¹ This article is a slightly expanded version of a communication read at the meeting of the Classical Association of Canada in Vancouver, June 1965.

² For the size and mobility of the Sicilian population, compare the operations undertaken by Gelon at Syracuse and Megara Hyblaea (Dunbabin, T. J., The Western Greeks, [Oxford 1948] 416–17Google Scholar); Hiero's treatment of the Catanaeans (Diodorus xi 49); Theron's of the Himeraeans (ibid.); and the somewhat earlier events at Zankle (Herodotus vi 22–3). Alcibiades' famous opinion on the Sicilians, expressed in Spring 415, may well, in fact, reflect the conditions of Aeschylus' Sicily; cf. Dover, K. J., Thucydides Book VI (Oxford 1965)Google Scholar, note on ch. 17.2.

³ Diodorus xii 53.

⁴ Especially inc. fab. 170 and 171 Kaibel.

⁵ For this guess see, e.g., Lorenz, A. O. F., Leben und Schriften des Koers Epicharmos (Berlin 1864), 146Google Scholar, citing G. Bernhardy; and Olivieri, A., Frammenti della Commedia Greca (Naples 1930), 56.Google Scholar

⁶ If we can trust the Anonymus de Comoedia, sec. 9 (Kaibel, C.G.F., 8, with note).

⁷ Most clearly seen in the twelfth Olympian, for Ergoteles of Himera. This ode is dominated by Tyche Soteira, a surprising figure in Pindar or his era; one would rather expect to find her in a Hellen istic poem. The only close contemporary parallel—and an interesting one, from our point of view—comes in a late play of Aeschylus, , Ag. 664 ff.Google Scholar

⁸ Cf. the second Olympian, for Theron of Akragas—Empedocles' city.

⁹ Fully discussed by Jebb, R. C., Bacchylides (Cambridge 1905), 13–24.Google Scholar

¹⁰ With the exception of item 9, all the testimonia here printed are already to be found scattered in F. Schoell's De Aeschyli Vita et Poesi Testimonia Veterum, prefixed to Ritsch, F.'s edition of the Septem (Leipzig 1875).Google Scholar

Nearly every book or long article on Aeschylus naturally contains some reference to Aeschylus' Sicilian visits. But for discussions of the basic evidence, I would refer particularly to: Hermann, G., Opuscula ii (Leipzig 1827) 144–162Google Scholar; Kiehl, E. J., ‘Aeschyli Vita’ in Mnemosyne i (1852) 361–74Google Scholar; van Leeuwen, J., ‘de Aeschyli itineribus Siculis’ in Mnemosyne n.s. xviii (1890) 68–75Google Scholar; Schmid-Stählin, , Gesch. der gr. Lit. i 2 (Munich 1934), 189–192Google Scholar; Stanford, W. B., ‘Traces of Sicilian Influence in Aeschylus’, in Proc. Royal Irish Acad. xliv, section C (1937–1938), 229–40.Google Scholar References to a number of other contributions will be given in later notes, at the appropriate stage.

I should note that this article does not attempt to evaluate or criticize all the numerous theories that have been constructed on the basic evidence; that process would take far too long. I am concerned here, primarily, to set out as clearly as possible what seem to be the extent and the limits of the knowledge afforded by the ancient testimonia; secondarily, to present my own interpretation of it. Something of that interpretation I owe to the excellent article by Körte, A. (‘Das Prometheus Problem’ in Neue Jahrb. für das Klass. Altert. xlv [1920] 201–213)Google Scholar, which explains the ‘modernity’ of the Prometheus as due to Syracusan conditions. But I have come to differ from him on many points, especially on the crucial one: the dating of the play.

¹¹ A good recent survey of the question of the productions of the Persae will be found in Broadhead, H. D., The Persae of Aeschylus (Cambridge 1960), xlviii ff.Google Scholar

The value of Eratosthenes' testimony has been questioned (e.g. by Schmid-Stählin, op. cit., 190, n. 3), but I can see no good reason, in method, for doing so. Everything we know about him suggests that on a question of fact he was among the most reliable of all the ancient scholars; compare, for instance, his meticulous treatment of the chronological problem raised by Ar. Clouds 553 (a rather similar problem to that in Frogs 1028), quoted by Starkie, W. J. M., The Acharnions of Arutophanes (London 1909), li–lii.Google Scholar

A different, and now almost forgotten, approach to the Frogs scholium was that of E. J. Kiehl, op. cit., 363–5, who pointed out that the word used in the quotation from Eratosthenes is δεδιδάχθαι not ἀνα-δεδιδάχθαι (ἀνα-διδάξάι; however, is found in the Vita, para. 18); Kiehl concluded that the original performance of the Persae took place in Syracuse, and hence that Aeschylus' first visit to Sicily should be placed before Spring 472, the date when the play was performed in Athens. This view deserves mention, because it is still not absolutely disprovable; but it seems far less likely.

¹² A convenient summary of the evidence for this problem, with a bibliography which largely concerns it, will be found in H. Lloyd-Jones' Appendix to Smyth, H. W.'s Aeschylus, ii (London 1957) 595 f.Google Scholar

¹³ This victory and its date are of course doubly attested, in the manuscripts (Hypoth. Ag., ad fin.) and on the stone of the Fasti (quoted, e.g., Murray-Maas, p. 206).

¹⁴ The date, again, is doubly attested (see Testimonia, 12), though on what ultimate authority we do not know (Timaeus, or some other Sicilian historian? Contemporary reference to Aeschylus' death in some lost Attic comedy, comparable to the references to the recent deaths of Sophocles and Euripides in the Frogs?).

¹⁵ The earliest contribution to the debate known to me is that of Hermann, op. cit., the latest that of Cataudella, Q., ‘Eschilo in Sicilia’, in Dioniso xxxvii (1963) 5.Google Scholar

¹⁶ Diodorus, xi 49.

¹⁷ Though the exact date of that event is uncertain; cf. the recent discussions in van Compernolle, R., Chronologie Sicilienne (Brussels 1959) 390–1Google Scholar, and Stauffenberg, A. F. Graf von, Trinakria (Munich-Vienna 1963) 261.Google Scholar

¹⁸ Bowra, C. M., Pindar (Oxford 1964) 409.Google Scholar

¹⁹ E.g., Wilamowitz, , Aischylos-Interpretationen (Berlin 1913), 242Google Scholar; Fraenkel, E., ‘Vermutungen zum Aetna-Festspiel des Aeschylus’ in Eranos lii (1954) 68.Google Scholar

²⁰ G. Hermann, for example, postulated three visits besides the last one. But this is achieved by assuming that almost all the ancient accounts of Aeschylus' motives for leaving Athens (to be considered in the next paragraph of this article) preserve some memory of historical fact. Thus the Suda (Testimonia, no. 4) will show that the poet fled to Sicily very early in the nineties, on the famous occasion of the collapse of the benches; the Vita, para. 8, will bring him to the island after his defeat by Simonides in the eighties, and also (supported here by Testimonia, no. 1) after the defeat by Sophocles. None of these visits can be positively disproved; though, since the discovery of the date of the Septem, the last becomes highly unlikely. But both the method and the general result are disquieting.

²¹ Diodorus xi 68 fin. (year 466/5); xi 72–3 (year 463/2); xi 76 fin. (year 461/0). This point is well put by Méautis, G., L'authenticité et la date du Prométhée enchaîné d'Eschyle (Geneva 1960) 67 ff.Google Scholar, though I do not find myself able to follow all his conclusions from it.

²² There is, at any rate, an obvious confusion in the Cimon passage. Like the Vita (para. 8, ad init.), it maintains that Aeschylus fled to Sicily because he was disappointed by the young Sophocles' victory, that is, not later than 468. Unlike the Vita, however, it brings him on this visit not to the court of Hieron, but to his own death and burial. Both accounts are probably worthless guesses, but the Vita's is at least chronologically possible; whereas that in the Cimon ignores Aeschylus' presence in Athens in 467 (Septem) and 458 (Oresteia).

Van Leeuwen (op. cit., 73 f.) offers an ingenious explanation of this passage: Plutarch has here misread his source, which will have referred not to the κρίσις of Aeschylus' play in 468, but to the κρίσις, i.e. the ostracism, of Cimon in 459. But there is not enough evidence to raise this above the status of an interesting guess. And the same is true of the suggestion, often discussed in the past (e.g. Van Leeuwen, p. 71; Schmid-Stählin, p. 190–1), that the famous trial for impiety was responsible for Aeschylus' departure from Athens.

²³ In fact, I believe we can say that he did have an authority; but one whose reliability, unfortunately, cannot now be controlled. Pausanias (Testimonia, no. 5) seems to be extracting from the same source for his own purposes. It will be noticed that both Plutarch and Pausanias start from the case of Euripides and work backwards, the word πρότερον being used in both passages. Plutarch, stressing the idea of ‘permanent exile’, includes Herodotus and Homer; Pausanias, intent on the association between poets and dynasts, omits those two, but includes Anacreon. One might surmise that a handbook of exempla, containing a chapter on ‘Famous Literary Exiles’, underlies both lists.

²⁴ See note 10 above; the article is to some extent a review of the evidence already collected by Aly, W., De Aeschyli Copia Verborum Capita Selecta (Berlin 1906), ch. iii.Google Scholar

²⁵ Stanford himself, of course, fairly admits the difficulties of demonstrating that any one word is or is not Sicilian, and presents his results with all due caution. To me it seems that eight of his instances (numbers 1, 3, 4, 7, 10, 15, 16 and 20 in his list) are probable; about most of the rest we can only say, at the outside, that the meagre evidence as it stands points towards Sicily or Magna Graecia. I cannot bring myself to share the outright scepticism of E. Fraenkel (on Ag. 1507), though he rightly corrects some of Stanford's details.

²⁶ Lobel, E. in P. Oxy. xviii (1941) 9Google Scholar; cf. Setti, A., Annali della Scuola … di Pisa, Classe di Lettere (etc.) xvii (1948) 32–3, 35Google Scholar; and Cantarella, L., I nuovi frammenti eschilei di Ossirinco (Naples, n.d.) 64–66.Google Scholar One of the five Dorisms concerned was already known through ancient quotation: θῶσθαι (Diktyoulkoi, line 815).

It is interesting to see that both Stanford (p. 231) and Cantarella (p. 66, note 1), writing before the discovery of the Supplices-didascalia, were exercised by the number of presumed Sicilianisms which they found in what, at that time, was supposed to be an early play, datable long before any likely visit by Aeschylus to Sicily. Their explanations of this phenomenon, which were not in themselves very convincing, are now unnecessary; and it may perhaps count, in some degree, as a confirmation of the late date implied by the didascalia.

²⁷ Cf. Fraenkel, E., ‘Vermutungen zum Aetna-Festspiel des Aeschylus’ in Eranos lii (1954) 61–2Google Scholar, and notes.

²⁸ I do not claim that the suggested quasi is certainly right, though it is not difficult to imagine the process of corruption—especially in minuscule script—which would change uirguasisiculus into uirutiqueskulus. But I am confident that utique is wrong.

²⁹ The deep impression that Aeschylus evidently made on the Sicilian theatre deserves at least a mention at this point. Although one might argue that this cannot in itself constitute proof that he resided for a long time in Sicily, it may perhaps be worth considering as a supporting argument to those given in the text.

Most of the evidence is collected in Focke, F., ‘Aischylos' Prometheus’, Hermes lxv (1930), esp. p. 302Google Scholar, and Bock, M., ‘Aischylos und Akragas’, Gymnasium lxv (1958), esp. pp. 412 and 439Google Scholar, with notes. It falls into two classes: (1) evidence that Epicharmus parodied Aeschylus fairly extensively; this is shown, e.g., by no. 11 of our Testimonia, and probably by the Epicharman title Persae. (2) Evidence that Sicilian interest in Aeschylus endured long after the poet's death; it is even suggested that he left a ‘school’ of Akragantine tragedians. I note with interest that even when the tyrant Dionysius of Syracuse sets out to write tragedy (Lucian, , adv. indoctum 15Google Scholar) he buys up the writing-desk of—Aeschylus!

³⁰ Cf. Phoenix xvii (1963), 195, note 55, (b).

³¹ The evidence is collected and fully discussed by A. Peiretti, ‘La Teoria della generazione patrilinea in Eschilo’, La Parola del Passato xi (1956) 241 ff. Whatever nuances of interpretation be put on it, two points emerge beyond question. First, the curious problem is not touched on by any other classical Athenian writer except Euripides (Or. 552—an obvious parody of the passage in Eum.—and, more doubtfully, inc. fab. fr. 1064 Nauck). Second, it was discussed, favourably or otherwise, by the following non-Athenians: (for), Diogenes of Apollonia, Hippon of Metapontum; (against), Anaxagoras (?), Alcmaeon of Croton, Parmenides, Empedocles.

³² Or, more precisely: with ideas like those of Empedocles. Cf. Phoenix xvii (1963) 192–5, and the references given there.

³³ They include Stanley, Bergk, Sikes and Willson, and Thomson.

³⁴ The evidence for the date is of course the Oxyrhynchus hypothesis; see, e.g., Lloyd-Jones,, op. cit. 595–8, for bibliography and discussion, and compare, perhaps, supra, note 26, end.

³⁵ This is not the place for a discussion of the date of the Prometheus. For the moment, the present writer would only say that the more familiar he becomes with the language, metre, composition and outlook of the play, the more meaning he sees in Körte's paradox (op. cit., 204): ‘were there no obstacles, one would certainly place the Prometheus ten or twenty years later than the Oresteia.’ Compare also C.R. n.s. xiii (1963) 5–7, and C.R. n.s. xiv (1964) 239 f.

³⁶ This collection was originally based on the material printed by F. Schoell, op. cit., but it has been entirely rearranged, one item has been added, and the texts have been revised against the more recent critical editions. I have not judged it necessary to give an apparatus criticus, since none of the variant readings affects the substance of what is said, at least from our point of view.

I should add that I have deliberately excluded Aristophanes, fr. 618, from the collection. As edited by Edmonds, J. M., The Fragments of Attic Comedy, i (Leiden 1957) 740Google Scholar, it might appear to be valuable fifth-century evidence for the death of Aeschylus at Gela:

But the fact is that neither the name of Aeschylus nor any reference to him is found in the source of the fragment (Plutarch, , Comp. Ar. et Men., 853 bGoogle Scholar). And the mere mention of Gela (itself due to emendation here) does not necessarily imply that Aeschylus is being referred to; cf. Acharnions 606.

³⁷ The end of this epigram is used by Müller, K. O., Dissertations on the Eumenides (Engl. tr., London 1853), 80Google Scholar, as an argument for supposing that Aeschylus' final emigration was due to unpopularity at Athens. But it seems likely to be no more than a repetition of a commonplace current since Euripides, and Plato's Apology.

³⁸ The ‘Rhetorical Lexicon’ is almost certainly that of Aelius Dionysius (temp. Hadriani), which is often appealed to by Eustathius. Compare Schwabe, E., Adii Dionysii et Pausaniae Fragmenta Lexicorum Rhetoricorum apud Eustathium laudata (Leipzig 1890)Google Scholar, esp. his comments on this passage, p. 230 and note 2.

³⁹ Cf. supra, p. 79.