. ’Iö&a occurs once in the fragments of Anaxagoras, where his orépuara are described as iödas mavroias éxovra Kai xporäs xal nöovas (fr. 4, FVS? 315). The form or shape of a orréppa is derivable by a simple step from the popular notion of the form or shape of a man or a toy (found in Herodotus, Hippocrates and Xenophanes, fr. 15, whom Prof. Taylor classes among the writers not affected by Pythagorean terminology, p. 253). In Empedocles the word eldos occurs seven times: in five of these it is used in a sense indistinguishable from the ‘ popular’ sense of the form of a living body (see p. 243), as Prof. Taylor himself allows in his renderings (fr. 23, 35, 71, 115, 125, quoted on pp. 250-1). In one we have alma xa) aAAms eidea capròs (fr. 98), an instance of the classificatory use frequent in the medical writers, where Diels’ ‘die sonstigen Arten von Fleisch’ (FVS? 200) will recommend itself to most readers rather than Prof. Taylor’s ‘ the stuff of which flesh is made’; in the seventh (fr. 22), Hate brings together things differing e/Seow éxpaxroiow, shapes or forms. Thus there is no question in Empedocles of any specially technical uses of the words, and yet Prof. Taylor comments as follows: ‘The iôéa are specially prominent in the philosophies of Empedocles (in whose school the term became technical for what later usage has taught us to call the four “ elements ”), etc. (p. 254).’ This is a good example of illegitimate transitions, to put the case mildly. What does Prof. Taylor mean by the iSéac? Does he mean the elements, ‘ monads,’ ‘ simple reals,’ Empedocles’ own four simple reals, or what? The argument purports to be based on terminology; there is not a scrap of evidence that iö&a in any absolute sense was used by Empedocles, who on Prof. Taylor’s own showing uses the word eldos in the ‘popular’ sense. The statement in parentheses rests on other evidence, which I will examine presently. Diogenes, fr. 5D, speaks of the idéa of &#a, either shape or nature, a ‘popular’ use; Melissus, fr. 8D of odd kaì aldıa Kai elôn re Kai ioydr &xovra, where the common opposition of péois and dvvayes is meant. Can we then regard the physicists of this period as giving a highly technical meaning to the words, mediated by mathematical ideas ? As for the supposed extension of the words to mean substance, thing, independent entity, I deny that there is any ground for it, either in the medical writers or in the fragments of the physicists. They are indeed applied in connexion with orougeîa; but the application is sometimes classificatory, sometimes the specific nature or form is meant, which everything, simple or compound, is supposed to have. There is no case in which the word is an absolute name; it always requires a dependent genitive to complete its meaning. There is no passage justifying such language as that vous is composed of a number of eidea in the sense of simple natures. The rootmeaning of form implies something that has the form, and this implication is

never lost. It is true that in one treatise elôos acquires through opposition with ôvoua the idea of reality; but when it is argued that the existence of a name establishes that of an eidos, the connecting link is the something named; if it has a name, it has a positive character of its own; this positive character need not be a ‘body,’ it may be the positive character essential to the idea of such a thing as an art. When a writer describes the four substances which enter into the composition of the human body, he uses the simple neuter, and immediately follows with the four eiöea of diseases therefrom proceeding. When the author of another work attacks the Empedoclean doctrine of the four elements, he does not say that they have not discovered an eldos, but a tt; when he goes on to describe it as hot or cold or sharing in no other eiöos, whether eiöos means form (substantial) or nature (referring to ùypóv, etc.), there is nothing to show that it means simple form. If I am right, there is no need to seek for an explanation of the way in which eiôos came to mean bodily element, because it never had that meaning. But even if we suppose that it had such a meaning, Prof. Taylor’s importation of Pythagoreanism is unsupported by any evidence. ‘The link of connexion, in the case of the eiön which are physical ‘‘elements” or “ opposites ” conceived as primary kinds of “stuff,” lies ready to hand in the notion that the övvaueıs or properties of a body flow in the last resort from its geometrical structure’ (p. 257). He has to show that this link was actual. To this end he collects the examples of the words in Empedocles, Anaxagoras, Melissus, Democritus. In no example from these, except Democritus, is eldos or idéa ever used with the meaning of ‘ monad,’ ‘simple body’: in Empedocles and Anaxagoras it means the shape or form of any object. Prof. Taylor in his interpretation of Hippocrates paints for us a highly imaginative picture of the early physicists discussing their fundamental problems under the form that búois manifests itself in many «idea or simples; simple substances in the pluralism of Empedocles, simple ‘qualities,’ vaguely conceived as ‘stuffs’ in monistic theories. The echoes of the discussion have penetrated into medicine from the philosophers. When we turn to the fragments of these philosophers we find no trace whatsoever of such a terminology. Prof. Taylor’s remark on p. 254 that in the school of Empedocles the term iôéa ‘became technical for what later usage has taught us to call the four “ elements”’ seems to rest on the authority of a passage in the- /atrica of Menon (Fragmente der Aerzte, Wellmann, i. 110). Philistion the Siceliot physician oleraı &x Terrapwv idewv cuvectdvat ids, Todt’ éotiv ex TETTAPWV oTouyelwv Tupds dépos Udatos yÿs. But this citation, away from its original context, is not enough to establish the point; the gloss orouyeiwr may be misleading ; iôeôv may be used in some such general sense as yevav in Plato, Timaeus 82A rerrápwv yap Övruv yevav, EE dv ouprémmye TO ca, yijs mupòs Üdaros Te Kai dépos, where orouyeiwr might be substituted without spoiling the general sense, but would not be an equivalent. Is yévn technical for the four elements in Plato? Moreover, a vague assertion of Pythagorean influence is of no value for Prof. Taylor’s NO. Il, VOL. VI.

conclusions; what he has to establish is the right influence. It is intrinsically improbable that either Anaxagoras or Empedocles should have taken over from that system a name for simple bodies implying a theory of these bodies not held by them; in point of fact, there is no evidence that they did. The medical writers, who are alleged to use the words in this sense, are well acquainted with Empedoclean and Ionian (Anaxagorean) views of matter; the Pythagorean element is inconspicuous, being mainly limited to such unscientific notions as the importance of seven and its multiples in human life. That Democritus called his elements (Séa: ‘shapes,’ is certain. But with him (Òéa is not a general word for aroıxeiov; it is a word expressing his own peculiar view of the nature of the orovyeta. And however much he may have been under Pythagorean influence, his technical application of iôéa is just as near to the common meaning ‘ shape of a solid body’ ’ as it is to the meaning “geometrical figure,’ which includes figures of two dimensions. The evidence of these dvorxoi is really against Prof. Taylor’s interpretation: idéa is used by Democritus for his own oroıxela, it is not used in the extant fragments of Empedocles and Anaxagoras and the other pvourol of the time in the sense of ‘ elements,’ ‘ monads,’ ‘simple bodies.’ The result of this examination is to show how unfounded is Prof. Taylor’s contention that Plato found the words already in current use with the specific technical sense of ‘simple beings,’ ‘monads,’ ‘ things-in-themselves,’ and merely applied them to a new kind of hyperphysical monad. It leaves without tangible proof the supposed universal influence of the Pythagorean idea of mathematical figures. I may point out that if the words did pass so completely into the current philosophical terminology as Prof. Taylor makes out, Plato’s own use of them cannot be held to show any special dependence on Pythagorean doctrine; he proves too much. My examination seems to indicate that in the time of Socrates the words eldos and idéa show two trends of meaning in the general vocabulary of science. The first is mainly physical, but without mathematical associations : including many gradations of meaning from the popular to the technical: the form of a bodily object—occasionally used for the bodily object itself, like our own words form and shape, but always distinct from capa: sometimes the outer visible form or shape : often the inner form, the structure, nature, búots, a specially physical conception: often extended to the nature of objects other than bodily: in one treatise of rhetorical character passing, by an easy transition, nearly, if not quite, into the metaphysical notion of essence. The second is semi-logical, classificatory; used especially in such contexts as ‘there are four forms, kinds’ of anything, whether a substance like the ‘moist’ or a disease or whatnot. Thus in Thucydides aca idéa has become so much a formula that it does not matter whether we translate form, mode or kind. In this line of development the later meaning of species is but a single step further. Prof. Taylor seems to have made out a case for the employment of elöos in the Pythagorean mathematics in the sense of geometrical ‘pattern’ or ‘figure.’ But there is

no evidence whatsoever to show that this highly specialized meaning was a determining factor in the other developments; it seems to have been a collateral growth, and eldos soon gave place to oyua as the technical term for a geometrical figure. The technical uses of eidos by Isocrates in rhetoric (the examples are collected in Varia Socratica, pp. 201-211) seem to me explicable by the current scientific uses, without calling in the aid of mathematical conceptions; what I have called the classificatory or divisory meaning is marked in many of the instances. In conclusion I must touch very briefly on the question, what light my results may throw on the origin of the Platonic usage of elôos and idéa. It must at least be admitted that their meanings in the Platonic writings show much greater affinity to the current scientific usage in both its tendencies than to the specialized mathematical meaning. If the development of scientific use was independent of Pythagorean mathematics, as I maintain that it was, then the linguistic evidence, so far as it goes, bears out the statement of Aristotle (Metaph. i. 987 a 31 sqq.) that the Platonic eiön were derived from another source than Pythagoreanism. Can we say why Plato gave the name of eiön to his principles? Not, according to the evidence of the associations of the word in current usage, because they were ultimate simple beings, but presumably to mark their specific character. Aristotle tells us (/.c.) that they were in their origin primarily the objects of definition, an aspect of them prominent in the dialogues themselves: were they called e/5n because of this logical function? Ithink not. The key to the name is to be found, I believe, in Cratylus 386E sqq. In this passage we have two formulae equated with each other. The first, aùrò à ort xepxis, represents the object of defining thought as opposed to the object of sense (Aristotle, Metaph. 987 b 5): it can be easily shown to have arisen from the dialectical question ré éorw; in this aspect the ‘idea’ is derived from Tv Ev roîs Aoyoıs oxéyev, as Aristotle puts it (ibid. 987 b 31). The second formula, rò ris Kepxidos etdos, uses eldos in the sense of nature, form, ¢vovs (a frequent synonym for it in the Cratylus), thus bringing it into close connexion with the scientific conception of eidos as form. We may perhaps express the difference thus: the ‘idea’ is aùrò è &orıv Exacrov or oùoia primarily in its epistemological and ontological aspects, eldos primarily in its scientific aspect as cause of the particulars, conceived on the analogy of causation in the arts. Thus the name elôos has nothing to do with the doctrine that the ideas are numbers, a doctrine which Aristotle, our only authority for it, always treats as concerned with the relation of the ideas to their elements.! 1 This result seems to be in agreement with that obtained by Ritter from his examination of the uses of the words in the Platonic dialogues themselves: see his Neue Untersuchungen ucbey Platon, p. 323. He there treats the logical and metaphysical meanings as developed from the original ‘outer appearance, form’ (äusseres Aussehen, Gestalt) through the intermediate meaning of ' inner structure’ (inneres Verhältnis, Verfassung, Beschaffenheit). Tue UNIVERSITY, LERDS.