The Cnidian treatises of the Corpus Hippocraticum

CAMBRIDGE LE, TS HA. \6$ The Cnidian Treatises of the Corpvs Hippocraticvm [The Cnidian Treatises of the Corpus Hippocraticum] Author(s): I. M. Lonie Source: The Classical Quarterly, New Series, Vol. 15, No. 1 (May, 1965), pp. 1-30 Published by: Cambridge University Press on behalf of The Classical Association Stable URL: http://www Jstor.org/stable/637845 Accessed: 30/10/2013 15:40 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Cambridge University Press and The Classical Association are collaborating with JSTOR to digitize, preserve and extend access to The Classical Quarterly. http://wwwjstor.org This cantent downloaded fram 199 87 31 20 on Wed 30 Met 9013 15-40-37 PM

7KH&QLGLDQ7UHDWLVHVRIWKH&RUSYV+LSSRFUDWLFYP>7KH&QLGLDQ7UHDWLVHVRIWKH&RUSXV +LSSRFUDWLFXP@ $XWKRU V ,0/RQLH 6RXUFH7KH&ODVVLFDO4XDUWHUO\1HZ6HULHV9RO1R 0D\ SS 3XEOLVKHGE\Cambridge University PressRQEHKDOIRIThe Classical Association 6WDEOH85/http://www.jstor.org/stable/637845 . $FFHVVHG Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. . Cambridge University Press and The Classical Association are collaborating with JSTOR to digitize, preserve and extend access to The Classical Quarterly. http://www.jstor.org This content downloaded from 192.87.31.20 on Wed, 30 Oct 2013 15:40:37 PM All use subject to JSTOR Terms and Conditions

CNIDIAN TREATISES OF THE CORPVS HIPPOCRATICVM GALEN in a celebrated passage! remarks that there were three ‘choirs’, xopot, in early Greek medicine: the choirs of Cos, of Cnidus, and of Sicily. The word is vague and suggestive, and we do well to keep it so. If we look in the Hippocratic Corpus for schools of medical theory, with distinct sets of doctrine marked off clearly from the doctrines of rival schools, we shall be lucky indeed if we can find them, and, having found them, succeed in convincing others of their existence. But we shall find, now and again, an individual voice proclaiming itself from among its impersonal surroundings with clarity and vigour: some treatises in the collection have their author’s personality stamped on them as distinctly as on a work of art. And we shall find yopoi: groups of writings which are akin to each other in their general attitude towards medicine, and to some extent in their theory and in their style and vocabulary. We may also find, within these larger groups, smaller groups of works, two or three or four, which clearly have a more precise and concrete relation to each other. One of these groups is formed by those works which are generally known as Cnidian. The term ‘Cnidian’ is an extremely elastic one. It was stretched to its furthest extent by Ilberg,2 who included among the work of the school no less than twelve treatises of the Hippocratic collection. We find the opposite extreme in L. Edelstein, who admits only three: 7. dtairns o€€wv 1 and IT; a. vouowv 11; r.r@v évros ral&v.3 L. Bourgey accepts only two: 7. vovowv 11 and Tr. T@v évros rabé@v.4 The question of the precise number of the Cnidian works, without being altogether idle, is perhaps unanswerable: one is faced with the problem of not being able to decide what Cnidian doctrine is, and what individual variations are compatible with it, until one has decided which works can rightly be called Cnidian; and not being able to call any works Cnidian until one has a clear idea of Cnidian doctrine. We have, it is true, some external evidence, but in itself it is not of much help. Applied rigorously, as by Edelstein,’ it means that we can call only three treatises Cnidian, and this seems to me, and I should imagine to most readers of the Corpus, an absurdity. Although one cannot accept all Ilberg’s attributions, one must agree with the correctness of his general assumption; there are in the collection a number of works which are strikingly similar, in form and in content, and this similarity is particular, extending as it does to verbal parallels, as well as general. Each of these works has unmistakable affinities with two or more of the other works in the same group, in such a way that not one of them shows 1 X,6K. and note. 2 J. Ilberg, ‘Die medizinische Schrift über die Siebenzahl und die Schule von Knidos’ (Festschrift für Lipsius [1894], pp. 33 ff); ‘Die Ärzteschule von Knidos’ (Berichte über die Verhandl. d. Sächs. Akad. Wiss. 1xxvi. 3 [1924]). 3 ITepi depwv und die Sammlung der hippocratischen Schriften (Berlin, 1931), p. 159 de 4 Observation et expérience chez les médecins la collection hippocratique (Paris, 1953), 5 Op. cit., p. 159 n.: “Zur sicheren Identifikation als knidisch reicht nur das objektive Moment der Zahlung der Krankheiten

throughout a dependent relation to any other single work. In trying to unravel these affinities, one feels that one is confronted with a ball of string which has neither beginning nor end. To explain such a peculiar and exasperating state of affairs, only Ilberg’s hypothesis of a common stock, itself lost, to which all these works are variously related, will serve. One must surely go so far with Ilberg. This common stock [berg found in the Kyiôiu yoSpar,! critized by the author of 7. Siairns dËéwv, and ascribed by Galen to Euryphon.? The grounds for Ilberg’s belief were the coincidence of the one fragment which Galen quotes from the Kvidia: yvouaı with 7. vovowv II. 68 (VII, p. 104 L.), and the remarkable correspondence of some of the subdivisions of diseases which Galen quotes, as characteristic of the Cnidian school, from the Kvidıaı yvópas, with the subdivisions which appear in 7. r@v Evrös madav. At the most, such correspondences could only show that the Kvidia: yrôua were related to the other treatises, not that the relation was one of priority. But Ilberg had in mind the implications of the treatise 7. diairns of€wv. The author of this work begins? his essay by alluding to the composers, of ovyypdibavres, of the Kvidtas yv@por, whose inadequacies he criticizes. He then remarks that of vorepov ériiaokevalovres showed some improvement (intpuxwrepov di Ti ErnAdov). Galen in his commentary on the work glosses the word Emiöraokevaleır as follows: émidteckevdcba Akyeraı BiBliov émi TO mporépw yeypappevw TO devrepov ypapév, Orav Tv Úrrodeoiv Exov THY adrmv Kal ras TMÄeioras TOY pijoewv Tas avTas TWa ev AbNpNuEVa TV ex TOO mporépov cvyypdpparos Ex, Tiva dé TpooKeipeva, rıva 9 varndAdaypeva (C.M.G. V. 9. 1, p. 118). If Galen’s gloss is appropriate to the Hippocratic use, the original Kvidıaı yvôua were improved not by the sort of marginal addition of which we can see the traces clearly enough in other Hippocratic works, and which were perhaps made by individual owners of the text, but by a substantial remodelling, undertaken under the school’s authority, to produce a new work which superseded rather than expanded the old. The original edition, however, was available to the author of rr. dvatrns ofewv, a situation which at least suggests that he is working in Cnidus, with access to the library there. Edelstein indeed goes so far as to suppose that the author was actually a member of the Cnidian school, although a somewhat recalcitrant member.4 But whether Cnidian or not, we must in any case assume that he has access to the fountain-head of Cnidian doctrine, 2 e.g. Berichte über die Verhandl. d. Sachs. Akad. Wiss. Ixxvi. 3 (1924), 9: ‘Die Abhangigkeit der genannten sechs Biicher [i.e. ar. vovowv I-III, 7. radüv, m. rüv evros ral&v, and 7. eBdouadwv] von den Kvidia: yraypar steht außer Frage, ihre nahe Verwandtschaft wird durch viele Parallelbilder von Krankheiten bewiesen, die oft wörtliche Übereinstimmung zeigen.’ 2 XVIIA,886K. 3 For the unity of 7. dcairns öfewv 1-3 with the remainder of the treatise, see O. Regenbogen, ‘Probleme um die hipp. Schrift De Victu Acutorum’ in Studies Presented to D. M. Robinson II (St. Louis 1953), pp. 624 ff. + Cf. Edelstein, op. cit., pp. 154 ff. Among the reasons which E. advances is the author’s acceptance of modvoyidin trav vovowv (I, p. 110, 8 Kuehlewein); that passage, however, is ambiguous, and a number of interpretations are possible. Against that of Edelstein, see, e.g., H. Diller, Gnomon xiv (1938), 303. In any case, the author’s attitude to medicine and to the nature of disease is fundamentally different from what we can reasonably conjecture of the attitude of the Cnidian school; and in the second part of the treatise (on the genuineness of which see R. Blum, ‘La composizione dello scritto ippocrateo epi duairys oféwv’? [Rend. d. Acc. d. Lincei (1936), 39-84]) there are some striking correspondences in doctrine with Epidemics I and III, and Epidemics II, IV, and VI.

and that he is not wasting his time in the criticism of a secondary and derivative treatise. If then, as Ilberg supposed, there was a common stock for the Cnidian works, then the Kvidiaı yvôua were that stock. The works which certainly derive from it, the Cnidian works of the Hippocratic collection, are mr. nav, m. vovowv I, 7. vovowv II. 1-11 and 12 ff, 7. vovowv III, and 7. rav evros radóv. Beyond these works, any ascription is uncertain, even, 1 should be inclined to say, unlikely. In the Cnidian treatises there are three ingredients : symptoms, treatment, and aetiology. Individual works vary in the amount of space they allot to each of these ingredients: symptoms and treatment appear in some, symptoms and aetiology alone in two; in others again, all three. ‘They vary also in the order in which these are presented: symptoms, treatment, aetiology, as in 7. radav; aetiology, symptoms, treatment, as in 7. r@v évros rad@v. Moreover, a bald list of symptoms may be followed separately by aetiology, or symptoms and aetiology may be mixed up together. In general, aetiology is regarded as the least important of the three. It may be omitted altogether, as in 7. vovowv III; or it may be confined to a sentence, which names the cause but explains little about its operation, as in 7. radóv. In those works which are the most detailed, m. vovowv II. 12 ff., 7. vovowv III, and 7. rav évros radóv, it is always upon treatment that the emphasis falls, with symptoms next in importance, and aetiology coming a long way behind. These differences in themselves would make an interesting object of study, with some considerable relevance to matters of more philosophic or scientific interest; and they redeem what, one must admit, is often comparatively dull reading. Because of this pragmatism, this emphasis on treatment, we tend to regard it as a characteristic of the Cnidian school, as opposed to the school of Cos, to give detailed rules of treatment which could be applied au pied de la lettre irrespective of the individual constitution of the patient, to which other works in the collection pay so much regard.' Certainly the physician who had mastered the detailed therapies of 7. rav évros radóv and carried them in his head would be armed against all contingencies. Prepared to apply his remedies swiftly and mechanically, he need never appear at a loss in the sickroom; and he must have aroused awe and admiration by the complexity of his prescriptions. But was this pragmatism, with its comparative indifference to aetiology, characteristic of the school at its earliest discernible stage, that is, in the Kvióa: yvôuu? Neither Galen nor the author of epi duairns dÉéwv mentions aetiology at all in connexion with this work: its authors may have paid as little attention to the subject as the authors of those works which derive from it. Such a question is connected with a more general question, relevant to the Hippocratic collection as a whole, about the relation between symptoms, cause, and cure. At first sight, nothing could be more obvious: the occurrence of such and such symptoms identifies the disease for the physician; they are the outward manifestation of inner factors and processes which cause the disease ; and upon his knowledge of these factors he prescribes a treatment which will counteract them. As a medical scientist, he will be interested in the relation of cause and symptom: this or that malfunction causes this or that symptom to appear. As a practising physician, he will be more concerned with the relation of symptom and treatment: the disease being identified by the symptoms, 1 Cf. the characterization of the two schools in K. Sudhoff, Kos und Knidos (Munich, 1927), P- 97-

he proceeds to the accepted therapy for that disease, bypassing the scientific question of cause (which may be, but is not necessarily, implied by the treatment). It was for such physicians that a work like 7. rv evros mabdv was written. This concept of the relation between cause, symptom, and treatment is, I suppose, modern, and we should be cautious about attributing it to the Hippocratic writers.! The scientific concept of cause is a sophisticated concept, much younger than the occurrence of disease and the necessity of treating it. Primitive medicine, even at the stage of magic, is sometimes aetiological : the incantation works because it expels the evil spirit. But humanity has more common sense than the ways in which it expresses itself sometimes suggest: if primitive medicine had been merely aetiological, it would not have worked; and herbs were used because to some extent they did work, however odd the reasons given for their effectiveness. In other words, long before scientific aetiologies were developed, diseases were distinguished and treatments recommended. At this stage of thought, the relation between treatment and cause is logically, as well as historically, prior. The introduction of the concept of medical aetiology therefore presents an interesting situation. While it is clear that the treatments recommended in the Cnidian works depend to at least some extent on the accepted aetiology, it is unlikely that no attention was paid to traditional remedies; unlikely, that is to say, that the introduction of an aetiology based on bile and phlegm involved a complete revolution in methods of therapy: the results would have been disastrous. For a time at least it must have appeared to the early medical scientists that the effectiveness of traditional treatments might serve as an argument for the truth of their proposed aetiologies: the treatment explaining the cause rather than the cause dictating the treatment. There seems to be a fossilized example of such an attitude preserved for us in a sentence in 7. &dvoav 9 (VI, p. 104 L.). The author, discussing the treatment of pain, says 810 67 kai Depuaivovres eEwlev, mupimuaor Teıpeovraı padOaooew Tov Tróvov" aparoupevov yap Uro Tis Beppacins TOD mupmuaros diépyerar TO mvedua Ota To owparos dore madAav Twa yeveodaı THY rróvwv. Admittedly, this occurs in a theoretical, not a prescriptive, passage; but the point is that a reason is offered for a treatment which is regarded as already established. However, as the system developed, the aetiology no doubt tended to modify traditional treatments: to the primitive method of purgation of the hostile element, which is still the main principle of treatment in the Cnidian works,3 1 It does not appear with full explicitness until the Alexandrian period, in the debate between the empirics and the dogmatics. Cf. K. Deichgráber, Die griechische Empirikerschule (Berlin, 1930), pp. 308 ff. seiner Gegner zuerst interessieren.” (My italics.) This attitude is true enough for the treatise K. is discussing, but we cannot take it for granted everywhere in the Corpus. 3 It is significant that in the earliest of 2 Cf. the view of the empirical school, as described by Celsus (De Med. Proem. 36): them, treatment of conditions other than nec post rationem medicinam esse inventam, sed post limited number of purgatives (7. Giairns otéwv I, p. 109, 15-17 Kuehlewein). Moreover, it is no objection to the antiquity of this inventam medicinam rationem esse quaesitam. Contrast the attitude implied by J.-H. Kühn, System- und Methodenproblem im Corpus Hippocraticum (Hermes Einzelschrift xi [1956]), pp. 29-30: ‘Wer in seinem Tun Tag fiir Tag vor die Frage nach der Krankheitsursache acute was confined to the application of a system to suppose, as I think we must, that what was purged according to the earlier Kvidiar yv@auar were the humours bile and phlegm. These are not, as Fredrich (Hippogestellt war, um richtig handeln zu kônnen, der kratische muBte 34 ff.) sich fiir die Krankheitserklarung Untersuchungen [Berlin, conjectured, a development of the

were added other therapeutic methods based on theories of counteracting Suvdpets; in other words, dietetic medicine. Like aetiology, dietetic medicine is also a late-comer, although, like aetiology, it no doubt developed to some extent from traditional practices and beliefs. But dietetic medicine in its full sense implies a developed physiological system, in a way in which the purely purgative, and traditional, aspects of Cnidian therapy do not: this is true of the system expounded in 7. ötairns,! true also of certain parts of the Epidemics; and Deichgraber seems to be right in arguing that Herodicus of Selymbria himself based his new system upon a scientific physiology. This theory concerned the heating, cooling, moistening, and drying powers of various foods and liquids; it could therefore easily be adapted to an aetiology which explained diseases by the operation of humours which are set in motion by changes in temperature, and which themselves cause changes of temperature and of humidity in the body. Thus just as the introduction and development of dietetic medicine in itself depended upon the existence of a physiological theory, so also did its adoption, more manifest in some treatises than in others, by the Cnidian school. By the time dietetic treatments are established side by side with traditional treatments, the relation between aetiology and treatment is the one which we tend to regard as natural.3 But these developments could ‘opposites’ of Alcmaeon, made by Euryphon and his followers, but are probably much the historical sketch given by A. Palm, Studien zur Hippokratischen Schrift IIepi Auairns older: there is good reason for supposing [Túbingen, 1933], pp. 110 ff.). It is possible that in that Euryphon was the first to combine these origin (cf. R. O. Steuer and J. B. Saunders, Ancient Egyptian and Cnidian Medicine [Berketween the earlier and later editions of the the Cnidian theory is Egyptian two elements, and that the differences beley, 1959], though their arguments are open Kvidvat yv®pai, noted by the author of 7. to criticism in a number of places). Euryphon was familiar with the term xoA% (Galen, XVII A, 888 K.), but the term itself in its Ouairns dééwv, are to be traced to the importation by Euryphon of dietetic medicine medical application is older than Euryphon account for his importance in that tradition. (Archilochus 96 Diehl); while into dAéyua the 1 Cf. Cnidian R. Joly, tradition. This Recherches sur du (Paris would le traité Katappéov appears to have been familiar to Hecataeus (cf. F. Heinimann, Nomos und Physis (Basel, 1945], p. 180. Heinimann also credits H. with a knowledge of dietetic medicine [ibid. 177-8], but the grounds here are less certain; while his suggestion that pseudo-hippocratique Herodotus’ remark [2. 77] about Egyptian medicine [vouilovres aro trav Tpepovrwv the first to do so: cf. Palm, op. cit., p. 111. a.Tiwy Tmaoas Tas vovaovs Tolar avOpwro.ct yiveodaı] reflects purely Greek ideas is unp. régime 1960), pp. 109 ff. 2 ‘Die Epidemien und das Corpus Hippocraticum’ (Abh.d. Preuss. Akad. d. Wiss., Phil. hist. Kl. 1933), pp. 60 ff., particularly p. 63. Herodicus, however, is unlikely to have been 3 It is clearly stated in 7. wafav 25 (VI, 236 L.). ‘Dysentery and leientery and diarrhoea are akin, and must be treated by likely to be true: cf. Steuer and Saunders, preventing or diverting the flow from the op. cit., p. 5). I would suggest tentatively that Cnidian aetiology as we possess it is head and upper cavity: tod yap vovonuaros a combination of two originally distinct ideas: the bile-phlegm pathology, and the pépuhera Tv Ovávosav.” The same author says theory of övvaueıs, on which Cnidian dietephysician must ask himself, after discovering tics are based. It is significant that bile and the symptoms from the patient, is whether the Y pvors évreülev yiveraı, Kal oddels oddév cou in 37 (p. 246) that the first question the phlegm in the Cnidian treatises show distinct disease is caused by bile or by phlegm or by signs of a double and inconsistent use: on both. Cf. also 7. dvavos avdpwrov 9 (VI, the one hand they are treated as Övvaueıs, Pp. 54 L.): rnv Bepareinv xpn moıdeodar évavwhile on the other they are opposed to the riovpevov TH mpodaceı THS vovoor. It is perhaps duvdpecs, which act upon them (cf. 7. radav I). It is in the latter case that they show their original character, while the dietetic use is to be traced back to Alcmaeon (cf. significant that a remark on dietetics has preceded: d7Aov örı Ta ÖLaıryuard éorw airsa ékaora ékaorowo. The thought is similar in rr. rexvns 11 (VI, p. 20 L.): ris yap avrÿs

only have been possible if the retention of traditional treatments governed and controlled the development of theory. If we could only document the stages of this process, we should possess a record of one of the most momentous adventures of the human mind. Aetiology comes to influence treatment: it may also influence the observation of symptoms. What are the relevant symptoms which the physician observes? Those which have an organic relation with the disease, that is, with the factors and processes which produce it.! Moreover, aetiology may influence the interpretation of symptoms: the observer ignorant of aetiology merely sees urine of a reddish colour: the scientific physician ‘knows’ that it is bilious urine, because the disease is one that is caused by bile. (He may also of course be influenced in his interpretation by what he expects to see: if the colour is dubious, he may decide it by his knowledge of the cause.) For these reasons, aetiology must have had an increasing influence upon symptomatology, as it did on treatment. The extent, therefore, to which aetiology was already developed in the Kvidiar yv@pai, and its importance in comparison with symptomatology and treatment, are questions which have considerable bearing on the early history of Greek medicine as a whole. In what follows, I am partly concerned with these questions, and partly with attempting to show some, though only some, of the relations existing between these writings. The historical suggestions outlined above are tentative only, and are intended as a framework of reference for the more detailed problems I wish to discuss. Finally, I shall attempt to show, in disagreement with Edelstein, that there is a distinct aetiological system which is shared by all these works. mr. vovaowv 112 rr. vovowv II contains two parts: 1-11, in which the symptoms and aetiology of eleven maladies are given; and 12 ff. where symptoms are given, followed Évvéaros eoriv, fomep TO eldévai TV vovawv TA n. 4. P. regards the aetiology of 1-11 as (a) airia, Kal TO Oeparrevew adras errioraoda: KrA. later than the source of the symptoms and treatment of 12 ff. (though these have been worked over by a later author on the basis of 1-11); and (b) largely dependent upon the treatise 7. ip7js vovoov (cf. also M. Wellmann, “Die Schrift 7. ¿phs vovoov’, Arch. Gesch. Med. The treatise 7. ¿pis vovcov is entirely governed by this attitude: see especially the initial and final chapters, and cf. the remarks of M. Pohlenz, Hippokrates (Berlin, 1938), pp. 85 ff. 1 Cf. 7. av évros maddy 35 (VII, p. 252 L.): torara: odv Y xoAy bro 7H deppari Kal ev TH kepalÿ, Ware evbds AAAoxpoeeı TO ca Kat yiverat wypdv k7A. The relation between cause and symptom is not usually stated so clearly in these works. 2 As will be seen, I regard 7. vovawv II. xxi [1929], 304 ff.). The considerations which I advance in the text will indicate my grounds for not accepting (a) ; as for (5), in view of the considerable dissimilarities between the aetiology of 7. vovowv II and of m. ipÿs vovoov (which P. himself admits, N.G.G. [1937], 94-95; to what he says 1-11 and 12 ff. as two separate works, both there I would add that the reason for the depending upon a common source, which contained the aetiology (which is therefore importance of the head [not the brain, which is not mentioned] in this section of 7. vovowv early) of 1-11, and the symptoms and treat- II, is simply that the arrangement is a capite ment of 12 ff.; 1-11, however, was probably ad calcem; moreover, the head here is not composed later than 12 ff. Thus the relation regarded as the seat of phlegm only, as P. is complicated, but seems to me to account for the facts best. For a different explana- Says it is), it seems to me that Wellmann’s hypothesis of a common source for both works tion, see M. Pohlenz, ‘Hippokratesstudien’, (op. cit., pp. 304 ff.) is at least likely to be N.G.G. (1937), 91 ff; and the same author’s nearer the truth, though based on somewhat Hippokrates (Berlin, 1938), p. 49 and p. 108, tenuous grounds.

by detailed prescriptions for treatment with no aetiology. Since the diseases discussed in 1-11 reappear in 12 ff. in the same order, with close parallels in the description of symptoms, we seem to be confronted by a conflation of two editions, in one of which symptoms and aetiology were given, and in the other, symptoms and treatment. The passage parallel to the passage from the Kvidvat yv@uar quoted by Galen occurs in the second part, at 68 (VII, p. 104 The beginning of 7. vovowv II. 1 1s clearly missing: the pattern of composition observed in 2-11 is a brief enumeration of symptoms, followed by a more extended account of their cause, and in this latter part the symptoms are usually repeated individually, followed in each case by a description of the process producing them, given in a temporal or causal clause: 3 (VII, p. 10 L.) is a good example. It is with this secondary description of symptoms that 1 begins.! In 12 (VII, pp. 18-20 L.), corresponding to 1, the author remarks: jv pev payí kara Tas pivas Y kara TA Wra Vdwp Kat BAevva, arradMarreral THs vouaou Kal sravera THs oTpayyoupins.2 Here the connexion between the symptom and the prognosis based upon it is explained by the aetiology of 1: the strangury is caused by a liquescence of phlegm in the head, which flows da rav dAeßav ai dyovow és TO aidotov. It is therefore prevented by an apostasis through the ears. In 13 (VII, p. 22 L.), corresponding to 2 (p. 8), the author recommends purgation of phlegm and bile for ulceration of the brain (dapparov of doüvaı, vd’ od phéypa Kat xoAn kadapetraı avw). In 2, the cause of the condition is given as dAecypa vroxodov. In 17 (VII, p. 30 L.) the author recommends treatment by warming the patient's head, and adds: mv yap payh da Tv Wrwv Y TAV pwdv Vdwp, ovTws ékpuyyaver Tov ÖAedpov. Here yap only has meaning for one conversant with the aetiology of the disease. It is a chance remark which has strayed into the prescriptive section from the aetiological section of the source. This is given in 4 (VII, p. 10 L.): úrrepepée (see below) doreeı orav és ras pAéBas xoAn 7 préypa ecéXOn...... OUTOS HY pév payh avrà és Tas pivas 7 és TO OTÓLA vdwp ral dr€ypa, byuns yiverat. A similar phenomenon is shown by another sentence in 17: Kal Ta Dra nxeei, Kal mvedpaTos eumimAartat, Kai axovet ovdev. Now the filling of the aural cavity with air has no place in a list of symptoms: it describes in itself nothing observable. Once again, it is a stray remark about cause; one, moreover, which appears to be a misunderstanding of the phrase dre odx Eveovros Tob népos ioomAndeos in 4 (p. 12). It is the forcing out of air from the aural cavity, rather than its filling, which causes deafness. The author was misled by the word tcorAn0éos and misunderstood the process. 1 The beginning was also missing in Galen’s text, of which the initial words coincided with ours: XVII A, 888 K. In view water is a liquescence of phlegm: és de ra ra Àemruvouevoy TO pdéyua diadidot. Cf. rr. maddy 19 (VI, p. 228 L.), d9wp 7 Aya, of the manner of composition in 2-11, in and the change of phlegm to water in dropsy, which symptoms are repeated, I cannot agree rr. Tav evros maddy 22 (VII, p. 220 L.). Prewith Ilberg (Festschrift für Lipsius, p. 37, n. 1) cisely the same process is described in a. that a comparison with 12 indicates that radúv 4 (p. 210), TO $déyua Aerrruvópevov. 1 is complete. rr. rabóv 4 is a compendium of the diseases of the head which are discussed in extenso in 2 For the term ßAevva and its relation to phlegm, see C. Fredrich, Hippokratische Unterthese chapters of 7. vovowv 11, and is presuchungen, p. 42, n. 2. For véwp, cf. 7. vovawv sumably abbreviated from the equivalent 11. 2 (VII, p. 8 L.) ex rav drwy Vdwp pet. The section in the Kvidia: yronaı.

The relation between 4 and 17 is complicated by the criticism of the term drepeuéew! made by the author of 4: although the term is commonly used, he says, it is in fact an impossible occurrence, and even if it did occur, it could not cause any harm. The author of 17, however, uses the term without comment ; a fact which suggests that the composition of 4 is subsequent to the composition of 17: this may be true of the whole edition represented by 1-11. The criticism disturbs the order of thought, and is an anomaly in the pattern symptoms: aetiology. It is clearly an addition, not very well integrated, made independently by the author of 1-11; it was presumably not present in his source. In 20 (VII, p. 34 L.) the author warns against washing with warm water or the application of fomentations, if the patient vomits blood or if blood flows through the nose. In the corresponding section 5 (VII, p. 14 L.) the reason is given: it is overheating of the veins in the head which causes blood to flow from the nose, and overheating of the veins along the spine which causes blood to flow into the body. Heating the patient would therefore aggravate the condition. In 22 (VII, p. 36 L.), where a form of apoplexy is described, the author mentions the contingency that the patient may vomit bile: jv de avıorauevos xoAnv éuén. Cf. 6 (VII, p. 14 L.), where the disease is caused by xoAn pédawa. At the end of the same section, the author remarks nv de ex Owpn£ıos raôra mn, macxe U0 TAV abréwv, Kat amoAAuTaL imo Tv adréwv, Kai Siadevyer imo Tv adréwv. In 22 adwvia caused by excessive drink is regarded as a separate disease altogether, differing in prognosis (the patient dies on the third, or fifth, instead of the seventh day), and in treatment. The disease is described again in 7. vovowv III. 8 (VII, p. 126 L.), where it is again treated as one disease, though the remark is added: #ooov de oi Er tis pébns Tovodrdv Te mahovres Kai dpwvot peivavres amoßvnokovanv. The subdivision of diseases characteristic of the Cnidian school was a developing process. The author of 7. Stairns 6€€wv commends some, but only some, of the apxatoı for recognizing the roAvoyiôtn of diseases : it was not universal. In the present instance we can see this subdivision in the process of growth. The common source contained one disease, with a remark that it may also arise through excessive drink; possibly it included a variation in prognosis. This variation in the form of the disease, slight enough for the author of. vovowv II. 6 to deny it altogether (the passage seems to be expressed as a criticism), becomes a distinct disease in m. vovowv II. 22. In 25 (VII, p. 38 L.) a treatment of warm water and fomentations is suggested: the treatment depends on the aetiology which appears in 8 (VII, p. 16 L.). Paralysis occurs when phlegm chills the blood and brings it to a standstill. In 26 (VII, p. 42 L.) (kvvayxn), with emi é€w rpamnra TO dAeypa 1 The word is clearly nonsense, and should be drepauéev: cf. LS] s.v. drepeuéw. Only then does the whole passage become intelligible. The author says, in effect: ‘It is ridiculous to talk of the veins having too much blood. What actually happens is that bile and phlegm enter the veins, and as evidence (onunıov, p. 12) that this is so, when you make an incision the blood is black and in a morbid state. Now if the expression vrreparuetv were literally correct (dixawov kar” oùvoua, p. 12) the blood should be red, its pure colour, not black and muddy.’ For the phenomenon itself, cf. 7. r@v évros wabdv 10 (VII, p. 190 L.), üreprAnoôÿ Ta pAéBia (with phlegm), and 18 (p. 210): ro ôè vovonpa yiverar pev amo pPdéyparos Kal yoÂÿs, oxdtav és Tas pAéBas auppuÿ. ai de phëBes abra: aluaróos cio mAnpees’ Hv odv TımapeAdg aAAotov és adras, voaoÿar (cf. voowöes, 7. vovowv II. 4 [VII, p. 12 L.]).

compare 9 (VII, p. 16 L.), where the inflammation characteristic of the disease is caused by phlegm.! The above may serve as examples of the relation between 7. vovowy II. 1-11 and 12 ff. The treatments recommended in 12 ff. depend on the aetiology described in 1-11. Moreover, some remarks occurring in connexion with symptoms and treatment in 12 ff. make better sense in an aetiological description than where they actually occur. Taken together, therefore, the two fragments indicate a common source which contained the symptoms common to both, the aetiology of 1-11, and at least some of the treatment of 12 ff. That is the pattern which is reflected in various ways by =. 7a0@v, m. r@v évros rabäv, and 7. vovowv III. In the source, the aetiological account must have been significant enough to determine the manner in which symptoms were described (cf. the double description of symptoms in 1-11), and the prescriptions for treatment. Since treatment is emphasized more strongly in the Cnidian works than symptoms and aetiology, it is strange to find a fragment such as 7. vovowv IT. 1-11, which contains symptoms and aetiology, but no reference to treatment.? It is possible that in the complete edition, the intention was to give treatment separately, in subsequent sections. We find traces of such an arrangement in the treatise 7. rórov kar” avdpwrrov (VI, pp. 276 ff. L.), a work which is excluded from the Cnidian group by its individual system of aetiology, but which is certainly Cnidian-influenced. The treatise, however, is theoretical, designed apparently to present a certain view on physiology and the cause of diseases, and the relation of the second half, which is in any case disordered, to the first half is puzzling: it does not seem to have been part of the original plan. A discussion of treatment does not really belong to the work at all; and even if it did, it is difficult to see the advantages of such an arrangement. A better analogy to 7. madav II. 1-11 might be found in the treatise 7. vovowv I. Here no attention is paid either to treatment or to symptoms: the author’s interest in the disease is purely theoretical. The reason lies in the nature and intention of the work as a whole; and it is succinctly given in its introductory sentence (VI, p. 140 L.): és av wept inovos é0éAn Epwräv re ophüs Kat epwravre atokpiveodai, Kai avriréyew opbds, évOvpéecbar ypn Tade TPÓTOV pév ad wvatvotoot yivovrar mäcaı Tolaıw avOpuwrrovow krÀ. The remaining subjects announced in this first chapter fall under the following rubrics: prognosis (okdoa avaykas exe THY vovonudrwv lines 4-8); chance as opposed to rational scientific procedure (émiruyin, eikaoin—oôkdoa axptB@s moréera Ev TH rexvn lines 9-19. This is of course founded upon aetiology, and is thus consistent with the general tendency of the treatise, announced at the beginning. Cf. 6 (VI, p. 150 L.): 6p0@s dé, ra Te vovonuara ywwoKew d TE €OTL Kat ap” 6rwv) ; ethics and deportment (edyerpin—ayerpin in line 19; the rest of the chapter is disordered and repetitive). If the physician wishes to expound or to defend his science, he must do so 1 On phlegm as a cause of inflammation, see Fredrich, op. cit., p. 38, and instances second edition inrpuwWrepov di Ti émrÿ}6ov from the Cnidian works quoted there. also die Therapie aus’. See, however, Galen’s 2 Ilberg, Festschrift für Lipsius, p. 38, n. 1, mepi Tv TpocoLoTéwy éxdorowor: ‘sie bildeten interpretation of zpooxarapabeiv: C.M.G. conjectured that 7. vovowv II. 1-11 might V. 9, I, p. 268. It is certainly evident that the go back to the first edition of the Kviérat authors of the Cnidian works went their own yv@pai, on the ground that the author of m. Stairyns Ofewv says that the composers of the way in treatment. unites them. It is aetiology which

above all in terms of aetiology. The context in which this is to be understood is the contemporary sophistic debate about the status of medicine as réxvn.! Apart from this question, however, the implications for treatment and symptomatology of æ. vovowv I are interesting. One would expect a physician’s claim to be scientific to rest above all on his ability to demonstrate that his treatment is the correct one under the circumstances. He can, however, only do so in terms of aetiology. In themselves, both symptoms and the effectiveness of a given treatment are irrational facts. A symptom is present or it is not; a treatment works or it does not: they can only be recorded or recommended. They become objects of rational discourse, which is the concern of the author of a. vovowv I, through their relation to an aetiology. Now the relation between aetiology and treatment is highly variable in the Cnidian works. Aetiology only influences treatment in the most general way. The accepted cause of a disease may suggest a treatment based on purging, or heating, or cooling, or on some combination of these; but it will not suggest, for example, the particular kind of purgative to be used, or the proportions in which a preparation is to be mixed. Prescriptions in the Cnidian works are given in the utmost detail, detail, moreover, which varies from treatise to treatise; and while it is true that some of these prescriptions might be shown to have a rational basis, it seems safe to assume that most of them have not, and are suggested either by traditional practice, or by individual experience—or even individual quackery. It is only in the strictly dietetic treatises, or in those parts of the Cnidian works where the influence of dietetic medicine is strong, that a detailed relation between aetiology and treatment begins to be worked out. There, a particular kind of meat will have a particular effect; and the effect will vary according to the way in which it is cooked, and to the degree and kind of exercise prescribed with it. These, however, are still comparatively modern developments: it is unlikely that they can account for most of the prescriptions in the Cnidian works. It is notable, in those works, how much the treatment varies in parallel passages. The physician, drawing upon a general handbook such as the Pharmakitis mentioned frequently by the author of 7. radúv, sets down his own individual combination of traditional remedies. He may favour prescriptions with wine for most conditions, while another author recommends ‘complete abstention from wine’ (dowéew de ro maumav) with monotonous regularity. Or he may prescribe cautery, but only when all other remedies fail, while another author prescribes cautery and bleeding without restriction. Much of this variation may depend upon personal experience: the physician has found this cathartic preparation to be more effective, in a particular condition, than that, in the past ; but he cannot give a reason, in terms of the aetiology of bile and phlegm, why it should be so.3 Hence variation in treatment cannot be used as an argument for detaching the Cnidian treatises from each other, or as evidence that a particular author is propounding an individual theory. 1 Cf. J.-H. Kühn, op. cit., particularly pp. 39 ff. reasoning” (45, VI, p. 254 L). His contrast amo TUxns-yvaun nicely expresses the dif- 2 It is a physician who possesses such an ference between the traditional treatments in aetiology, one who can Adyov è.dova., that the Cnidian works and those which have Plato has in mind in his portrait of the been influenced by dietetic medicine. Indeed, physician of free men (Laws, 790 d). it is precisely this difference that the author, 3 The author of 7. radúv freely admits that the discoveries of pharmacology are himself ‘more a matter of chance than of scientific familiar with the dietetics, may have in mind. principles

To a lesser extent, one notices a similar variation in symptomatology ; this variation is characteristic of the Hippocratic collection as a whole. The symptoms used to identify a disease were often of a very general kind. The description of icterus, for example, shows throughout the Corpus only the following common symptoms: yellowness of eyes and complexion, yellow evacuation, and, of course, the absence of positive symptoms indicating a different disease. Fundamentally, diseases seem to have been identified in a negative way: the disease is such and such because it is not anything else. There is often something aprioristic about these identifications: the Epidemics, particularly, give the impression that the visiting physician has already made up his mind about the nature of the disease before he notes the symptoms. The attitude, and sometimes the expression itself, is ‘in such a phrenitic case, these symptoms occurred’.! Within the group of the Cnidian works themselves, variations in symptoms are found even between clearly parallel passages. Here again, the explanation probably lies in the personal experience of the physician : along with the cut-and-dried account of symptoms which are part of his professional training and equipment, he may note variant symptoms which he has personally observed. But the relation between symptoms and aetiology is closer, in the Cnidian works, than the relation between treatment and aetiology. It acts as a control on such variation. Since, then, treatment and, to a lesser extent, symptomatology are largely empirical matters, which must, in order to become amenable to scientific treatment, be related to aetiology, it is with aetiology that the author of a. vovowv I is concerned. Once in possession of it, the physician will be in a position to debate the significance of symptoms, and to justify treatment: the treatise throughout aims at outlining essential information, rather than at completeness. It is a theoretical epitome. It is possible that the work from which ar. vovowv 11. 1-11 comes may have had a similar purpose: as a handbook, not to aid the physician in the practice of his art, but to equip himin its theory. It would have drawn from the more comprehensive work, which included symptoms, aetiology, and treatment; it would presumably have had a general introduction; and it may have treated, like 7. vovowy I, a variety of subjects. m. vovowv II and 7. trav évros rabdv The treatise 7. r@v évros radóv follows on directly, in its reflection of the source, from +. vovowy II, as berg claimed.? L. Edelstein} disagrees : according to him, the two treatises follow altogether different principles of arrangement. a. vovowv 11 takes as its starting point the factors, bile and phlegm, which cause disease; and it arranges its material accordingly. Thus 7. vovowv II. 1-37 is devoted to a discussion of diseases caused by phlegm, while the chapters 38 ff. describe diseases caused by bile. The author of 7. 7úv evrös mad@v on the other hand arranges his material according to the organ affected. While it is true that the relation between 7. vovowv II and 7. r@v evros malay is not as clearly discernible as Ilberg perhaps suggested (we must reckon with the problem of misplaced chapters in all these texts: they are ! Cf. the case histories in the epidemic of phrenitis recorded in Zp. III (III, pp. 60, 116, 136, 140, 142, 146 L.). 2 Berichte über die Verhandl. d. Sachs. Akad. Wiss. Ixxvi. 3 (1924) p. 7.: ‘diese beiden ergänzen sich inhaltlich in der Reihenfolge der Krankheiten a capite ad calcem und weisen auf einen gemeinsamen Archetypus, das waren zweifellos die Kvidia: yvouas. 3 Op. cit., pp. 162-3.

liable to such disarrangement), I cannot agree that the arrangement in 7. vovowv II amounts to a system according to aetiology, or that it is any different from the principle followed in all these works, that of a capite ad calcem. The diseases described in the early chapters of 7. vovowv II are not placed there because they are caused by phlegm,! but because phlegm is the cause of diseases of the head, ears, nose, and throat, all of which are placed here in accordance with the principle a capite. One need merely compare 7. rabo, which is clearly following, in abbreviated form, an archetype in which maladies were discussed in the same order as in 7. vovowv II. The diseases are condensed to two sections (the condensation of an elaborate discussion of polyps, reflected in 7. vovowv II. 33-37, to 7. wa8dv 5, is particularly striking). These diseases in 7. ra0@v are also, as they are in 7. vovowv II, caused by phlegm. Now the author of 7. rab&v, finishing his discussion of this part of the body, like 7. vovowv 11, with a section on polyps, then remarks (5, VI, p. 214 L.): ratra pèv oa amo Tas Kehads Pverat vovonuara TAV obBaAumv. He then begins the following section with repi de r@v kara KkoıAinv vovonudrwv. The principle is clear. Moreover, in view of the absence of aetiological remarks in m. vovowv II. 12 ff. it is rash to assert that all the diseases in 12-37 are caused by phlegm, while those in 38 ff. are caused by bile. We do not know; and the description of symptoms and treatment suggests that this is not the case. The explanation of the section beginning 40 (VII, p. 56 L.), ruperoi azo xoAns, is probably a complex one; we must reckon, I think, with conflicting principles of classification in the Cnidian works. However, this is a question which I wish to discuss later. The treatise 7. r@v évros rabdv is the most systematic Cnidian work which we possess, superior to the other treatises in language, arrangement, coherence, and inclusion of detail. Since the latter part of 7. vovowv II deals with affections of the upper cavity, 7) avw kodin, while the beginning of 7. rüv évros ralàv discusses some of these before passing on according to the Cnidian sequence to diseases of the lower cavity, 7 xdrw kodin, i.e. affections of the liver, spleen, and kidneys, there are certain parallel passages between the two works. An examination of these suggests that in some points 7. r&v évros mad@v stands closer to the original source than 7. vovowv II. rr. vovowy II. 55 (VII, pp. 84-86 L.) > ? > # x > a. r@v évros ta0&v6 (VII, p. 180 L.) / Epvoiredas ev TÀevpovi: Mv épuoirelas "Hv > yivera Oe pddiota azo olvodAvyins Kat , 7 \ »” 4 x ev mAevpovi yevntai, Bné exer, Kat TO U > [4 \ 3 € [4 Ka otadov arromTve TrovAd Kai Üypov, olov 3 A ? LL aro Bpayyov, ear: 6° > >} € ~ oùy aiuar@des, ¿deyuovy U 3 yaorpınapyins 2 èv TAevuovı yévnTas, 2 tyQvwv e eyxedvwv: Tadra 7 x Kkepadwv yap Kat \ Tv ampedAnv Kai 6dUvn LOXEL TO ETAPpevoy Kal TOUS rodemwrdrnv exe Trpos THY pvow Tod Kevedvas avOpwrou: Non de TO vovonpa Eyevero Kal 4 Tas [4 Aamdpas, \ > , Kal Ta # E x \ 307 e A \ Ca \ o [2 y A 4 / 9 # amo pA€éyparos, oKdTav puyev TO aluarı ÉTUPPÜN eri Tov mÀevpova* TMPOUTITTEL olov d€0s, Kai TOUS dddvTas ALUWOLÁ, Kal muperös Kal piyos kai dita AauBave, Kal Orav Ti Bay, mi Tolar orÀdyyvouot [4 \ > [4 > , \ e ? pote, Kat épevyerar o€U, Kal Y Kowrin \ [4 1 omdyyva puter, Kat éuée damnv Kal + 2 \ ld a x 3 ~ 2 € / de Kai ek kpenpayins kal é£ vdaTos peraAn La o / 7 3 Cai BoAns. Tade oùv tacoyer Bhoce ioyupas, x Kai TO ciadov ATTOTTVEL Üypov Kal Toddov, Tpiler, Kal TO O@pa vapKa, Kal Ora moddakis dé Kal Tayv Kal Aeukóv, olov ao 1 In any case, the first eleven diseases pp. 22-24), and 6 (p. 14) (= 22, pp. 36- in 7. vovowv II are not exclusively caused by phlegm: cf. 2 (VII, p. 8L.) (= 13, [4 A A 38): pélaiwa yo). \ \ /

éuéon, doxéer pdwv elvar Orav Oe un éuéon, Amiovons TAS Mpépns, oTpopos Kal 6dUvn éyyiverar ev TH yaoTpi, Kal ATOTATOS ÚYPOS yevopevos ÖLEXWPNOEV. ‘H Se [4 vovoos \ padtota yiveral ? / \ ~ ” 4 rs \ 3 [4 [4 3 13 td ? A Bpayxov: Kat odvvn méber ofeim és Ta A oTnfea Kal TO peragpevov Kat TOUS A 4 Keve@vas Kal Ta mAeUpa, Kal epevyerat 3 ? \ 9 a [4 \ , OËV, Kal Er TeV OTNOEwV Kal TÀevpovwv 7 x ex olov yaornp Tpuddibet, Kat eueeı arm 9 o£einv, Kai TO épueopa Hv Exxéns yapale, Ever THY yhv womep déos Erriyeovri, Kal Owpnéiwv Kat éx kpendayiéwv kai éé e 4 » Udaros peraPodîs' loyer de Kat addws. / \ >? / TOUS dddvTAaS ALUWÓLA, Kai piyos Kal x \ / ” 3 [4 rs» TTUPETOS Kal dijja Eyer LOXUPN, Kal NV TL [4 x [4 / A \ OéAn Aurapov dayéew, pvler mpos Ta omAdyyva Kal éperov aye, Kal TO Ua dmav vapkn exe OKOTAV dé ÁTTEMEOA, > ET’ 9 tA 94 LA € / 7 „ odAtyov doreeı pawy elvar Ereita erreidav THS Muépns oyirepov yevnrat, LA € f \ [4 \ Bpéper n Kowdin Kat orpépe kai Bop- Bopüte. 7. TÓv evros adv begins its description of the disease with an account of the originative cause: over-indulgence in wine, kedadoı, and eels. The author adds the reason: the nature of the fat of these fish. Then the aetiology of the disease is given, followed by a remark, which is strictly out of place, on two other originative causes. Then the author begins his account of the symptoms. This order of treatment, originative cause / aetiology, followed by symptoms which are often introduced with the phrase rdde odv wdoxet, is characteristic of this treatise throughout.! 7. vouowr 11 on the other hand follows its usual arrangement: recital of symptoms, followed by prescriptions for treatment. The remarks on the originative causes of the disease, which this treatise does not normally include, occur at the end of the section on symptoms. It was probably characteristic of the common source to begin its account with a description of symptoms. This is the arrangement in the passage from the Kvidta: yvGpa: quoted by Galen ;? and it is common to 7. vodowv II. 1-11 and 12 ff., 7. vovowy III, and 7. radav. The order originative cause / aetiologysymptoms is therefore presumably due to the author of 7. rüv èvros rabav. Such individual characteristics in the treatise admittedly make comparison difficult. Another characteristic of this author is to add detail; the presence of such detail in 7. rv évros radóv does not necessarily indicate its presence in the source. Here, for example, while 7. vovowv II notes as a symptom pain in the back and flanks, 7. r@v évros radiv adds the chest and ribs as well. The detail may have appeared in the source; but it may equally well have come from the author’s personal experience—or even from a desire for literary fullness. There are, however, apart from such ambiguous instances, details which suggest an origin in the source. 7. vovowv 11 gives as originative causes excessive drinking, meat-eating, change of water, and then adds ioye de kat aAAws. The causes which this author does not specify are preserved in 7. 7@v évros rabóv, which refers to kedaAoı and eels. Similarly, whereas 7. vovawv II notes that the patient vomits Aazrnv Kai olov Öfos, Tr. Tv evrös madav remarks that when poured on 1 Cf, e.g., 18 (p. 210); 22 (pp. 220-2); (p. 284), and many more. The order, how- 2 XVIIA, 888 K. It is clear that Galen quotes the passage from the beginning, since the rubric is given: Jledás: muperös ioye ever, is not invariable. 23 (p. 224); 25 (p. 230); 40 (p. 264); 48

the ground, it erodes it like vinegar. Possibly this early example of a pathological test is an independent addition, but it seems more likely that 7. vovowv II abbreviates here, while the other work preserves the fuller version of the source. This suspicion is supported by the appearance of the same illustration in another context in 7. vovowv II. 73 (VII, p. 110 L.) (péAaiva voúcos): the patient vomits dre de Ôpuuv ofov dos, ore dé ciadov Kal Admmv! . . . Kal TO épeopa nv yÿvaiper. Such illustrations were perhaps characteristic of the Kvidiaı yvdpar: a similar one appears in 7. vovowv IT. 48 (VII, p. 72 L.), rrúovri én’ avOpaxas Bapv din TO aiadov. The author of 7. r@v évros rradiv presents symptoms with a coherence which is notably absent from 7. vouowv II. In general, when symptoms are recorded in the Hippocratic works, they are of two kinds. Either the symptom is noted down by itself, in a purely enumerative list; or it is connected with some preceding event, which may be external or internal (i.e. aetiological), or with a definite stage or crisis in the disease: when x occurs, y follows. I suspect that many of the symptoms recorded in the Cnidian works are, originally, of the latter kind, but that they have been taken from their context and set down side by side for reasons of brevity. In the present instance, the author of 7. vovowv II has used both methods, producing the careless repetition ra omAdyyva piler—emi roto. omAdyyvoior pose. The symptom is set down in inventory form in the first case, and occurs in context in the second case. While both works reflect an orderly arrangement of symptoms in temporal sequence (7. vovowv 11: órav ri dayn... piler...70 copa vapka... orav éuéon . .. Orav de un épéon ;? 77. Tv evros Taba: nv Ti HeAN... pÙler. . . OxdTav dé arreuéon ... ereita KTÀ.), it is carried out more consistently in 7. r@r évros ma0&v, and the unnecessary repetition of 7. vovowv II is avoided. Similarly, while 7. vovowv II has the ambiguous 7 «oıAin pie, the corresponding passage in the other work has the necessary precision: ex trav orybéwv... rpvAAileı. Moreover, the remark in 7. vovowv II occurs out of place in the order of symptoms. Either the author of 7. r@v évros maôüv represents the original more faithfully, or he imposes coherence, consequence, and precision upon a haphazard collection of symptoms. It seems to me that the first alternative is more likely. This relation between the two works in their description of symptoms is well illustrated by a comparison of a further two parallel passages: 7. vovowv II. 53 (VII, pp. 80-82 L.) r.róvevros ma0&@v 1. (VII, p. 166 L.) Aprnpin Tpwbeîoa: nv TPWON Y aprnpin, “Fr, Bné exer, Kal ala Biocerai, Kai AavBdver y Pdpvyé miumrdapevn Tod aipatos, Tt pay Tv dAefiwv Tv AerTÓV, Tv \ | 4 4 1 9 [4 La € ~ À / > / >| vn) TOD mAeupuovos aprnpin 05 ld EAKWwBR 7 KaTakpepapévwy és TOV TTAEÚLLOVA, Y TOV Kat erBaAdeı ÜpouBous Kai odvvn yiverai ovpiyywv tev dia TOD TÀeVpovos TETA- EX TOD oTmÿbeos és TO peragpevov ofen, pévwv, Kal ouppayéwow és amas Kat A \ LÀ a x la [4 \ ae [4 9 > A oe Kat TO ciadov yAtoxpov Kal TTOAU, Kal 7) atuaros mAnodwor, Ölaonavrai Te Kai papvyË Enpn, Kal muperos Kal plyos > La A [4 € [4 émAauBave, Kat kepyverar 4 papuyé katappynyvevrat dia TAode tas alrias La iI , x , páliora: did ralaırwpinv, dia Spdpous, olov Úrro Aurapob: és TE pev TrEVTEKALdua mrwuara, Sia mAnyds, di dera nuépas madoxyer Toraîra* pera de 1 On Aarn, see Fredrich, op. cit., p. 39 n. 2 Contrast 7. rav évros ra0@v. It is difficult to tell which represents the correct so 9 épérovs Buatous yivouévous, dia muperous- Tadde version, but 7. r@v évros ray reads more naturally: em’ oAiyov doxéer... Ereıra.

atov mre, Kal ola ÉÂKEOS KpOTwvas, + A + , 933¢7 [4 % 1 Kat adds PnË, Kal éppayn oùv ro aipa, Kal META TO TÚOV TAXUTEPOV TTVEL, Kai O mvperòs LOXUPÓTEPOS yiverai, Kat TEAEVTÁ 4 ~ ve > 4 Les € 4 és mAevpova, kai Kadéerat pyyuarias mAEUpOVOS. odv maoyeı TO pev prov Pf toyes Enon éreura OAiyw Vorepov aromTue TO [4 A x La / ciadov Ühauov, Tore dé Kalapov. % 4 x ~ x 4 [4 4 ~ 4 Otros nv pev Ev Tayeı mavonrar THs co / or ld vovoou: nv dé un, mpoiovros TOD Xpovov TO ala mAeîoy ywpéei, Eviore ev 4 € A 4 > y 4 kabdapóv: ¿ori & Ore Kat ÜTocarpoy moAlakrıs dì Kai 7 Papuyé Aavdaveı Le , aipatos mymdapévn: meta, OpópBovs aluaros ékBpäocere kar” 6dlyov Capua: [4 ~ ’ 9 > , A V9 > / éviore Kal 00 Bapein am adtav yivex m [4 4 Tat, Kai Y papuyé Eorıv OTE TIVös dayvns miutdarar, Kal plyos Kal mvperòs Émt2 tal AapBdaver, kar” apxas pev THs vovaov A Lé A 4 oo , adddpa, mpotovons de BAnyporepov kai # A # 3 LA x adore Kat aAore émAauBave, Kat ddvvn Eviore Eykeıra ev roto. armBeoı kal Ev T@ peradpevw Kat ev Thor mievpfiot, Kal ókóTav TO ala TavonTaL arrúwv, olaloy moÀÂoy Amorrüeı vypov, > / \ À # la) x > éviore dé Kai yAioxpov. Tatra pev oùv OÙTW mdoxEl, HEXPL Teocapeckaidera e LA LA A A [4 n muépar mapédborev pera de Tavras NY un madonraı TO vovonpa, deridas amo Ths dprnpins anoßnoowv arrocrá, olas > A LÀ A > [4 > mep amò pAukravidwv, Kal oduvn épninreı es Ta orndea Kal és TO perddpevoy Kai és TO mAeupôoy, Kal Tüv Ümoxovöpiwv ws EAkos avdpevos aAyeeı. [4 La ~ Tovrw Évubépe novyinv ws padiota TH VOVONUATL Exeiv €ow, TV OÙTWS Exp” NV yap TL TTOVÑON, O Te Trévos OËUTEPOS Kal Y \ a n 1 là , \ A Be mäAdov 7 TO mporepoy mele: Kal TO t / € y , > [4 plyos Kai 6 Truperos paddov exer, Kal NV Ca A a 4 mTapy, Y odvvn ofein mpooeßadev: aÀyéer de Kal ev TH EUVT, OKOTAV Tepi- 4 a a orpednrau. T [4 x [4 OUTW xpr Trpoodepeiwv KTA. The order of treatment in 7. rüv evros rradóv is the same as in the previous example: since it is a mechanical rupture that is in question, no aetiology as such is necessary, and only the causes of the rupture are recorded. They are omitted in 7. vovowv II, along with the anatomical detail of the other treatise. Symptoms are identical, but are more detailed in 7. rév Evrös radóv, which also records possible variations and contingencies. Both passages divide symptoms into two groups, those occurring before, and after, the crisis on the fourteenth day. But while 7. vovowv 11 sets down the symptoms in an indiscriminate list, 7. rüv évrôs ma0@v carefully follows the disease through its several stages, and apportions out the symptoms to these stages: 70 ev 7p@rov 2... ÉMELTA . . . mpowvros TOU xpóvov ... €rreira krÀ. For example, the pharynx

fills with blood only after the condition is already advanced; and it is clear from 7. trav évros rabdóv, as it is not from 7. vovawv II, that the clots of blood are a consequence of this (Erreıra OpouBous atuaros ékBpaooeru). Of this division of the disease into stages, 7. vovowv II has preserved only the mention of the crisis. Again, whereas 7. r@v evros madav records the intensification of pain, cough, and fever, which occurs after the crisis, as a consequence of movement at this stage, they are set down unconditionally in 7. vovowv 11, despite the practical necessity for clarity here. Here again, as in the previous example, while it is possible that the greater coherence of 7. t@v évros radav may be due to the author’s own experience, it is more plausible to suppose that the author of a. vovowv II has simply summarized the symptoms by lifting them from their context—preserving, one notices, the same order. In the parallel passages 7. vovowv II. 62 (VII, p. 96 L.) and 7. rúv èvros radav 8 (VII, p. 186 L.), which describe rupture of the breast or back, rr. vovowv II after recording the symptoms remarks: udAıora de radra TWACKEL, HY TROL yepoi TL movnon Y eb apakav emßn y EP” immov. This detail is embedded more closely in the account of 7. r@v evros rad@v : there, it occurs in isolation from the rest of the symptoms, in the part devoted to treatment. The author recommends cauterization and a milk diet, with as much rest as possible: ro de Aoımov novxalwv TH owuarr palora Öarrnodar nv yap Ti Tovnon y emi Auafav avaßas Y EP” immov, 7 Totow œuorow ev axdeı adros radarTwpnon, Kıvövvevoeı mAdw ÜmoTpomacaı N voDoos, Kai, TV yévgrai, Kivduvos dapdaprivas: TO yap vovonpa droorpapér uäAAov midlet Y KaTapxas. Here, the link between the forbidden activities and the increased severity of the symptoms, namely, the possibility of a relapse, has been entirely left out by the author of 7. vovowv II. In its account of the symptoms of $déyua Aeukôv, m. Tv Evrös madav 21 (VII, pp. 218-20 L.) corresponds closely to 7. vovowv II. 71 (VII, p. 108 L.) with coincidences in phraseology. Here there is a further parallel in 7. ra0@v 19 (VI, p. 228 L.). All three note that the severity of the patient's condition varies within the same day. 7. vovowv II says otros TAS adTAS muépns Tore ev pdwv yiveral, rote dé kakıov Loye. The corresponding sentence in 7. radúv reads rijs adris muépns Tore Ev dokéer páwv elvar, tore dé draupdtepos. TT. Tv evros madav has a fuller form: odros ris adris muépns more pev yiverai pawv, more de moveeı éfarrivns, kai Soxéet amodaveeodaı. In connexion with the same disease, metastasis to dropsy is noted by 7. radúv, 7. vovowv I. 3, p. 146 (cf. Aphorisms vii. 75, IV, p. 604 L.), and 7. rúv évros rabdóv: it must have stood in the original source. Here again, 7. r@v evros radóv seems to stand nearer to that source: the account of ¿Aéyua Aeuróv is preceded by an account of ‘seasonal’ (émômov) or ‘recent’ (vewrarov) phlegm (20, p. 214), and succeeded by the long section 22—26 which discusses various forms of dropsy : this section begins with an account of the metastasis from ¢Aéypa Aevrov to dropsy. The arrangement is therefore systematic. Admittedly, there is no reflection of this arrangement in 7. 7a0@v or in 7. vovowv II, while in 7. vovowv I the remark on metastasis occurs in a compendium passage, which is clearly composed of extracts from another work or works. But 7. ra0@v is highly condensed and selective, while the section of 7. vovawv II, from 66 to the end of the treatise, in which the chapter on $Àéyua AevKdv occurs, seems to have suffered disturbance. If, then, 7. r@v évros madav does conserve the original order, it is not surprising that these other works should show no trace of it.

The question of the order in which diseases were arranged in the Avidcat yv@pa and the related question of the principle of arrangement are difficult and probably insoluble. I have already given my reasons for supposing, against Edelstein, that the arrangement a capite ad calcem can be definitely traced in 7. radóv, 7. vovowv II. 1-11 and 12 ff; it is somewhat less clearly discernible in 7. vovowv III.! In the case of 7. vovowv I, Littré’s suggestion (VI, p. 138) that the diseases mentioned are given exempli gratia seems to me to be plausible. The elaborate discussion 11-17 of varieties of empyema, which is disproportionate in such a work, might perhaps be explained on the hypothesis that the varieties were multiplied by a later owner of the text, for his own convenience: the work as we have it seems to me a hybrid, arrested in the process of developing from a theoretical treatise to a hand-list of diseases. Whatever the history of this section of the text, it is interesting to notice that the discussion of empyema follows the arrangement a capite: 15 (VI, p. 164L.) rYv de avw Koıdinv éuTvor yivovrat ToAAaxas . .. 17 (P. 170) : THY dé KaTW KoLAinv EuTVOL YLVOVTAL .... Apart from these considerations, the arrangement a capite seems to have been widespread: it occurred certainly in Egyptian medicine, and possibly in Babylonian.? There are a priori grounds as well for supposing that the arrangement is archaic. Some arrangement of diseases must have been in use, as a simple necessity of teaching. Since such an arrangement must have antedated the development of physiological and anatomical knowledge, it cannot have been based either on a system of causes, or on a detailed anatomical system. A simple arrangement of head, upper cavity, lower cavity, and extremities, seems to be the obvious answer; and given the importance of the head, often the source of life in primitive thought, it would be natural to start the enumeration from there. Difficulties, however, arise in connexion with 7. r@v evros ral&v. It is best to begin with Galen’s information about the Kvidia: yv@ua, in his commentary on 7. duairns ofewv.3 The Cnidians, he says, described seven diseases caused by bile; twelve diseases of the bladder; four diseases of the kidneys (cf. 7. r@v evros nadav 14-17) ; four stranguries; three forms of tetanus (cf. ibid. 52-54) ; four forms of icterus (cf. ibid. 35-38); and three forms of phthisis (cf. ibid. 10-12).* The heading ‘seven diseases caused by bile’ suggests an arrangement of diseases according to cause: bile, phlegm, bile and phlegm. Edelstein suggested that this was in fact the principle adopted by the author of 7. vovowv 11.5 I do not agree that this was the main principle adopted in 7. vovowv II, or that this treatise is inconsistent with 7. r@v évros rabdóv. But it was, I think, one principle adopted. Now 7. vovowv II. 40 (VII, p. 56 L.) has the rubric rruperol amo xoAjs.® There follow in 40-43 four varieties of fever, of which the first two are caused by bile. The second two are tertian and quartan fever: while no indication of their cause can be deduced from the account of 7. 1 On this treatise, see Ilberg, Festschrift may have appeared in the Kvidta: yvanaı. für Lipsius, pp. 34-35. 2 For Egypt, cf. H. E. Sigerist, | 5 The obscure remark in 7. A History (VI, pp. 226-8 L.) radav 18 may point to such a ofMedicine, i (New York, 1961), p. 354, where classification: «ai yiveraı ev 6 rpuratos Kal 6 the author suggests a religious origin; for TeTapratos ume yodîîs Kai pdéyparos: didri dé Babylonia, ibid. p. 419 and notes. 3 CM.G. V. 9, 1, p. 118. yéypartai. + No argument can be based upon the order of Galen’s citations: he had no reason for citing them in the order in which they 6 rpitatos Kal 6 Terapralos érépwbi 6 Cf. the xoAwdees muperoi of Ep. IV. 20 (V, p. 156 L.).

vovowv II, which is purely therapeutic, according to x. 7a0@v 18 (VI, pp. 226-8 L.), tertian and quartan fever are caused by bile and phlegm. In 44 the author discusses pleuritis, and the diseases in 44-65 are all concerned with the lung and upper cavity: kavowôns, in 63, belongs in this group since it is partly discussed in relation to its metastasis to peripneumonia. In 66, a disease called avavry is discussed, which is an affection of the digestive system. There follow in 67 and 68 two maladies, both of which have as symptoms fever and the vomiting of bile. It is tempting to connect 67-68 with 40-41, and to see in these mvperoi amo xoAns four of the seven x0A%s vodoor mentioned by Galen. Moreover, 38-43 are out of place, since they interrupt the order a capite: 37 brings the section on polyps to a close, and 44 begins diseases of the chest (cf. the arrangement in 7. ra@@v). I suggest that the original order was something like 12-37 (diseases of the head) ; 44-65 (upper cavity) ; 40, 41, 67, 68, and three others (bilious fevers). 66 and 69 are isolated chapters, which have wandered in after the original displacement, having been added here because of their bilious nature. The same applies to 38 and 39, which discuss two forms of icterus. There follow 70 and 71, both of which are diseases caused by phlegm ; while 72-75 are haphazard additions, a common phenomenon at the end of such texts as these. In other words, the diseases in 7. vovowv II were originally discussed in the following order: diseases of the head, of the upper cavity, of the lower cavity, fevers caused by bile, diseases caused by phlegm. These two categories of diseases caused by bile and by phlegm must also have been present in 7. 7@v évros radav. At the beginning of 20 (VII, p. 214 L.) the author says zrepi de rod dA€ypatos tas adras yvwpas exw, ds Kal wept yoAÿs, ideas adroû moAAas elvar.! What this remark introduces is a series of diseases caused by phlegm. By analogy we may suppose that the passage on ‘forms of bile’ (not present in our texts) to which the author apparently refers similarly contained a series of diseases caused by bile. These may even have been Galen’s seven diseases, although it is not of course necessary to suppose so. The evidence therefore suggests that in the common source of 7. vovowv II and 7. rav évros radóv some diseases were categorized according to whether they were caused by bile or by phlegm. Unfortunately, a large number of the diseases in 7. r@v évros raß@wv are not arranged according to cause, but according to the organ affected, as Edelstein? observes. Moreover, according to the evidence of Galen, this arrangement seems to have been also present in the Kvidiar yvdpai (twelve diseases of the bladder; four diseases of the kidney). Of the four diseases of the kidney in 7. r®v évros radóv the first is stated by the author (14, p. 202) to be caused by phlegm. The cause of the second variety is not given; but it is the general practice of the author in these subdivisions to omit the cause of a variety when it is identical with the preceding variety. The third has black bile assigned as a cause; and the fourth, bile and phlegm. The three affections of the liver in 27-29 are stated to be caused by black bile, bile, phlegm and blood, respectively. The five diseases of the spleen 30-34 are caused by bile, bile (yivera: pev ano rv avróv ab’ dv Kai y mpöodev 31, p. 246), blood (see below on blood as a cause of disease in this treatise), phlegm, black 1 idéas might = ‘forms of disease caused by phlegm’ rather than ‘forms of phlegm’; cf. 21 (p. 218) where in ldenv aMoinv Exew Tod émômuiou doxeira, idénv refers to the appearance of the symptoms. See, however, Fredrich, p. 47. In any case the author’s main interest is in the different forms of disease which different kinds of phlegm cause. 2 Op. cit., p. 162.

bile. Icterus, throughout the Corpus, is a bilious disease, characterized by yellowness of complexion and of urine: 7. rad 32 (VI, p. 244 L.) assigns the cause to bile; in consistency with the brevity of this treatise, only one variety is given. Of the three varieties in 7. rüv Evros ra0@v 35-38, the first is caused by bile. For the second, no cause is stated, so that one assumes it to be bile also, and this is confirmed by a comparison with 7. vovowv II. 38-39. There, two varieties are given: the first (38, VII, p. 54 L.), although it does not correspond in all symptoms with the second variety of 7. rüv évros rabüv, shares in its symptoms the combination of yellowness with blackish veins, and a. vovawv II notes that the urine is bilious.! For the third kind in 7. r&v évros ra0&v the cause is not stated; but the fourth kind lacks the yellowness characteristic of the other three, and is caused by phlegm. Of the three varieties of phthisis (10-12), the first is caused by phlegm; the second, with symptoms similar to the first, presumably has the same cause; the third is caused by phlegm and bile. Of the five varieties of the disease known as typhus (39-43), the first and the third are caused by bile, the fifth by black bile, while the second and fourth have mechanical causes, an excess of moisture and bad coction of fruit respectively, into which bile and phlegm do not enter. Finally, the causes for the four varieties of mayéa (47-50), are bile and phlegm, bile, phlegm, white phlegm. The situation, then, is that 7. vovowv II uses two principles of classification : the principle a capite, and the principle of bile and phlegm. 7. té&v évros madav has, as well as the principle a capite (1-13 = latter portions of 7. vovowv II, upper cavity; 14 ff. organs of lower cavity), a principle of classification according to the organ affected. This is then further subdivided on a different principle, that of bile and phlegm. This principle of subdivision is again used in diseases which are not classified according to the organ affected, i.e. icterus, phthisis, typhus, and the raxéa. In other words, there are two principles of classification in 7. vovowv II, and three in 7. tv evros rado; moreover, these three are used in combination. Finally, since not only some of the diseases cited by Galen correspond to those in z. r@v évros tab dv, but the actual numbers of the subdivisions correspond as well, it is probably safe to assume that the principle of subdivision, upon which the numbers are based, was also the same. The situation in 7. r®v évros radúv reflects the system in the Kvidcat yodpat. A little reflection will show that such a combination of principles, so far from being surprising, is only to be expected. Whereas a rough classification according to very general regions of the body may have sufficed at a stage when knowledge of anatomy was rudimentary, and complex varieties of disease were not yet recognized, it was no longer adequate when knowledge and theory advanced. In fact, each of the three systems used in the Cnidian works has disadvantages which would obviously make it impracticable if used in isolation. A classification according to organs is useless for a disease such as icterus, which is not connected with any particular organ, or indeed with any particular region of the body, since its varieties affect both upper and lower cavities. It falls most easily under the rubric of the humour which causes it, 1 Treatment in both works is fairly similar: parsley, and incision of the two symptoms of itching caused by the white wine, patient’s clothes, and his irritation at being blackened veins, but there are, as usual, considerable differences in detail. Cf. also 7. vovowyv III. 11 (VII, p. 130 L.), where the spoken to, reappear in the second and first forms of 7. 7úv évrôs radóv respectively.

and may be grouped, as it probably was in 7. vovowv II, among diseases caused by bile. But even so, a sub-variety may be detected, as in 7. m@v evros madar, which is caused by phlegm :! it would be an obvious absurdity to separate it from its congeners on this account. Again, a grouping according to bile and phlegm, apart from cutting across classification according to organs, which has advantages of its own,? is too wide to be of practical use, since it offers anly three categories. Used, on the other hand, as a principle of subdivision among diseases classified according to the organ affected, its advantages are obvious, as far as exposition and teaching are concerned, while from the point of view of practice it is entirely necessary, since treatment varies according to cause. It seems likely then that in the Kvidvat yv@ua not one but three principles of arrangement were used; and that this classification is reflected in the Cnidian treatises which derive from that work.3 ar. ra0wv and n. vovowv I Some further light is thrown on the common source of the Cnidian works by a comparison of certain passages in 7. waf@v with passages in 7. vovowv I. In 7. ra@&v 11 (VI, p. 218 L.), among the symptoms of kaöoos the following is included: ra pev ew buxpos yiverai, tà de éow Bepuos (cf. ibid. 14, VI, p. 222 L. on kavowôns: Mv de rà ev éÉw un rupuwóns % odddpa, ra de Eow). For the formulation ééw-écw and the combination of heat and cold, there is a parallel in the account of kadoos in rr. vovowv I. 29 (VI, p. 198 L.) where the following explanation is given : ra évdov, being moist, do not dry out completely ; while ra ¿£w, being naturally dry, are dried out even more and therefore become cold and dry to the touch. Here the aetiology given in 7. vovowv I has suggested the formulation of the symptom in 7. ra0@v.4 There are other parallels between the descriptions of kadoos in 7. vovowv I and 7. radúóv. The author of 7. ra0@v says ‘the tongue becomes rough and black through the heat of the breath’ (706 rveÿuaros dé ye Bepuòrnros : the MS. 8 has úrro OepudryToS Tod mvevpatos). The corresponding passage in 7. vovowy I. 29 reads 1) Te yAdooa Kai y papuyé rpnyüverai Te Kal adaiverai ÚTTO TOD TVEÚATOS Tot évdov kai THs Bepuornros. In 7. vovowv I ‘the heat of the breath inside” is perfectly understandable: it is explained by the aetiological description which has preceded. The blood, heated by bile, in its turn heats the inner parts of the body, which being naturally moist, retain the heat. The remark in 7. ra0@v 1 It is interesting that Ep. II. 1. 10 (V, in logic, like Plato’s divisions, nor as exp. 82 L.) recognizes two varieties of icterus, haustive descriptions of a field, like Aristotle’s; but as means towards the twofold end of teaching, and of medical practice. 3 Since most of the organs recognized by one coming from the liver, which produces a pallid (viródevxkov) complexion, and which is classified with dropsy and $Aéyua AevKov; and the other coming from the spleen, which produces a dark complexion. The first corresponds to the fourth variety in 7. TÓv evros rabiv 38 (VII, p. 260 L.) (not to the first: though the complexion there is de- Cnidian anatomy are in the lower kotiin, this sufficiently explains the predominance of classification by organ in 7. r@v évros ray, which is mainly concerned with this region. 4 For the expression cf. also 7. Ötairns scribed as wypdAevxov, the disease in Ep. II o£ewv (I, p. 149, 11 Kuehlewein); 7. 7a0@v is definitely phlegmatic); and the second perhaps to the first variety in 7. vovowv II. 28, p. 240; the Cnidian-influenced 7. rorwv Kat’ avOpwrov 45 (VI, p. 340 L.); Ep. VI. 38 (VII, p. 54 L.). 4.12 (V, p. 310 L.); and 5. 15 (V, p. 322 L.). 2 In discussing these classifications, it is It is coupled with heat and cold in the last necessary to remember that their fractical passage, on which see Deichgräber, op. cit., importance PP. 54-55- was the prime consideration. ‘They were not embarked upon as exercises

on the other hand is obscure: no aetiological description has so far been given (and when it is given, it omits reference to the process of heating) ; while the corresponding symptom, ra de éow depyos, comes after the mention of depuörns. Here again the key to the account of symptoms in 7. radav is provided by the aetiology of 7. vovowv I, which 7. raôüy gives in an abbreviated and incomplete form. The passage in 7. raôüv is another example of the way in which a symptomatic passage is written up from an aetiological passage, with unfortunate results for its intelligibility. (Of. the relation between 7. vovowv II. 1-11 and 12 ff.) In connexion with the same disease, the author of 7. ra@@v notes that kadoos often changes to peripneumonia. Metastases between the acute diseases are common enough in the Cnidian works, but kadcos-peripneumonia does not, as far as I know, occur elsewhere except in the corresponding passage of 7r. vovowv I. There, a physiological explanation is given (VI, p. 200 L.) for the metastasis of both pleuritis and ravoos to peripneumonia: the heating of the upper cavity in these diseases causes an attraction of matter to the lung. rr. nadav adds that when this change does occur, few survive it. The equivalent remark in 7. vovowv I again occurs in an aetiological context: dre aodevees NON éovres, Kal, Erepns vovoov Kawîs Eemiyevvndeions, od Suvdpevor TAS nuépas duarehéew, äypis où TO oiadov Terravbi} Ev TH mAevuovı. The relation between the two works is similar to the relation between 7. vovowv II and 7. trav èvros radav: it is more reasonable to suppose that an isolated symptomatic or prognostic remark has been lifted from its context, than that the remark has been subsequently provided with an aetiological context. For a similar instance, compare 7. radúv 7 (VI, p. 214 L.), the description of pleuritis. The author prescribes: tod prev mAevpod ris oôvyns diddvat, 6 Te dmoornoeı amo Tob mAeupoÿ TO Te Pdeypa kai THY xoAnv. No aetiology has been given, and when it is, it refers merely to the external cause of the disease. In m. vovowv I. 26 (VI, p. 192 L.), however, the following explanation is given: the phlegm and bile contained in the flesh are separated off, and stick (apoomyvurai) to the side, causing acute pain (odvvnv rapéyxe ioxvpnv). In the same passage of 7. rabüv, the brief aetiological remark is a fair example of the careless and indifferent attitude of this author to aetiological matters: yivera de y vodaos avTn padiora pev EK mociwv, OTav Tis, Vypalovros Tot GwpaTos, Y peOtuv Y vndwv prywon. Compare this with the account in 7. vovowv I: 7) peév mAevpêris, OKOTav TOGTLES ddées Te Kai toxupal Kapta AdBwou StabeppaiveTrae yap TO oda Amav Úrro TOD oivou Kal Dypaiveraı (whereas 7. ra0@v does not make it clear that the moistening is a consequence of the heating) padAvora dé Y TE YOAH Kai TO BAcyya... ÖKOTav cv TOUTWV KEKLYNMEVWY TE Kal dtvypacpevwy Evyxupnon wore pıy@oaı peWvovta Y vndovra kTA. One notices in particular the careful attention paid to temporal sequence in the description, an attention which is also characteristic of 7. r@v évros ra@v, and, I believe, of the Kviöiar yvönaı.! In the case of kaöoos, the symptoms which 7. rradóv records are clearly based on the lucid and precise physiological account which is preserved in rr. vovawv I. When, however, the author of 7. za8av comes to give the aetiology, I Nétis eipouevn, which may have been m. vovowv I has imposed a later literary characteristic of the Kvidia: yrôua (cf. 7. elegance on his source, but the structure of vovowv II. 11, p. 18) is of course particularly the thought remains. suited to such descriptions. The author of

he writes: 7) de vodoos aury yiverat Úxro xoAns, Orav kwmbeïoa évros TOV owuaros karaornpi£n. This is inconsistent with the aetiology of 7. vovowv I, which ascribes the cause to bile indeed, but uses it as an agent to heat the blood, and says nothing of its becoming “fixed”. It is, moreover, inconsistent with the account of 7. ray itself. In view of the comparative indifference of the author of 7. radav to questions of aetiology, the inconsistency is not surprising: the sentence is a vague formula which would in fact cover a number of diseases in these works. There are other examples in 7. wav. In 16 (VI, p. 224 L.) the author gives an account of the causes of pain: a condensation or collection (£vveornköra) and fixation (or) of bile and phlegm, which causes acute pain and discomfort, movov re Kai dduvnv ioyupmv rrapexeı, in whatever part of the body it settles. Again, in 20 (p. 230): xoAn 7 Pdéypa . . . és Tov OTAÑVA KataoTnpien; and 30 (p. 242): karaornpilovaw ai ddvvat (followed immediately by treatment with 4vyuara); ibid: kiuwndévra (bile and phlegm) és ra apôpa karaornpién: a. vovowv I also makes use of karaornpıkıs: 15 (VI, p. 164 L.) oxorav... ÉAkwO TO mAeupor ÚTTO TOD mpoorenTrwkôros héyuaros re kai xoAnjs; ibid., p. 166 yivovrar dé éurrvoi, Kal y bAeyna ex THs Kehadys puev pos TO TÀevpov Tpoorayÿ Kat cam); 17 (p. 170) : sepsis of the lower cavity occurs örav dréypa 7 yon avoTh des peonyù THs capKos Kai Tod Oépuaros; 19 (p. 172) óxkorav Pdéypa n xoAn Evorpabn, onmerau.... oKdrav de TeraivnraL, odvvn yiverat kai mpoobev Kal ömıodev ofen (here the armpı£ıs is the cause of pain as in 7. rabüv) ; the word itself occurs in 20 (p. 178) : ömn dv rod owyuaros ... KaraoTypien (SC. ro vypov). This explanation in terms of coagulated deposits of bile or phlegm is common to all the Cnidian works: it is an essential part of Cnidian pathology, and its most extensive use is in 7. r@v evros radóv.! One interesting example is in the account of phrenitis in 7. rad&v 10 (p. 218). There the author says 7 de vodoos yiverat Úrro xoAñs, Grav Kıyndeica mpos ra omAdyxva Kai Tas dpevas mpooibn. In the corresponding account in 7. vovowv I. 30 (p. 200), although the disease is referred to bile, nothing is said either of the dpéves or of a deposit of bile: the disease comes about, as in kadoos, when bile overheats the blood. The ¢péves are, however, mentioned in 7. vovowv III. 9, p. 128, where pain in the @péves is included among the symptoms. Apart from these two passages, the $péves are not associated with phrenitis elsewhere in the Corpus.? Now one of the most frequent uses of orjpifis of bile or phlegm is as an explanation of pain. As well as 7. radav 16 (VI, p. 224 L.) and the account of pleuritis in 7. vovowv I. 26 (VI, p. 192 L.), the following examples from 7. r@v évros malay are 1 3, p. 174 (mAeupovis) oxdtav.. . Evorpabnj «ai Évurayÿ (sc. alpa); 14, p. 202 (first disease of kidneys) óxdrav 6 veppòs és éwurov avalaßwv dréypa un adin madw, addr’ adtod Evprwpw8f; 16, p. 204 (third disease) óxórav xXoAN és Ta BAEßıa ouppÿ Ta Teivovra és Tov vedpov, Kai OKkôTay o7 ÿ, ¿Axoí ra pAéBia; 35, p. 252 (first icterus) ioraraı odv y yo) by drying and rés of the fluid in the joints. It must be admitted, however, that ornpı£ıs is not confined to the Cnidian treatises, though its use there is systematic. Cf. Ep. IV. 35 (V, p. 178 L.); VI. 2. 6 (V, p. 280 L.). It is related to the doctrine of ardoracis, a doctrine characteristic of the Epidemics, in n. xvuóv 7 (V, p. 488 L.) and in 7. dcalrns vrro TO déppari Kat ev TH kehaAñ; 36, p. 256 otéwv (I, p. 162, 9 Kuehlewein). Thucydides (second variety) 70 dé dypôy TO ev TH Owyarı nmyvuras ev TH Öcpparı; 43, p. 272 (fifth todos) OKOTAV TO ÚYPOY TO EV TH OWUATL ovuray Kai dvafnpavOn u&)ov rod Katpot; 48, p. 284 (vouonua maxv) yiverat pev dro xoAñs, uses the term, presumably as a familiar one, éxoTav xoA7 eri TO Harrap émppuÿ Kai Kara THY Kehadnv kaTraoTÿ; 51, P. 292: ioxıas is caused in 2. 49. 2 mr. ma0&v notes pain in the vroxdvòpia, particularly on the right side towards the liver. Cf. the case of Apollonius, Ep. III (III, p. 136 ff. L.); tension of the belly on the right side in Ep. IV. 45 (V, p. 186 L.).

significant: in 28 (VII, p. 240 L.) the fixation of bile on the liver causes acute pain; 41 (p. 266) caraornpilerat (7) yon) . . . Kal ddvvas mapexeı ogelas. Particularly interesting are cases where ddvv7 itself is the subject of evornpilew, «araornpilew etc.; 49 (p. 288) és rrv kebaAnv ééarivns oduvy ornpiler oËein ; 19 (p. 214) és de rov arrAñva ddvvy évornpiler d€ein with two more examples further down; 7. rad@v 29 (VI, p. 240 L.) ornpilovoa y 68uyn. In view of the aetiology, ornpitew etc. in these passages must, I think, be taken literally, as referring to the substance or deposit causing the pain: no distinction is made between the sensation and its cause. For an example of similar concreteness of thought, cf. 7. t&v èvros rabóv 50 (VII, p. 292 L.): Kai oxdrav orf (sc. 76 vovonpa: i.e. the dAeyua which causes the vovonua) ev adri (sc. 7H Kowin).! In this case, the aetiology of 7. radav and the symptom noted in 7. vovowv III may go back to a common source. Otherwise, however, the accounts of phrenitis in 7. rad&v and 7. vovowv III have nothing in common (no aetiology is given in 7. vovowv III), except that both note a milder and a more acute form of the disease; and neither has any point in common with 7. vovowv I,? except that the disease is also caused by bile in 7. adv. There is, however, a connexion in aetiology between 7. vovawv III. 15 (VII, p. 140 L.) and 7. vovowv I. 7 (VI, p. 152 L.). In the latter passage, which is a compendium of beneficial and harmful effects, the author notes as harmful a flux occurring in pleuritis or peripneumonia (cf. Aphorisms, IV, p. 566 L.). Such aphorisms might come from any source, but it happens that the same observation is made in 7. vovowv III, and in that passage it is attached to a coherent explanation: the effect of a flux is to dry out the upper xowAin, and so prevent expectoration.3 The aetiology contained in 7. vovowv I illustrates a passage in 7. vovowv II. 44 (VII, p. 62 L.). There, among the symptoms of pleuritis the author mentions pain in the groin and urination of blood. Urination of blood aiso occurs among the symptoms of ‘dorsal pleuritis’ (7) és ro v@rov wXevpiris) in am. vovowv III 16 (VII, p. 144 L.). The reason appears in 7. vovowv I. 26 (VI, p. 194 L.): pain in the side is caused by the chilling and coagulation of blood in the hepatitic and splenitic veins, the anatomy of which is described by the author. Often the pain moves (rpdrrnraı 6dUvn: the same concrete use 1 The pathological principle of orjprfis true, is another instance of the eclectic is behind the frequent recommendation in character of 7. vovowv I. But we should have these works of fomentations, yxÀidopara, as a to conclude that 7. vovowv II. 3 (VII, p. 10 L.) also goes back to the same doctrine: means of relieving pain. Cf. 7. radúv 16 (VI, p. 224 L.): bile and phlegm, when £vveornköra, cause pain, but on the application of fomentations, they become dLarexvpeva, and the pain disappears. (Contrast 7. romwv Kar’ avOpwrov 42 [VI, p. 334 L.], where pain is explained as the effect of heat and cold on their contraries.) The remedy itself is of course traditional, even instincmapagpovéer de órav TO alpa TO Ev TH kepaÀñ vò xoAñs bAéyuaros vrepdepuavdn. (Incidentally, this passage contradicts the doctrine of 7. ipñs vovoov, and therefore weakens Pohlenz’s theory about the source of the aetiology of rr. vovawv II. 1-11.) 3 Each of these works has its own characteristics: 7. vovawv III, while it gives no tive; certainly much older than the aetiology aetiology as such, is noticeably careful to which explains it. 2 The difference was noted by Wellmann, give an explanation for every stage of the treatment it recommends. With the present Fragmente der sizilischen Aerzte (Berlin, 1901), pp. 17 ff, who argued that 7. vovawv I goes passage, compare the account of pleuritis in 16 (VII, p. 142 L.). The clarity and logical back to the ‘Sicilian’ doctrine that blood is consequence with which the author of 7. the seat of dpdvyats, while 7. ra? @v represents vovowv 111 describes treatment are remini the view that its seat is in the gpéves. This, if scent of the style of 7. r@v evrös ral.

as was noticed before) to the lower parts, and is conveyed through the veins to the bladder, causing much urination of a bilious nature. The explanation of pain and its movement over the body by a orypiéts of blood in the splenitic and hepatitic veins lies behind other passages in the Cnidian works.! 7. vovowv III. 15 (VII, p. 136 L.) (peripneumonia): oöVVn Úrro THY wuorÄdrnv Kal eis THY KAnlda Kai es rov rurbov: here the pain follows the course of the splenitic and hepatitic veins. The same distribution of pain occurs in 7. r@v évros rradóv 25 (VII, p. 230 L.), in which dropsy originating from the spleen is described : ööVvaı re yap d€etar Ev TO omAmvi Kabtoraou, perarimrovor de Kal és TOV Duov Kal és THY kAnida Kal és Tov TuTOOV Kai és rov Aaydva. In 32 (p. 248) (third disease of the spleen) the same distribution is given, including the shoulder blade (cf. 7. vovowv 111 above). Moreover, on p. 250 in connexion with the same disease the author recommends bleeding azo rot aykóvos Tod apıorepov THs etow dAeßos (i.e. from the splenitis) ; this is paralleled in 7. ra0@v 20 (VI, p. 230 L.), where the advice is given, in conditions of the spleen, 77» PAéBa TV orAnvirw adrevar murva.2 Although we do not possess the part of 7. rav evros mad@v which dealt with diseases of the upper xo:Ain we can perhaps infer that in its account of pleuritis the same description of pain was given. In 27 (VII, p. 236 L.), in discussing hepatitis, the author advises as a remedy for pain the same fomentations as for pleuritis. Earlier in the passage the symptom has been included of pain in the liver, the lower ribs, shoulder, collar-bone, and under the nipple. Thus the doctrine on this point is identical in 7. vovowv I, II, III, 7. rabüv, and 7. T@v Evros rabüv. CNIDIAN AETIOLOGY The Cnidian works are linked by a common aetiology, which appeared in the Kvidiat yv@uau. It is time to examine this aetiology in more detail. In attempting to reconstruct it, one must remember that, with the exception of a. vovowv I, these treatises are not primarily interested in aetiology, and their remarks on the subject are necessarily incomplete; while 7. vovowv I itself, a handbook written for a special purpose, is abbreviated from a more extensive source. Even so, it presents a coherent system, with which the other works agree on essential points. 1 For the vascular system to which the author refers, cf. Fredrich, about ornpı£ıs here. Peripneumonia occurs Hippokratische when the lungs swell so as to touch the sides ; Untersuchungen, pp. 57 ff. and his remark on if one side only is touched, the disease is p. 61: ‘Ein Unterschied zwischen koischer und knidischer Schulanschauung lässt sich lung collapses and sticks to the side. It is pleuritis; ‘dry’ pleuritis is caused when the nicht feststellen.’ The author of 7. rorwv interesting that a remark in Polybus’ account Kat’ avdpwrrov 3 (VI, pp. 280-2 L.) appears of the vascular system implies to have a system of his own (cf. also Fredrich, aetiology of pain as in the Cnidian works :. p- 60); in this respect it is significant that the aetiology which this work gives for periemuTndevew dì xp TAS TOMAS ds mpoowTaTw pneumonia (VI, p. 302 L.) a. uoios avOpwrov 11 (VI, the same p. 60 L.): and pleuritis rapvew darò Tay xwpiwv, ¿vda dy ai ddvvar (ibid. and p. 316 L.) is quite different from penaßmkwoı yiveodaı Kal ro aipa EvAAdyeodaı. the aetiology which appears in 7. vovowv 2 Apart from this passage, the actual I, and which is implied in 7. 7ra0@v and mr. vovowv II. In these works, the pain name orAyvíris is not used elsewhere in the associated with the disease is caused by a apart from later writers, such as Rufus of Corpus, so far as I am aware; although arnpı£ıs of blood in the hepatitic and splenitic Ephesus, Aristotle uses it in connexion with veins, where they touch the side; 7. rörwv Diogenes of Apollonia (Hist. An. 3. 3, 512% kat’ dvOpwrrov on the other hand says nothing

The system is summarized in 7. vovowv I. 2 (VI, p. 142 L.): At pev odv VOVOOL yivovrat ÁTATAL, Tv pev Ev TH OWUATL Eveövrwv, amo Te XoAns Kal pAéyuaros, Tv de E£wdev, amò rróvov Kal Tpwudrwv, Kal do Tod Depuoú drepleppalvovros, Kal TOD ıhuxpod drephüyovros, Kat Tob Enpod vrrepénpaivovros, Kat 708 vypod Urepuvypaivovros. Kal y pev xoAn Kal TO hAéypa yivopevoLol TE ovyyiverat, Kal évi alel Ev TH owuarr Y mAEOV éaooov: TAS dé vodaous mapéyet, TAS pév AO ouriwv Kai Torûv, Tas de darò Tod Beppoò Urrepdeppaivovros Kai amo Tod Yuypod ürephéyovros. The parallel passage in 7. 7a0@&v (VI, p. 208 L.) reads as follows: Novonuara roioiw avOpmrovow yiveraı Amavta Úrro vos Kat dr€éypatos: 7 dé xoAn kai TO pdéyua TAS vovoovs Tapexet, OTAV Ev TH OWpaTt Y VTEpENpaivyTat, N) Örepvypalvmraı, Y Örepdepuaivnrat, Y dmepydynrat. ITaoyer de radra To pAéypa Kal Y XoAN Kal ATO OLTÍWV Kal TOTÜV, Kai drrò TOVWY Kal TPWUATWY, KA ATO Gaps Kal akoÿs Kal dos Kal Aayveins, kai amo Tod Beppod Te Kal puxpod: maoxeı dé, OTav TOUTWY ExaoTa THY eipnuévwv Y UN Ev TH S€ovTe mpoodepnra TH cwpart, Y un Ta eiwbora, y mAeiwm TE Kal loyupôrepa, Y éddoow TE Kal aodeveorepa. In 7. vovowv I the manuscript 0 omits the words kai Tod Enpod drepénpaivorTos, Kai TOD Üypod ÜUrrepvypaivovros!, nor does it use them at the end of the chapter. On the ground of this omission, Edelstein? argued that the systems in the two works are not identical, but that 7. ra0@v has its individual modification of the doctrine in 7. vovowv I. Such an omission would in any case be insignificant, in view of the abbreviated character of both works; but whether the actual words occurred or not, the doctrine itself is confirmed in 7. vovowv I. 26 (VI, p. 192 L.). In that passage, giving the causes of pleuritis and peripneumonia, the author writes: dvafepyaiverar yap TO cpa dmav Úrro TOD olvov Kal vypaivetat, padiora dé Y TE xoAN Kai TO dreypa dadeppalveral TE Kal vypaiverar. Cf. 18 (p. 172): erysipelas occurs in the lung örav ürep£npav6n: the causes are kaûua, rruperoi, raAautwpin, akpaoin. So far 7. vovowv I and 7. rradów are consistent: cf. the latter, 17 (VI, p. 224 L.) : órav 70 cpa Umo Tod nAtov HaAbO7, dypaiverar: dypawopevov de vooée, Y TAv, Y és O TL Av karaornpiËn TO Phéyua Kal Y xoAN. It is more significant that 7. vovowv I has the words kai 7 pev yo Kai TO Pdéypa yivopevotol Te ovyyivetat, Kal eve alei ev TH compatti? TA€ov 7 €Aacoov. Nothing of the sort is said in 7. ra0@v, and yet I venture to think that it is implied there, as well as in 7. vovowv II and 7. ta&v évros 7abdv. First we should look at some passages in the remainder of 7. vovowv I. The passage in 26 {p. 192) quoted above, 76 oûua arav...pddora dì y TE xoAN Kal To phéyua means quite simply ‘the body as a whole...and in particular its bile and phlegm’. It confirms the doctrine expressed in 2. It is similarly confirmed by the expression in 15, p. 168: oxórav yap SuadepuavOr (sc. 7) kebaAn) bio THs Kodins (i.e. the upper cavity, 7 dvw KotAin: the doctrine as in the description of kadoos in 29 [p. 198]) EAreı és ewurnv ex TO owuaros TO AeTTTOTATOV rod ddéyparos. Read in the context of the whole chapter, the implication is that léyua is a permanent ingredient. One may compare also 26 (p. 192) öcov ev adrÿ TH capri evi xoAfjs Kai pAéyuaros. There is nothing in 7. vovowv I to contradict its initial doctrine, and its language tends rather to confirm it. Now 7. ra0@v, rr. vovowv II. 1-11 and 12 ff., and 7. rav évròs radóv, not being primarily concerned with aetiology, take bile and phlegm for granted. Their remarks on the subject range from the vagueness of ‘this disease is caused by 1 Cf. Littré, VI, p. 142 n. 17. 2 Op. cit., p. 161, n. 1.

phlegm’, through the greater precision of ‘phlegm flowing from the head’, to ‘phlegm flowing from the head, to which it is drawn from the rest of the body’. As in 7. vovowy I, there is no passage anywhere in these works which contradicts the doctrine that bile and phlegm are permanent ingredients in the body; and there are numerous passages which tend to confirm it, if only through significant silence about their origin. Fortunately, however, we are not obliged to rely on passages of dubious interpretation. The doctrine is implied by another aspect of the theory.! m. vovowv I. 2 divides the causes of disease into internal causes, namely bile and phlegm, and external causes :? the latter are fatigue, wounds, heat, and cold (dryness and moisture). This is again stated succinctly in 11 (p. 158): TOV pév Ev TO OWparı evedvTwy, amo Yodfs kai drAéyparos, Tv bE eLwlev, amo [rod népos Eriuyvuuevov TH ouubürw Depp, arap kai amo] movwv Kal Tpwudrwv.3 To this list of external causes, the author of 7. radóv adds food and drink, smell, hearing, sight, and sexual intercourse. He also makes it clear, as the author of 7. vovowv I does not, that these two categories of causes are connected: the external causes act upon the internal causes, bile and phlegm, as agents causing dryness, moistening, heating, or chilling. This relation between the two kinds of cause is illustrated in 7. 7a0@v 27 (VI, p. 240 L.), a passage which provides a neat commentary on the two passages we have been considering. There, the author says yiverat dé radra ra a) yuara (i.e. cholera and diarrhoea) dca Ex mooiwv yivera 7 €£ edwyins, OTav Ta ouria kai Ta morà TAdw TOD elwbóoros és THY Kodinv éoedbn, Kai Ta E€wlev ciwhdra Ürrepdepuaivewv TO o@pa Kwen xoAnv Kat dAéyua. The patient overeats, and is exposed to heat which sets in motion the bile and phlegm already within him. Thus a comparison of 7. vovowv I and 7. radóv indicates a coherent and unified system, upon which both depend, and which has reached a sufficient degree 1 The four humours of 7. vovowv IV are also connate (VII, p. 542 L.), but they are as well taken in with food and drink (ibid., p. 544: for the differences in doctrine in this respect between 7. vovawv IV and 7. duaios avdpwrrov, cf. S. Hornstein, ‘Untersuchungen zum hippokratischen Corpus’, Primitiae Czernovicienses ii [1911], pp. corresponds to the explanation given by Menon for Philistion’s mapa Ta orotyeia (xx. 35). The second operates through ‘surfeit or deficiency of nutriment’. But neither Alcmaeon nor Philistion has the dichotomy of external and internal cause which is characteristic of, and used systematically in, 76-77). This a. radóv | m. vovawv I. There are three classes may represent a combination of two originally of cause according to Philistion-Menon, while two different principles of classification are used by Alcmaeon (since ud’ od, é£ où, distinct doctrines; traces of both are found in the Cnidian works. 2 The category of external cause is comand ev ofs cannot be regarded as subdivisions mon. Alcmaeon seems to have been the of ‘internal causes’). On the other hand, first to distinguish it (Diels-Kranz 14 B 4: it must be admitted that in 7. vovowv I. 23 eyyiveodor dé... Kan Tv étwéev airidv, which are then specified as waters, places, fatigue, and dvayxn), followed by Philistion (Anonymus Londiniensis, xx. 30 ff. Diels: (VI, p. 188 L.) there is something like the vp’ où | é£ où of Alcmaeon: Beppaiverai de y xoAn kai ro phéypua (= vd’ od) Evöoder pèv ano diseases arise rapa Ta orouyeia Or mapa Tv amò (wounds, fatigue, excessive heat and cold). Contrast 7. madav, where ouria and mora are definitely external causes. Here, as in other matters, 7. vovowv I shows a certain eclecticism. Ty cwparwy Sidfeow or apa Ta éxrós, where the last are further specified as wounds, excess of temperature or change of temperature, and change to an unsuitable and harmful diet). Alcmaeon’s other causes are ws airiwy Kat mordy (= éÉ 00)... éEwlev 3 @ omits the words in brackets. Connate vd’ od and ws é£ od (‘Ursache’ and ‘Veranheat lassung’: Diels). The first is described as Cnidian works. operating ‘by excess of heat or cold’, which Se is not mentioned elsewhere

of sophistication to be formulated in terms of categories of causes.! The substance of the theory is given a logical form. Moreover, such a distinction of internal and external causes makes sense only upon the assumption that bile and phlegm are permanent ingredients in the body. The effect of heat, cold, etc., is to set in motion, arouse (xwéew) the bile and phlegm (as one ‘arouses’ a sleeping dog, not one which has just been let in by the door). This notion of kivnoıs recurs throughout the Cnidian works, with the regularity and the precision of a technical term. Compare the following passages: 7. vovowv I. 27 (VI, p. 194 L.) KEKLVTLÉVOU Kal Beppawopévov TOU phéyuaros Kat Tis xoAns; 7. nadav 11 (VI, p. 218 L.) y de vodoos aurn yiverai dò xoAjs, órav Kıyndeica Evros TOD owparos karaornptËn; ibid. 12 (p. 220) 7 de perdoracis aùr@v (fevers which occur in winter) rovadrr yiverar: órav, TMV Odo Kexwnpevwv Tod pdhéyparos TE Kal THs xoAñs, un Ta Evupépovra mpoodepnrau TO owuarı, cvaTpepopeva aùrà mpòs EwuTa TO TE Pdéypia Kal Y xoAn MpooTintet TOO cwpatos À av rúxy KTA. (Here un ra Évupépovra have been specified above as wine or fatigue: once again the distinction between external and internal cause.) Ibid. 15 (p. 224) (¢ArA€ypa) orav kıyndev mpooréon pos THY kapôimr. Cf. ibid. 30 (p. 242) (xtwn0évra) and 32 (p. 244) (xoAn kırmdeica). m. Tv evros radóv 30 (VII, p. 244 L.) yivera de 7% Tovadry vodoos dia Beppacinv Tod nAtov xoAns kırmdeions; ibid. 39 (p. 260) Emiaußaveı de Dépeos év wpn, ¿óxkórav 6 KUwY TO doTpov EmireAAn, xoAns kiwnbeions ava TO O@pa. m. vovowr II. 5 (VII, p. 12 L.) ayuvyéer de drav mpooiornrau mpos Tv Kapdinv Pdéypa 7 xoÀn: mpocioracba dì avayKn, Kkekivmuévwv Kai dypacpevwy (!); ibid. 6 (p. 14) 1 Hornstein, however (op. cit., p. 68), attempts to distinguish 7. ra@@v from n. vovowv I on the ground that the former does not explicitly mention the distinction of causes in its first chapter. But it is clearly implied in the wording of that passage, just as it is actually used in the passage I have mentioned. Cf. also 7 (VI, p. 214 L.) where the author, discussing pleuritis, remarks yiverar dé y voücos aurn padiota pev ex moolwv.... Here the external cause is given without the internal cause, as it occasionally is in these works. The internal cause, however, is implied by what the author has said viously the author has given a long account of the disease in terms of its internal cause, in the passage (p. 280) beginning öxorav dréypa Kal xoAn pix0î Kara TO odpa. Clearest of all is the passage on ¿oxeds in 51 (p. 292), which begins dro r@vde rHv adoppov yiveraı, the causes being sunbathing and the consequent drying (ava&npavdn: cf. n. vovowy I. 2 and 7. rada» 1) of the fluid in the joints. So much for the external cause. Then, shortly after, the author says (p. 294) yiverac de TO vovonpa amo yoÀÿs yivera dé Kal amo pPhéyuaros Kai amo atparos KTÀ. It must be admitted that occasionally the earlier in the same passage. The parallel two categories of cause seem to be conpassage on pleuritis in 7. vovowv I. 26 (VI, p. 192 L.), which I have already quoted, distinguishes more clearly between external and internal cause (the latter being bile and phlegm). This precision appears most clearly, of all these works, in 7. r@v évros ralàv: eg. 28 (VII, p. 240 L.) yivera de (sc. hepatitis) dao xpendayins Boelwv kpedv kai €£ oivodAvyins: Tara yap mavra modepudrara Taurnmv Thy «Wpyv TÒ rar, Kal xoAnv udAıora TpocioTnoı mpös TO frap; 48 (p. 286) : aùrn 7 vodoos mpoominre ualıcra ev addodnpin Kal nv Kou Epnumv ddov Badily Kat fused: 7. vovowv I. 20 (VI, p. 176 L.): ev de o previously veloped: on these, see M. Wellmann, Die (p. 284) the author has said yivera pév azo pneumatische Schule (Berlin, 1895), pp. 153 ff., dößos atrov Aaßn... where TO mievpà yiverar pév $upara Kal ane phéyuaros Kal amo xoAñs . . . yivera dé Kal aro róvov KTÀ. Cf. 7. rüv évros taba 17 (VII, p. 206 L.): ro uev vovonpa yiverar amo xoAns Kai pdéyparos ... yiverat dé Kal amo Aayveins y vodcos. But such passages are a minority. A similar distinction of causes is pointed to by the remark in Zp. II. 4. 5 (V, 126 L.): aM’ emi Tv mpdédacw Set eAdeiv Kai THs mpodao.os Tyv apyynv. I suspect that it was from this distinction that the later altia mpokarapkrıy and ovvexrixy dexoAns ; 47 (p. 282) roûro TO vovanpa yivera dia and the definitions from Dioscorides quoted deppacinv rod HAiov «ai Vöpomocinv where prethere.

péAauwa yo ev Ti kebañ kıyndeica; ibid. 9 (p. 16) : ev rH Kedar dréeypa Kırmdev. In the last two passages it might appear that the author of 7. vovowv II. 1-11 regards the head as the seat of phlegm. That this is not the case is shown by 3 (p. 10): Eueeı de yornv dre Kexwnpevns aùrÿs ev TH owpart, kai 1 Kepadr vrro THs Deppacins EAkeı Ed Ewurnv, Kal TO pev TaxVTaTov Eueei, TO de Aerrórarov éAke és ewvriv. With the last words, and indeed with the whole passage, compare 7. vovowv I. 15 (VI, p. 168 L.): the head, being hollow and placed above, when it is heated by the lung, draws 70 Aerrórarov Tod dAéyuaros towards it. In the passage from 7. vovowv II, the head has been heated, not by the upper cavity as in 7. vovowv I, but by the bile and phlegm which are already present in the head. This is not inconsistent, since it is the doctrine of these works that bile and phlegm are present in all parts of the body. This duplicate process, if one may call it such, in which the initial action of a humour has as its consequences the attraction of the same humour in still greater quantities, is characteristic of 7. vovowv I. Compare the process of empyema in the lung, 7. vovowv I. 15 (VI, pp. 164-6 L.) and 13 (p. 160) ; and ofpleuritis 26 (p. 192) and peripneumonia 27 (p. 194). In the last passage, the words are eAkvon o mAevuwv Úrro Beppacins ed éwvrov aro THY TANOLOV ywpiwv mp ds roîouv ÜTapxovouv ev ewvTÚ. There seems to be a contradiction here with the account of peripneumonia in 7. radóv 9 (VI, p. 216 L.) in which phlegm flows not from the contiguous parts, but from the head. That the two accounts are not necessarily inconsistent is indicated by the process described in the passage above, 7. vovowv II. 3 (VII, p. 10 L.); and also in 11 (p. 18): ro de pléyua amo ris kepadis karaßaiveı y de kedado ex TOD owuaros EAkeı: EAKet de OTav diaPeppav0ij; drabeppaiverai de Úrro citiwy Kai TAlou Kal rróvwv Kai rrupós: orav OLaDdepuavOh, EAke TO Aerrrórarov és EWUTNV Éx TOD cwpaTos: órav de eipvon, karaßalveı Kat TaAW és TO O@pa.! According to 7. vovowv I and 7. wafév, the internal causes of disease are bile and phlegm; 7. radóv emphatically adds in a later chapter (37, VI, p. 246 L.) that the cause must be either one, or the other, or both. The emphasis of 7. radav, however, raises a difficulty in respect of 7. rav èvros rabóv, where on several occasions blood is mentioned as causing this or that disease: 4 (VII, p. 178 L., with the reading of 0); 7 (p. 182); 12 (p. 192); 29 (p. 242) (this is, however, mechanical, since phlegm and blood together cause a rupture of the gall bladder) ; 32 (p. 248) ; 51 (p. 294). The relation between blood and bile or phlegm in the Cnidian works is usually that of patient to agent: blood is heated, or chilled, or corrupted, by the action of the humours upon it (e.g. rr. vovowv I. 24 (VI, p. 188 L) piyos; rr. radóv 19 (VI, p. 228 L.) dilution of the blood through phlegm ; 7. vovowv II. 6 (VII, p. 14 L.) chilling of the blood by bile in apoplexy). The idea also occurs in 7. tév Evros 7afdv itself: in 6 (VII, p. 180 L.) we are told that inflammation of the lung occurs through phlegm, órav puyev TH aipare mppuñ mi tov mAevuova, and in 41 (p. 266) that a form of typhus occurs ökörav Y xoAn oametoa puyf TÓ alpari ava ras pAéBas kai Ta apOpa. That a corruption of blood is the consequence of such mingling is indicated by similar passages in the other works: 7. radúów 23 (VI, p. 234 L.) 7 de ! Cf. also 7. vouawv I. 19 (VI, p. 174 L.): aro Tis xepadîjs te Kat TOD AAAov owuaros $Aeyna xarappéov. If we should look for verbatim fragments of the common stock of leaps to the eye from its surroundings. It is all the more significant that it is detailed and full (contrast the inferior aetiological remarks of 7. ra@&v); and it observes clearly these works, we might do worse than find the distinction between external and internal one in this passage from 7. vovowy II, which cause.

voticos yiveraı, erreidav yon Kal dréypa karaornpién és ras pAéBas kai THY KotAinv. VOTÉEL pevTOaipa..... ; 7. vovowv I. 20 (VI, p. 178 L.) Kai óxorav pry OR TO TE alua Kal TO Üypov TO amo Ths capkos, TaxUveraı TO ala and it becomes voowdéotepov and oracıuwrepov. The process is in fact identical with the one to which the term úrepayuetv was wrongly applied: 7. vovowv II. 4 (VII, p. 10 L.) Ureparuéewv doreeı OTav és tas bAEBas xoAn n phéypa éoédbn. The process is also mentioned in 7. vovowv II. 12 ff., deficient as that is in aetiological passages: 41 (p. 58) 7 votcos aurn AapBaver, qv tbrepyoAnon TO aipa. Here the form of the word is interesting in the light of the earlier passage in 7. vovowv II. It is doubtful whether the use of blood as a factor producing disease in 7. Tv evrös radów is anomalous with its use in the remaining Cnidian works. The first case, 4 (VII, p. 178 L.) is dubious. Blood is there mentioned only according to Mack’s collation of 8; Littré did not accept the reading. Even if it were correct, blood is used here not independently, but in combination with black bile; this need mean no more than the normal process by which the humour acts on blood. In 7 (p. 182) oedema of the lung is due to an attraction and retention of blood by the lung. Presumably, although the author is not explicit, the blood suppurates. Such independent suppuration of blood is a contingency which occurs in other Cnidian works: cf. 7. vovowv I. 4 (VI, p. 146 L.) kai nv aîua Er rpwuaros Y dAeBos pui) és Thy avw Kodinv, avayKn TobTO arúos yeveodaı: the process is described in 14 (p. 162) and in 17 (p. 170). In m. Tov évros mad@v 12 (VII, p. 192 L.) blood again acts in concert with bile; while 32 (p. 248) and 29 (p. 242) are simple cases of rupture, of spleen and gall-bladder respectively, caused by a superabundance of blood. 7. r@v évrôs radóv is to a large extent concerned with internal organs, and it is not surprising to find such ruptures mentioned here, and not in other works. In 51 (p. 294) ioxıas is caused by ornpı£ıs of blood in the joints. ornpı£ıs of blood as well as of bile and phlegm is normal in the Cnidian works; it is used to explain this same disease by 7. rad@v 29 (VI, p. 240 L.). There, ioxıds is caused by a ormpı£ıs of blood, and the arppı£ıs is further caused by bile and phlegm. The words in 7. r@v Evros mal&v are yiverai de TO vovonpa amo yoXîjs: yiverar Oe Kal amo pléyuaros Kal darò aiuaros. There is no reason to infer from them that the author conceives of blood as acting independently of the humour here, any more than it does in the parallel account in 7. ra0@v. Finally, one may compare a passage in 7. ra0@v, the work which is so insistent that diseases are caused by bile or phlegm or both: 34 (VI, p. 244 L.) dvpara doa duerai, mavra do phéyuaros y alparos dvera.! I conclude that on this point the aetiological system of 7. r@v evros radúv is no different from that of the other works. Such was the Cnidian system. Its most distinctive feature is the doctrine that bile and phlegm are connate and permanent ingredients of the body, ‘inborn in us’, in the words of the author of 7. vovowv I. This is in turn closely related to, and implied by, the distinction between internal and external 1 Moreover, p. 172 L.) in blood 7. vovowv I. 18 (VI, is in fact the cause of humours, as a humour causative of disease, while in the two latter it is regarded as erysipelas in the lung: Kai oxórav ürepfnnormally preserving health, though liable to pavOy, éÂke TO aia Eb’ éwurdv. Hornstein corruption by bile and phlegm. In view of the passages cited above, I do not think that (op. cit., p. 73) argued that the doctrine of mr. Tov évros mal@v was distinct from that of this distinction between a. wabav{m. vovowv I, since in the former be maintained. blood is regarded, analogously to the other treatises can

cause. If I am right in arguing that this theory is common to the authors of a. vovowy I, II, 7. radóv, m. rüv évros adv, and possibly 7. vovowv III, then it follows, I think, that it was also the theory of the Kvidia: yvöuaı, the common source of these works. All the more interesting, then, that it is not the theory ascribed to Euryphon of Cnidus by the author of the Menon Papyrus.! University of Sydney 1 iv. 31 ff. Diels. When the same author contrasts ‘Aristotle’s’ report of Hippocrates’ I. M. LoNIE Fredrich (De libro mepi duüaios avOpwrov pseudippocrateo, pp. 28-31) that precisely the doctrine with that of ‘Hippocrates’ himself, two points of internal and external cause, it is precisely upon the decisive difference between the humours as repioowuara, and as innate ingredients of the body, that he lays his emphasis (vi. 42 ff.): [xai w]s p(ev) 6 ApıororeAns oiera/mepi ‘“Im[rokparolus Taira * * ws Sé/adros “Imroxparns déya, yilveodaı) ras vooovs/[m(apa] ras Scadopas T(Dv) ev ri auordole dúseaw. (The theory of meptoowpara and düoæ has preceded.) In what follows, the Anonymus is relying upon 7. vovowv I. 2, and it is interesting to see, even without the supplements suggested by and the congenital nature of the humours, are emphasized by the abstracter (vii. 5): radra (bile and phlegm) éri 67 yıvopevoıs; ibid. 11: rade y(ap) ékrôs...môvwv UTeEppérp[wv]. In these passages the author is of course following out his own doxographic principle (iv. 26-28: of pév...aapa ra mepiocwpara ... ci dé mapa ra orouxeia). On the other hand, it was surely the ease with which 7. vovowy I. 2 fits into this doxographic scheme that led the author to select it as characteristic of ‘Hippocrates’.