Acknowledgment. I am most grateful to Miss R. Howorth, whose lucid and penetrating comments helped to clarify the ideas in this essay. I am no less grateful to Mrs K. M. Burnett, Mr B. Harries and Miss P. Smith for reading the essay and suggesting a number of improvements.

¹ Certain general features of Empedocles' style of simile relevant to frr. 84 and 100 are considered separately in note 1 pp. 154–7 below.

² Timaeus 45b–46c, cf. 31b and 67c–68d.

The originality of Plato's theory is considered separately in note 2 p. 157 below.

³ Theophrastus, , De sens. 7–8 Google Scholar (DK 31A86; these references are to Diels-Kranz, Die Fragmente der Vorsokratiker 5th edn onwards). For the theory of good and bad vision, see ‘The relation of Anaxagoras and Empedocles’, JHS lxxxviii (1968) 110–13. As an indication of the completeness of Theophrastus’ account, note especially the concluding sentence of the report on vision,

Theophrastus' account does of course include mention of fire which leaves the eye (see p. 144 below), but Theophrastus does not give this as in any way the cause of vision. Beare distorts when he writes that in his account of Empedocles Theophrastus introduces us to ‘vision by means of fire issuing forth’, Elementary cognition 20.

⁴ Aristotle, De sensu 437b10–438a5.Google Scholar

⁵ A synopsis of earlier views is provided in note 3 pp. 157–9 below. References to works cited in note 3 are given elsewhere in an abbreviated form.

Bignone and Cherniss are exceptional in denying any part to outward-flowing fire in Empedocles' explanation of the act of vision, see p. 145 n. 28 below, and cf. note 3 p. 159 below.

Miss Millerd and Professor Guthrie are exceptional in allowing outward-flowing fire a place in Empedocles' explanation of vision, but in refusing to synthesise this with an explanation in terms of effluences flowing from the object seen, see note 3 pp. 157–9 below, and cf. p. 142 n. 9 below.

Doxographical evidence for Empedocles' theory of vision, other than that in Aristotle and Theophrastus, is considered separately in note 4 pp. 160–1 below.

⁶ De sensu 437a22–438b30.

⁷ De sensu 437a22–b10. The precise nature of the phenomenon to which Aristotle alludes in this passage is considered separately in note 5 pp. 161–2 below.

⁸ I have done no more than transcribe the text of the fragment given by Ross in his edition of the Parva naturalia (except for one misprint), without intending to endorse the various interpretations of detail implied therein.

⁹ Miss Millerd and Professor Guthrie are exceptions, see note 3 p. 159 below.

¹⁰ Bignone exaggerates when he writes, 249 n. 2: ‘Aristotele … dica che Empedocle … spiega la teoria della vista … per mezzo del fuoco che esce dall’ occhio e si congiunge col fuoco esterno, come nella dottrina del Timeo di Platone' (my italics). Aristotle does not attribute to Empedocles the idea that outward-flowing fire mingles with fire outside the eye.

Likewise, there is no need to suppose that Aristotle's later criticism, 438a29 ff., is directed specifically against Empedocles, as von Prantl, supposes, Aristoteles über die Farben 45.Google Scholar

¹¹ Ross, , in his edition of the Parva nat. 190.Google Scholar Karsten, 486.

¹² Bonitz, H., Index Aristotelicus 263b24–5.Google Scholar Other instances of the same construction, quoted immediately before and after this reference, have a clearly different sense.

There are of course a number of instances of ἐοικέναι with an infinitive, only some of which carry the connotation which is present when there is a dependent participle. A good example (not indexed by Bonitz) is De caelo 30531–4, which concludes Comparison with De gen. et corr. 325b19–25 shows that Aristotle was not at all certain that the view in question could properly be attributed to Empedocles.

¹³ 1042b11–15.

¹⁴ ACP 97 n. 409.

¹⁵ 1042b15–1043a28.

¹⁶ 724b34–725a1.

¹⁷ Chapter 1 = vii 470 Littré.

¹⁸ Loeb edition of the De gen. anim. 78.

¹⁹ 725b4–8.

²⁰ 725a21–4.

²¹ There has inevitably been a tendency to suppose that the two factors in the act of vision were harmonised in some part of the poem now lost to us, e.g. by Beare, , Elementary cognition 19 n. 3Google Scholar, and Lackenbacher, , WS xxxv (1913) 42–3.Google Scholar But Aristotle gives a fairly clear impression, it seems to me, that the simile of the lantern was his only evidence for the notion of fire leaving the eye. Alexander makes it fairly clear that he too has taken Aristotle's words in this way, De sensu 23.8–10, cf. 24.2–3. The lack of other evidence is also indicated by the implied completeness of Theophrastus' account, cf. n. 3 above.

²² De sens. 7 (DK i 301.26–35 = Doxographi 500.19–29).

²³ On the text of this passage see note 6 p. 163 below.

²⁴ In Plato's account of Empedocles' theory of vision in the Meno 76c–d (DK 31A92) there is a twofold division. Plato first outlines the general theory of pores and effluences, 76c7–d2. He then applies this theory to the process of vision, 76d2–5. But Plato so abbreviates the application of the theory to vision that he gives no more specific account of the structure of the eye than that it is ‘symmetrical’ to effluences from the object seen.

²⁵ I have avoided calling these funnels ‘pores’. This is probably an unnecessary scrupulosity: for the function of these ‘funnels’, to keep back water and let through fire, is directly analogous to the function of the ‘furrows’ (ἄλοξιν fr. 100.3) which in the process of breathing keep back blood and let through air; in his paraphrase of fr. 100 Aristotle speaks of the furrows as ‘pores’, De resp. 473b1–5.

The composition and function of the funnels and membranes is considered further in note 6 pp. 163–6 below, where I conclude that in fact fire and water are the only percipient elements in the eye.

²⁶ Aesch. Prom. 356.

²⁷ Theoer. Id. xxiv 18–19. These and other examples, from human and non-human eyes, are quoted (in the course of a different argument) by Verdenius, , Studia Vollgraff 161–2.Google Scholar

²⁸ Bignone, 249 n. 2 and 381 n. 1, and Cherniss, , ACP 317 n. 106Google Scholar, both take the simile to explain some kind of flashing from the eye, whether as an account of the structure of the eye (Cherniss), or as an indication that the eye was made of fire (Bignone). Verdenius, Studia Vollgraff 156 n. 5 and 159, objected that this rendered the description of otiose. Since Cherniss, versions of the theory that we see by outward-flowing fire have been repeated by Verdenius, Guthrie and several other scholars, as cited in note 3 pp. 157–9 below.

Two loose suppositions could have served to attach the notion of outward-flowing fire to the act of vision. First, the fact that there are pores of the right size for fire to leave the eye through naturally implies that there are pores of the right size for fiery effluences to enter the eye through. Secondly, Empedocles may conceivably have thought that fire must leave the eye in order to make room for fiery effluences from the object seen.

In neither case would the fire which leaves the eye have acted as an organ of vision, so that both suppositions would be compatible with Theophrastus' silence and with the explanation that I offer of Aristotle's implied charge of inconsistency.

²⁹ It would be wrong to set limits to Aristotle's ingenuity, but it would perhaps be difficult to see what other grounds of argument he could have employed without resorting to dissection, which in this context would have been untypical of Aristotle's method. As it is, Aristotle does once cite an instance from the battlefield, 438b11–16.

³⁰ Tim. 77c–79e. On the element of purpose in Plato's account cf. note 7 pp. 166–9 below.

³¹ De resp. 473a15–474a24. Throughout this essay I have used the convenient periphrasis of ‘lungs and chest’, taken from the Timaeus 79C2, simply in order to avoid attributing any too detailed anatomical knowledge to Empedocles.

³² Both these suggestions on the purpose served by Empedocles' theory of breathing are intended to be speculative. They are considered further in note 7 pp. 166–9 below.

³³ References to modern interpretations of Empedocles' theory of breathing are given separately in note 8 pp. 169–71 below.

If we abandon the theory of cutaneous respiration for Empedocles, the question arises: how original is Plato's theory of respiration? This question is considered separately in note 9 pp. 171–3 below.

³⁴ As with the lantern I print the text from Ross, without intending to commit myself to the details of interpretation implied therein.

³⁵ The element of ambiguity in and ῥινός is considered further in note 10 pp. 173–6 below.

‘Partly filled with blood’ paraphrases λίφαιμοι (line 1). Aristotle writes, 473b2–3: Aristotle's qualification may be based on no more than the lines before us: the veins are not full of blood because blood moves up and down in them, periodically leaving room for the entry of air.

³⁶ The nature and workings of Empedocles' clepsydra are excellently described by Last, H., ‘Empedokles and his klepsydra again’, CQ xviii (1924) 169–73.CrossRefGoogle Scholar The clepsydra in question is a vessel with perforations at the bottom and a vent at the top. By blocking and unblocking the vent at the top, liquids can conveniently be transferred from one container to another.

Various misunderstandings connected with the workings of the clepsydra are considered separately in note 11 pp. 176–9 below.

³⁷ Furley, D. J., ‘Empedocles and the clepsydra’, JHS lxxvii (1957) 31–4.CrossRefGoogle Scholar Furley is more thorough going than most other writers in his pursuance of a Platonising interpretation for the simile. But his primary suggestion, 32, that the upper vent of the clepsydra corresponds to the nose or mouth, while the perforations correspond to pores, is not original, as both Furley himself, 31–2, and Lloyd, , Polarity and analogy 329–30Google Scholar, seem to suppose. Precisely this correlation was put forward by Winnefeld, , Philosophie des Empedokles 38.Google Scholar Before that, the same correlation had been put forward, and rejected for its deficiencies, by Lommatzsch, , Die Weisheit des Empedokles 223–4.Google Scholar

³⁸ The air which presses on the perforations from below, but which does not enter the clepsydra, is taken into account below, pp. 153–4, see also note 11 pp. 176–9 below.

³⁹ Furley, 33.

⁴⁰ The half admission is in the footnote, 33 n. 5. Such repetition is of course a regular feature of Homeric simile, cf. Fränkel, Hermann, Die homerischen Gleichnisse (Göttingen, 1921) 4–5.Google Scholar

⁴¹ Furley, 33.

⁴² Furley, 33. The observation is taken from Taylor, , Timaeus 560 Google Scholar: ‘unfortunately nothing has been left to show how Empedocles worked the mouth and nostrils into his account’.

⁴³ Furley, 33.

⁴⁴ Furley, 33. The same idea occurs in Lommatzsch, , Die Weisheit des Empedocles 224 Google Scholar: ‘denn so wie, wenn die Hauptmündung geschlossen ist, der jedesmalige Zustand der Wasserglocke unverändert bleibt, so würde dann auch dasselbige in Beziehung auf den Athmungsprocess wohl als empedocleisch gelten, nämlich bei geschlossener Nase und Mund der Athmungsprocess gleichfalls still stehen’.

We might expect Furley to argue that being gagged was explained by the clepsydra's being full of water, not full of air. But the picture he has in mind is that ‘blood cannot leave the surface of the body to make room for air, because the air cannot escape through the nose and mouth’ (p. 33). In other words, Furley supposes that the clepsydra's being full of water is equivalent to there being both blood and air in the body. It is true that the writer of the Problemata 915a4–24 (in part DK 59A69) explains the retention of water in the clepsydra by the presence of air wedged in the neck of the clepsydra. This explanation has been applied to Empedocles' clepsydra by Diels, and recently by Wilkens, see note 11 p. 176f. below. But this is not the explanation of the clepsydra's behaviour which Furley has adopted on the preceding page of his article.

⁴⁵ Furley, 34.

⁴⁶ Booth, N. B., ‘Empedocles' account of breathing’, JHS lxxx (1960) 10–15.CrossRefGoogle Scholar

⁴⁷ Cardini, M. Timpanaro, ‘Respirazione e clessidra (Empedocle fi. 100)’, La parola del passalo xii (1957) 250–70.Google Scholar

⁴⁸ Guthrie, , History ii 220–6.Google Scholar Lloyd, , Polarity and analogy 328–33.Google Scholar For Verdenius, and for Seeck's recent article, see note 8 pp. 169–71 below.

⁴⁹ Booth, 13. The equation of blood with air first appears in a very confused form in Freeman, , Pre-Socratic philosophers 195.Google Scholar Within the space of a few sentences Miss Freeman first implies the equation of blood with air, and then implies the opposite equation, of blood with water and of air with air.

The explicit equation of blood with air and of air with water is also made in the course of some very brief remarks by Webster, T. B. L., ‘From primitive to modern thought in ancient Greece’, Acta congressus Madvigiani = Proceedings of the second international congress of classical studies ii (Copenhagen, 1958) 35.Google Scholar

Bollack equates both air in the clepsydra and water with air in breathing, while blood, he thinks, is represented by the girl's hand, Empédocle i 244, see further note 11 pp. 176–9 below.

The equation of blood and air is already beginning to breed its own mythology. In Studi Torricelliani 155–6 Timpanaro Cardini writes that the equation shows ‘come Empedocle avesse osservato il funzionamento della clessidra senza un' interpretazione preconcetta’.

⁵⁰ Booth, 12–13.

⁵¹ It is in favour of Booth's interpretation (although he does not take up the point) that the two descriptions of aether (line 7) and (line 24), contain words commonly used of a liquid, see LSJ s.vv.

On the other hand, τέρεν is used three times, twice of blood (lines 6 and 22) and once of water (line 11). This tells, if only very slightly, in favour of the other correlation, of water with blood.

In fact I should be loth to lean at all heavily on these slight similarities of language. For example, eyes are ἀτειρέα in fr. 86. Fire or light flows from the lantern fr. 84.6. But I do not take the repetition of die adjective as an indication that we see by fire flowing from the eye.

⁵² Timpanaro Cardini, 257 and 269–70, see also Studi Torricelliani 155–6. Lloyd, , Polarity and analogy 310–1.Google Scholar

⁵³ Note 11 pp. 176–9 below.

⁵⁴ The same verb, τετρήατο (a virtually certain emendation), is used for the ‘wonderful funnels’ in the eye, fr. 84.9.

It should be noted that πυκ(ι)νός, the adjective applied to pores and perforations in fr. 100, is also once used of air (line 14). This diminishes perhaps, if only very slightly, the idea that the word is intended to indicate a parallelism of pores and perforations.

⁵⁵ This point, which is obscured in Lloyd's account, is considered further in note 11 pp. 176–9 below.

⁵⁶ Furley, , JHS lxxvii (1957) 32.Google Scholar

⁵⁷ Polarity and analogy 330–1.

⁵⁸ Harry, and Thornton, Agathe, Time and style, a psycho-linguistic essay in classical literature (London, Methuen, 1962) 23.Google Scholar

⁵⁹ The different directions in which air moves in the clepsydra seem to puzzle Otto Regenbogen, , ‘Der Klepsydravergleich des Empedokles’, Beilage iv of ‘Eine Forschungsmethode antiker Naturwissenschaft’, first published in Quellen und Studien zur Geschichte der Mathematik Astronomie und Physik, Abteilung B Studien, Band i (Berlin, 1931) 181 Google Scholar, reprinted in Kleine Schriften (München, 1961) 193.

Timpanaro Cardini runs two points together, La parola del passato xii (1957) 257 sub finem. First, air moves in opposite directions. Secondly, it is the same air which keeps water out of the clepsydra and which then leaves the clepsydra; while the air which keeps water inside the clepsydra is different from the air which then enters through the neck of the clepsydra.

It would of course be possible to remove this second anomaly by attributing to Empedocles a theory of ἀντιπερίστασις, whereby the air which pressed on the perforations from below moved around and entered the clepsydra from above.

⁶⁰ These two movements of air, and the whole difference between pores in the body and perforations in the clepsydra, are considered further in note 11 pp. 176–9 below.

⁶¹ Kranz, W., ‘Gleichnis und Vergleich in der frühgriechischen Philosophie’, Hermes lxxiii (1938) 100–9.Google Scholar

⁶² Snell, Bruno, Die Entdeckung des Geutes, Studien zur Entstehung des europäischen Denkens bei den Griechen, 3rd edn (Hamburg, 1955) 284–98.Google Scholar

⁶³ References to other studies may be found in Bibliographie zur antiken Bildersprache unter Leitung von Viktor Pöschl, bearbeitet von Helga Gärtner und Waltraut Heyke (Heidelberg, 1964) 150–2.

⁶⁴ Snell does have some remarks pointing in this direction, 286–7, but his purpose is to contrast Empedocles and Homer.

⁶⁵ I think it is improbable that Empedocles' earth rotates, cf. ECC 52 n. 3. Even if it does, it will not be the same as the water in the κύαθοι, for except on a Pythagorean system the earth would rotate in the same place, while water in the κύαθοι is carried round in a circle.

⁶⁶ Pp. 150–1 above

⁶⁷ Plutarch speaks explicitly of Empedocles' having connected curdling with tears, Quaest. nat. 917a (DK 31A78). But tears do not match happily the emphasis on hardening in ἐγόμφωσεν καὶ ἔδησε.

Hardening could be accounted for in Empedocles' embryology, for Aristotle says that Empedocles explained the sterility of mules by the mixture of male and female semen being too hard, ‘like copper mixed with tin’, De gen. anim. 747334–b10 (in part DK 31B92). Now Aristotle several times himself compares the action of fig-juice or rennet on milk with the effect of male sperm on matter provided by the female, De gen. anim. 729a9–14, 737a12–16, 739b20–6, 771b18–27, 772a22–5. This analogy therefore, although it is not attributed to Empedocles by name, may perhaps provide a better context for fr. 33.

Both applications of the simile are mentioned (with less evidence) by Zeller, ZN 991 n. 2.

⁶⁸ For ἔδησε the manuscripts also have ἔπηξε, Plut. De amic. mult. 95a.

With ἐγόμφωσεν cf. γόμφοις fr. 87. Other metallurgical images are:

(i) copper and tin in fr. 92, mentioned in the preceding footnote.

(ii) χόανοι ‘hollows for melting metal’ in fr. 96.1, cf. fr. 84.9.

(iii) the mixing of four elements compared to the mixing of four metals, Galen, , Hippocratis de natura hominis i 2 Google Scholar = xv 32 Kühn (DK 31A34).

(iv) the comparison of stars with nails in Aetius ii 14.3 (DK 13A14). (For the attribution to Empedocles, see JHS lxxxviii [1968] 117 n. 25: the mention of nails indicates that πέταλα in the next entry may be metal plates, and not, as is usually assumed, leaves.)

(v) perhaps the comparison of hot rivers or springs with some kind of underground heating system, Seneca, Quaest. nat. iii 24.1–2 (DK 31A68). (Only perhaps—for it is possible to read the passage as though the comparison were Seneca's own.)

There may conceivably be a secondary metallurgical connotation in κολλήσας ‘welding’ in fr. 34 (cf. κόλλησιν. 96.4) and in ‘hammered’ or ‘beaten out’ in fr. 30.3.

⁶⁹ Arguments for taking these two fragments together are listed by Bignone, 427–8. Love's formation of animal parts at the beginning of her zoogony (for which see ECC 200–3) provides the simplest context for the fragment.

⁷⁰ The manuscripts have λοχάζετο and ἐχεύατο, Arist. De sensu 438a1. Λοχεύσατο is Förster, A.'s emendation, ‘Empedocleum’, Hermes lxxiv (1939) 102–4.Google Scholar

⁷¹ See note n pp. 176–9 below.

⁷² CQ n.s. xvi (1966) 263.

⁷³ Apuleius, , Apologia 15 Google Scholar (DK 47A25).

⁷⁴ Cherniss, , ACP 317 n. 106Google Scholar, infers Plato's originality solely from the reference to Archytas. But there is little need for Apuleius' report to carry this implication, unless it is taken in conjunction with the passage from Theophrastus.

⁷⁵ Cf. p. 145 n. 28 above.

⁷⁶ Cf. p. 145 n. 28 above.

⁷⁷ Aristotle, , De sensu 437b10–438a5.Google Scholar This attitude is exemplified by Siebeck, Geschichte der Psychologie i 1 p. 270, and by Beare, , Elementary cognition 17 n. 4.Google Scholar

⁷⁸ See for example Arist. De sensu 438a12 (DK 68A121), Alexander, De sensu 24.19 and 22, 56.12 (in part DK 67A29), and Aet. iv 13.1 (DK ibid.).

⁷⁹ Lur'e, supposes that Empedocles here anticipates the Atomists, Essays in the history of ancient science 76 Google Scholar = Luria, , Anfänge griechischen Denkens 85.Google Scholar

⁸⁰ De sensu 23.5–24.9, especially 23.8–10 and 24.2–9.

⁸¹ Aet. iv 13.5 (not in DK). [Galen, ] Historia philosopha 94 Google Scholar (not in DK = Doxographi 636).

⁸² Studia Vollgraff 156.

⁸³ Aet. iv 13.9 (cf. DK 28A48).

⁸⁴ The whole passage runs from De gen. et corr. 153.22–154.2. For the technical use of the verb προσβάλλειν see Mugler, Charles, Dictionnaire historique de la terminologie optique des grecs, in Études et commen taires liii (1964) s.v.Google Scholar

⁸⁵ De gen. et corr. 324b26–35 (DK 31A87). Joachim, in his edition of the De gen. et con. 157–8, takes the passage in the way that Philoponus has done. But there is no ground for this interpretation in Aristotle's text.

⁸⁶ Examples are given, ECC 203, 207–9, 212–13.

⁸⁷ [Philoponus] De gen. anim. 217.13–25 (not in DK). Cf. Arist. De gen. anim. 779b15–20 (DK 31A91).

⁸⁸ For the context of this passage see p. 140 above.

⁸⁹ J. I. Beare, Oxford translation of the Parva nat. ad loc.

⁹⁰ Ross, W. D., edition of the Parva nat. 188.Google Scholar

⁹¹ ACP 316.

⁹² Magnus, , Augenheilkunde der Alten 97.Google Scholar Ross, G. R. T., edition of the De sensu and De memoria 134.Google Scholar Taylor, , Timaeus 279.Google Scholar Siegel, , Bulletin of the history of medicine xxxiii (1959) 147.Google Scholar

⁹³ Polarity and analogy 326 n. 1.

⁹⁴ SirNewton, Isaac, Opticks: or, a treatise of the reflections, refractions, inflections and colours of light 4th edn (London, 1730) 321–2Google Scholar (= Book iii Query 16).

Johannes Müller, Handbuch der Physiologie des Menschen: of the fourth and latest edition I have been able to obtain only the French translation by Jourdan, A. J. L., Manuel de physiologie ii (Paris, 1845) 253–9.Google Scholar

Helmholtz, Hermann L. F., Handbuch der physiologischen Optik 3rd edn ii (Hamburg und Leipzig, 1911) 6–11 Google Scholar, cf. 19: translated as Helmholtz's treatise on physiological optics ii (Menasha, Wisconsin, 1924) 5–11, cf. 20.

Young, Thomas, ‘Observations on vision’, Philosophical transactions of the Royal Society of London 1793 ii 178–80.Google Scholar

Purkinje, Johann, Beobachtungen und Versuche zur Physiologie der Sinne Bändchen i Beiträge zur Kenntniss des Sehens in subjektiver Hinsicht (Prag, 1819 Google Scholar: ‘zweite unveränderte Auflage’ Prag, 1823) 176 pages: conveniently available in Purkyně, Jan E., Sebrané spisy = Opera omnia i (V Praze, 1918) 1–56.Google Scholar

Further references to literature on the subject may be found in these works.

⁹⁵ This is a rather less common experience than the other two, and the point about the eyes being closed is not stated quite as explicitly as I would have wished in the works I have quoted; but Professor Sir Vincent Wigglesworth informs me that in his own experience the point is as I have stated it.

⁹⁶ 437a29–b9.

⁹⁷ Young, 178 (cited above n. 94), speaks of the eye being ‘rubbed or compressed’. But I think it is clear that this means more or less the same as Newton's reference, quoted in the preceding paragraph, to moving one's finger ‘with a quavering Motion’.

It should also be noted that ‘rubbing the eyes’, in the conventional sense, usually at least produces no more than faint blobs of light, which are much less vivid than the effect described as the result of pressing the eye.

⁹⁸ Doxographi 500.24.

⁹⁹ SBB (1884) 354 n. 2.

¹⁰⁰ Poetarum philosophorum fragmenta 99.32 (repeated in Diels-Kranz i 301.31).

¹⁰¹ Karsten, 484, has Panzerbieter has precisely the same text as Diels, , ‘Zu Empedokles’, Zeitschrift für die Altertumswissenschaft iii (1845) no. 111 col. 883.Google Scholar

¹⁰² EGP 246 n. 2.

¹⁰³ CQ n.s. xvi (1966) 262 n. 2 and 263 n. 2.

¹⁰⁴ CQ n.s. xvi (1966) 262 and 264, cf. 263 n. 5.

¹⁰⁵ Studia Vollgraff 155, cf. 163.

¹⁰⁶ See LSJ s.v. ὁράω and Stephanus s.v. ὄπτω.

¹⁰⁷ Aristotle: De anima 404b8–15, Met. 1000b3–9. Theophrastus: De sens. 10 (DK i 302.21–2 = Doxographi 502.9–10).

¹⁰⁸ This point is made by Bignone, 372 n. 1, 476.

¹⁰⁹ Aet. i 15.3 (DK 31A92). Long, , CQ n.s. xvi (1966) 264 n. 1.Google Scholar

¹¹⁰ Winnefeld, , Philosophie des Empedokles 42–3.Google Scholar Prantl, Von, Aristoteles über die Farben 41–2.Google Scholar Kranz, W., ‘Die ältesten Farbenlehren der Griechen’, Hermes xlvii (1912) 126–8Google Scholar, cf. Hermes xlvii (1912) 41–2 and Empedokles 61. Siegel, , Bulletin of the history of medicine xxxiii (1959) 152–3Google Scholar (where 152 n. 31 is misplaced, and 31B32 should read 31B23).

¹¹¹ Doxographi 222.

¹¹² De sens. 17, 59, 73, 76, 79 (DK 31A86, 68A135).

¹¹³ This confusion can be seen at work in all the writers cited above in n. 110. Kranz starts off by accepting Theophrastus' testimony, but he abandons it in effect in the course of his exegesis.

¹¹⁴ De sens. 7 (DK 31A86), quoted above p. 144.

¹¹⁵ Michael Ephesius remarks in passing that the organ of vision for Empedocles is made of the four elements [Philoponus], De gen. anim. 217.13–14 and 17. But he seems to think that fire alone is the active element in vision, 217.14–16.

¹¹⁶ The description of air outside the eye as may seem impossibly exaggerated. But Theophrastus, in his account of Empedocles, does once use ὕδατος for the dark air of night-time, De sens. 8 (DK i 302.6 = Doxographi 501.8).

Alexander is evidently led to his interpretation by taking the lantern to equal the whole of the eye, so that whatever is outside the lantern must represent whatever lies outside the eye: contrast the interpretation which I offer below.

Verdenius, , Studia Vollgrqff 159–60Google Scholar, rightly compares Empedocles' with the report on Alcmaion in Theophrastus, , De sens. 26 Google Scholar (DK 24A5): If, as seems most likely, Alcmaion's water is inside the eye, then this is a powerful argument against Alexander's view. Unfortunately, it is possible to take Alcmaion's water as being outside the eye: this is the view of Taylor, apud Stratton, George M., Theophrastus and the Greek physiological psychology before Aristotle (London and New York, 1917) 176 Google Scholar, expressed more cautiously, Timaeus 282. Since disagreement is possible, I have thought it best not to use Alcmaion's theory as evidence here for Empedocles.

¹¹⁷ Magnus, , Augenheilkunde der Alten 97.Google Scholar Taylor, , Timaeus 280 n. 1Google Scholar, cf. 277 and 282. Lloyd, , Polarity and analogy 326.Google Scholar Also Frenkian, , Études ii 59.Google Scholar

¹¹⁸ Cf. 78A2–6.

¹¹⁹ Panzerbieter, , ZAW iii (1845) no. 111 coll. 883–4.Google Scholar Diels, , SBB (1884) 354.Google Scholar Burnet, , EGP 248.Google Scholar Lacken, bacher, WS xxxv (1913) 39–40.Google Scholar This also seems to be the interpretation of Winnefeld, , Philosophie des Empedokles 41–2.Google Scholar

It is not possible to determine with certainty the view of those who simply translate ‘keep out’, e.g. Millerd, 83, Ross, , Parva nat. 190 Google Scholar, Guthrie, , History ii 235.Google Scholar

¹²⁰ De sens. 7 (DK 31A86): quoted above 144.

¹²¹ Polarity and analogy 326.

¹²² For this correlation of two different elements, wind and water, see p. 155 above.

¹²³ P. 146 above.

¹²⁴ Timaeus 569.

¹²⁵ Parva nat. 312.

¹²⁶ Tim. 77c8–9, 78e3–5, cf. 70c–d and 80d.

¹²⁷ Tim. 79b1, c1, cf. 80C3.

Aristotle mentions Plato's avoidance of a vacuum, 472D16; but he does not of course count this as a final cause.

He also considers, and rejects, Plato's theory that breathing is τροφῆς χάριν, 473a3–14.

It is true that in Plato's account of breathing the element of purpose is not given nearly as much prominence as it is in Aristotle. This, and the inadequacy of Plato's account in Aristotle's eyes, lead to the exaggeration that on the question of the final cause in breathing Plato and his followers οὐθέν εἰρήκασιν.

Aristotle also complains of the lack of a final cause at the conclusion of his criticisms of Anaxagoras and Diogenes, 471b23–9, and at the beginning of his account of Democritus, 472a1–3.

¹²⁸ Gilbert, Otto, Die meteorologischen Theorien des griechischen Altertums (Leipzig, 1907) 343–4Google Scholar, cf. 339 and 380–3.

¹²⁹ Longrigg, J., ‘Empedocles's fiery fish’, Journal of the Warburg and Courtauld Institutes xxviii (1965) 314–15.CrossRefGoogle Scholar

¹³⁰ philistion: Galen, , De usu respirationis 1 Google Scholar = iv 471 Kühn = Wellmann, , Sikelischen Ärzte 112 Google Scholar (where the word ἀνάψυξις is missing).

Philolaus: Anonymus Londinensis xviii 8–29 (DK 44A27).

Hippon: Arist. De anima 405b24–9 (DK 38A10). The attribution is from Philoponus, De anima 92.2–11 (DK ibid.). The etymology which Aristotle alludes to is made explicit without attribution, in the Cratylus 399d–e (not in DK).

Diogenes of Apollonia: Aet. v 15.4 (DK 64A28).

The same association of breathing and cooling occurs in two Hippocratic treatises: 4 = vi 368 Littré, , and περὶ καρδίης 5 Google Scholar = ix 84 Littré.

In the περὶ σαρκῶν 5–6 = viii 590–4 Littré, πνεῦμα feeds the heart. This implies cooling, since the author remarks both that the heart is and that

In the περὶ φυσῶν 7–8 = vi 98–104 Littré, an excessive amount of breath taken into the body with food and drink cools the blood and causes shivering and fevers.

Galen attributes the association of breathing and cooling jointly to Plato, and to ‘Hippocrates’, De Hippocralis et Piatonis placitis viii 9 Google Scholar = v 713 Kühn.

¹³¹ Longrigg is wrong to add as evidence for the association of breathing and cooling the passage from Theophrastus, , Hist. plant. v 9.6 Google Scholar (DK 32A3):

Longrigg, 315, interprets this as meaning that ‘ivy is fiery and … has the fastest and most copious rate of respiration’. But it is at once evident from the context (not supplied by Diels-Kranz) that πυρεῑα here is ‘kindling’, and that the verb ἀναπνεῑ has the sense of burning, or as we might say of ‘drawing up’ (see LSJ s.v.).

From the fact that it makes good kindling it does of course follow for Menestor that ivy is fiery, De caus. plant. i 21.5–7 (in part DK 32A5), cf. i 22.5 (not in DK) and Hist, plant. v 3.4 (DK 32A3a). But there is no mention of breathing: at De caus. plant. i 21.7 is represented as

Longrigg's primary reason for attributing the idea of breathing and cooling to Empedocles is the report that Empedocles spoke of fish moving to a cool element in order to counteract an excess of internal heat. This and the contrary notion, that birds have a lot of fire and move upwards through the attraction of like for like, seem to me to be best explained as part of two zoogonies in the cosmic cycle, see ECC 189–95.

¹³² Empedocles did allow for changes of temperature in the blood, for sleep is the result of a partial cooling of the blood, Aet. v 24.2 (DK 31A85), cf. v 25.4 (DK ibid.). Theophrastus' two kinds of unintelligence, De sens. 11 (DK 31A86), are also to be explained, I think, in terms of a difference of temperature, as well as of texture, in the blood. (I intend to develop this point in a future article.)

¹³³ iv 22.1 (DK 31A74). v 15.3 (not in DK).

¹³⁴ Burnet, , ‘L'expérimentation et l'observation dans la science grecque’, Scientia (=Rivista di scienza = Rivista internazionale di sintesi scientifica) vol. xxxiii anno 17 (1923) 94–5Google Scholar, cf. EGP 27, 228–9, 266–7. Farrington, Benjamin, Science in antiquity, in die Home University Library series (London, 1936) 76–8Google Scholar, and Greek science, its meaning for us (Thales to Arstotle) in the Pelican series (Harmondsworth, Middlesex, 1944) 51–3. Lloyd, , Polarity and analogy 331–2.Google Scholar

¹³⁵ 309a19–21 (DK 59A68), cf. 305b16–18 (DK 68A46a).

¹³⁶ 309a16–21 (in part DK 39A68). Cf. [Arist.] Probl. 914b9–915a24 (in part DK 59A69), where again Anaxagoras' name alone is mentioned.

¹³⁷ 294b13–30 (in part DK 13A20).

¹³⁸ This is of course not quite the same as the observation described in the later passage of the De caelo, for there it is air trapped inside the clepsydra which is relevant, while here the idea appears to be that the air outside the clepsydra prevents the heavier element, water, from falling through the perforations, in the same way that air, allegedly, prevents the earth from falling.

¹³⁹ De sudore 25–6, repeated in an abbreviated form and without Theophrastus' name in [Arist.] Probl. 866b9–14. There is the opposite theory in the περὶ διαίτης ii 64 = vi 580 Littré.

¹⁴⁰ E. S. Forster, Oxford translation of the Problematu ad loc.

¹⁴¹ Polarity and analogy 331–2.

¹⁴² References to works that have already been cited in note 3 pp. 157–9 above are given here in an abbreviated form.

¹⁴³ Cf. p. 150 n. 49 above.

¹⁴⁴ Hermes xcv (1967) 50–2.

¹⁴⁵ In quoting from this work I have transcribed the text from Diels, Supplementum Aristotelicum iii (Berolini, 1893), without distinguishing the additions made by Diels to the original text of the papyrus.

The στρατιώτης is spoken of also in Pliny, , Nat. hist. xxiv 18.105 § 169Google Scholar, in Dioscorides, , De materia medica iv 101 Google Scholar = ii 256.5–257.5 Wellmann, and in Galen, , De simplicium medicamentorum temperamentu ac facultatibus viii 40 Google Scholar = xii 131 Kühn.

Two alternative identifications are offered by Lewis and Short, s.v. ‘stratiotes’: the straliotes abides, water aloe or water soldier, and the lemna polyrrhiza, or greater duckweed (which they appear to confuse with the pistia stratiotes, mentioned below). A comparison with duckweed is offered also by W. H. S. Jones, to illustrate the passage from the Londinensis, Anonymus, The medical writings of Anonymus Londinensis (Cambridge, 1947) 39.Google Scholar The identification with the water soldier is offered by several other writers, in particular Max Pohlenz, who concludes that the fact that the leaves of the water soldier ‘liegen nicht flach auf, sondern sind steil emporgerichtet und recken sich jedenfalls zur Blütezeit in die Luft empor’ is intended as an indication that ‘der Mensch seinen Geist erst dann voll entfaltet, wenn er sich über die feuchten Regionen des Bodens in die reine Luft erhebt’, Hippokrates und die Begründung der wissenschaftlichen Medizin (Berlin, 1938) 73.

Neither identification is likely to be correct.

(i) The water soldier. This is described as having leaves with ‘teeth and points very sharp’, in Sowerby, James, English botany vi (London, 1797) tab. 379.Google Scholar The leaves would hardly have been suitable therefore as a cold compress, the use prescribed by Pliny, Dioscorides and Galen. It is a further disadvantage that in Europe the water soldier is rare in the southern part of the continent.

(ii) The greater duckweed. According to Pliny and Dioscorides, the stratiotes has leaves like the sempervivum, but larger. The comparison is probably with the sempervivum tectorum, or common houseleek, for according to the commentary in Sowerby this toe was used for cold compresses, xix tab. 1320: ‘The bruised leaves are by rustic surgeons used as a cooling external application, but their virtues are inconsiderable’. Other haemostatic and curative properties attributed to the common houseleek in Syme, John T. Boswell, English botany 3rd edn iv (London, 1865) 61 Google Scholar, are similar to those claimed for the stratiotes by Pliny, Dioscorides and Galen. The leaves of the greater duckweed are from ¼ to ⅝ inch across, according to Syme, ix 24. The leaves of the common houseleek are more than twice as large as this, Syme iv 61.

The stratiotes is identified with the pistia stratiotes or water lettuce by Sprengel, Kurt, Geschichte der Botanik ‘neu bearbeitet’ i (Altenburg und Leipzig, 1818) 155 Google Scholar, by LSJ s.v., and by Gilbert-Carter, Humphrey, Glossary of the British flora 3rd edn (Cambridge, 1964) 79.Google Scholar (I owe this last reference to the kindness of Di S. M. Walters of the Cambridge Botany School.)

This identification may well be right. The water lettuce floats on the water and has leaves larger than the common houseleek. It also approximates to Pliny's condition, ‘in Aegypto tantum et inundatione Nili nascitur’, for its presence in the Upper Nile at least is noted by Hope, C. W., ‘The “Sadd” of the Upper Nile: its botany compared with that of similar obstructions in Bengal and American waters’, Annals of botany xvi (1902) 495–516 CrossRefGoogle Scholar, especially 506. Pliny's inundatione may indeed reflect the flooding caused by accumulation of vegetation called the ‘sudd’ or ‘sadd’.

The water lettuce is described and illustrated in (Curtis's) Botanical magazine lxxvii, series 3 vii (1851) tab. 4564. There are a couple of fine specimens in the Cambridge botanical gardens.

¹⁴⁶ Deichgräber, K., ‘Die Epidemien und das Corpus Hippocraticum, Voruntersuchungen zu einer Geschichte der Koischen Ärzteschule’, Abhandlungen der Preussischen Akademie der Wissenschaften, philosophisch-historische Klasse (Berlin, 1933) no. iii 154–5.Google Scholar Pohlenz, , Hippokrates 71–4.Google Scholar

¹⁴⁷ Especially chapters 3–4 and 7 = vi 366–8 and 372–4 Littré.

Pohlenz comes close to attributing cutaneous respiration to the He writes, Hippokrates 71–2: ‘Bei der Atmung spricht die Schrift über die Heilige Krankheit (Kap. 7) freilich nur vom Mund und Nase als den Hauptwegen; aber das geschieht im beiläufiger Erwähnung und schliesst die Hautatmung “durch den ganzen Leib” … keineswegs aus.’

¹⁴⁸ This is also the view of Guthrie, , History ii 223.Google Scholar It was of course also Wellmann's view, but joined in his thesis to a theory of cutaneous respiration for Empedocles, , Sikelichen Ärzte 70–1.Google Scholar

¹⁴⁹ Die Epidemien und das Corpus Hippocraticum 74–5.

¹⁵⁰ The sentiment in question is expressed on pp. 140 and 146–7 above.

I prefer not to rely on ProfessorGuthrie, 's suggestion, that Plato is following Empedocles in his avoidance of void, History ii 223–4Google Scholar, for the addition of this feature to Empedocles' theory can be only speculative, see pp. 166–9 above.

¹⁵¹ Pp. 146–7 above.

¹⁵² Polarity and analogy 328–30, especially 329 n. 2.

¹⁵³ Parva nat. 124.14–127.8.

¹⁵⁴ Parva nat. 125.19–22, cf. 4–9.

¹⁵⁵ Parva nat. 124.18, 125.4–5 and 19.

¹⁵⁶ Pp. 146–7 above.

¹⁵⁷ Il. v 291, xiii 616. Od. iv 445, xviii 86.

¹⁵⁸ Il. xiv 467, xvi 349, 503, xix 39, xxiii 395, 777. Od. v 456, xxi 301, xxii 18, 475, xxiv 318.

¹⁵⁹ Il. v 308. Od. xiv 134, xxii 278.

¹⁶⁰ Il. vii 248, x 155, 262, 334, xvi 636, xx 276. Od. v 281, xii 423.

¹⁶¹ Il. iv 447, vii 474, viii 61, xii 263, xiii 406, 804. Od. i 108, xii 395.

¹⁶² Seeck, , Hermes xcv (1967) 49.Google Scholar Cf. Cardini, Timpanaro, La parola del passato xii (1957) 259 n. 2.Google Scholar

¹⁶³ Karsten, 248, quoted in support of ῥινῶν meaning skin Il. xix 39, where, to preserve Patroclus' corpse, Thetis pours ambrose and nectar κακὰ ῥινῶν. But the meaning here is nostrils, cf. Herodotus ii 86.

¹⁶⁴ It is interesting to note that some confusion between skin and nose seems to have arisen in ancient times. In a gloss on Il. xiii 616, ῥινὸς ὑπὲρ πυμάτης, which clearly means ‘above the bridge of the nose’, Apollonius Sophistes wrote Lexicon graecum Iliadis et Odysseae s.v. ῥινός.

¹⁶⁵ iv 22.1 (DK 31A74).

¹⁶⁶ v 15.3 (not in DK).

¹⁶⁷ Sikelischen Àrzte 72.

¹⁶⁸ Plato's cosmology 319 n. 1, cf. 306–7.

¹⁶⁹ Empédocle i 242.

¹⁷⁰ Millerd, 72. Bignone, 359 n. 3. Booth, , JHS lxxx (1960) 14.Google Scholar

¹⁷¹ Apart from the Homeric passages cited above, ῥινός is moderately common, as both singular and plural, in poetry. It does not seem to occur at all in prose.

The meaning of ‘skin’ or ‘skins’ for ῥῑνες is not merely unfamiliar, as Professor Guthrie observes in his note on this passage from Aetius, History ii 223 n.3. It is unknown.

¹⁷² Karsten, 479 n. 275. Panzerbieter, , ZAW iii (1845) no. 111 col. 886.Google Scholar Reiche, Harald A. T., Empedocles' mixture, Eudoxan astronomy and Aristotle's connate pneuma (Amsterdam, 1960) 67–9.Google Scholar

Diels started off by agreeing with Karsten, , Doxographi 411.Google Scholar But he later thought better of it, Poet. phil. fragm. 96.20, repeated in DK i 298.9.

¹⁷³ For these two features of Empedocles' zoogonical theory, cf. ECC 200–3 and 50, 209–10.

Diels, , Poet. phil. fragm. 96.21 Google Scholar, repeated in DK i 298.10, rightly notes that is used by Aetius in a zoogonical context in v 7.1 (DK31A81).

Reiche, Empedocles' mixture 67–9, argues that Empedocles cannot have spoken of ‘the first breath of the first animal’, apparently on the ground that the whole-natured creatures were the first animals, and these, Reiche supposes, had no air in their composition.

In fact there seems to me no good reason for supposing that air was missing from the composition of the οὐλοφυῆ, see ECC 203–4 and 206.

It is true that whole-natured creatures had no voice (fr. 62.8), and it may be that they did not breathe. If so, then ‘the first breath of the first animal’ would simply be intended to refer to the first breathing animal, i.e. to the first animal of the kind that we know now. An abbreviation of this kind would be entirely natural in a doxographical compilation.

¹⁷⁴ 640b4–15.

¹⁷⁵ On this passage cf. ECC 213, and for the verses forged to match this context, ECC 346.

¹⁷⁶ A somewhat similar process for the formation of ‘channels of air’ may be found described in the περὶ διαίτης i 9 = vi 484 Littré.

It is unfortunately not wholly clear whether in Aetius the liquid which withdraws is (i) the amniotic fluid, which on birth fills the mouth and nostrils, and most of which leaves the body as soon as pulmonary respiration begins, or (ii) mucenum, which at birth fills the lower part of the ileum and the whole of the great intestine, and which is passed out of the body during the first three or four days after birth, which is also about the time that the lungs take to become fully distended. If only the former, which admittedly seems more probable, then the parallel with the passage in Aristotle is less exact, for in Aristotle the fluid must presumably pass down through the body in order to fashion the belly.

¹⁷⁷ Cf. p. 148, n. 36 above.

The difficulties inherent in the earlier confusion of Empedocles' clepsydra with a water-clock are well exemplified by Taylor, , Timaeus 554–5Google Scholar, and by Powell, J. U., ‘The simile of the clepsydra in Empedocles’, CQ xvii (1923) 172–4.CrossRefGoogle Scholar The confusion goes back at least as far as Dionysius Petavius, whose attempt to distinguish Empedocles' clepsydra from a water-clock is only partially successful, in his edition of Synesius (Lutetiae, 1612) Motae 21–2.

A useful collection of texts on both kinds of clepsydra, with many illustrations, is provided by Schmidt, Max C. P., Kulturhistorische Beiträge zur Kenntnis des griechischen und römischen Altertums Heft ii Die Entstehung der antiken Wasseruhr (Leipzig, 1912) 84–113.Google Scholar Unfortunately Schmidt's own comments, 24–30, on fr. 100 consist of a lengthy and really rather ridiculous attempt to show that Empedocles' clepsydra was used as an egg-timer.

The chief texts which describe a clepsydra of the kind in use in fr. 100 are as follows:

[Arist.] Probl. 914b9–915a24 (in part DK 59A69).

Hero, , Opera i Pneumatica et automata i 7 Google Scholar = 56.12–60.3 Schmidt.

Byzantinus, Philo, De ingeniis spiritualibus 11 Google Scholar = 480.21–482.15 Schmidt.

Aphrodisiensis, Alexander, Probl. phys. i 95 Google Scholar = i 33.6–15 Ideler.

Simplicius, De caelo 524.17–525.4, Phys. 647.26–30.

Two scholia on Aristotle, 's De caelo, printed in Aristotelis opera ed. Academia regia Borussica iv (Berolini, 1836) 506b17–22 and 23–43.Google Scholar

There is also a competent short account of Empedocles' clepsydra by Michael Ephesius Parva nat. 123.24–124.11 (reading at 124–2), cf. 125.25–126.14.

In the twelfth century a similar device was used for washing one's hands under. It is described by Adelard, of Bath, , Quaest. nat. 58.Google Scholar

¹⁷⁸ Loeb edition of the De caelo 226–9.

¹⁷⁹ Taylor, , Timaeus 554.Google Scholar

¹⁸⁰ Opera i Pneumatica et automata i 7 = 56.15–16 Schmidt.

¹⁸¹ Presocratic philosophers 341, 342 n. 1.

¹⁸² Polarity and analogy 331.

¹⁸³ To suppose, as does Guthrie, , History ii 222 Google Scholar, that a certain amount of air follows the water through the strainer, when the clepsydra is being emptied, would be possible perhaps, but fanciful.

¹⁸⁴ Polarity and analogy 331.

¹⁸⁵ Studi Torricelliani 156, cf. La parola del passato xii (1957) 257 and 269–70.

¹⁸⁶ It is true that Aetius uses verbs which denote pressure in his account of Empedocles' theory of breathing: ὑπαναθλίβοντος and ἀναθλίβοντος, iv 22.1 (DK 31A74). This has perhaps helped to mislead Lloyd.

In Aristotle's account, De resp. 473b1–8, the verbs are again (as in Empedocles) simply verbs of movement: and ἰόντος of blood, εἰσρεῑν and ἐκπίπτειν of air. Only ἐκπίπτειν (as ἐμπίπτοντος in Empedocles) might perhaps denote pressure.

¹⁸⁷ For this reason Aristotle isolates only two factors as required for Empedocles' account of the process of breathing: the movement of blood, and the presence of pores, De resp. 473b1–8. There is no mention of pressure, because pressure is required solely for the workings of the clepsydra.

¹⁸⁸ In criticising Lloyd in this way I am conscious that I may be attributing to him too careful and deliberate a distinction between pressure and movement. But if we consider the two halves of the simile simply in terms of movement, then the alleged comparability of air with blood and of water with air seems to me not at all clearly marked, certainly not sufficiently well marked to be able to oust the obvious comparison of air with air and of blood with water, cf. pp. 150–1 above.

¹⁸⁹ Empedócle i 244. There is essentially the same idea in Lommatzsch, , Die Weisheit des Empedokles 223 Google Scholar: ‘Bei der empedocleischen Vergleichung selbst nun entspricht … das Spiel des Mägdleins, welche die Wasserglocke einsenkt und wieder hervorhebt, der ein und ausstrebenden Kraft des Blutes selbst’.

¹⁹⁰ De resp. 473b5–6.

¹⁹¹ The question of movement caused by fire in the blood has already been considered in note 7, pp. 166–8 above.

¹⁹² Empédocle i 244.

¹⁹³ Wilkens, K., ‘Wie hat Empedokles die Vorgänge in der Klepsydra erklärt? Bemerkungen zur Fragment B 100’, Hermes xcv (1967) 129–40.Google Scholar

¹⁹⁴ 915a4–24 (in Part DK 59A69).

¹⁹⁵ Poet. phil. fragm. addenda 270. Wilkens finds Diels' note on fr. 100 ‘ganz unverständlich’ 133 n. 2. He has evidently not consulted the addenda.

¹⁹⁶ Wilkens makes this second point, 133, but he does not note in this connexion the stronger point, the usage in the Problemata.

¹⁹⁷ Pp. 153–4 above.

¹⁹⁸ Wilkens, 133. This argument comes initially from Karsten, 252. It was repeated by Powell, , CQ xvii (1923) 174.Google Scholar It seems to be echoed in Regenbogen, , Quellen und Studien i 182 Google Scholar = Kleine Schriften 194, from whom Wilkens has taken it.

¹⁹⁹ CQ xviii (1924) 173.

²⁰⁰ See note 1, p. 157 above.