Aeschylus’ Suppliants

Stanley Ireland

Seeing that the completion of this work, a full edition with commentary, is not to be expected for some years, I make this excerpt public here. The draft introduces novel ideas with regard to the hypothetical reconstruction of the trilogy and individual features of text and meaning. I ask that any reader who might study the material forgive the typical deficiencies of a work in progress, not least the absence of relevant citation, which should not be interpreted as disparagement of the works overseen. The first half of this draft is now revised and updated as part of this publication: https://www.academia.edu/56534342/Aetiology_and_Justice_in_the_Danaid_Trilogy Abstract: While just as hypothetical as the other alternatives, an alternative reconstruction of the trilogy is possible in which Danaus is cast in the role of hero rather than villain, helping to justify the crime of the Danaids as an act of self-defence, legitimize the change of regime in Argos in the finale, and construct an aetiology of the "Danaan people" that would be palatable to the contemporary Greek audience. The aetiological purposes of the drama are highlighted in the choral ode 524–99.

The motivation behind the Danaïds’ flight from marriage to their cousins in Aeschylus’ Supplices has long been debated. For more than a century, prominent classicists have entertained the notion that the Danaïds consider such a marriage incestuous, despite the difficulty of justifying such a claim on the basis of the Danaïds’ language. This article will argue that since Darwin’s discovery of the deleterious effects of inbreeding, the incest taboo has been differently constituted, and that thinking native to our own culture has persistently intruded on attempts to interpret Aeschylus’ play. Having surveyed these problems in some detail, I use conceptual metaphor theory to eliminate the “flight from incest” theory, and posit that closer attention to the cultural biases I identify might help decode the myth’s enigmatic significance to Roman authors such as Vergil.

Ten years after producing the Persians in 472 BC, in which Greeks and barbarians are locked in conflict with each other, Aeschylus in the Suppliants explores the inextricable intertwining of Greekness and barbarity. While even in the Persians Aeschylus recognizes the ultimate ‘kinship' between Greek and barbarian (the women of Atossa's dream – one wearing Persian robes, the other Dorian – are described as ‘sisters of one race': Aesch. Pers. 180–7), in the Suppliants the poet develops this theme and casts it in sharper relief. In this later play, now generally accepted (despite archaic or archaizing elements) to have been produced in the late 460s, Aeschylus is more actively interested in the ways in which kinship both intersects with and is complicated by cultural polarity, and at the same time undercuts and complicates ‘Otherness'.

Aeschylus’ Supplices Introduction and Commentary on vv. 1–523 Aeschylus’ Supplices Introduction and Commentary on vv. 1–523 B by Pär Sandin Corrected edition SYMMACHUS C LUND 2005 Aeschylus’ Supplices: Introduction and Commentary on vv. 1–523 / Pär Sandin. Corrected edition: Lund, Symmachus publishing 2005 First edition: Gothenburg, Göteborg University 2003 (doctoral dissertation) © Pär Sandin 2003, 2005 Printed in Sweden by CO-print Professional, Nässjö 2005 ISBN 91-628-6401-7 Symmachus publishing Karl XII-gatan 14c SE-22220 Lund Sweden Aeschylus’ (525–456 B.C.) drama the Suppliant women (Greek Hikétides, Lat. Supplices) is certain to be the first in a trilogy of tragedies with an appurtenant comic epilogue, ‘satyr-play’. The other two tragedies and the satyr-play have been lost except for a few lines preserved in quotations and, possibly, papyri. The dissertation contains an introduction to the entire drama, a translation and commentary on the first half of the text (verses 1–523), and an excursus. The Introduction deals with the date of the theatrical production, the literary theme, the mythological background, the hypothetical reconstruction of the trilogy, and the contemporary Athenian theatre. The Commentary constitutes the major part of the work, being primarily philological, but also literary and historical, dealing with matters of scenic production and the nature of the chorus, where some new hypotheses are proposed, and with Greek mythology, religion, politics, and history in general as these become issues of particular passages of the text. The constitution of the text is a major concern. The Supplices is based on virtually only one manuscript: the Florentine Laurentianus graecus 32.9 (‘Codex Mediceus’) from the 10th century. There are five apographa from the later Middle Ages and the Renaissance, but there is no evidence to suggest that any one of these has independent authority. The text is often in need of reconstruction by emendation. The approach has been moderately conservative. About thirty new conjectures of varying probability are proposed and discussed; the reading of the extant manuscripts is defended in fourteen places against a majority of recent editors. The Excursus deals with a general problem of textual criticism in versified texts, the displacements of verses. The conclusion is that there has been an abuse of this conjectural measure in several editions. Aeschylus, Hiketides, Supplices, Suppliants, Suppliant women, tragedy, Greek drama, Greek theatre, textual criticism Preface The amount of work that remained to be done on the Supplices came as rather a surprise to me, seeing that, at least in my own country, the belief prevails that ‘everything has been done’ on the authors at the high end of the Classical canon. The great edition and commentary of Holger Friis Johansen and Edward Whittle, renowned for its exhaustiveness, was published a mere twenty years ago. It was followed by a number of long and learned reviews; then Martin West’s monumental Teubner edition with the accompanying Studies in Aeschylus appeared in 1990. One might have thought that things had been put to a relative rest in the absence of further evidence. Nevertheless, the present dissertation, originally intended as a collection of critical notes on discrete passages from several Aeschylean dramas (‘Studies on the Text of…’), turned out after a few months’ work to be a growing commentary on the Supplices, with gaps that needed filling. I thus set aside my notes on the other plays, publishing some material that was reasonably finished (Sandin 2001, 2002), and set to work on the Supplices. The gaps that needed filling were not only spatial, but conceptual: a modern commentary is expected to offer more than text-critical notes, and I have done my best to meet this demand, if sometimes only with references to the works of specialists. Certainly a large portion of the present study is devoted to textual criticism, which is inevitable in the case of a work notorious for the corrupt state of its text. Needless to say, Friis Johansen–Whittle’s commentary lay open by my side at virtually all times whenever and wherever I worked. The huge amount of information contained in it turned out not to be an obstacle, by virtue of its exhaustiveness, to further research; rather it was a great source of inspiration and a spur: when wrong, to attempt to disprove the commentators’ theses; when right, to advance further argument. Inevitably ‘pace FJ–W’, ‘rightly FJ– W’ and the like will occur repeatedly in my text—not, I hope, to the consternation of the reader. My views on theory and method are set out in a postscript to an article in Eranos 100 (Sandin 2002, 155–57). The present dissertation should be regarded as a preliminary study: my intention is to publish a full commented edition of the Supplices and of the fragments of the lost parts of the trilogy (Aegyptii, Danaides) with appurtenant Satyr-play (Amymone). I owe my heartfelt thanks to all the people and institutions who have v guided me through the alternatingly idyllic, tragic, farcical, ecstatic, and unbearably dull process that is post-graduate studies. My tutor, Professor Staffan Fogelmark, has supported me throughout my academic career in Lund and Gothenburg. He has patiently read the drafts of my dissertation and supplied invaluable observations and criticism, often discussing Aeschylus with me in the company of scholars such as Turnebus, Stephanus, and Casaubon, in the surroundings created by his marvellous library. Besides—a debt that will be even harder to settle—he was the one who taught me Greek in the first place, sharing, during a happy period of my life, his expertise and his love for the beauty, precision, and cogency of the Greek language as mastered by the best authors. I shall be forever grateful. Professor Martin West generously supplied me with a copy of his unpublished repertory of sixteenth- and seventeenth-century emendations in the Supplices, based on his own collations of scholars’ marginalia in copies of the early editions (see his Studies in Aeschylus, pp. 358–65). He also took time to discuss a palaeographical detail in the Laurentianus Graecus 32.9 (‘Codex Mediceus’) with me. In the course of a stimulating correspondence, Dr. Sir Charles Willink discussed several details of textual criticism and metre in the first choral ode of the Supplices; he also supplied me with a draft of his own notes on the entire cantica of the play, and I have had reason to re-evaluate and correct my views in several places in the light of his observations. If I happen to disagree with either of these scholars in a few instances in my commentary, this in no way diminishes my opinion of their stature, and in particular of Professor West’s unsurpassable contribution to Aeschylean studies. Two stipendiary visits abroad offered superb opportunities for research and much inspiration. In the spring and summer of 2000 I worked at the London Institute of Classical Studies, with the financial support of the Swedish Foundation for International Cooperation in Research and Higher Education (STINT) and Birgit och Gad Rausings Stiftelse för Humanistisk Forskning. During my stay, Professor Richard Janko took time—amidst massive commitments of his own—to tutor me for free, reading and commenting on drafts of parts of the dissertation. The Director of the Institute, at that time Professor Geoffrey Waywell, and the staff were most kind and helpful in every way. My second sojourn was in Rome in 2002–3, at Svenska Institutet (Istituto svedese di studi classici), where I spent an unforgettable year having been awarded the ‘grand scholarship’ in philology. The Director, Professor Barbro Santillo Frizell, and the staff were exceedingly helpful. vi The Greek seminar in Gothenburg has endured several sittings devoted to Aeschylean textual philology, and supplied valuable criticism. In particular I would like to thank Dr. Karin Hult, who has also read all the Greek passages in the book and most of the English, correcting a number of errors; furthermore she advised me on several practical details concerning the production of the book. Professor Marianne Thormählen has corrected my English with firm hand and unfaltering judgement; and Ms. Katarina Bernhardsson undertook to read the final typescript in full, saving me from a multitude of typographical embarrassments. Apart from the grants and scholarships mentioned above, I am grateful for a considerable grant from Adlerbertska stipendiefonden, and, towards the costs of printing the book, one from Längmanska kulturfonden. A grant from Kungl. Vetenskaps- och Vitterhets-Samhället i Göteborg enabled me, in the late spring of 2003, to make an excursion from Rome to Bologna in order to examine in situ the manuscript Bononiensis Bibl. Univ. 2271. During my time as a doctoral student I have also received grants from Stipendiefonden Viktor Rydbergs minne and Stiftelsen Dagny och Eilert Ekvalls premie- och stipendiefond. Finally, I owe thanks to my family and friends for their support and understanding. τοιῶνδε τυχὼν ἐκ πρυµνῆς φρενὸς χάριν σέβοµαι. Lund, December 2003. A renewed grant from Längmanska kulturfonden allowed the printing of this corrected edition. The text has been reset, but the pagination remains intact —a few words or lines may have been shifted into neighbouring pages. I am deeply indebted to Professor James Diggle for his critique of the first edition, presented, orally and in writing, at my public disputation in Gothenburg, 27 January 2004. Formal errors noted by Professor Diggle and others have been corrected here: the scholarly errors and misjudgements will have to remain for the present. I hope to be able to correct a few in the not-too-distant future. Lund, January 2005 vii Contents Preface Contents Introduction I. The date II. The fable 1. Myth 2. Other dramatic productions 3. Aeschylus’ trilogy 4. The meaning III. The early theatre of Dionysus 1. The shape 2. Stage building 3. Raised stage 4. Elevation 5. Altar 6. Terrace wall Translation Commentary Excursus: Transposition of lines References I. Ancient works II. Modern works 1. The Supplices: editions, translations and commentaries (separately or with other plays) 2. Editions, translations and commentaries on other ancient works 3. Scholarly works other than editions, translations and commentaries Indices I. Index locorum II. Index coniecturarum Aeschylearum III. Index verborum IV. Index parva nominum v ix 1 1 4 5 8 9 12 13 13 14 15 15 16 19 20 36 210 212 213 214 215 217 223 239 239 248 249 251 ix Introduction The standard work on the Supplices is still Alex Garvie’s Aeschylus’ Supplices: Play and Trilogy (1969; referred to here as ‘G.AS’). It is complemented by the commented edition of Holger Friis Johansen and Edward Whittle (‘FJ–W’; i. 22–55 on the drama as a whole). Other noteworthy general studies of recent years are Kraus on the Danaid trilogy (1984) and Sommerstein’s compact introduction to the drama (1996, 135–68). Special aspects of the Supplices are comprehensively treated in, for instance, Taplin (1977) 192–239, Rash (1981), Sicherl (1986), Court (1994) 145–80, Bakewell (1997), Rohweder (1998), Gödde (2000),1 Bachvarova (2001), and Turner (2001). After the commentary of FJ–W, critical notes on large portions of the text have been published by Diggle (1982), Verdenius (1985, 1990), Griffith (1986), West (W.SA 128– 69), and Liberman (1998), all of which are repeatedly cited in the Commentary. I. The Date The history of the dating of the Supplices is also interesting as an instructive example from the history of scholarship. The prominence given to the chorus in the play induced scholars of the early twentieth century to believe in a very early date, well before the Persians (472), in the light of Aristotle’s statement that tragedy evolved from the choral lyric.2 This in turn led to a number of assertions concerning the allegedly immature and archaic style and character of the play. Then a piece of external evidence turned up: a fragment of a didascalia, first published by Lobel (pp. 30–31 = POxy 2256.3) and conservatively edited by Radt:3 1 Rohweder and Gödde present interpretations of the entire drama in accordance with their preferences among recent scholarly fashions, ‘polis’ and ‘ritual’, respectively. For sceptical views in general on the former fashion, see Griffin (1998); on the latter, Scullion (2002b); for a positive but balanced account of the possible origins of tragedy in ritual sacrifice, Lloyd-Jones (1998). 2 The seminal work was Müller (1908): see further G.AS 88–110 with refs. 3 A. test. 70; cf. also Snell’s edition in TrGF i. 44–45 (Didasc. C 6), and West’s in the Teubner Aeschylus (p. 125). 1 ἐπὶ α̣.[ ἐνίκα [Αἰ]ϲχύλο[ϲ ∆αν[αΐ]ϲι, Ἀµυ[µώνῃ δεύτ[ε]ρ[ο]ϲ Σοφοκλῆ[ϲ, τρίτοϲ Μέϲατοϲ Ν.[.].[ Βάκχαιϲ Κωφοῖ[ϲ Ποι]µέσιν Κύκ.[ ϲατυ 5 1 ἄρ[χοντοϲ vel Ἀρ[χεδηµίδου Lobel, ἄρ[χοντοϲ Κόνωνοϲ Luppino (1967, 211), Ἅβ[ρωνοϲ vel Ἀκ[εστορίδου Radt, Ἀρ[χίου vel Ἀρ[ιµνήϲτου Tronskij (1957, 159), Ἀρ[ίϲτωνοϲ Stoessl (1979, 9) 2 [Αἰ]ϲχύλο[ϲ τεθνηκὼϲ vel -ο[υ τεθνηκότοϲ Tronskij (1957, 155–56), [Αἰ]ϲχύλο[ϲ Ἱκέτιϲιν, Αἰγυπτίοιϲ Snell Sophocles is said to have competed for the first time in 468/694 and, perhaps less plausibly, to have been victorious at the debut.5 If the first of these claims is true, we have a terminus post quem for the Supplices, a quarter of a century later than what was previously thought to have been the approximate date of the production (the 490s). Garvie (G.AS 29–82 passim) then thoroughly demonstrated that most of the alleged signs of an archaic or immature style and composition were pure fantasy: the more tangible ones (the prominence of the chorus, the frequent use of ring-composition) might as well indicate the author’s design for this particular play and have nothing whatever to do with its date. The late-twentieth-century orthodoxy, then, which was based on the assumption that the first line of the didascalia-papyrus has to be supplemented with an archon’s name beginning Ἀ(ρ), basically left room for the year 463 only, under Archedemides (See G.AS 1–2, 10–11). However, it has been shown—on analogy with another fragment from the same papyrus and by the same hand, containing the didascalic data on the Laïus–Oedipus–Septem trilogy6—that ἄρ[χοντοϲ is the most likely supplement in the first line of our 4 Apsephion was the archon: Plu. Cim. 8.8 (= A. test. 57, S. test. 36). Plu. ibid. and also Marm.Par. A 56 (= S. test. 33), confirming the date of Sophocles’ first victory but saying nothing about the time of the debut. A later source, Isid. Chron. 174 Mommsen, claims that Aeschylus, … Sophocles et Euripides … celebrantur insignes in 477. 6 POxy 2256.2 = A. test. 58b, Didasc. C 4 Snell. 5 2 didascalia.7 Accordingly, the date of the Supplices could be any time between Sophocles’ debut and Aeschylus’ death.8 As for Sophocles’ debut, the evidence is not as certain as one might have hoped. Apart from the fact that ancient biography is unreliable (see in particular Lefkowitz 1981), there are various conflicting statements as to the debut and the first victory in the different fragmentary versions of Eusebius’ Chronicon (see S. test. 32a–d). According to Chron.Pasc. 162A, the debut took place as early as 486 (the third year of the 73rd Olympiad); at that point, though, Sophocles would have been about eleven years old, if our information about his birth is to be trusted. In two other versions cited by Radt (S. test. 32a–b) the debut is alleged to have been in the second year of the 77th Olympiad (470/71), a date which was accepted by Snell in his edition of the didascalic charters (TrGF i. 5, cf. 51). In the light of this, Scullion has taken up the case for a relatively early date (2002a, 87–101). He argues against the reliability of the evidence for Sophocles’ debut and, on the basis of internal stylistic and structural evidence, thinks it probable that the Supplices is indeed our earliest extant play by Aeschylus, and that it should be dated to the mid-470s, some years before the Persians. His arguments, which especially concern the feature of ring-composition, the prominence of the chorus, and particle-usage, are noteworthy if not positively convincing, and they will have to be weighed carefully against the reliability of the sources for Sophocles’ debut and first victory.9 In Scullion’s defence, we may—for what it is worth—add that the earlier date 7 As noted by West, p. 125. The didascalia-fragments are likely to have been identically phrased (the second lines of both begin with ἐνίκα Αἰϲχύλοϲ). If thus ἐπὶ is restored from our fragment (2256.3) in the beginning of 2256.2, ἄρχοντοϲ becomes a certain supplement in the latter on account of the space available (see Scullion 2002a, 87, n. 24 for details). Analogy then requires that the same word be restored in the first line of our fragment. 8 Some (e.g. Tronskij 1957, Stoessl 1979) have argued that the didascalia refers to a posthumous production, which would invalidate all pertinent external evidence both as to the authoring and as to the original production of the play. Fortunately this is a very unlikely alternative (see especially FJ–W i. 23, G.AS 21). 9 One feature that argues for the Persians being Aeschylus’ oldest extant play is the metre: the use of trochaic dialogue is not found elsewhere outside comedy and satyrplay, and the very sparse occurrence of the dochmiac metre also suggests an early stage in the development of the drama. See G.AS 38–40 with refs. 3 would be in accordance with the fashion of ethnographic comparison, and of polarisation between Greek and Barbarian, which arose and peaked during the first three decades of the fifth century.10 This is one of the most pertinent themes of the present drama, as well as of the Persians. On the other hand, the prominence of Argos in the play and the hints about its democratic traditions (cf. 365–69, 398–99) would make a date in the late 460s attractive: at that time an alliance between Argos and Athens took place, which is alluded to in Eu. 289–91, 669–773, and 762–74;11 besides, around the same period there was an Athenian expedition to Egypt in support of the Egyptians against the Persians (Th. 1.104, D. S. 11.71.4–6, 11.74.3– 6) which is likely to have fuelled public interest in things Egyptian, a major theme of the Supplices. Sommerstein (1997, 74–79) brings up another event in the late 460s which he thinks may have influenced Aeschylus’ story: an actual situation with a Spartan help-seeker or suppliant, Pericleidas (= Danaus), seeking help from Athens’ strong man Cimon (= Pelasgus) against revolting Helots at Ithome. This probably took place in 462, resulting in a war with ignominious consequences for the Athenians and the ostracism of Cimon (= the victory of the Aegyptiads and the probable death of Pelasgus: see below, II 3). The suggested parallels are hardly striking, however. II. The Fable The most important constituents of the fable of the Danaids as we know it from Greek and Latin sources, and those on which virtually all sources agree, are that (1) the fifty daughters of Danaus flee Egypt and come to Greece (Argos), trying to escape marriage to their cousins, the sons of Aegyptus; (2) they are forced to marry anyway but kill their husbands on the wed10 See Hall (1989) 1–19 passim and esp. 59–76. The Persian wars were the kindling flame of this interest, and Hecataeus was probably one of the seminal exponents (see 220–21n., text for n. 381, and 284–86n.). We know of a large number of tragedies from the early fifth century which dealt with ‘barbaric’ or ethnical matters; in fact, most of the preserved titles from Phrynichus’ dramatic production imply such a theme: Aegyptii and Danaides (from the same trilogy?), Antaeus/Libyes, Dikaioi/ Persae/Synthokoi, Miletus capta, Pleuroniae (see fr. 5), Phoenissae. As for Aeschylus’ own dramas with ‘ethnical’ titles, we know little of the dates. 11 See Sommerstein ad locc. and pp. 25–32; and Th. 1.102.4 with the notes of Hornblower and Gomme. 4 ding night; (3) the sole exception is Hypermestra, who spares her husband Lynceus. Only the first motif is treated in the present drama, and none of the Danaids is named. However, we may safely assume that the murder featured in the second or third part of the trilogy (see II. 3 below). 1. Myth. The motif of the brothers’ wooing of their relatives has a close parallel in a Hittite story of thirty young men with thirty sisters, whom they unwittingly intend to marry.12 The earliest witness is a bronze tablet that was probably inscribed as early as the 15th or 16th century B.C. (Otten’s ed., p. 1). A detail that might otherwise have been thought incidental to the core myth also occurs in the Hittite narrative, namely the motif of the exceptional: one of the brothers refuses to sleep with his sister. Possibly he somehow becomes the founder of a line of kings,13 perhaps through sleeping with the Sun-goddess (?). Apparently the fable is a very old folk-tale, presumably of Indo-European origin, as we find that similar stories appear later in various European oral traditions14—Calvert Watkins also notes a close verbal similarity between the beginning of the mentioned Hittite story and a line of the Rigveda.15 Burkert (1991, 534) argues that some form of the Danaid myth has been used from the very beginning, i.e. probably the Bronze Age, as an aetiology for the Greeks’ settling in Hellas, as against the opinion that the myth of Danaë (four generations later in the same family tree) is the original eponym (e.g. West 1985, 145, 146–51 passim). It is certainly tempting to regard the story as we have it as being connected with the widening of the mythical landscape that took place during the Greeks’ colonial and other geographical exploits, to include Egypt and the rest of the Mediterranean world and the Middle East.16 12 See West (1997) 446–47, Burkert (1982) 719, Burkert (1991) 534. The Hittite text has been edited and translated into German by Otten; an English translation is published by Hoffner. 13 So the fragmented ending is interpreted by Burkert (1991, 534). 14 See Laistner (1889) ii. 87–109 on ‘Menschenfressersagen’, with which he (p. 89) connects the Danaid myth; more concisely and with more relevant parallels Bonner (1900) 30–33, Bonner (1902) 149–52; cf. also Megas (1933), G.AS 175–76. 15 Watkins (1989) 796–97, cf. Watkins (1995) 53. 16 West (1985) 145–51; cf. Gantz (1993) 202–3, Hall (1996) 338–39 (138–139). See further G.AS 171–76 with refs for theories on the origin of the myth, and Auffarth (1999) for an interesting discussion about the role of the peculiar Danaid myth in the ‘social memory of the polis’. 5 The Danaids have been linked genealogically with Io, one of Zeus’s consorts, who is taken to be their great-great-great-grandmother. This link at any rate must have been a recent conceit by Aeschylus’ time, as earlier versions of Io’s myth appear to have taken her wanderings to end in Euboea, not Egypt.17 Her exile in Egypt must be an integral part of her being linked with the Danaids, who are connected with this land in Greek sources from the very beginning (the Danais, fr. 1, PEG p. 122: see below). In any case, the story as we have it presents the Danaids’ kinship with Io as being of crucial importance for their purpose in coming to Argos, as she is their link to Argos and Greece (see especially 274–325 with notes). Io’s legend varies in the sources:18 Aeschylus presents a version (narrated in 291–315) in which, a priestess of Hera in Argos, she was seduced by Zeus and then transformed into a cow by the jealous Hera, who also appointed a watcher, the all-seeing Argos.19 Argos was slain by Hermes, but Hera instead sent a gadfly which drove Io into exile. Coming at last to Egypt, Io was impregnated by Zeus who begot a son Epaphus, the great grandfather of Danaus and Aegyptus.20 The little that is known of the persona of Danaus seems to have a connection with the geographical-colonial motif (see above), which of course does not mean much: any embellishments of his character may be late additions to the myth. In any case, he is mentioned in literature as the inventor of shipbuilding and introducer of important knowledge to Greece (from Egypt).21 17 Cf. ?Hes. Aegimius fr. 296 (ap. St.Byz. s.v. Ἀβαντίς, p. 3 Meineke), West (1985) 145–46. 18 See, e.g., Gantz (1993) 198ff., R. Engelmann in Roscher ii. 263–69 (s.v. ‘Io’) and FJ–W i. 44–45, ii. 234–56 passim; also FJ–W’s notes on 291–92, 295, 299, 303, 311. 19 So already in ?Hes. Aegimius (fr. 294): καί οἱ ἐπὶ σκοπὸν Ἄργον ἵει κρατερόν τε µέγαν τε | τέτρασιν ὀφθαλµοῖσιν ὁρώµενον ἔνθα καὶ ἔνθα, | ἀκάµατον δέ οἱ ὦρσε θεὰ µένος, οὐδέ οἱ ὕπνος | πῖπτεν ἐπὶ βλεφάροις, φυλακὴν δ’ ἔχεν ἔµπεδον αἰεί. 20 The story of the cow that was loved by a god has parallels in Near Eastern mythology, even to the point that some of the wording in Aeschylus is similar: see West (1997) 442–46, Bachvarova (2001) 52–64. 21 Shipbuilding: Marm.Par. 15–16, [Apollod.] 2.1.4, ΣD Il. 1.42 (Heyne), Eust. on the same passage (i. 60–61 van der Valk), Σ A. Pr. 853a (Herington), Σ A.R. 1.1–4e, Hyg. Fab. 168, Plin. HN 7.206, Σ E. Med. 1, Lactant. in Stat. Theb. 2.222. Irrigation: e.g., Hes. fr. 128 (see below), Plb. 34.2.6 (= Str. 1.2.15), EM 681.5 (s.v. Πολυδίψιον Ἄργος); cf. Luc. Dmar. 8.1–2. See further O. Waser in RE iv. 2095 (s.v. ‘Danaos’). 6 The earliest sources that mention him are (allegedly) Anaximander and Hecataeus, who appear to have been cited by Apollodorus of Athens as support for the claim that Danaus brought the letters to Greece from Egypt.22 The Danaids are not mentioned in Homer, nor is Io. A fragment of (possibly) Hesiod names the Danaids (∆ανααί) in a rather different context from the present one, namely as—apparently—the introducers of irrigation to Greece (fr. 128): Ἄργος ἄνυδρον ἐὸν ∆ανααὶ θέσαν Ἄργος ἔνυδρον. Thus in Str. 8.6.8; a variant reading gives the credit to Danaus himself.23 This story is to be connected with another one: that of Amymone, the only Danaid apart from Hypermestra to stand out from the crowd. She became the lover of Poseidon, who showed her the hidden springs of Lerna and thus the means of watering Argos. Water sprang from a rock hit by Poseidon’s trident, which he threw as he rescued her from a satyr (Σ E. Ph. 185, Hyg. Fab. 169a); cf. E. Ph. 186–89, where the waters are called Λερναία τρίαινα, Ποσειδάνια Ἀµυµώνια ὕδατα (see Mastronarde ad loc.), and also Luc. DMar. 8.1–3.24 The affair was certainly the subject of the satyr-play that accompanied the Danaid trilogy (see below); and it is not unlikely that the version of the myth involving one or several satyrs stems from there. Amymone’s union with Poseidon was known before Aeschylus, however: in a contemporary story with a folkloristic touch (see Gantz 1993, 206), Pindar (P. 9.112–22) numbers the virgin Danaids as forty-eight, which implies that the fate of Amymone as 22 ΣVat. D.T. p. 183 Hilgard = Apollodorus fr. 165 FGrH (no. 244, ii B p. 1092); Hecat. fr. 20 FGrH (no. 1, i. 12; Fowler pp. 133–34); ?Anaximand. fr. 3 Fowler (p. 38 = i. 90 Diels–Kranz): Πυθόδωρος δὲ {ὡς} ἐν τῷ περὶ στοιχείων καὶ Φίλλις ὁ ∆ήλιος ἐν τῷ περὶ χρόνων πρὸ Κάδµου ∆αναὸν µετακοµίσαι αὐτά φασιν· ἐπιµαρτυροῦσι τούτοις καὶ οἱ Μιλησιακοὶ συγγραφεῖς Ἀναξίµανδρος καὶ ∆ιονύσιος καὶ Ἑκαταῖος, οὓς καὶ Ἀπολλόδωρος ἐν νεῶν καταλόγῳ παρατίθεται. 23 ∆αναὸς ποίησεν ἔνυδρον (Eust. i. 729 van der Valk). Cf. above, text for n. 21. 24 On Lerna, see also, e.g., [A.] Pr. 652–53, 676–77 with Griffith’s notes, E. Ph. 613, Str. 8.6.7–8, Paus. 2.15.5. It appears to have been the name of a river as well as a swamp (Str. 8.6.2); according to Pherecyd. fr. 31b FHG (ap. Σ Pi. O. 7.60) it was also a city. The place is elsewhere mentioned in connection with the Danaids: Paus. 2.24.2 claims that the murder of the Aegyptiads took place in Lerna, whereas according to [Apollod.] 2.1.5 their heads were disposed of there (cf. Zen. s.v. Λέρνη κακῶν, Apostol. s.v. Λέρνη θεατῶν). The Danaids may also have played a role in the Λερναῖα, an Argive festival to Demeter and Dionysus (see G. Baudy in NP vii. 81–83). Wilamowitz (1914, 9) dismisses the idea of any geographical precision in Aeschylus as to the landing-place of the Danaids. 7 well as that of Hypermestra was familiar to him and his audience. Pindar’s ode relates a running contest which is held by Danaus, in which the line-up are allowed take turns to pick out his daughters for wives. Pherecydes is also familiar with the union of Poseidon and Amymone, and with their son Nauplius, the founder of Nauplia (fr. 13 FHG, ap. Σ A.D. 4.1091). The earliest evidence for any action on the part of the Danaids which is relevant to the present drama appears to be a fragment of the Danais, an epic by an unknown author usually taken to be earlier than Aeschylus, perhaps from the sixth century (PEG p. 122): καὶ τότ’ ἄρ’ ὡπλίζοντο θοῶς ∆αναοῖο θύγατρες πρόσθεν ἐυρρεῖος ποταµοῦ Νείλοιο ἄνακτος. Clement of Alexandria, who is our sole source for the verses (Strom. 4.19. 120), quotes them as an example of female valour. Nevertheless, it would seem to be more in accordance with what we know about the myth to take ὡπλίζοντο as ‘made themselves ready (for the flight)’.25 The situation is described as taking place by the Nile, which suggests that the girls are simply getting ready to sail and make their escape to Greece, not preparing for a fight. Certainly no other source supports the notion of the Danaids ever going to battle, or that of a battle taking place in Egypt before the flight of the Danaids.26 Clement, like us, may well have read the verses out of context, for instance in a florilegium. A survey of the later sources, who are in chaotic disagreement about the details of the entire myth, is found in FJ–W i. 47–55. 2. Other dramatic productions. Phrynichus wrote an Aegyptii (frr. 1–1a) and a Danaïdes (fr. 4), of which we know next to nothing—not even if they are part of the same trilogy; or whether they were staged before or after Aeschylus’ versions. The one scrap of information we have tells us that Phrynichus let Aegyptus come together with his sons to Argos in the Aegyptii (fr. 1, ap. Σ E. Or. 872). 25 So Meyer (1892, 82, n. 3). Contra e.g. G.AS 179, Vürtheim p. 13, Wecklein (ed. 1902, p. 2). 26 In Melanipp. fr. 1 (ap. Ath. 14.651f), however, the Danaids are depicted as Amazonlike women: ἐν ἁρµάτεσσι διφρούχοις ἐγυµνάζοντ’ ἀν’ εὐήλι’ ἄλσεα πολλάκις θήραις φρένα τερπόµεναι. Cf. 287–88 of the present drama. 8 3. Aeschylus’ Trilogy. The evidence suggests that the plays of the trilogy went under the names of Ἱκέτιδες (Supplices), Αἰγύπτιοι and ∆αναΐδες, and that the satyr play was the Ἀµυµώνη.28 The Danaïdes is certainly the ending play, unless it is used as a title for the entire trilogy in the didascaliafragment (test. 70; see above, ch. Ι, the Date): this is unlikely, as the catalogue of Aeschylean dramas (test. 78) mentions each of the three tragedies as a separate play. As for the previous two dramas, the scholarly consensus has long been in favour of the Supplices being the first, a view that has seldom been seriously questioned after Hermann (1846–47, 123–27 [180–84]). The parodos, which seems to contain all the necessary information about the events previous to the depicted action (cf. 6–10n.), certainly has the appearance of an introduction to the entire trilogy. The strongest argument, however, is that a second place would mean that far too many important events would have to be crammed into the finale—several of the most dramatic events and conflicts, which have been anticipated by many hints in the Supplices (cf., e.g., G.AS 181–82), would have to be recounted in a prologue. If the Supplices is the second play of the trilogy, the last play will have to contain or recount the arrival of the Aegyptiads, Pelasgus’ death (probably), Danaus’ acceptance of the Aegyptiads’ claim to the Danaids, the marriage, the wedding-night murder and, presumably, some sort of reconciliation including Hypermestra and Lynceus being hailed as progenitors of a future royal lineage. It has also been argued that Hypermestra stands trial in the last play, a scenario that may be suggested by Dan. fr. 44 (see G.AS 205–8 with refs). It is hardly possible that the wedding night would be included within the timeframe of a single drama —especially as the chorus, certainly consisting of either the Danaids themselves (the play being the Danaïdes) or their bridegrooms, must be absent, not being able to fill the time with a choral ode. If any sort of dramatic unity 27 27 A trilogy it is, certainly, even if some have not wished to exclude the possibility of a dilogy: cf. Gantz (1979) 297–98, Hermann (1820) 6 (310). 28 The last two titles are found in the famous didascalia-fragment (test. 70, see above); all four of them are in the catalogue of Aeschylean dramas (test. 78). Cf. also frr. 5, 13–15, 43–46. Hermann (1846–47) 123–27 (180–84) suggested that the θαλαµοποιοί was identical to the Aegyptii, which is not impossible, as the former title is not found in the catalogue and could hence be an alternative title; but there is no positive evidence of any kind for this, and the drama has been suspected to be a satyrplay (see Radt ad loc). One source, the Etymologicon Gudianum s.v. Ζαγρεύς (cf. 156n.), speaks of Αἴγυπτος instead of Aegyptii. 9 is to be attained, the Danaïdes will have to begin after the murder: if Supplices is the second drama we will then have to suppose that the entire dramatic conflict that leads to the most dramatic event of the story—the wedding-night murder—would be recounted in a prologue, and that the audience will not see the Aegyptiads alive again (having met them already in the first drama). Such indeed is the view of the play held by Wolfgang Rösler (1993, cf. also Rösler 1992), who has made an ambitious case for the Supplices being the second part of the trilogy. Rösler starts with an assumption made before by Sicherl (1986, 88–101, 108, passim), namely that in order to make the Danaids’ refusal to marry their cousins explicable, the motif of the oracle which fore-told that Danaus would be killed by one of Aegyptus’ sons must have featured explicitly somewhere. Sicherl assumed that the oracle would have been mentioned in the last play (ibid. p. 98); Rösler (1993, 7) argues, with FJ–W i. 47, that it would have had to be mentioned in the first, which would then take place in Egypt. He argues further (pp. 17–20) that Pausanias 2.19, where Hypermestra is said to have stood trial in Argos, accused by Danaus of not obeying his command, is derived from the last play of Aeschylus’ trilogy. However, the ‘evidence’ is circumstantial to say the least, and also somewhat circular: for instance, Rösler takes Pausanias’ mention of Danaus’ fear of Lynceus as an allusion to the oracle, and thus deriving from the Aeschylean trilogy. Moreover, if the oracle did feature in the Aegyptii it would be unsatisfactory, indeed impossible, for the trilogy to end without also including its fulfilment: hence yet another motif would have to appear in the last play, namely Danaus’ death at the hands of Lynceus. This has been noted by Sommerstein, who still follows Rösler in assuming that the oracle featured explicitly and that the Supplices was the second play of the trilogy.29 He suggests that the Danaïdes began after the death of the Aegyptiads, and that it featured Lynceus as protagonist. But surely the audience, tension having been built up throughout the Supplices with expectations of war and bloody murder, would feel cheated by this. After the Supplices, one expects the conflict between the Aegyptiads and the Danaids to appear on stage—and to culminate (off-stage) in the wedding-night slaughter. If all the action was actually recounted in a prologue in the last drama, it is hard to believe that Aeschylus would have won the first prize in the contest. 29 Sommerstein (1995), Sommerstein (1996) 141–51. 10 To sum up, by far the most likely scenario is that the Supplices was the first play of the trilogy. If the oracle does appear, which is not impossible, it would have to be mentioned in the second or third play. It might for example be disclosed to the Danaids by their father in combination with his giving orders that they kill the Aegyptiads (a wedding being inevitable). The premise that the oracle is a necessary feature of the trilogy is hardly tenable, however.30 First, there is no hint whatsoever about an oracular response or about any fear for the life of Danaus in the Supplices. Secondly, in the large majority of the accounts of the myth, there is no mention at all of the oracle—it features in a few late sources only,31 and we would expect several of the authors who write about the myth to mention it if they had known about it. If the oracle said that any future son-in-law would kill Danaus,32 this would certainly be incompatible with the one mention of the Danaids we have that is contemporary with Aeschylus, namely the one in Pindar’s Ninth Pythian (see 30 A secondary argument of Sicherl and Rösler is Σ A. Supp. 37 ὧν θέµις εἴργει] ὧν τὸ δίκαιον ἡµᾶς εἴργει διὰ τὸ µὴ θανατωθῆναι τὸν πατέρα. As noted by LloydJones ap. G.AS 216, n. 6, the use of the verb θανατωθῆναι in the aorist tense is somewhat odd: thus Sicherl l.c. p. 92 takes it to mean not ‘because the father is not (yet) dead’, but ‘in order that the father must not die’. This may indeed be so: we may not even have to supply, with Sicherl, µὴ <βούλεσθαι>: cf. Σ rec. A. Pers. 353 βουληθέντων τῶν Λακεδαιµονίων … ἀπελθεῖν πρὸς τὴν ἑαυτῶν πατρίδα διὰ τὸ µὴ παραδοθῆναι ταύτην τῷ Ξέρξῃ. On the other hand the traditional interpretation can hardly be said to be impossible: cf. Σ A. Th. 130b Παλλὰς] Ἀθηνᾶ διὰ τὸ φονεῦσαι Πάλλαντά τινα, Ps.Nonn. Schol.myth. (Comm.in Gr.Naz.Or.) 4.7 ἡ Ἰφιγένεια οὖσα ἐν τοῖς Ταύροις, διὰ τὸ µὴ ἐπιγνωσθῆναι παρὰ τῶν ἐπιξενουµένων τίς ἐστιν. In either case, even if the scholium should allude to the oracle, it does not automatically follow that the oracle did feature in the trilogy: the scholiast may well have drawn on external sources without considering the fact that the oracle is not explicitly mentioned by Aeschylus. 31 Σ E. Or. 872, Σ A. Pr. 853a, Apollod. ap. Σ Il. 1.42 (cf. [Apollod.] 2.12), Eust. i. 60 van der Valk, Σ Stat. Theb. 2.222, 6.269. According to Paus. 2.19.6, Danaus took Hypermestra to court τοῦ τε Λυγκέως οὐκ ἀκίνδυνον αὑτῷ τὴν σωτηρίαν ἡγούµενος. 32 Thus, e.g., Sicherl (1986, 93), Sommerstein (1995, 114, text for n. 17), Turner (2001, text for n. 9), but in the ancient sources actually only at Σ Stat. Theb. 2.222: the other sources name the Aegyptiads as such as the sources of the danger. 11 above).33 If Danaus thought that he was in danger of being killed by a future son-in-law he would hardly, unless suicidal, make sure that forty-eight of the foremost athletes in Hellas should marry his daughters.34 4. The Meaning. An oracular response may or may not have featured in the second or third play as an indirect reason for the obstinate chastity of the Danaids. To be sure, it is hard to think of many other reasons that would be acceptable or understandable for a contemporary Greek audience. Why should the Danaids not want to marry their cousins, a marriage that would seem perfectly natural to contemporary Athens?35 Indeed what is the meaning of the Supplices and the dramatic trilogy: what, if any, moral lesson did Aeschylus think he could draw from this, as it seems to us, wholly amoral folk-tale? The matter has been discussed at length in countless books and articles. Is Aeschylus on the side of the Danaids or the Aegyptiads? Are the girls averse to marriage as such or only to this particular marriage, and why? For a thorough discussion see G.AS 212–24 with refs, and on the last-mentioned issue see my note on v. 82. Presumably the matter was developed and resolved in the second and third plays of the trilogy, of which we know next to nothing. However, in the absence of an oracle, I believe that one of the likelier scenarios would be that Aeschylus took some sort of power-struggle between Aegyptus and Danaus as being at the heart of the conflict—perhaps mixed with pseudo-ethnic sentiments, Aegyptus and his sons having become more Egyptian in their ways than Danaus, who still holds on to some of his Greek origins. This would explain the reluctance of the Danaids, and it is certainly their sentiment before the Egyptian herald and his stooges in the (albeit badly corrupt) scenes in 825–910, and that of Pelasgus in his speech in 33 Winnington-Ingram (1969, 12–13) actually suspected that Aeschylus’ trilogy might have been influenced by the Ninth Pythian. 34 On the trilogy see especially G.AS 163–233; see also, apart from already cited works, Winnington-Ingram (1961), Taplin (1977) 194–98, Gantz (1980) 141–42, FJ–W i. 40–55, Radt pp. 111–12, W.SA 169–72. 35 Indeed there was even a law that stated that the next of kin of fatherless girls had the right to marry them (see Harrison 1968, 10, 132–33, and, e.g., Is. 10.4–5, Arist. Pol. 1304a8, FJ–W i. 34–37 with refs). Thomson (1973, 289–93; 1971) took the sense of the trilogy to be just this: the conflict between exo- and endogamy, the Danaids representing an older, inflexible exogamist view which is overturned in the end through Hypermestra’s marriage to Lynceus. This narrow view has not found much favour with later critics, and it is well refuted by G.AS 217–20 and FJ–W l.c. 12 911–53, which is chauvinistic and replete with ethnical slurs. But certainly Aeschylus is not partisan on the side of the Danaids: their questionable behaviour is thoroughly demonstrated in the play, and whereas we are probably meant to feel some sympathy for the headstrong girls, the fault of hybris and the sin against Aphrodite are apparent and certain to have unfortunate consequences, all the more so as we know what the final outcome will be: the hideous slaughter of the Aegyptiads. We also know that Hypermestra will marry Lynceus, and that this is likely to be the good and conciliatory outcome of the dramatic conflict. How the moral conflicts are to unfold and be resolved, and how the guilt of the Danaids and the defilement of the weddingnight murder are to be cleansed, remains somewhat of a mystery—although perhaps in the end not more so than Orestes’ acquittal in the Eumenides after murdering his own mother. III. The early Theatre of Dionysus The comprehensive scholarly output on the subject of the theatre of Athens in recent decades seems mainly to have gone to show that we know nothing for certain about virtually any feature of the early theatre. Even things that have long been taken for granted, such as the shape of the orchestra and the position of the altar, have been shown to rest on inconclusive evidence. On a few matters, scholars agree in their guesses; on others, opinions vary greatly. I restrict myself here to a short survey of the opinions in a limited selection of pertinent works, mainly from the last three decades. Only features relevant to the production of the Supplices will be mentioned. See Green (1989, 1995) for a detailed bibliography of the period 1971–95. 1. The shape of the early orchestra has since Dörpfeld–Reisch (1896, 26 ff.), or indeed since Vitruvius (5.7), usually been assumed to be circular. However, a case for a rectangular, trapezoidal or irregular shape gradually built up during the second half of the last century; it is impartially summed up by Ashby (1988), a revised version of his article appearing in Ashby (1999) 24–41. The idea was, I think, originally presented by Carlo Anti.36 The positive case for a rectangular orchestra is successfully countered by, for instance, 36 Anti (1947) 55–72. The view is defended or re-stated by, e.g., Anti–Polacco (1969, 129–59), Butterweck (1974), Gebhard (1974), Pöhlmann (1981), Polacco (1990, 101– 4), Bees (1995), Moretti (2000). 13 Hammond, Scullion, and Wiles, who show that the archaeological remains may be interpreted as favouring an early circular orchestra.37 Hammond and Scullion also present some circumstantial evidence in favour of a circular shape. Martin Revermann lately (1999) pointed to an all-but-forgotten piece of literary evidence, Heniochus fr. 5 PCG, a poet of the Middle Comedy, who clearly refers to a circular orchestra, which is not unlikely to have been that of Athens.38 The fragment is presumably from the first half, probably the first quarter, of the fourth century;39 it is thus the oldest mention we have of a circular orchestra, being pre-Lycurgean (> 338) and also prior to the foundation of the theatre of Epidaurus (330). At some point in history the theatre of Athens certainly did assume a circular shape, as is evident from the present remains. There is no definite evidence either for or against a circular orchestra at the time of the Supplices, but I am inclined to take the evidence as slightly favouring a circular shape. 2. Stage building. The same applies to the existence of a stage building, or σκηνή, in the earliest extant plays (before the Oresteia). There is no archaeological evidence, and the received opinion has long been that there is nothing in the texts of the three oldest surviving dramas of Aeschylus to suggest the presence of a building (so first Wilamowitz 1886, 606–11). There were always dissenting voices,40 and an ambitious case in favour of an early skene was recently stated by Bees (1995). He is at his most convincing as regards the Persae, where at least one controversial passage (140–43) appears to suggest the presence of a building. In the Septem and in the Supplices, however, a house has no place in the drama, and the existence of one would have to be ignored by the audience. The early plays thus present a conspicuous contrast to the Oresteia, in which the palace of the Atreidae is a notable feature, and to the plays of Sophocles, Euripides, and Aristophanes, where references to palaces, caves, temples, and indeed to the skene itself (Ar. Pax 731) abound. For further arguments against the existence of a skene in the 37 Hammond (1988) 8–9; Scullion (1994a) 19–21, 24–28, 38–41; Wiles (1997) 46–52; cf. also Sommerstein (1996) 35–36. 38 The circular orchestra (of Athens?) is depicted, in a direct address to the audience, as Olympia: τὸ χωρίον µὲν γὰρ τόδ’ ἐστὶ πᾶν κύκλῳ | Ὀλυµπία. 39 On the dating of the fragment, see Hunter (1979) 35, n. 61. 40 E.g. Arnott (1962) 4 ff., Dale (1969) 260–63, FJ–W on 1 ff., Polacco (1990) 161–62. 14 early plays, see especially Taplin (1977) 452–59. Pers. 140–43 does not necessitate the presence of a house, according to Hall ad loc.41 One may ponder Hammond’s (1972, 425–27) assertion, following e.g. Pickard-Cambridge (1946, 10), that the early orchestra made use of a movable skene, a façade. A changeable façade will accommodate any type of scenery suggested in the extant plays: a house in the case of Pers. (and also perhaps in Phrynichus’ Phoenissae, produced four years earlier: see Hammond 1972, 426); an open place with a sanctuary in the Septem and the Supplices.42 Each dramatist might then produce and bring his own scenery. 3. Raised stage. Hourmouziades (1965, 59–61) presents a case for an early stage. According to Hammond (1972, 411 ff.), Aeschylus introduced an impermanent ‘stage’, the ὀκρίβαντες, with the production of the Oresteia.43 The scant archaeological evidence that exists for an elevated stage—in the form of allegedly supporting cuttings for one, found in the stone foundation of the theatre—is most probably from the fourth century B.C., however (Rehm 1988, 279–81, with refs). There is no internal evidence from the early plays for a separate area for the actors; however, there is clearly an elevation of some kind, serving, for instance, as a lookout post for Danaus in 713 and perhaps at the beginning of the drama (see my note). FJ–W ii. 4, following Arnott (1962, 22), take this elevation to be in fact the raised stage, which should then have been at least a metre high;44 but there are other alternatives: see III. 4–6 below. 4. Elevation. Several passages in Aeschylus’ dramas indicate the presence of a heightened area on the orchestra, and it seems unlikely that this feature would simply be left to the imagination.45 One theory that has recently been 41 Cf. also Sommerstein (1996) 33–35, W.SA 13, Rehm (1988) 281–82. See also W.SA 48, 170. According to Polacco (1983, 74–76), even the altar and the gods in our play were ‘mostly’ painted images. 43 Cf. Arist. fr. 7 Bagordo (ap. Them. Or. 26.316d), Philostr. VS 1.492, Hor. Ars P. 279. 44 Cf. Taplin (1977) 441, Sommerstein (1996) 41. 45 The particularly relevant passages in the early plays are Pers. 659, Th. 240, Supp. 189, 713 (and implicitly in 508), but references to a rock abound, for natural reasons, in the Prometheus as well. See also my note on 351–52. There are also several references to an altar and/or a sanctuary that appears to occupy a demarcated space in the orchestra. See Melchinger (1974) 90–100, Hammond (1972) 416–25 for a detailed survey. 42 15 popular, and controversial, is that the repeated mention of a πάγος, ὄχθος, σκοπή, etc., in the texts of the early plays refers to an actual rock outcrop which stood at the north-east side of the orchestra, until possibly levelled at an (alleged) reconstruction of the theatre of Dionysus around 460.46 The theory was developed by Hammond (1972), independently also by Melchinger (1974, 20–22, 82–111, passim), with a forerunner in Flickinger (1930, 90 and fig. 6).47 It is embraced by, for instance, Taplin (1977, 448–49) and West (1979, 135–40), and denied by, for instance, Scullion (1994a, 42–49). Here, too, the evidence is inconclusive both ways. Poe (1989, 118–20) and Sommerstein (1996, 37–39) argue that the elevation was in fact the altar (see below). 5. Altar. We cannot say for certain that an altar was a permanent feature of the early stage. In respect of this issue, too, Clifford Ashby presents a comprehensive summary of the evidence and the scholarly debate;48 he also argues (1991, 18–21) that the altar was probably situated at the rear, not the centre, of the orchestra. As for archaeological evidence, there are no remains from the Theatre of Dionysus to suggest that the early orchestra was permanently equipped with a central altar: the hole found in the centre, which earlier archaeologists regarded as evidence for this feature, appears to relate to a Christian basilica from the fifth century A.D.49 On the other hand, archaeologists have found centre stones and similar things that might be interpreted as support for central altars in the orchestras of theatres outside Athens (Ashby 1991, 9–13). These are all from the fourth century and later, however; moreover, Ashby argues that ‘almost certainly these were building bench marks, not altar bases’ (ibid. p. 18). Some archaeological evidence from other theatres points to an altar located at the side or the rear, not the centre: so, for instance, a presumed altar base in the sixth-century theatre of Thorikos. 46 See, e.g., Melchinger (1974) 12–47, Taplin (1977) 449, 457. Cf. also Hammond (1988) 6–7 and passim on its possible use in Pers. and the post-Supp. plays. 48 Ashby (1991), a revised version appearing in Ashby (1999) 42–61. The erratic translation (not by Ashby’s own hand) of a relevant passage in the Suda does not diminish the overall usefulness of the article, although it ought to have been corrected in the second version. 49 Ashby (1991) 9, following Travlos (1971, 538, 549). 47 16 The painted vases give little reliable information,50 and the external literary evidence is scant and confusing, in all likelihood concerning itself with the later classical period, after the (alleged) mid-fifth-century reconstruction (see above, text for n. 46)—or even with the Lycurgean and/or Hellenistic theatre.51 The internal literary evidence confirms the presence of an altar in most 50 For two interesting examples of the possible depiction of tragic choruses dancing by an altar, see Poe (1989) 139. 51 This evidence involves two concepts of uncertain meaning, θυµέλη and ἀγυιεύς, which are spoken of by ancient scholars (Pollux and the Suda) as permanent features of the orchestra. Poe (1989) makes much of the latter term, taking it to refer to a column-shaped altar; but this is not relevant to the early plays, being used first in the Agamemnon, according to Poe (1989, 135). The θυµέλη is said by the Suda s.v. to be in fact the altar of Dionysus. The lexicon puts it ‘behind’ (µετὰ) the ‘orchestra’; however, the term ὀρχήστρα here means the raised stage, being opposite to the κονίστρα, viz. τὸ κάτω ἔδαφος. Thus the altar, according to the Suda, is placed in the middle of the orchestra (taken in its usual meaning), before the raised stage (on which see above, III 3). As for the term thymele, it usually refers to the orchestra or the stage as such (LSJ s.v. II b–c). It should not be used indiscriminately as a technical term for something which we do not really know existed and which, if it did exist, we do not know was actually so called. Phrynichus the Atticist claimed that the term, which in his time denoted the stage or the entire orchestra, was contemporary Greek, and not at all a theatrical term in Classical Attic (Eclog. 135, cf. PS 74): Θυµέλην· τοῦτο οἱ µὲν ἀρχαῖοι ἀντὶ τοῦ θυσίαν ἐτίθεσαν, οἱ δὲ νῦν ἐπὶ τοῦ τόπου ἐν τῷ θεάτρῳ, ἐφ’ οὗ αὐληταὶ καὶ κιθαρῳδοὶ καὶ ἄλλοι τινὲς ἀγωνίζονται. σὺ µέντοι, ἔνθα µὲν τραγῳδοὶ καὶ κωµῳδοὶ ἀγωνίζονται, λογεῖον ἐρεῖς, ἔνθα δὲ οἱ αὐληταὶ καὶ οἱ χοροί, ὀρχήστραν· µὴ λέγε δὲ θυµέλην. Pollux (4.123) writes that the thymele is a feature on the orchestra, εἴτε βῆµά τι οὖσα εἴτε βωµός. Thus he is not, pace Arnott (1962, 43–44), certain about what the thymele actually is, but apparently makes two conjectures with the aid of the literary sources available to him. If Pollux could not with any certainty identify the thymele as an altar, he probably did not have access to more crucial evidence than we, or the Suda. As is shown by LSJ s.v. θυµέλη II, our identification with the altar is actually based on a single passage (apart from the Suda), Pratin.Trag. fr. 3 (= Pratin.Lyr. fr. 708), where a satyric chorus is indignant at the emphasis on flute-playing in the orchestra: τίς ὁ θόρυβος ὅδε; τί τάδε τὰ χορεύµατα; | τίς ὕβρις ἔµολεν ἐπὶ ∆ιονυσιάδα πολυπάταγα θυµέλαν; If we take this fragment in isolation, however, θυµέλαν does not seem to refer to an altar, but rather to the entire orchestra—the sense common in later times —being the realm of Dionysus. The epithet πολυπάταξ, ‘very noisy’, suits this sense better than that of an ‘altar’. 17 dramas; and in the present one there actually are some indications that it may have been situated at the rear of the orchestra, unless the πάγον ἀγωνίων θεῶν in 189 is indeed located at the centre.52 Thus Sommerstein who argues, following Poe, that the supposedly central altar was raised on a mound and served as the elevation mentioned above.53 However, the mound must have been a considerable one if it were to contain, in the present play, twelve busts or statues (high enough to be able to hang oneself from), one altar, and thirteen persons sitting down (see 204–24n.). The juxtaposition of altar, gods, actors and chorus in the relevant scenes also becomes hard, not to say impossible, to visualise if taking place in the middle of the orchestra. It seems more plausible that the gods were situated on an elevation of some kind at the rear of the orchestra, and that the altar stood on the level ground in front of this elevation—or possibly on the elevation itself, in case this consisted of the raised stage (see above). The latter arrangement would in fact accord rather well with the stage settings described by the Suda s.v. θυµέλη (see above, n. 51). The exact details are unclear, but some arrangement of this kind does appear to serve as the sanctuary in which the Danaids sit as suppliants. See also my notes on 189, 204–24, 222–23, 345, 351–52. We have no definite evidence that an altar to Dionysus was a permanent feature of the orchestra: Ashby (1991, 20) points to some circumstantial evidence suggesting that the Dionysic sacrifice and ritual of the festival might have taken place at the nearby sacred precinct of Dionysus, which contained two altars, and not in the theatre at all. Accordingly the altar of the theatre may simply have been a stage-prop, and movable as well as removable in case no altar was needed for dramatic purposes. What about the romantic notion of a central altar around which the chorus danced in a circle? There is at least one piece of pertinent internal literary evidence: A. fr. 379, noted by Hourmouziades (1965, 75), where a chorus of women are ordered βωµὸν τόνδε … | κύκλῳ περίστητ’ ἐν λόχῳ τ’ ἀπείρονι 52 To complicate matters, the orthodox view has long been that the supposed central altar, hallowed to Dionysus, was not, on account of its religious sanctity, used as stage-property. So, e.g., Pickard-Cambridge (1946, 34, n. 2, 130–31), Arnott (1962, 45, 53), Hourmouziades (1965, 75). However, Rehm (1988) disputes this (as well as Tucker on the present drama, 196n.) and has been followed by most subsequent scholars expressing an opinion on the matter. 53 Sommerstein (1996) 39; Poe (1989) 118–20. 18 | εὔξασθε. This fragment was actually taken by Hermann (1820, 6 [324–25]) to belong to the Danaïdes. It is clear from the expression ‘boundless troop’ that κύκλῳ is not used extendedly to mean ‘half-circle’. The image of a chorus dancing in a circle around an altar recurs often in Greek literature,54 although perhaps never elsewhere in explicit connection with the Athenian theatre. The so-called ‘cyclic chorus’ is the chorus of the dithyramb. It is sometimes mentioned as distinct from the tragic chorus (Ar. fr. 156.10 PCG, Ath. 5.181c), which, however, only shows that the circular movement was seen a defining feature of the dithyramb. The tragic chorus might or might not move in a circle. 6. Terrace wall. A ‘terrace wall’, estimated at the height of about one metre, is supposed by Scullion (1994a, 28) to have existed at the back of the early orchestra (i.e. in the place of the later skene) in order to protect the actors against a steep fall that lay behind.55 This might have served as the abovementioned elevation. The only argument for this feature is the inference that something, in the absence of a skene, ought to have protected the dancers from the fall at the back of the orchestra. 54 To the refs of LSJ s.v. κύκλιος II we may add E. HF 925–27 (circular chorus around an altar), IT 428–29, Hel. 1312–13, IA 1055–57, Ar. Th. 954–59, 968 (circular chorus in general). 55 Cf. Noack (1915) 3, Dörpfeld–Reisch (1896) 26, Pickard-Cambridge (1946) 8, Melchinger (1974) 85 ff. 19 Supplices 1–523: Translation The translation is as literal as possible (perhaps in some cases even more so). Epithets of gods are usually not translated, but transliterated and put in italic type: please refer to the commentary for explanations. Personifications of abstract qualities are usually translated. Ὕβρις, a very central concept of the drama, is rendered throughout as Hybris. Footnotes indicate all the places where the translation is based on a text that differs significantly from West’s Teubner edition. Angled brackets indicate that the corresponding words have been supplied in the Greek text; braces indicate that I regard the words as interpolated or misplaced, and cruces that the text is too corrupt to make sense of. An asterisk by the interlocutor’s name means that the (change of) speaker is not indicated in the mss. by name or paragraph. Parodos Enter Chorus of Danaids in single file, chanting anapaests as they order themselves on the stage. Danaus probably comes last, entering the stage at v. 11 and climbs an elevation, watching for followers. The anapaestic periods are indicated with paragraphs. *Chorus: 5 10 15 —May Zeus Aphictor benignly oversee our nautical expedition, which set out from the soft-sandy mouths of the Nile. Having left God’s country, with pasturages that border on Syria, we flee, not a flight of banishment because of bloodshed, sentenced by the voting-pebble of the State, but a self-chosen flight from men, denouncing marriage to Aegyptus’ sons as both impious <and dishonourable>.56 Danaus, our father, head of counsel and of faction, arranged the gaming table and brought this to pass, best of sufferings: to flee unbridledly by the ocean wave, and land on the earth of Argos, even the place whence our race, asserting to be of the touch and breath of Zeus upon the gadfly-driven cow, was created. 56 20 ἀσεβῆ τ’ ὀνοταζόµεναι <καὶ ἄτιµον>. At which land more benign than this could we arrive with these suppliants’ tools, these wool-wreathed boughs? <O paternal gods of Argos,> to whom the city, to whom the land and the clear water belong; you high gods, and you chthonians of heavy vengeance, possessing the tombs, and Zeus Soter as the third, house-guardian of pious men, may he57 receive the suppliant female expedition with a reverent air from the land: but before the outrageous manfilled swarm begot by Aegyptus put foot on this muddy dryland with swift-rowing coach send them seaward: may they perish there in the stormbeating hurricane, meeting with the thunder, the lightning and the rain-bringing winds of the savage sea, before they can mount the unwilling couch that Right prevents, having appropriated this cousinhood. 20 25 30 35 First Ode (stasimon) Str. 1 Ant. 1 Calling now on the calf of Zeus, avenger from beyond the sea 40 and flower-grazing son of the ancestral cow by the breath of Zeus: in the name, the fated time significantly fulfilled his 45 touch [ephapsis]:58 he begot Epaphus. Having called him forth now in the grass-pasture haunts of his 50 ancient mother; having mentioned the former woes, what I shall now show forth will appear as sure proof, albeit being unexpected to the inhabitants of the land: but one will under- 55 stand in the length of the tale. 57 δέξαιθ’. ἶνίν τ’ ἀνθονόµον | τᾶς προγόνου βοὸς ἐξ ἐπιπνοίας | Ζηνός· ἔφαψιν ἐπωνυµίᾳ δ’ ἐπεκραίνετο. 58 21 60 65 Str. 2 Ant. 2 70 Str. 3 75 If some native seer of birds nearby happens to hear lament, he will believe to hear some voice 59 of the sad thought of Tereus’ wife, the hawk-driven nightingale, who, debarred from the lands and rivers,60 cries a new lament over her old haunts: she tells the fate of her child, how it perished through murder by her own hand, meeting with dysmaternal wrath. Thus I too, much-lamenting in Ionian strains, tear my soft sunwarmed cheek and my tear-inexperienced heart. I don a garland of laments, fearing on this friendless flight from the Aerian land, lest a guardian should not appear. 85 Ancestral gods, listen well and see that which is just: by not giving youthful prime to be had in fulfilment beyond what is allotted, and truthfully hating Hybris, you should be fair in regard to marriage. There is, even for fugitives torn by war, the altar, a ward against ruin, reverence of the deities. Str. 4 95 May Zeus’s—if really truly Zeus’s—desire set things right. It is not easily tracked: for rugged and shadowy do the paths of his heart extend, impossible to descry. Ant. 4 90 But safely and not on its back does a matter land, if by the nod of Zeus it is destined to be fulfilled. Everywhere it blazes forth, even in black darkness,61 with fortune for the mortal folk. Str. 5 It hurls mortals from high-towering hopes to utter ruin, but does not array force: all is effortless for the divine. Seated above, it fulfils a thought 62 completely from afar, from the holy seat. 80 100 Ant. 3 59 60 22 δοξάσει τιν’ ἀκούων ὄπα. ἀπὸ χώρων ποταµῶν τ’ ἐργοµένα. 61 κἀν σκότῳ κελαινῷ. 62 ἥµεν’ ἄνω φρόνηµά πως. Ant. 5 Str. 6 Ephymn. a Ant. 6 Ephymn. a Str. 7 Ephymn. b Ant. 7 Ephymn. b 63 64 May it look at mortal Hybris, how it juvenesces,63 a stock that 105 blooms in transgressing ill-purposing minds through our marriage with frenzied intention as a goad inescapable, deceiving 110 the unclean with infatuation.64 Such miserable sufferings do I wailingly recount, shrill, burdensome, tear-inducing! Ah! ah! conspicuous by lament, while still 115 alive I pay my own respects with wailings. I beseech the grace of the Apian hilly land: well do you know, Earth, my barbarian speech. Repeatedly I fall upon my Sidon- 120 ian veil with linen-tearing rent. To the gods sacred offerings stream forth, if things go well, where death be absent. Oh! Oh! Oh toils inscrutable! Where will 125 this wave carry me off? I beseech the grace of the Apian hilly land: well do you know, Earth, my barbarian speech. Repeatedly I fall upon my Sidon- 130 ian veil with linen-tearing rent. The oar and the linen-stitched house that keeps the sea out of the 135 hull sent me free of storm with the breezes, and I do not find fault: but may our all-seeing Father in time render the outcome gently: 140 that the great seed of the revered mother escape the beds of men, oh! unwedded, unconquered. May the holy daughter of Zeus, safe with reverend countenance, 145 willingly behold me who wills it; and †… with all her might against our followers, may she become the unconquered saviour 150 of us as are unconquered: that the great seed of the revered mother escape the beds of men, oh! unwedded, unconquered. ὕβριν | βρότειον οἵα νεάζει. ἄτᾳ δ’ ἀπατῶν ἀνάγνους. 23 155 Str. 8 160 165 170 175 If not, we the black sun-beaten race shall come before the Earthen, the Much-hospitable Zeus of the Deceased, with our boughs, the nooses by which we die, without having met with the Olympian gods. Ah Zeus! for Io, oh! inquisitorial wrath from the gods. I know 65 Mesode the heaven-conquering malice of the wife: for out of harsh wind comes storm. Ant. 8 And shall not Zeus, then, be subject to rightful censure as having dishonoured the son of the cow, he whom he himself once created by engendering, if his gaze is averted now in our prayers? May he listen well from above being called. Danaus addresses his daughters from the top of the hill: 180 185 190 195 Danaus: —Children, there is need for prudence. You have arrived with this your prudent, dependable old shipholding father; I have now also taken precautions as regards the dry land, and I urge you to retain my words writing them down in your heart. I see dust, speechless messenger of a host: axle-driven hubs are unsilent. I see a shield-covered, lance-wielding crowd, with horses and rounded chariots. Perhaps the princes of this land come to us as onlookers, having heard news from messengers. However, whether unhostile or whetted with savage rage it speeds this array, the best is, on all accounts, girls, to sit nearby the rock of these Gods of the Assembly: greater than the fortlet is the Altar, a shield unbreakable. Now go as fast as possible and, with suppliants’ boughs wreathed in white, sacraments of Zeus Aidoios in your left hands, exchange pitiful, plaintive and not useless66 words with the strangers, as beseems foreigners, clearly recounting this bloodless flight. Attached to the voice should be, first, that which is not bold: may also nothing inane come out of faces intelligent with quiet eyes. Be 65 24 γαµετᾶς οὐρανόνικον. 66 κοὐκ ἀχρεῖ’. Ch: *Dan: *Ch: *Dan: *Ch: *Dan: *Ch: *Dan: *Ch: *Dan: *Ch: *Dan: *Ch: *Dan: *Ch: not talkative, nor laggard in your speech: the people here are very easily offended. Remember to yield! You are in need, a stranger and a fugitive. Bold talk befits not those that are inferior. <See to this and exchange words this way, so that this matter may prevail happily for you.>67 —Father, prudently you speak to prudent people. I shall take care to remember these sage admonishments of yours: may Zeus Gennetor behold!68 —Now do not tarry! Let there be strength to carry out the plan. —I will presently take a seat nearby you. O Zeus! look to us and have pity, that we are not destroyed. —May he indeed behold with gentle eye. —Him willing, this will end well. —Now call upon yonder bird of Zeus. —We call the delivering rays of the Sun. —And holy Apollo, the god that was a fugitive from heaven. —Knowing this fate, he would take pity on us mortals. —May he indeed take pity, and graciously stand by. —Whom among these deities should I call upon further? —I see yonder trident, the sign of a god. —He brought us here well, and well may he receive us upon the earth. —This next one is Hermes,69 according to the custom of the Greeks. —Let him announce good news to people free. 200 205 210 215 220 67 … οὐ πρέπει τοὺς ἥσσονας. | 232–33 <σκοπεῖτε κἀµείβεσθε τὸνδε τὸν τρόπον | ὅπως ἂν ὕµιν πρᾶγος εὖ νικᾷ τόδε> | — πάτερ, φρονούντως. 68 205–12 as follows: φυλάξοµαι δὲ τάσδε µεµνῆσθαι σέθεν 205 κεδνὰς ἐφετµάς. Ζεὺς δὲ γεννήτωρ ἴδοι. <∆αναός> µή νυν σχόλαζε, µηχανῆς δ’ ἔστω κράτος. <Χο.> θέλοιµ’ ἂν ἤδη σοὶ πέλας θρόνους ἔχειν. ὦ Ζεῦ, σκοπῶν οἴκτιρε µὴ ἀπολωλότας. <∆α.> ἴδοιτο δῆτα πρευµενοῦς ἀπ’ ὄµµατος. 210 <Χο.> κείνου θέλοντος εὖ τελευτήσει τάδε. <∆α.> καὶ Ζηνὸς ὄρνιν τόνδε νῦν κικλῄσκετε. 69 Ἑρµῆς ὅδ’ ἄλλος. 25 *Dan: 225 230 —Now pay reverence to the common altarship of all these lords. Then sit down in the sanctuary like a cluster of doves in fear of hawks, alike winged, kin enemies, defiling the race. How could a bird that eats of a bird be pure? How could one take an unwilling woman as bride from an unwilling father-inlaw, and be pure? No, not even in Hades shall he who did this escape responsibility after death for his abysmal acts. There, they say, among the deceased does another Zeus adjudge final sentences for misdeeds. {See to this and answer this way, so that this matter may prevail happily for you.}70 Enter Pelasgus, lord of Argos, with a retinue of armed guards. *Pelasgus: —Of what nation is this un-Hellene dressed crowd that we address, that revels in foreign clothing and head-bands? For these women’s dress is not Argolic, nor of the lands of Hellas. How you fearlessly dared to tread the land, unheralded and 240 without patron or guide, this is a wonder. However, boughs from you, in the custom of beseechers, rest with the Gods of the Assembly: only in this shall the land of Hellas agree with a guess. As to the other matters, it would have been reasonable to make more guesses still, had it not been that the one stand245 ing before me has a voice that can explain. Dan: —You speak without falsehood as regards the habit. But may I address you as a citizen, as a warden with holy staff, or as a leader of the state? Pel: —To this request <I shall present trustworthy information; you, in turn,> answer me and speak without fear:71 for I am 250 the son of the Earth-born Palaechthon, Pelasgus, leader of this country. Reasonably named after me, as the ruler, the race of Pelasgians harvests this land. All the country, through 255 which runs the holy Strymon, do I rule, the part on the side of the setting sun. My border is the earth of the Paeonians; 235 70 232–33: see after 203. πρὸς ταῦτ<α δείξω µὲν τὰ χρὴ τεκµήρια· | σὺ δ’ αὖτ’> ἀµείβου καὶ λέγ’ εὐθαρσὴς ἐµοί. 71 26 the land nearby Pindus, close to Perrhaebiae;72 and the mountains of Dodona. The border of the wet ocean fences us in. Of this do I rule what is on the hither side. As for the Apian land itself, this plain has long been called thus on account of a healer: for Apis, a seer-healer and son of Apollo, came from the opposite shore of Naupactus and cleansed this land of mandestroying beasts, which the earth had released †wrath ...† as it had been stained by ancient blood: an unbearable dragoncrowded co-existence. Blamelessly effecting remedies from this for the Argive land by knife and solvent, Apis then as reward received mention in prayers. Now that you have the information from me, you may proclaim your race and speak further: however, the state does not love long speeches. *Ch: —The story is short and plain: we claim to be of Argive race, seed of the cow blessed in children. And for the truth of this I shall give full account.73 Pel: —Unbelievable for me to hear, o strangers, do you speak, that this race of yours should be Argive. You are rather more like Libyan women, and in no wise like the native ones. The Nile might also foster such a plant. {A Cypriot mark in female forms has been beaten alike by male craftsmen.} And I have heard that there exist female Indian nomads who traverse the land led on camels that go like horses, neighbouring on the Ethiopians. I would certainly have likened you to the husbandless meat-herding Amazons, had you been carrying bows. I ought to understand this better if informed, how your race and seed can be Argive. *Ch: —They say that Io once was the Keyholder of Hera’s house in this Argive land. *Pel: —She was that indeed, and widely the tale prevails. *Ch: —Is there not another story, that Zeus consorted with the mortal? 260 265 270 275 280 285 290 295 72 256–57 ὁρίζοµαι δὲ τήν τε Παιόνων χθόνα | Πίνδου τε τἀπέκεινα, Περραιβῶν πέλας. 73 χὡς ταῦτ’ ἀληθῆ πάντα προσφύσω λόγον. 27 300 305 310 315 74 —Yes, and this †…† was not hidden from Hera.74 —How did this quarrel of royals end? —The Argive goddess turned the woman into a cow. —Does not Zeus yet again approach the well-horned cow? —So they say, in the form of a cow-mounting bull. —What then did the mighty wife of Zeus in answer to this? —She set the warden that sees everything over the cow. —Who is this all-seeing single-heifer herdsman of whom you speak? *Pel: —Argos, the son of Earth whom Hermes slew. *Ch: —What then did she contrive yet more for the ill-fated cow? *Pel: —An urging, cow-driving gadfly… *Ch: —By the Nile they call it Oestrus. *Pel: — … for indeed it drove her off the land in a long course. *Ch: —All this which you have said accords with my claim. *Pel: —Yes, she came also to Canobus and to Memphis. *Ch: —And Zeus Ephaptor engendered offspring with his hand. *Pel: —Who then does Zeus’s calf by the cow claim to be? *Ch: —Epaphus, in truth named after the seizing. <*Pel> <—Who was begot by Epaphus?> . *Pel: *Ch: *Pel: *Ch: *Pel: *Ch: *Pel: *Ch: 296–13 as follows: κοὐ κρυπτά γ’ Ἥρας ταῦτα †παλλαγµάτων <Χο.> πῶς οὖν τελευτᾷ βασιλέων νείκη τάδε; <Πε.> βοῦν τὴν γυναῖκ’ ἔθηκεν Ἀργεία θεός. <Χο.> οὐκοῦν πελάζει Ζεὺς ἔτ’ εὐκραίρῳ βοΐ; <Πε.> φασίν, πρέποντα βουθόρῳ ταύρῳ δέµας. <Χο.> τί δῆτα πρὸς ταῦτ’ ἄλοχος ἰσχυρὰ ∆ιός; <Πε.> τὸν πάνθ’ ὁρῶντα φύλακ’ ἐπέστησεν βοΐ. <Χο.> ποῖον πανόπτην οἰοβουκόλον λέγεις; <Πε.> Ἄργον, τὸν Ἑρµῆς παῖδα γῆς κατέκτανεν. <Χο.> τί οὖν ἔτευξ’ ἔτ’ ἄλλο δυσπότµῳ βοΐ; <Πε.> βοηλάτην µύωπα κινητήριον. <Χο.> οἶστρον καλοῦσιν αὐτὸν οἱ Νείλου πέλας. <Πε.> τοιγάρ νιν ἐκ γῆς ἤλασεν µακρῷ δρόµῳ. <Χο.> καὶ ταῦτ’ ἔλεξας πάντα συγκόλλως ἐµοί. <Πε.> καὶ µὴν Κάνωβον κἀπὶ Μέµφιν ἵκετο. <Χο.> καὶ Ζεύς γ’ Ἐφάπτωρ χειρὶ φιτύει γόνον. 28 296 298 300 305 310 311 313 —Libya, who reaps the greatest <harvest> from the earth. —And who do you say is the next offshoot of this woman? —Belus with two sons, the father of my father —Tell me now his all-sagacious name. —Danaus. And there is a brother who has fifty sons … —Reveal also the name of that one with ungrudging speech. —Aegyptus. Knowing of our ancient race, you should act to take our Argive party in.75 *Pel: —You do appear to me to partake of this land of old. But how is it that you ventured to leave your father’s house? What fate did strike? *Ch: —Lord of Pelasgians, human evils are shifty: nowhere will you see toil of like plumage. For who would have thought that this engagement, a former blood-relationship, would strand its unexpected flight in Argos, changing plumage76 through loathing of the wedding bed? *Pel: —Why do you mean that you beseech these gods of the Assembly with newly plucked boughs wreathed in white? *Ch: —So that I may not be a slave to the race of Aegyptus. <Pel:> <—But it is customary for women to subject themselves to their husbands. <Ch:> —We will choose death before our cousins’ unclean beds>77 Pel: —Because of enmity, or do you mean that it is not proper? *Ch: —Who would think their owners78 to be friends? *Pel: —In this wise strength will increase among mortals. *Ch: —And riddance of the unfortunate is easy. Pel: —How then may I act piously towards you? Ch: —By not releasing us again to the sons of Aegyptus when they demand it. Pel: —You speak grievously, of taking on a new war. Ch: —But Justice stands by her allies. Pel: —If she was a party to the matter from the beginning. *Ch: *Pel: *Ch: *Pel: *Ch: *Pel: *Ch: 320 325 330 335 340 75 πράσσοις ἂν ὡς Ἀργεῖον ἀνστήσαις στόλον. µεταπτερωθὲν. 77 lacuna after 335. 78 τίς ἂν φίλους οἴοιτο. 76 29 345 Ch: —You must revere the stern of the state crowned in this wise. *Pel: —I shudder as I see these shadowed seats. Certainly the wrath of Zeus Hikesios is great. Second Ode (kommos or amoibaion) 350 Ch: Child of Palaechthon, hear me with gracious heart, o lord of Str. 1 Pelasgians. See to me, the suppliant fugitive, wandering like a wolf-hunted heifer on a high cliff, where, trusting its protection, she bellows, telling the herdsman of her hardships. Pel: 355 360 —I see, shaded by newly-plucked boughs, the †... party of the gods of the Assembly. May this matter of citizen-foreigners be free from harm, and may not conflict arise for the State out of matters unhoped for and unforeseen: for the State is in no need of that. Ch: May indeed Themis Hikesia of Zeus Klarios oversee an unAnt. 1 harmed flight. But you, learn from the late-born, thinking old: if you revere the approaching suppliant you shall not want: the gods’ receipts <come forth very> acceptant of sacrifice from a man who is pure.79 —You do not sit at the hearth of my own house. If the state is being defiled in its commonality, the people must work together to find a cure. I would not effect a promise before, but only after sharing this with all the citizens. 365 Pel: 370 Ch: You are the State, you are the Public: a prince unaccountable Str. 2 do you govern the altar, hearth of the land, with your singlevote nods: in your single-sceptre throne do you accomplish every charge. Beware the defilement! 375 79 30 οὐ λιπερν<ής ποτ’ ἔσῃ· πρόσεισίν γε µάλ’> ἱεροδόκα θεῶν λήµµατ’. Pel: —May defilement come over my adversaries! But you I cannot ward without incurring blame. Nor is this prudent, to dishonour these prayers. I am at a loss and my mind is held 380 by fear, of acting and of not acting; of seizing fate Ch: See to him who sees from above, guardian of much-suffering Ant. 2 mortals who, sitting before their neighbours, do not meet with lawful justice. The wrath of Zeus Hiktaios awaits those who 385 are hard assuaged by the wailings of the sufferer.80 *Pel: —If the sons of Aegyptus do govern you according to the law of the State, claiming to be the closest in birth, who would want to oppose them? You must plead according to the laws 390 of your homeland, that they do not have any authority over you. 80 81 Ch: Str. 3 May I never be subject to the power of men! I make the stars the limit of the means for my flight81 from a malicious wedding. Take Justice as an ally and adjudge that which is pious before 395 the gods. Pel: —The judgement is not easy to make: do not choose me as judge. As I said before, I would not do this without the commons, even if I do rule, so that the people might not say, if 400 somehow something less fortunate should occur, ‘honouring immigrants, he destroyed the State’. Ch: Ant. 3 Zeus oversees both parties of the kindred ones in this matter, weighing each and reasonably administering that which is unjust to the wicked and that which is hallowed to the lawful. While these do balance the scales equally, why do you agonize 405 about doing what is right? µένει … δυσπαραθέλκτοις. φυγᾷ. 31 Pel: 410 415 420 425 —There is need for a deep delivering thought, for a perceiving eye to go like a diver into the depth, not overly intoxicated, so that all this, first, may be without harm to the city, and also end well for ourselves, and that neither Strife may lay hold of spoils, nor that we, handing you over when settled thus at the seats of the gods, make Vengeance, the alldestroying god our grave neighbour, he who does not even in Hades release him who has died. So does there not seem to be need of a delivering thought? Ch: Be thoughtful and be an all-justly pious patron: do not betray Str. 4 the fugitive, she who sped from afar due to godless castings-out. Do not look on as I am dragged from the seats of many gods, o Ant. 4 you who have all the power over the land: recognise the Hybris of men, and beware of the wrath. 430 Suffer not to see the suppliant led away from the gods’ images, Str. 5 in violation of the law, in the manner of a horse, by the headbands, and seizures of my many-threaded robes. 435 For know this: whichever you establish, that shall remain for Ant. 5 your children and your estate, to pay † … † equal retribution. Consider this. Zeus’s justice prevails. Pel: 440 445 82 —I have indeed considered. Here it founders: to raise a great war, either with these or with those, is entirely necessary. The hull is bolted, being thus drawn forth with seamen’s turning windlasses. There is nowhere to halt without misery. If goods are plundered from an estate, that which someone carries off filling up the greatest cargo may be replaced by other goods by the grace of Zeus Ctesios:82 444–45 ἅ τις φέρει µέγιστον ἐµπλήσας γόµον, | γένοιτ’ ἂν ἄλλα Κτησίου ∆ιὸς χάριν. 32 *Ch: Pel: Ch: *Pel: *Ch: Pel: Ch: *Pel: Ch: Pel: Ch: Pel: Ch: *Pel: 83 and in case the tongue shoots forth something unseasonable, <painful darts that agitate the soul>, word may become the healer of word {painful, that sorely agitate the soul}. But in order that kindred blood may not be shed, there is dire need to sacrifice and for many oracular offerings to fall to many gods, remedies of calamity. In truth I am lost in this conflict. And truly I would rather be inexperienced than wise of evils. May it go well, contrary to my expectations. —Hear the final end of many pitiful words. —I hear, and you may speak: it shall not escape me. —I have bands and girdles that hold my robes together. —Perchance these are things that are seemly for women. —Know now of a fine device which is made out of these. —Tell me, what sound is this that you will utter. —If you will not make a pledge of trust to this party… —Then what is it that your device of girdles will effect? —Adorn these images with new votive tablets. —Your words are like a riddle: speak clearly. —To hang ourselves, as swiftly as possible, from these gods. —I heard a speech that scourged the heart. —You understand. For I presented you with a clearer sight. —†…† hard-wrestled matters on every side.83 The mass of ills comes forth like a river. I have gone down into an unfathomable, far from easily travelled sea of ruin, and nowhere is there a harbour from the ills. For if I do not exact this charge to you, you spoke of a defilement that is unsurpassable. Then again, if with your kin, the sons of Aegyptus, I shall come to stand before the city-walls for the purpose of battle, how cannot the loss be bitter, that men make bloody the ground for the sake of women? Yet one must revere the wrath of Zeus Hikter: the greatest fear among mortals. You, aged father of 450 455 460 465 470 475 480 †καὶ µὴν† πολλαχῇ γε κτἑ. 33 485 490 495 500 these maidens,84 quickly take these boughs into your arms and put them on the other altars of the gods of the land, so that all the citizens may see the sign of this beseechment, and that word may not be cast against me: people are fond of accusing the leadership. And perchance someone may take pity seeing these, and come to loathe the Hybris of the party of men, whereas towards you, the people may be more favourably inclined. For everyone carries good-will towards those that are inferior. *Dan: —Highly is this valued by us, to have found a reverent patron. Do send with us attendants and guides out of the local people, so that we may find the temple-front altars of the gods that guard the city and their †... seats, and that there may be safety for us as we walk through the city: the nature of our appearance is not attired similarly to here: for the Nile does not foster a race alike to the Inachus. Take care that boldness does not engender fear: it happens that people will kill even a friend because of ignorance. Pel: —You may go, men: for the foreigner speaks well. Show him the altars and the seats of the gods. And it is important not to speak much to those that you meet, leading this seafaring suppliant of the gods. Exit Danaus with the Argive soldiers. 505 —You spoke to him, and he may go as ordered: But I, what shall I do? Whence will you assign boldness to me? Pel: —As for the boughs, you may leave them here, as a sign of your hardship. *Ch: —And so I leave them, subject to your words. *Pel: —Now turn down toward this level grove. *Ch: —And how should a grove that may be trodden ward me? Ch: 84 34 No lacuna after 480. *Pel: *Ch: *Pel: *Ch: *Pel: *Ch: *Pel: —We will not surrender <you> to the rape of winged beasts. 510 —But what of those that are more loathsome than malicious dragons? —May the word of her who has been auspiciously addressed not be inauspicious. —It is no wonder if one is distressed though fear in the heart. —to relieve excessive fear is a matter for rulers.85 —May you gladden my heart in deed as well as in word. 515 —Your father will not be abandoning you for a long time. As for me, I shall call together the people of the country and try to persuade them,86 so as to induce the community to be benevolent: and I shall tell your father how he ought to speak. You stay here and wait for this, and ask the gods of the coun- 520 try for that which you desire to receive. I shall go and attend to these matters: may Persuasion follow, and Providence of Action. Exit Pelasgus. 85 86 λύειν ἀνάκτων ἐστὶ δεῖµ’ ἐξαίσιον. ἐγχωρίους | πείσω. 35 Commentary The lemmata in bold type represent the text as I would read it (in the cases where cruces are added, as the mss. read it). The line-enumeration is conventional (the same as appears in, for instance, Page, FJ–W, and West) and ultimately based, I believe, on Wilamowitz’ edition (see West pp. liii–liv). In the lyrical passages it is often at odds with modern colometry (see ibid.). 1–39: Parodos The action takes place in the orchestra, which was possibly, but not certainly, circular in shape. There are a number of man-high busts or statues representing Greek deities present (see 204–24n., 220–21n.), and an altar. There is also an elevated area of some kind, on which the idols, if not the altar, are placed: a natural rock, the wall at the back of the orchestra, or possibly a wooden ‘stage’ of some kind. There may or may not have been a stage building at this time, alternately a movable façade, decorated with suitable scenery (according to some the idols were painted images only).87 Enter chorus of presumably twelve women (as for Danaus’ entrance, see below).88 The girls have a dark complexion (i.e. masks indicating as much; see 154–55n.) and ‘barbaric’ costumes (120–22 = 131–33, 234–37, 432), and they carry suppliants’ boughs wreathed in white wool (21–22, 191–93, etc.). 87 See chapter III in the Introduction for a short, mainly bibliographical exposé on the early theatre. 88 As for the size of the tragic choruses, the evidence is scant, late, and mostly seems to be based on inferences from the text of the extant tragedies. An exemplary exposé, because it includes references to the ancient evidence, is found in Haigh (1907) 288–90, who, however, puts too much trust in authorities like Pollux (see Taplin 1977, 437). It is unlikely that the tragic chorus ever consisted of fifty members (Poll. 4.110): see E. Reisch in RE iii. 2390–91 (s.v. ‘Chor’), Taplin (1977) 47. The Suda s.v. Σοφοκλῆς (= A. test. 2) has it that Sophocles increased the number of choreutae from twelve to fifteen (Lesky 1971, 314, suggests that this increase was advocated in his lost theoretical treatise περὶ χοροῦ, mentioned in the said Suda article); Taplin (l.c., 323, n. 3) suspects that they always numbered fifteen. Some internal evidence from the Supplices may favour the number twelve: see 204–24n. The peculiar stichomythia in Ag. 1343–71, where each individual member of the chorus speaks in turn, may be interpreted as evidence for either twelve or fifteen choreutae: see Hammond (1972) 419, n. 58. 36 Possibly half or a third of the chorus are marked as being of inferior rank, and perhaps not carrying boughs: they would be the Danaids’ handmaidens, but will not be revealed as such until towards the end of the drama (954, 977). See further 204–24n. The Danaids chant anapaests as they march into the orchestra,89 a metre which is always employed when the opening of a drama is performed by the chorus, and also by single actors in the beginning of Euripides’ Andromeda (fr. 114 Nauck) and the probably interpolated (so Diggle in the latest OCT) dialogic opening of the Iphigenia in Aulis.90 It is unlikely that a choral opening had an ‘archaic’ flavour at the time of the first staging of the Supplices (cf. Taplin 1977, 61–64). The poet still restricted himself to two actors, and the psychology of economy which must accompany this restriction would make him sparing of minor parts:91 Persae has four actor’s parts in all, Septem contra Thebas three, Supplices three. The numbers are doubled in the Oresteia, in which a third actor has been introduced: its parts have six, seven, and six personae respectively. If for dramatic reasons the poet wanted to delay the entrance of the main character(s), he would probably have been less inclined, under the restriction to two actors, to let a minor figure speak the prologue, letting the chorus begin instead (so in the Persae; in Septem contra Thebas Eteocles, the main character, speaks the prologue).92 There is no indication of Danaus’ entrance, and there is some controversy as to where, exactly, it took place: at the very start, together with the chorus, or at the beginning of his own speech in v. 176. The former alternative is preferred by modern critics, and Mazon, among others, is criticised by Taplin (1977, 194) for the suggestion that Danaus may have ‘entered at some point during the song or even at the end of it’. But that is not really Mazon’s view; 89 On ‘marching’ anapaests, see, e.g., Masqueray (1892) 119–23, Nestle (1930) 72, Dale (1968) 47–48, West (1982) 53–54, Scullion (1994a) 73–74 and n. 18. 90 ‘Prologue’ as a technical term is reserved for opening passages spoken by actors; opening choral anapaests constitute a parodos. See Arist. Po. 1452b15–20 and cf. 40– 175n. and n. 159 below. Choral parodoi are found in Pers., Supp., Myrm. (fr. 131), Pr.sol. (fr. 190–92), probably Niobe (cf. Ar. Ra. 911 ff. [test. 120]), and [E.] Rh. See further Taplin (1977) 63–64 and Nestle (1930) 14–17. 91 Cf. Taplin (1977) 215, Easterling (1997) 153. 92 On the other hand Phrynichus is said to have let a eunuch perform the prologue in the Phoenissae (fr. 8); this, however, is contested by Arnott (1962, 70). On the prologue of Phrynichus’ play, see Scullion (2002a) 97–98 with refs. 37 Mazon only states that Danaus is ‘entré dans l’orchestre derrière ses filles’ and that he has been standing on a hill for a long time observing the horizon (p. 19). Obviously all thirteen participants cannot have entered the orchestra at exactly the same moment. The most likely scenario is that they walked in single file, with Danaus either at the front or, more probably, last (this is Mazon’s view as I understand it), to cover their backs from the pursuing enemy and to accentuate his subordinate role in the drama. This calls for some speculation. At the first mention of Danaus in v. 11, the anapaestic stretch has comprised exactly forty feet (if we accept the probable notion that the last two feet have fallen out at the end of v. 10, q.v.). Each foot will coincide with a step taken by the members of the chorus (see above, n. 89). One possible scenario may then be that the chorus enters and takes its place in the middle of the orchestra during the first forty anapaestic feet: the number of steps fits in reasonably well with the space they would have to travel,93 supposing that they stationed themselves in three rows of four persons each. Then, delayed for a little while, Danaus comes in behind his daughters, so that his entrance coincides with, or takes place just before, their mention of him in v. 11 (∆αναὸς δὲ κτἑ). Since he appears behind their backs, there is no typically deictic phrase like ἀλλ’ ὅδε γὰρ ∆αναός at his entrance; neither should we want one, since this would create an expectation of a speech. Instead, he silently takes his place behind the chorus, probably standing on the elevated area (see 189n.), cautiously observing the horizon.94 1. µὲν is to be taken as primarily ‘inceptive’ (D.GP 382–84), but δέ in v. 11 (q.v.) may perhaps contain a hint of a response. Ἀφίκτωρ: probably an Aeschylean innovation, here apparently instead of Ἱκέσιος (347, 360, 616, cf. also 385 Ζηνὸς … Ἱκταίου, 478–79 Ζηνὸς … Ἱκτῆρος). ἱκνέοµαι is used in the sense of ‘supplicate’ in Homer and the drama,95 and Aeschylus has allowed derivatives of ἀφικνέοµαι to take this mean93 With two steps measuring slightly more than one metre and the orchestra having a radius of 12–13 metres. On the size of the orchestra, see Scullion (1994a) 17–28 for a survey of scholarly opinion from Dörpfeld–Reisch (1896), who first estimated the radius at 12 m, and onwards. 94 On silent and multiple action in drama see now Slater (2002–3), who however concentrates on the post-Oresteia period (p. 346) and in particular on Sophocles and Euripides. 95 LSJ s.v. II.3. For an attempt to outline the relationship between ἱκνέοµαι and ἱκετ-, see Létoublon (1980). 38 ing in the Supplices, too (also at 241 ἀφίκτωρ, 483 ἄφιξις). The explicit sense ‘Zeus the Suppliant’ should be avoided, however, being incompatible with the notion of Zeus in general, and the Aeschylean Zeus in particular.96 The omnipotence of Zeus is particularly stressed in the present drama (see 86–103, 590–99, 816, 1048–49 with notes).97 Whereas gods often take epithets from the activities of their worshippers (cf. Verdenius 1985), these usually come with a blander suffix, usually -ιος, meaning simply ‘the god connected with this or that activity’—for instance Ζεὺς Ἀγοραῖος, Ἀρότριος, Κτήσιος and Ὅρκιος.98 Here, however, the suffix -ωρ necessitates taking the epithet as referring to an activity actually pursued by Zeus.99 It should perhaps be taken in the broadest possible sense, ‘he who arrives’: Zeus, by extension, becomes a protector of those ἀφίκτορες who come as suppliants, without being a suppliant himself. Dobias-Lalou (2001) suggests that ἀφίκτωρ should be taken as semantically akin to the verb ἀφικετεύω and the noun ἀφικετεία found in inscriptions from Cyrene, Rhodes, Cnidus, and Cos from the fourth and third century B.C.:100 ‘intercede for a suppliant’ and ‘intercession’, LSJ (Supplement). This would be 96 Notwithstanding Ζεὺς Ἱκέτης appearing in a Spartan inscription: IG v 1.700 = GDI iii 2.4407 = DGEE 1; cf. also SEG 43.134. On this inscription see C.Z. ii. 1096, n. 1. It is not Ζεὺς Ἱκέσιος as FJ–W claim, but ∆ιοhικέτα (= ∆ιὸς ἱκέτου, probably). LSJ reject the opinion of Eust. ii. 129 Stallbaum (on Od. 16.422) that ἱκέτης may mean ἱκετευθείς: Cook (C.Z.) l.c. suggests that ‘Suppliant Zeus’ in the inscription refers to ‘the very primitive notion that a stranger suddenly appearing in the midst of the community may well be a god on his travels’. For the evidence for a cult of Ζεὺς Ἱκέσιος, on the other hand, see IG xii 3.402–3, Inscr.Cos. 149 (= SIG iii 929), SEG 33.244d, 45.1447, C.Z. i. 143, n. 12, ii. 1093, 1096–98, Alessandrì (1995) 88–90, E. Fehrle in Roscher vi. 631–32 (s.v. ‘Zeus Ἱκέσιος’). 97 See, e.g., Lloyd-Jones (1956) 55, 57–59 (238–39, 243–46) with refs. 98 For an explicit identification of Zeus with his worshippers Ζεὺς Γεωργός (IG iii 1.77, C.Z. i. 176, n. 2) comes closest to forming a parallel, actually appearing to mean ‘Zeus the Husbandman’: this is far from making Zeus into a suppliant, though. On Zeus ἀµφιθαλής in Ch. 394, wrongly adduced by Rose as an example of a god taking an epithet belonging to his worshippers, see Garvie ad loc. 99 The suffixes -τωρ and -τήρ almost always imply nomina agentis: see S.GG i. 530– 32. The exception is words denoting kinship, e.g. πατήρ, µήτηρ, µητροπάτωρ, and also the names of a few utensils and other inanimate objects, e.g. κρατήρ, ζωστήρ, where the force of nomen agentis has been lost (see Buck–Petersen 1945, 302). 100 SEG 9.72.132, 138; 38.812a.6; 39.729; IKnidos 220.6. 39 ﬁtting for Zeus here, and if Aeschylus knew about these religious terms he may have been influenced accordingly. However, the other instances of the stem in the present drama cannot mean anything but ‘supplication’. It is also hard to see how the audience would be able to connect the hapax ἀφίκτωρ with ἀφικετεύω and not with ἀφικνέοµαι (which verb, incidentally, appears in 20). 2. νάϊον … στόλον go together: ‘shipping’, ‘nautical expedition’ (not as LSJ, ‘course’). On the significance of the ship, see 134–35n. 3. λεπτοψαµάθων: de Pauw’s emendation of λεπτοµαθῶν (M) is easier than Friis Johansen’s (Friis Johansen–Whittle 1975) λεπτοψάµµων, but the latter may seem to conform better to Aeschylean language.101 The sand of the Nile was soft (Plin. HN 35.167), especially compared to the Greek shores, and Verdenius (1985) observes that the epithet has a ‘didactic’ ring: ‘The abundant supply of geographical details in Aeschylus (e.g. 5 σύγχορτον Συρίᾳ, 75 Ἀερίας) obviously met a corresponding studiousness of his audience, just as in the case of Herodotus.’ Aeschylus’ source may be Hecataeus: cf. my notes on 220–21 (with n. 381), 256–59, 284–86. 4–5. ∆ίαν … χθόνα: cf. Hes. Th. 866 ἐν χθονὶ δίῃ.102 The epithet carries a narrower sense here: Egypt is the land of Zeus, who was identified with the Egyptian god Amun.103 The oracle of this god that was famous in antiquity was located in the middle of the desert, at the Siwa oasis in Libya.104 It was assumed, however, that the cult was Egyptian in origin,105 and the main sanctity of Amun-Re was still the temple in Thebes, of which the oracle may originally have been a branch (so Hdt. 2.54–58; cf. the refs in n. 104). 101 πολύψαµµον by emendation in 870, †ψαµµίαϲ in Ag. 985, cf. Pr. 573. Also Orph. H. 55.22, fr. 224b, Orac.Sib. 1.393, etc. 103 E.g., Pi. P. 4.16, 4.56, Hdt. 2.42, 2.55. Amun had long since amalgamated with the sun-god Re into Amun-Re, king of the gods (see Silverman 1991, 35–36, 39–40, Quirke 1992, 17). 104 See C.Z. i. 361–90, Pi. P. 4.16 with Braswell’s note, Lloyd on Hdt. 2.42 (ii. 195– 200), Dunbar on Ar. Av. 619 and cf., e.g., Hdt. 1.46, 2.32, 2.55, Plu. Lys. 20.7, E. Alc. 116, El. 734, Pl. Alc. 2.148e–149b. 105 Hdt. 2.42, cf. Pl. Phdr. 274d–275d, Pi. P. 4.52. Plato’s narrative is rather confusing in this context, as he identifies Amun not with Zeus, but with the Egyptian king Thamus (on Plato’s reluctance to accept the Egyptian gods as identical to the Greek, see 220–21n., n. 379). However, this Thamus does play a role that is somewhat reminiscent of Zeus in the myth of Prometheus (see Rowe ad loc.). 102 40 δὲ has no responsive force to µὲν in v. 1, pace FJ–W, who suggest a ‘virtual anaphora’ in Ζεὺς – ∆ίαν.106 The explanatory δέ (D.GP 169) suggested by Verdenius (1985) and the scholium (ὁ δὲ ἀντὶ τοῦ γάρ) is better. 5. σύγχορτον: ‘with joint pasturages’. It is uncertain whether much stock is to be put in the -χορτος part of the word; Mt. Casius, which constitutes the border between Egypt and Syria, is sandy and sterile (Lloyd on Hdt. 2.6; ii. 42). This may not have bothered Aeschylus, or he may not have known it, or -χορτον is used in the broadest metaphorical sense only (cf. the expression χόρτος οὐρανοῦ in Hsch. χ 652). 6–10. The extensive corruption has not managed to obscure the general meaning of the passage. The Danaids stress that they are not suppliants in the Homeric sense, seeking purification for a committed murder (see LSJ s.v. ἱκέτης), but fleeing of their own volition from unwanted suitors (see 8n.). Aeschylus’ disposition of the necessary information is economical and elegant: in but a few lines, we learn (1) that the girls are not polluted, (2) that they are not sentenced to exile, but (3) flee of their own free will, (4) from men, (5) who desire an impious marriage. The first point is probably an ironic foreboding of the deed for which the Danaids are notorious: the murder, on the wedding night, of their husbands. This kind of foreboding appears several times later on in the drama.107 The audience also receives the necessary information that in the present story this murder did not take place in Egypt, as it does in another version of the myth. All this speaks in favour of the Supplices being the first drama in the trilogy (see further the Introduction, II 3). 106 For the ‘anaphoric’ µέν … δέ (as categorised and exempliﬁed in D.GP 370, cf. 163– 64), the rule appears to be that either the subject or the verb (or both) must be the same (or understood to be, if not explicit) in both clauses. Most often the subject: from the examples given in D.GP, e.g., S. Tr. 229 ἀλλ’ εὖ µὲν ἵγµεθ’, εὖ δὲ προσφωνούµεθα, Il. 1.288, Hdt. 1.45.3, Th. 1.85.2, Pl. Lg. 697d, Ant. 5.62. The verb only in four places: in Th. 1.126.12 this is the actual word which is repeated in the anaphor: ἤλασαν µὲν οὖν καὶ οἱ Ἀθηναῖοι …, ἤλασε δὲ καὶ Κλεοµένης. In Th. 6.20.4, Pl. Lg. 739c and D. 19.84 the verb is (or is understood to be) some form of εἶναι. In the present case both subject and verb are different in the two clauses, which does not leave any relation between them strong enough to justify an anaphoric µέν … δέ connection. 107 See especially Gantz (1978), FJ–W i. 37, and cf. my notes on 21–22, 63–64, 123– 24, 196, 287–89, etc. Cf. also Stanford (1936) 145–46 on a similar foreboding in Ag. 314. 41 6–7. Auratus’ δηµηλασίαν for the ms. δηµηλαϲίαι is certain, as we cannot take οὔτιν’ as a dative: elision of the case ending -ι, is hardly found in classical verse.108 δηµηλασίαν should be taken ἀπὸ κοινοῦ with both φεύγοµεν (see 8n.) and γνωσθεῖσαι (‘sentenced’), which, as καταγνωσθῆναι often does, takes an internal accusative. Schmidt (1860, 163; cf. Schmidt 1863, 225) suggested γνωσθεῖσαν, as the normal construction of the verb requires the passive voice to be construed with the sentence passed (i.e., δηµηλασίαν), not the person sentenced (LSJ s.v. γιγνώσκω A.II.1). There is at least one safe parallel for the latter construction, however: Anaximen.Lampsac. Rh. 15.3 (= [Arist.] Rh.Al. 1431b30).109 8. αὐτογενῆ φυξανορίαν has been adopted, rightly, by the latest editors.110 The accusative is to be taken as internal with φεύγοµεν. The expression principally refers to the fact that the girls have decided to flee of their own free will (see, e.g., Sicherl 1986, 86 with refs), forming a contrast to and refutation of ἐφ’ αἵµατι δηµηλασίαν in v. 6, and to the suggestion that they are banished by ‘the vote of the city’ (ψήφῳ πόλεως). αὐτογενής, αὐτογέννητος, etc., elsewhere almost always take this sense (‘self-generated’), but the compounds are not attested until late (Philo, etc.). On the other hand, the only other extant classical example, κοιµήµατα αὐτογέννητα in S. Ant. 864, appears to take 108 See Maas (1962) 74, West (1982) 10, Jebb on S. OC 1435–36 and appendix pp. 289– 90. However, elision of iota here and in S. OC 1436, Tr. 675, E. Alc. 1118, fr. 21.5 Nauck, is defended by Brennan (1893), the latter passages also by Müller (1966) 259–64. 109 LSJ’s (A.II.2) other alleged parallel, IG i2 10.29 (= IG i3 14.30–31; Meiggs–Lewis 40.30), is too frivolously supplemented. The inscription is given by Meiggs–Lewis as τεθ[ν]άτο ἐὰν [γν]οσθ᾵ι [– – –] γ̣[ν]οσθ᾵ι̣ φ̣ευγέτο with an unknown number of letters missing in the middle. The subjects of the subjunctive clauses may as well be ‘it’ (θάνατος, τόδε, φυγή) as ‘he’. 110 The former emendation by Turnebus, Anon.Ald.; the latter by Ahrens (1832, 34): adopted by e.g., Page, FJ–W, West. Others (e.g., Wilamowitz, Murray) adopt Hermann’s (1820, 11 [330]) φυξανορίᾳ with Bamberger’s (1839) αὐτογενεῖ as a dat. modi, which is also defended by Kraus (1984, 93–94), ‘da wir Aischylos und nicht Gorgias vor uns haben’: he argues that the repeated construction with an internal accusative is too suave for Aeschylus. The dative is less economical, however (Kraus’s palaeographical explanation is far-fetched), and even Aeschylus might see the need for giving the audience some cue to the understanding of two highly semantically charged, five-syllable hapaxes presented in streaming anapaests. The rhetorical figure would hardly be Gorgianic, anyway. 42 the meaning ‘of one’s own kin’, referring to the incestuous relationship of Oedipus and his mother. Others have thus assumed (with LSJ) that this is the sense of αὐτογενῆ here, referring to the fact that the Aegyptiads are the Danaids’ cousins: ‘we flee men of our own kin’. But it is doubtful whether such an interpretation is possible. As FJ–W observe, the attribute would have to refer to the weaker (adjectival) part of the compound (-ανορίαν) only, and not to the word as a whole—an awkward conceit which lacks a safe parallel.111 Furthermore, this interpretation in our passage would spoil the rhetorical antithesis between banishment and voluntary exile (FJ–W), which is important since a voluntary flight on the part of the Danaids will naturally make the inhabitants of Argos more kindly disposed towards them than an expulsion due to a crime. Kraus (1984, 95, n. 39) notes that the adjective in Sophocles may actually be said to mean (in a transferred sense) ‘self-generated’: ‘von ihr selbst geboren’. Griffith (1986) suggests that the word in Aeschylus could take on both meanings simultaneously, as well as the sense suggested by Wilamowitz, ‘innate’, with φυξανορία = ‘misandry’.112 This would be a rather extreme case of verbal ἀµφιβολία, but perhaps not entirely impossible: see Stanford (1936) 144–49 for some striking examples in the Agamemnon. In such cases, however, it is probably necessary to recognise one sense as ‘basic’ and the others as subordinate, or mere hints.113 As for the textual corruption, Mpc presents the unmetrical reading αὐτογένητον φυλαξάνοραν, with the variant φυξάνοραν added in the margin, and φυ[..]ξανορὰν Mac. The marginal and ante corr. versions are not only closer to the true reading, but probably also represent a purer tradition than the reading of the ‘diorthotes’,114 which may be conjectural, notwithstanding the fact that λα is added in a rasura. The erased letters need not have represented a traditional reading: the erasure may simply have been a correction of a scribal 111 FJ–W note παιδοκτόνους σούς in E. HF 1381, which is not entirely convincing as a parallel, seeing that the possessive pronoun makes for a special case (cf. K–G i. 263, Anm. 2–3). At that passage the phrase may be influenced by the common use of the possessive pronoun as an objective genitive (K–G i. 560, Anm. 11): ‘children-killers by you’. 112 So also Mazon and Conacher (1996, 81, text for n. 15). 113 See also my notes on 21 ἐγχειριδίοις, 23 βαρύτιµος, 42 τιµάορ’, 146 ἐνώπια. 114 The Byzantine scholar who added the scholia in M: see FJ–W i. 57–65. When I refer to a reading as Mpc, it is to be understood that the correction is made by him. 43 error such as dittography.115 The loss of the ι in φυξανορίαν, turning the word into what appears to be an adjective, may have occurred in connection with the disappearance of a word in v. 10 (q.v.): φυξανορὰν would be taken as an attribute to γάµον in that verse. As for αὐτογένητον, Wilamowitz suggested that -τον was mistakenly copied from the beginning of v. 11. 9–10. τ’ seems to indicate that something has fallen out at the end of this clause (cf. 8n.). So also the metre which, together with the new subject introduced in 11 ∆αναὸς δέ (a very strong syntactical stop, see note on 1–39 above), calls for period-end and catalexis (it would also give 5–13 a neater order with three periods of twelve feet each116). Acatalectic period-ends are not found in recited Aeschylean anapaests, and elsewhere only at change of speaker.117 The scholium on the verse, ἀσεβῆ] ὃν οὐ σέβοµεν ἡµεῖς οὐδὲ τιµῶµεν, may imply that the missing part is <καὶ ἄτιµον>.118 The unusual (and obscure as to its exact nuance) epic word ὀνοτάζω could have been chosen to suggest ὀνοµάζω, so that the adjectives are predicative: ‘reproaching the wedding with the sons of Aegyptus as both impious and dishonourable’. This solution also conveniently introduces an important dichotomy which reappears in some places throughout the drama. The marriage with the Aegyptiads is, from the point of view of the Danaids, a twofold outrage, being at the same time impious, unholy, i.e. hateful to the gods (ἀσεβής here, ὧν θέµις εἴργει 37), and, on the secular side, an offence against the honour, dignity, and law115 See examples in FJ–W iii. 377. The γρ variant in Md φυγαξάνοραν may either be conjectural or an error due to ‘quasi-dittography’ (FJ–W l.c.), in which case it could be the actual reading of M before the rasura. 116 Could this have any connection with the arrangement of the choreutae in three rows? Cf. on 1–39 above. 117 West (1982) 95. Possible exceptions would be S. OC 188 and E. Andromeda (fr. 114 Nauck); in these cases, however, hiatus without period-end might perhaps be allowed after the vocatives παῖ and νὺξ ἱερά, respectively. Cf. West ibid. 15, n. 24, where he appears to accept hiatus without period-end in S. OC 188, in contradiction to his own statement on p. 95, where he claims a period-end in the same place. On the other hand, *Musgrave’s παῖς has been adopted here by the latest Oxford and Teubner texts of Sophocles. Period-end seems likelier than synapheia in the Andromeda, where a dramatic pause would be appropriate after the stately opening ὦ νὺξ ἱερά. 118 Rather than κἄτιµον, since correption is to be preferred to a contracted last biceps in recited anapaests, which, n.b., appeared only two lines above. Cf. West p. xxxiii, West (1982) 95. 44 ful rights of the young women (expressed by the concepts of τιµή and, more often, δίκη).119 An especially succinct expression is found in 395–96, q.v. Cf. also 82n. The lemma of the scholium only contains ἀσεβῆ (cf. Σ776b), but this may have been added after the disappearance of the end of verse 10. Weil’s supplement <διάνοιαν> may receive some support in the appearance of the same phrase, ἀσεβεῖ διανοίᾳ, in Th. 831, and perhaps also in the mention of the διάνοιαν µαινόλιν of the Aegyptiads in 109 of this drama. FJ–W suggest that the disappearance of διάνοιαν might be explained by its vicinity to the somewhat similar word ∆αναός. If we are unwilling to accept a lacuna, the τ’ must be considered corrupt: suggested remedies are Tucker’s ’ξονοταζόµεναι and Whittle’s (ap. FJ–W) ’πονοταζόµεναι, both of which verbs are unattested elsewhere, as is prodelision in Aeschylean anapaests (according to FJ–W). It is more likely that the τ’ itself carries some responsibility for the corruption, which may have arisen in connection with that in v. 8 (q.v.). The τ’ would appear to connect ἀσεβῆ with what was seen as another attribute to γάµον in that verse (φυξανορὰν, losing the ι), resulting in the disappearance of what followed. 11. ∆αναὸς δέ: I think Danaus enters on this cue (see 1–39n. above). He is an old man—a γέρων (177, 480, 775)—and presumably dark-skinned and exotically dressed like his daughters (cf. 496–98). Possibly he is wearing an outfit that would characterise him as a skipper (see 134–35n., 503n.). On his character and function in the drama, see on 176–78, 246–48; on his scant mythological background see the Introduction, II 1. δέ may answer to Ζεὺς µὲν in v. 1, contrasting the Danaids’ Heavenly Father and protector with their less eminent earthly one. 12. στασίαρχος: ‘leader of the faction’. Not just ‘chief of a band’ (LSJ): στασί- refers to the fact that Danaus and his daughters have broken with the family in Egypt, and that discord has arisen. So FJ–W, but ‘sedition’ is not the appropriate English equivalent in this case: Danaus and his daughters have probably not rebelled against any legal authority (FJ–W i. 47–48). In 119 Cf. 82, 343, 378, 429–30, 644–45, 1071, and also FJ–W i. 30. On ἄτιµος in general, see FJ–W 614n. For the juxtaposition of τιµή and σέβας, cf. also, e.g., 706–7, 990 (corrupt), Pers. 166, Eu. 545–46, S. Ant. 514, 516, 744–45, OC 1007, E. Alc. 998, Ph. 1321, Ba. 1009–10, Sthen. 15–16 (TrGFS p. 130; Pap.poet. fr. 16), Ar. Nu. 293, X. Mem. 43.13, Pl. Lg. 729c, 841c, Isoc. Busir. 26., [Pythag.] Carm.aur. 2. 45 classical (Attic) Greek, στάσις means ‘discord’, ‘faction’, ‘party-strife’, or even ‘civil war’, rather than ‘sedition’.120 ‘Sedition’, ‘uprising’, in the sense of an illegal movement to overthrow the government, is usually denoted by the compound term ἐπανάστασις.121 Nor is στάσις ever completely without political implications when referring to a group of people. In the examples in Aeschylus where it seems to mean only ‘group’ (LSJ s.v. B.II), it is always a question of a clearly defined party with a special purpose, never just any group of people: in Ch. 114, 458 the στάσις consists of the members of a conspiracy against unlawful tyranny;122 in Eu. 311 it refers to the Erinyes with their very well-defined agenda (Manolopoulos 1991, 92). The grammatical construction of τάδε πεσσονοµῶν κτἑ is somewhat unclear (cf. on 15–18). I would take πεσσονοµῶν as transitive with τάδε (pace FJ–W), which is thus taken ἀπὸ κοινοῦ also with ἐπέκρανεν. 13. κύδιστ’ ἀχέων ἐπέκρανεν: κύδιστα means ‘worthiest’, ‘noblest’, ‘greatest’, or (pace FJ–W) simply ‘best’ (with connotations of all of the above). One should note here as well as for E. Alc. 960, Andr. 639 that the original meaning of κῦδος was not ‘honour’ in its secular sense, but ‘power’, ‘force’, ‘glory’ (cf. DE s.v.); and that κυδαίνω and κῦδος almost always have the notion of cheering and giving (or having) strength in Homer: see especially Il. 5.448 ἀκέοντό τε κύδαινόν τε and 7.205 βίην καὶ κῦδος ὄπασσον.123 It is not the outer effects of ‘honour’ (if we are to use that word with its modern connotations) such as ‘fame’, ‘repute’, or ‘distinction’ which is denoted by κῦδος, but the inward ones: self-confidence, energy, health, power, strength. κῦδος is something real and concrete, which actually makes a person better and stronger. The few extant instances of the stem in Aeschylus include this meaning: Pers. 455, Th. 317 (difficult, but actually seems to mean ‘give strength’: cf. Hutchinson ad loc.).124 The notion of a remedy should probably be included here, and perhaps one should translate ‘fulfilled it as best for the sufferings’, ‘with regard to the sufferings’, with an objective rather than a 120 Manolopoulos (1991) 35–45, 74–80, passim. The stem may have a distinctly positive value, as in Lys. 2.61 ὑπὲρ τῆς δηµοκρατίας στασιάσαντες. 121 Manolopoulos (1991) 109–35, 279–81. 122 Cf. Manolopoulos (1991) 91, Dover on Ar. Ra. 1281. 123 Also, e.g., 20.42, 15.595, Od. 14.438, 16.212. 124 The neutral, formal address κύδιστε also appears in fr. 238. 46 partitive genitive (notwithstanding 1069 and Il. 17.105): cf. Ag. 1339–40: ἄλλων ποινὰς θανάτων ἐπικρανεῖ, ‘for other deaths he will fulfil punishment’, and 744–45: ἐπέκρανεν | δὲ γάµου πικρὰς τελευτάς, ‘she accomplished a bitter end of the marriage’. 14. ἀνέδην: with connotations of desperation as well as of unbridled freedom, contrasting with both the maidenhood and the nobility of the girls (but suggestive of bacchanals, cf. AP 6.172). Pace FJ–W, these connotations are rather effective as a contrast to ‘the authoritative and deliberate nature of Danaus’ decision’ in the previous lines (for the contrast between sensible manliness and emotional femininity, cf. Th. 78–263). Pl. Prt. 342c is not to be taken as a justification for a neutral or dispassionate use of ἀνέδην here: συγγενέσθαι ἀνέδην (‘converse freely’) can have little in common with φεύγειν ἀνέδην.125 κῦµ’ ἅλιον: the β which has been mysteriously introduced into the text of M (κυµβαλέον, β added in a rasura) and Md is possible evidence for a minuscule source for M: see 110–11n. with n. 276. The correct reading is found in Hsch. s.v. ἀνέδην. 15. δ’: not simply continuative (pace FJ–W), but stresses the opposite natures of φεύγειν ἀνέδην and κέλσαι (cf. 331, D.GP 165 ff.). Paley’s (ed. 1844) τ’, printed by Hermann, is not only unnecessary, but a considerable impairment. Ἄργους γαῖαν: as is evident from 254–59 (qq.v.), the kingdom of Argos at this time includes all of mainland Greece. Later Aeschylus also makes Argos the seat of power of Agamemnon (Ag. 24, etc.). Rather than, or perhaps in addition to, drawing on Athens’ being on friendly terms with Argos at the time of the Oresteia (and perhaps of the present play: see the Introduction, chapter I, The Date), this may, at least in our case, be an inference from the Homeric use of Ἀργεῖοι as a designation for the Greeks as a whole. 15–18. ὅθεν δὴ … εὐχόµενον τετέλεσται: there is some syntactical confusion. The two verbs are put at the end, after an assertion of ancestry consisting of three distinct claims, each stated as an adverbial. The Danaids claim for themselves Argive heritage (Ἄργους γαῖαν, ὅθεν δὴ), descent from Io (τῆς … βοὸς), and descent from Zeus (ἐξ ἐπαφῆς … ∆ιὸς). It is uncertain 125 Cf. Russell on [Longin.] 21.2: ‘ἡ ἐλευθερία τοῦ δρόµου is a much more positive concept than the English “freedom of movement”, which implies only the absence of impediment.’ 47 which verb is to be taken with what adverbial, and the syntactical function of βοὸς is furthermore unclear: is it an objective genitive of ἐξ ἐπαφῆς κτἑ, or does it go with εὐχόµενον? This verb, for which cf. 536 ∆ῖαί τοι γένος εὐχ. εἶναι, has plenty of parallels, especially in Homer, for construction with each of the previous elements: it may take an adverb ‘whence’, a simple genitive of origin, or ἐξ + genitive.126 The passive τετέλεσται, on the other hand, needs an agent, and thus semantically seems to fit only the last, prepositional assertion, ἐξ ἐπαφῆς … ∆ιὸς: created by the touch of Zeus (cf. 45–46). The emendation τετελέσθαι, suggested by de Pauw (who discarded it) and Schütz (comm. 1797), would clear up the syntax considerably, allowing this verb to govern the prepositional phrase, which otherwise, because of the word order, apparently would have to go with εὐχόµενον. On the other hand, one should as far as possible avoid taking incoherent syntax as a ground for textual emendation in Aeschylus.127 There may be no absolutely fixed grammatical structure here, and the reader/spectator is invited to take τετέλεσται either with ἐξ ἐπαφῆς κτἑ or as absolute, the participle εὐχόµενον with one or several of the three separate assertions of ancestry, and τῆς … βοὸς either with εὐχόµενον or with ἐξ ἐπαφῆς κτἑ, or with both simultaneously. 16–17. τῆς οἰστροδόνου βοὸς: the great-great-great grandmother of the Danaids, Io. See the Introduction, II 1. 20. ἀφικοίµεθα: cf. 1n. 21–22. ἐγχειριδίοις, ἐριοστέπτοισι κλάδοισιν: the ‘wool-wreathed (olive) boughs’ are traditionally carried by Greek suppliants (see FJ–W for refs); this apposition here explains Aeschylus’ singular use of ἐγχειρίδιον in its basic sense, ‘thing held in hand’ (cf. 314, 378 with nn.). Elsewhere in classical Greek the word always means ‘dagger’ (but later ‘handbook’: see LSJ s.v.). As well 126 Cf., in the first case, Od. 1.406–7 ὁππόθεν οὗτος ἀνήρ, ποίης δ’ ἐξ εὔχεται εἶναι | γαίης, 17.373 πόθεν γένος εὔχεται εἶναι, h.Ap. 470, E. Tel. I 3 (fr. 696.3 Nauck; TrGFS p. 132); in the second Il. 21.187 γενεὴν µεγάλου ∆ιὸς εὔχοµαι εἶναι, 6.211 ταύτης τοι γενεῆς … εὔχοµαι εἶναι, 20.241, Od. 14.204; in the third Il. 14.113 πατρὸς δ’ ἐξ ἀγαθοῦ καὶ ἐγὼ γένος εὔχοµαι εἶναι, Od. 14.199, 20.192–93, 21.335, 24.269, Pi. O. 7.23. 127 On anacolutha and other types of ‘syntactical impressionism’ in Aeschylus, cf. 27, 32, 33–36, 40–175, 44–46, 52–55, 60–62, 74–76, 78, 134–35, 186–87, 209, 254–55, 276, 287–89 with notes; also, e.g., Berti (1930) on Aeschylean anacolutha, and West (1990) on Aeschylean logic and grammar. ‘Aeschylus’ is a language of truth, perhaps, rather than logic’ (West 1990, 12). Cf. also Garvie (2001) 1–3 and below, text for n. 163. 48 as hinting at the underlying (passive-)aggressiveness of the suppliants—which becomes apparent later in the drama, with accusations against the gods as well as threats of divine vengeance and suicide128—it also hints at the murders of their husbands that will soon take place (cf. 6–10n.). ἐριοστέπτοισι is Scaliger’s129 probable correction, well defended by FJ–W, of the ms. ἱερο-. For the explanatory apposition, a common feature of Aeschylean poetry, cf., e.g., 41–44, 156–58, 415.130 22a–23. West (W.SA) offers the most credible explanation of the problem with v. 23: a line has fallen out just before, in which the correlate to the relatives ὧν … ὧν will have appeared. As West l.c. and FJ–W observe, Robortello’s conjecture ὦ … ὦ is unlikely simply because the ms. reading is so very much difficilior lectio. It also makes ‘Zeus as the third’ in 26 incredible: he would be the sixth. With West’s supplement, ἀλλ’ ὦ πάτριοι δαίµονες Ἄργους, the general address ‘gods of the land’ (πάτριοι δαίµονες) is divided into three categories in 24–25, thereby returning to Zeus his rightful position as the third (see ad loc.).131 23. βαρύτιµοι: a somewhat problematic compound, as shown by FJ–W, but certainly not impossible.132 It recurs around the beginning of the Christian era, meaning ‘of heavy value’.133 FJ–W, observing that the meaning of -τιµος in compounds elsewhere is not ‘vengeance’ or ‘punishment’, except in a few instances of ἄτιµος (‘unavenged’, ‘unpunished’),134 conclude that the vulgate interpretation, ‘of heavy vengeance’, is impossible (or at least ‘puzzling’). The problem is eliminated, however, if one recognises a broader sense of τιµή as 128 See 154–75, 381–86, 455–66, FJ–W i. 37–38. Also in the margin of Portus’ copy of Victorius (see Portus in the references section), but possibly not by his own hand (Professor Martin West). 130 And see FJ–W, Headlam ap. Thomson on Ag. 4–7, Wærn (1951) 49 ff. 131 FJ–W defend the ms. reading, taking the traditional relatives ὧν … ὧν to refer forward to the gods (so also Haupt), but the resulting asyndeton is harsh, and the invocation becomes awkward (see Griffith 1986, Diggle 1982). Others have taken the relatives to refer back to γαῖαν in v. 15 or χώραν in 19; but, as FJ–W rightly point out, χώραν … ὧν γῆ is an unlikely expression (pace Verdenius 1985). 132 Cf. Verdenius (1985). Hermann suggested βαθύτιµοι, a word unattested elsewhere, which seems rather flat by comparison. For βαρυ- and its exchangeability with βαθυ- in compounds (in Hellenistic literature), see Chryssafis on Theoc. 15.110, James (1970) 66–68. 133 Str. 17.1.13, Ev.Matt. 26.7, and several times in Christian literature. 134 LSJ s.v. II.2–3, Ag. 1279 with Fraenkel’s note. Cf. also Hsch. ἀντίτιµα· τὰ ἄποινα. τὰ ἀντέκτιτα [*Musurus, *Pearson: αντιτια … ανεκτιστα codd.]. 129 49 ‘due’. ‘Vengeance’ comes in the same category as ‘reverence’, ‘honour’, the difference being that the former is the due of the wicked, the latter of the honourable. βαρυ- makes it clear that the sense is intended in malam partem. Neither Aeschylus nor his audience would have had any problem understanding the noun in an active (βαρέως τιµῶν) rather than passive (βαρέως τιµῶµενος) sense: the active sense is found in, for instance, the compounds φιλότιµος in 658 and ξενότιµος in Eu. 546. Cf. also the compounds ending in -δικος, which are mostly passive in sense, as in ἄ-, ἔν-, ἔκ-, πάνδικος, but when active take the meaning ‘judging’ or, as in the Aeschylean βαρύδικος (Ch. 936), ‘punishing’.135 It is perhaps possible that βαρύτιµος refers not only to the heavy vengeance of the deities, but that -τιµος is simultaneously to be taken in its more common, positive sense of ‘honour’ (on such verbal amphiboly, see 8n.). The reference would be to the dignity of the position of the gods or heroes, the ‘weighty office’ they hold (cf. Fraenkel on Ag. 514 τιµάορον) and the entitlement they have to solemn respect and worship from mortals. This sense easily mingles with that of ‘vengeance’, part of the job of heroes as well as of chthonic gods (see on 24–25) being just that: to execute punishment on humans. Thus βαρυ-, with its sinister implications, is still appropriate as a designation for the deities’ office or dignity, carrying the notion of judicial sternness (cf. Pers. 828, LSJ s.v. βαρύς II.1) and severe punishment. 24–25. χθόνιοι θήκας κατέχοντες is more difficult than it appears at first sight. Being opposed to ὕπατοι … θεοί, which cannot mean anything but ‘Olympian’ (cf. Ag. 89), χθόνιοι must take the sense ‘of the underworld’: chthonic (LSJ s.v. I.1). This is always the meaning of χθόνιος in Greek literature when contrasted to Ὀλύµπιος, always the meaning when, as here, it is substantival, and invariably the sense of the adjective in Aeschylus.136 135 Cf. also the other Aeschylean compounds on βαρυ-, which always occur in pessimam partem: βαρύδικος, -δότειρα, -κοτος, -µηνις, -πεσής, -στόνως, -ταρβής. 136 Pers. 628, 641, Th. 522, Ag. 89, Ch. 1, 124, 359, 399, 476, 727, fr. **273a.8–9. (The unmetrical χθονία in Th. 735 is generally emended to γαΐα after Dindorf ed. 1841, ii. 640; cf. Hsch. s.v. γαῖα. A dubious fragment of Aeschylus [488] speaks of Χθόνιος as the name of one of the Spartoi, the men sown by Cadmus from dragon’s teeth. A more general sense of χθόνιος, not referring to a deity [‘subterranean thunder’], is found in Pr. 994.) Cf. E. fr. 868 Nauck, Pl. Lg. 717a, 828c, Arist. Mu. 401a, Aristox. fr. 13, Plu. Num. 14.3, Aet.Rom. 266e, 290d, Orph. H. 1.2, 3.8, 4.5, 7.9, 38.2. The distinction between chthonic and Olympian divinity has been questioned; see 50 θήκας κατέχοντες, on the other hand, suggests that the reference is to the local Heroes, ‘possessing the tombs’, i.e. being worshipped at their alleged place of burial. Similar expressions are found elsewhere as referring to the dead in their graves, the nearest parallel being Ag. 452–54 οἳ δ’ … θήκας Ἰλιάδος γᾶς … κατέχουσιν.137 This is how most commentators from Hermann onwards have understood the words. The problem with this interpretation is that there is little or no ground for numbering the heroes among the chthonic deities: in fact, nowhere else in extant literature are heroes referred to as χθόνιοι in this sense.138 Chthonic deities and heroes and their respective cults are repeatedly spoken of as distinct, and the heroes are often pictured as being separate from the underworld.139 Also, it has been decisively shown by Schlesier (1991–92) for the most radical viewpoint, and also the discussion, with refs, of Ekroth (2000, 310–13). On Schlesier’s article, see also Scullion (1994b) 119. Judging from the consistent and frequent use of the terms in extant literature, the Olympian–Chthonian dichotomy must be considered as an actually held belief, not a 19thcentury construction. As noted by Burkert (1985, 202, text for n. 38), the polarity is especially conspicuous in the tragedies of Aeschylus (cf. 156–61 with notes). It is true that the distinction may be more relevant on a theoretical, ‘theological’ level than in actual ritual practice (see Ekroth l.c.): this does not, however, make it less relevant. 137 See Fraenkel ad loc., and cf. also Ag. 1539–40, Th. 731–33, Pers. 404–5, S. Aj. 1166–67, OC 1763, X. Cyr. 2.1.1. 138 In Pi. P. 4.159 χθονίων does not refer to a hero (pace Tucker 24–25n.), nor to the dead, but, as the scholium takes it, to the infernal deities (see Giannini ad loc.). In one place in extant literature, Σ Pi. O. 2.104b, heroes are referred to as καταχθόνιοι: the scholiast speaks of characters in drama pouring libations to the καταχθονίοις ἥρωσι, praying for aid, for instance Electra to Agamemnon. This has little to do with the cult of heroes as part of Greek religion: Electra’s sacrifice is not to Agamemnon qua hero, but to the spirit of her father; if anything, it is an illustration of the cult of ancestral spirits (on which see, e.g., Harrison 1922, 55–76). On Aeschylus’ Agamemnon, see further below, n. 146. 139 The heroes are explicitly spoken of as distinct from the chthonic deities in Pl. Lg. 717a–b, [Pythag.] Carm.aur. 2–3, Plu. Aet.Rom. 269f–270a (cf. also 272d–e), Artem. 2.34, 2.39–40, Porph. Antr. 6.18. As regards the eschatological lot of heroes, we find that in literary sources they are separated from the chthonic world at an early stage. In Homer, the ‘heroes’ go to Hades just like everybody else (apart from Menelaus, who goes to Elysium: Od. 4.563–69), but in the Homeric context they have not yet ascended to the status of cultic deities: the notion of a hero-cult is absent from Homer’s universe (apart possibly from a few hints of things to come: see Janko on Il. 16.666–83 [pp. 371–73], Auffarth 1999, 41–42). See Albinus (2000) 57–66 on the 51 Ekroth (2000, see 310–325, and passim) that the ritual practice of archaic, classical, and Hellenistic Greece provides no rationale for including the heroes among the chthonian deities.140 discrepancies between the Homeric Hero and the later cultic one. However, as early as Hes. Op. 166–73 and some fragments of early lyric (Ibyc. 291, Sim. 558, Carm. conv. 894), several heroes are granted a happy afterlife on the Isles of the Blessed, possibly all heroes in Hesiod: cf. West and Verdenius on Hes. l.c. (The Hesiodic picture is complicated by the spirits of the Silver Race, referred to in Op. 141: ὑποχθόνιοι µάκαρες θνητοὶ καλέονται. Peppmüller’s [1882, 2–3] θνητοῖς makes the expression less awkward: it is not true that the dative of agent with passive verbforms is unparalleled in epic, pace West ad loc. and Schoele [1960, 257]: S.GG ii. 150 exemplifies with Od. 4.663–64 [16.346–47], 8.472: cf. also Il. 5.465, 8.244, 18.103, 21.556– 57 and, for the particular expression, Pl. Lg. 715b λέγεται … ἡµῖν. As West l.c. notes, however, this is an ad hoc definition for something that as yet has no defined place in mainstream religion. The ‘blessed subterranean’ are deities worshipped in graves who, at the time of Hesiod, were still anonymous and distinct from the heroes [West l.c. and 121–26n.]. Verdenius l.c. notes that ὑποχθόνιοι means not ‘in Hades’, but ‘under the earth’ in a concrete sense: residing in the graves [cf. the Pindaric examples below]. The notion of grave-sanctuaries belonging to the spirits of the Silver Race did not catch on: the graves came to be identified with the more illustrious epic heroes [West 121–26n.].) In Pindar, the heroes are also spared the chthonic underworld, although the poet will not usually hear of any Elysium (the exception is the mysterious second Olympian, on which see Lloyd-Jones 1985). Instead, we find here for the first time the expressed notion that the heroes are spiritually present at their sanctuaries. In P. 4.159 (cf. n. 138), the soul (ψυχή) of the hero Phrixus is to be brought back home from Aea by a ship; that is, the relics containing the hero’s soul are to be returned to his native country and buried in a sanctuary (see Farnell and Giannini ad loc. ‘Home’ is presumably Phthiotian Halos or Orchomenus in Boeotia: see Türk in Roscher iii. 2458, s.v. Φρίξος). In N. 7.45–47, Neoptolemus is imagined to reside at his shrine, watching over the ἡρώϊαι ποµπαί. The separation of the heroes from the chthonic world is even more explicit in P. 5.93–101, where the immortal lots of Battus and his successors on the throne of Cyrene are contrasted. The former, being the founder of the city, has become a ἥρως … λαοσεβής, while the latter, being in Hades, will hear of the honours of their ancestor with a ‘chthonic’ mind, as it were: ἀκούοντί ποι χθονίᾳ φρενί. Battus, then, is not, as opposed to his successors, in the underworld (cf. Giannini ad loc.). The heroes are also not in Hades according to Plato (R. 392a): περὶ δαιµόνων τε καὶ ἡρώων καὶ τῶν ἐν Ἅιδου. 140 Cf. also van Straten (1974) 176, Nock (1944) 141–48, Fairbanks (1900) 248–49, passim. As shown by these authors and even more decisively by Ekroth (2000), the character of ritual does not determine whether a certain deity is to be considered chthonic or Olympian. 52 Most critics have not reflected upon the sense of χθόνιοι in our passage. One exception is Haecker (1861, 230), who argued that the contrast between θεοὶ ὕπατοι and χθόνιοι which appears in Ag. 88–89, as well as the reference to Ζῆνα τῶν κεκµηκότων in Supp. 154–61 (his 136–43), indicates that χθόνιοι must refer to the chthonic gods, among which he did not count the heroes (so also the scholium and Fairbanks 1900, 244). Haecker’s emendation θάκους κατέχοντες becomes flat without an epithet for θάκους, however.141 The close parallels for the expression θήκας κατέχοντες, especially Ag. 452–54 (see above, text for n. 137), also indicate that these words are sound, and that they refer to graves of the dead, not to gods. As evidence for a reference to the heroes, we may also note a fragment from the Epigonoi (fr. 55) where a libation is taking place: λοιβὰς ∆ιὸς µὲν πρῶτον ὡραίου γάµου Ἥρας τε * * * τὴν δευτέραν γε κρᾶσιν ἥρωσιν νέµω * * * τρίτον ∆ιὸς σωτῆρος εὐκταίαν λίβα Here the second cup is offered to the heroes and the third to Zeus Soter, the same order of invocation as in our passage.142 (On Zeus Soter see below, 26n.) Nevertheless, χθόνιοι is at odds with the interpretation ‘heroes’. Scholars who have accepted this meaning have usually not reflected on the sense of the adjective: for instance, Smyth and Mazon translate ‘nether powers’, ‘dieux souterrains’, but state in their respective notes that the reference is to heroes. In NP our passage is taken, without comment, as the sole example of ‘chthonic’ being used of heroes (R. Schlesier, ii. 1187, s.v. ‘Chthonische Götter’). Scullion (1994b, 93; cf. Scullion 2000) also takes our passage as evidence for his thesis that the heroes were chthonians proper, but as far as I can see he is refuted by Ekroth. 141 Combined with Portus’ χθονίους (see below), it would become more attractive. The idea that the ‘second cup’ is due to the heroes is attested elsewhere, although perhaps without any evidential value independent of A. Epig.: cf. Plu. Aet.Rom. 270a, Poll. 6.15, Ael.Dion. s.v. τρίτου κρατῆρος, Hsch. s.v. τρίτος κρατήρ, Σ Pi. I. 6.10, Σ Pl. Phlb. 66d, Apostol. 17.28. 142 53 Other critics have taken the word in a more general sense, either referring to the graves, ‘powers … filling tombs within the earth’ (Headlam), or ‘indigenous’ (‘di questa terra signori’, Untersteiner ed. 1946)—senses which overlap.143 As we saw, however, the Aeschylean use of the adjective elsewhere, as well as the context—the contrast to ὕπατοι—does appear to require a theological implication, seeing that the reference is to deities. One way out of the dilemma would be Portus’ χθονίους.144 As an epithet to θήκας, the adjective loses most, if not all, of its theological impact, and it becomes possible to read the expression the way Headlam and Untersteiner have done. The attribute to θήκας is also not unwelcome from a stylistic point of view, and it makes the parallel to Ag. 452–54 (see above) closer, corresponding to the genitive attribute Ἰλιάδος γᾶς in that passage. The looser, ‘non-theological’ sense of χθόνιος is unparalleled in Aeschylus, however (unless χθονία is sound in Th. 735, which is unlikely). Keeping the ms. reading will mean that Aeschylus (or, strictly speaking, the Danaids) bluntly states that the heroes are chthonic deities. In the light of the available evidence, this appears to be heresy. Then again, the dogmata of Greek religion were not strict. The hero-cult, still relatively young at the time 143 Cf. also Untersteiner ed. 1935, Foucart (1918) 74, the translation of Friis Johansen (ed. 1970), and S. OC 1726 τὰν χθόνιον ἑστίαν, 947–48 Ἄρεος … πάγον … χθόνιον ὄνθ’, Ar. Ra. 1148–49, perhaps E. Hec. 79 χθόνιοι θεοί (so LSJ), and, it seems, Trag. adesp. 274 χθονίους Ἰναχίδας. χθόνιος in the sense of ‘indigenous’ is found once or twice as referring to heroes and heroines in Hellenistic and Roman times: A.R. 4. 1322–23 χθόνιαι θεαί … ἡρῶσσαι Λιβύης and D.H. 1.64.5 who translates a Latin inscription on an ancient Lavinian sanctuary, taken by local contemporaries to be the tomb of Aeneas, as πατρὸς θεοῦ χθονίου, ὃς ποταµοῦ Νοµικίου ῥέµα διέπει. Dionysius translates the Latin term indiges with χθόνιος: see, e.g., Verg. Aen. 12. 794–95, Livy 1.2.6, Serv. on Aen. 1.259, Castagnoli (1972) 65–66. Originally, however, this inscription, which was probably as old as the 7th century B.C., referred to the local deity sol indiges (Fromentin ad loc., pp. 59–60, Castagnoli 1972, 92–93, 110). In S. Aj. 202 χθονίων … Ἐρεχθειδᾶν, the epithet may refer to the fact that the Erechthidae were αὐτόχθονες, born out of the earth, which makes them indigenous in the most concrete sense (see Jebb ad loc. and cf. my 250–51n.). 144 Better than Auratus’ χθονίας: the feminine ending is not found in Aeschylus apart from the apparently corrupt χθονία in Th. 736 (see n. 136 above). The adjective is also found with two generic endings in Sophocles and Euripides. 54 of Aeschylus,145 lacks a secure metaphysical and eschatological foundation, intrinsically having no self-evident affinity either with the chthonic or with the Olympian sphere. It is also true that Aeschylus in his expressed views on the afterlife adheres more closely to the Homeric picture than to that of his contemporaries: the important dead, even ‘divine’ kings like Darius, reside in the underworld, albeit with special status.146 Thus Aeschylus’ view of the afterlife of heroes, and of their place in the divine cosmography, is far from clear. It is perhaps just conceivable that he regarded the heroes as ‘chthonic’ and belonging to the domain of Hades. It is important to note, however, that this would be a controversial theological statement, at odds with contemporary and later held beliefs. The evidence will not allow us to determine for certain whether χθόνιοι is corrupt, but we should note that it is remarkable, and keep Portus’ χθονίους in mind. 26. Ζεὺς Σωτὴρ τρίτος: Zeus Soter properly gets the third cup in libations of wine: cf. the fragment from the Epigonoi cited above (24–25n.) and 145 Heroic sanctuaries are identified from the end of the eighth century (Burkert 1985, 203). 146 See Pers. 691 with the notes of Hall and Broadhead. In the Choephoroi, the dead Agamemnon is apparently regarded as ‘chthonic’ in a sense (cf. esp. 489), although he is never explicitly given that epithet (476 is controversial, and it is uncertain whether the µάκαρες χθόνιοι of that passage are meant to include the dead souls, including Agamemnon’s, or just refer to the chthonic gods; evidence for the former view might possibly be found in the parallel expression in Hesiod’s ὑποχθόνιοι µάκαρες θνητοί: see above, n. 139). On the whole, the picture presented in the Choephoroi does more to mystify than to clear things up. In 354–62, the chorus speak of Agamemnon as πρόπολος τῶν … χθονίων … τυράννων, but this seems to be an unreal wish rather than a statement of fact (pace Garvie ad loc.). In other places in the same drama, the dead Agamemnon appears to be pictured as spiritually present in his grave (cf. 324 ff., 400 ff., Garvie p. xxxiii), which belief regarding heroic afterlife was commonly held among Aeschylus’ contemporaries (see above, n. 139). It is also not clear whether Aeschylus regarded Agamemnon as a hero in the cultic sense; at least he appears not to have received this honour immediately after his death, i.e. at the time of the action of the Choephoroi. In fact, Aeschylus may imply that Agamemnon’s ignominious death prevented him from receiving certain posthumous honours, including heroic elevation: cf. the mysterious hints in 345–79, and also 483– 85, where the vengeance wrought upon Agamemnon’s killer appears to be given as a requisite condition for his future ritual worship. 55 S. fr. 425.147 Cook regards this aspect of Zeus as chthonic, but there is no conclusive evidence for such a categorisation, rather the opposite:148 certainly Pindar must have regarded Zeus Soter as Olympian in O. 5.17 σωτὴρ ὑψινεφὲς Ζεῦ, and in fr. 30 Θέµιν οὐρανίαν … Μοῖραι ποτὶ κλίµακα σεµνὰν ἆγον Οὐλύµπου … σωτῆρος … ἄλοχον ∆ιὸς ἔµµεν. 27. δέξαιθ’: most editors and critics think that emendation is necessary (δέξασθ’ Heath 1762; δέξαισθ’ de Pauw) as πέµψατε, a direct second-person address, follows in 33 without any apparent change of subject. Wilamowitz, FJ–W and Verdenius (1985) defend the ms. (and Σ) reading. An exact parallel for a change from 3 pers. sg. opt. to 2 pers. pl. imper. has not been found, but similar changes of number, person, and mode all appear separately.149 If sound, the incongruity is best viewed as an attraction of the verb to the third person singular, induced by four different factors: (1) the new metrical period begun at 26 καὶ Ζεὺς; (2) the long apposition attached to Zeus; (3) the fact that Ζεὺς, Σωτήρ and τρίτος all take the nominative, not the vocative case. Fourthly, Zeus is by far the most important of the deities and has been referred to as Ἀφίκτωρ just before, being thus intimately connected to ἱκέτην 147 Also, e.g., Ch. 1073, Eu. 759–60, Σ Pi. I. 6.10 (whence the fragment of the Epigonoi and S. fr. 425), Hsch. s.vv. Σωτῆρος ∆ιός and τρίτος κρατήρ, with further references in FJ–W, C.Z. ii. 1123–25 and Rutherford (2001) 50, n. 60. In Eust.Macr. Hysm. 1.14.3, Zeus Soter is idiosyncratically given the fourth cheer. 148 C.Z. ii. 1123–25: ‘the sequence [sc. of libation] suggests that this final offering was in its essence simply drink for the soul of a dead man.’ C.Z. l.c. cites as evidence for ‘the chthonian character of the god’ the present passage of the Supplices and Ag.1386– 87: (πλήγην) τρίτην ἐπενδίδωµι τοῦ κατὰ χθονός | ∆ιὸς (Enger 1854b, 13: ᾅδου vel ἅδου codd.), νεκρῶν Σωτῆρος εὐκταίαν χάριν. But to claim that Clytaemnestra’s ‘libation’ to the ‘chthonian Zeus’ in the Agamemnon means that Zeus Soter is a chthonian deity is to miss out on the dark irony. While the third libation of wine is offered to Zeus Soter—the saviour of the living, of course—Clytaemnestra offers the third pouring of blood from the slain Agamemnon to Zeus under the Earth, ‘saviour’ of the dead. The utterance is scornfully blasphemous, and has nothing to do with the actual Zeus Soter: indeed the Ἅιδου of the mss. may be kept without any detriment to the irony (pace Fraenkel ad loc.) which consists in giving the epithet Soter, ‘Saviour’, to the Lord of the underworld, and in the adding of νεκρῶν. Neither passage supplies any evidence (pace Garvie 1970, 80) that the actual Zeus Soter was regarded as a chthonic deity. 149 Cf. 33–36, 656–709, K–G i. 79–81, 86–88, Headlam on Herod. 4.1. 56 (see 1n.). The adverbial αἰδοίῳ πνεύµατι also has a special relation to Zeus: cf. 192 Αἰδοίου ∆ιός, the references to Zeus’s ἐπίπνοια in 17, 44, 577, and those to Zeus Οὔριος (‘the sender of fair winds’, LSJ) in 594. Zeus is such an important figure in this drama that there is no reason to wonder at his being singled out as the sole subject for δέξαιθ’. The clause δέξαιθ’ … χώρας, then, might attain as it were a semi-parenthetical character, referring to Zeus alone. The general address to all the gods is taken up again with the new clause beginning ἀρσενοπληθῆ δ’. 28–29. ἱκέτην τὸν … στόλον: Weil deleted τὸν for the purpose of effecting catalexis and period-end after χώρας. This is attractive for at least two reasons: it divides the idiosyncratically long metrical period in 26–32 into two shorter periods, fairly equal in length to the surrounding ones,150 and it stresses the rhetorical contrast between ἱκέτην … θηλυγενῆ στόλον and ἀρσενοπληθῆ … ἑσµὸν ὑβριστήν (see Tucker, FJ–W), making the new period begin with ἀρσενοπληθῆ. Tucker observed that the predicative character of ἱκέτην which τὸν confers makes this contrast somewhat asymmetrical, as ὑβριστήν is attributive. The absence of dieresis effected by the deletion (αἰδοίῳ πνεύ¦µατι χώρας) is admissible, diaeresis not being mandatory in the catalectic clausulae of anapaestic systems.151 The word-end after αἰδοίῳ is problematic, though: in the anapaestic sequence ''''$$''|| word-end after third position is found in tragedy per emendation only.152 31. χέρσῳ … ἀσώδει: FJ–W note that the muddiness is mentioned as a contrast to the sandy shores of Egypt dwelt on earlier (3–4), and that the reference may be to the marshy region of Lerna south of Argos: on its significance in the myth of the Danaids, see the Introduction, II 3. The oxymoron (‘muddy dry land’) produces an almost comical impression; comic relief, or at least burlesque, is found a few times elsewhere in Aeschylean tragedy: cf. Th. 245–63, Ag. 1343–71, Ch. 750–60.153 150 The metrical periods of the prologue contain 8, 6, 6, 6, 10, 7, 8, 14, 8, and 6 metra respectively. 151 Dale (1968) 48. In recited anapaests usually only one short syllable overlaps between the metra of the clausula (e.g., 36 ἁλὸς ἀντήσαν¦τες ὄλοιντο); only very occasionally does an overlap of two syllables appear: Pers. 28 ψυχῆς ἐν τλή¦µονι δόξῃ S. Aj. 220 κείνου χρηστή¦ρια τἀνδρός, 1416 (perhaps interpolated), Ant. 161, Tr. 1263. 152 Rupprecht (1950) 23–24, Parker (1958) 84–85. 153 See also W.SA 153, n. 20, Schmid (1934) 283, n. 1. 57 32. ξὺν ὄχῳ ταχυήρει: the audience might have been as uncertain as we as to whether this goes with the temporal or the main clause.154 Aeschylus may actually have been deliberate in leaving the matter unclear (cf. on 15–18). 33–36. The exact construction of the datives (if there is one) is disputed (see FJ–W). One solution would be to take λαίλαπι χειµωνοτύπῳ as a local dative (further qualifying the demonstrative ἔνθα),155 and the following datives as dependent on ἀντήσαντες: ‘there in the storm-beating vortex, meeting with the thunder, the lightning and the rain-bringing winds of the savage ocean’. But ἀντήσαντες may also go with the genitives, in which case the datives become more diffuse grammatically.156 34–35. βροντῇ … ἀνέµοις: adapted from an Hesiodic formula: cf. Th. 140 Βρόντην τε Στερόπην τε καὶ Ἄργην ὀβριµόθυµον, 845–46 βροντῆς τε στεροπῆς τε πυρός τ’ ἀπὸ τοῖο πελώρου | πρηστήρων ἀνέµων τε κεραυνοῦ τε φλεγέθοντος.157 37–39. The distress of the Danaids is reaching its peak, and they finally give expression to their worst fear: sexual intercourse with the Aegyptiads. Then they start to sing. 38. σφετεριξάµενοι πατραδελφείαν: about the accent on the noun, see FJ–W. ‘This unclehood’ or lit. ‘father-brotherhood’, denoting the Danaids in relation to their suitors, is a sort of ‘patronymic abstract’, actually meaning ‘cousinhood’.158 The point of using the suggestive πατραδελφεία instead of the mundane ἀνεψιαί is to emphasise the incestuous quality of the ‘appropriating’ of the cousins. 154 With θεῖναι Hermann, Weil, Wecklein–Zomaridis, Wilamowitz, Bassi, Friis Johansen and, in fact, M, who has a colon after ταχυήρει: most editors, however (e.g., Murray, Page, FJ–W, West), take the adverbial with πέµψατε. 155 Cf. 219 δεξάσθω χθονί, FJ–W 219n. and 69n. 156 So Verdenius (1985), and West implicitly, putting a comma after ἀνέµοις. 157 Cf. also Hes. Th. 286, 504–5, 691, 707, 854, [A.] Pr. 1083–84, Ar. Av. 1745–46, 1751. 158 FJ–W object to the meaning ‘uncle’s offspring’ that ‘“offspring” is not denoted by any part of the compound’: this is not necessary, however. The collective is naturally named after its father, just as—for instance—∆αναΐδες after ∆αναός; but an abstract singular is used here instead of the plural (cf., e.g., Pers. 1 τάδε), and the father’s capacity as an uncle of the Aegyptiads is used instead of his name to form the ‘patronymic’. Cf. ἀδελφιδοῦς, nephew, and ἀδελφιδῆ, niece, formed from the word for ‘brother’ by means of a simple suffix, and παρθένιος, ‘born of a virgin’ (see Janko on Il. 16. 179–81). 58 39. ἀεκόντων: West’s comma after the adjective is somewhat confusing, the construction being λέκτρων … ἀεκόντων ἐπιβῆναι. The uncontracted form (ἀκόντων Hermann), being unusual in tragedy and in Attic prose, is—as has been noted (e.g., Blaydes 1902, FJ–W)—well suited to anapaestic metre and also found in this metrical context in S. Tr. 1263. Cf. also 227n. 40–175: First Ode159 The first choral ode is conceived as a prayer to Zeus and the other gods, but at the same time it is reminiscent of a deliberative speech, urging alliance (cf. on 147). Early deliberative rhetoric was heavily influenced by forensic speeches (Kennedy 1963, 204), and FJ–W’s observation that the parodos and first ode exhibit a high frequency of words and expressions from legal terminology is important.160 The rhetorical character of the ode is indicated early 159 ‘Parodos’, according to most editors and critics. This is perhaps the received modern term, but it may be based on a misunderstanding of Arist. Po. 1452b20–25: χορικοῦ δὲ πάροδος µὲν ἡ πρώτη λέξις ὅλου [ὅλη *Westphal] χοροῦ, στάσιµον δὲ µέλος χοροῦ τὸ ἄνευ ἀναπαίστου καὶ τροχαίου. Aristotle (if the passage is authentic) restricts the term parodos to the first speech (λέξις) of the chorus: as Lucas ad loc. points out, elsewhere in Aristotle λέξις is contrasted to µέλος, ‘song’. This contrast is amplified by the explicit mention of µέλος … τὸ ἄνευ ἀναπαίστου καὶ τροχαίου: these metres, as observed by Dale (1950, 15 [35]), refer to the recitative: ‘the anapaestic dimeter and tetrameter and the trochaic tetrameter catalectic, which here convey the meaning “recitative metres” in general’. Thus the ‘stasimon’ is explicitly lyrical, whereas ‘parodos’ in the Aristotelian sense would appear to refer not to a choral ode, but to the recitative (anapaestic) passages to which choruses sometimes enter (see 1–39n.). This is often considered an archaism (see ibid.), but the practice may have been revived before Aristotle’s time: so in the beginning of the probably fourth-century Rhesus. Aristotle’s definition of a stasimon better fits the first choral ode of the Supp., as does its etymological sense, ‘standstill’. It has been labelled thus by Paley 1–39n. (who refers to Σ E. Ph. 202 [his 210]) and Tucker p. xxxvi; cf. also the refs in LSJ s.v. στάσιµος II.3.b and Rode (1971) 89. There are several cases, particularly in Sophocles, where the first utterance of the chorus is lyrical: in many, perhaps most, of these cases the chorus has already entered silently, and stands still in the orchestra, wherefore they ought not to be labelled ‘parodoi’. On the Aristotelian terms in general see also Taplin (1977) 470–76, who concludes (p. 475) that ‘the chapter [12 = 1452b14–27] is totally inapplicable to fifth-century tragedy’ and thinks that ‘its authorship must be seriously in question’. 160 Cf. 6 ἐφ’ αἵµατι, 38 σφετεριξάµενοι, 57 λόγου … ἐν µάκει, 147 παντὶ … σθένει, 171 ἀτιµάσας with mine and FJ–W’s notes, and also 53–54 ἐπιδείξω πιστὰ τεκµήρια, a commonplace in orations of all kinds. On elements of rhetoric in the Supplices 59 on by phrases common to a rhetorical exordium (49–57, q.v.). The chorus thereafter skilfully argues its case, enumerating reasons why the gods should intervene on their behalf: they are kin to Io and Zeus (strophical pair 1, ephymnion 2, antistrophe 8); they are pitiable (strophes 3, 6); righteousness is on their side (strophe 3); their enemies are evil (antistrophes 3, 5); the gods are mighty and holy (strophical pair 4, strophe 5); rewards for sacrifices come from benevolent gods (antistrophe 6); and, finally, unfortunate consequences and shame will come to Zeus if he should be disinclined to help (strophical pair 8).161 This bold stance clearly indicates the pretensions and confidence of the Danaids: contrast the attitude of the Theban women in the similar prayers in Th. 108–81, 417–630 passim, where the tone is humble and there is hardly any coherent argument at all,162 and also the fairly conventional prayers in Ch. 306–509 passim, 783–806. The Danaids use the same strategy here towards the gods as later towards the king of Argos (see 341–465 with notes). The language is peculiar at times, even by Aeschylean standards; perhaps a deliberate means to depict the desperation as well as the foreignness of the Danaids (cf. 118–19). It would certainly be a mistake to try to emend away all anacolutha and ungrammatical passages, but it is hard to determine what is corruption and what is actually intentional. A very loose working principle for textual criticism here (and perhaps generally in Aeschylus) might be to very forgiving of breaches of grammar and syntax, even unparalleled ones, but to try to emend unacceptable sense. Cf. 15–18n. and n. 127 for a list of passages from the first half of the drama in which the grammar is ambiguous or ‘incorrect’.163 see also Buxton (1982) 67–90, Gödde (2000) 177–218. For a modern perspective on tragedy and rhetoric, see for instance Halliwell (1997, 141): ‘we can and should read the rhetoric of tragedy in ways which go beneath the surface of style or technique to the latent patterns, and the lurking anxieties, of a cultural mentality which sustained and mistrusted rhetoric in equal measure.’ 161 Cf. Th. 1.35.4 for a similar rhetorical strategy, towards the end of the speech of the Cercyrean embassy to Athens: having first enumerated the beneficial consequences for Athens of an alliance against the Corinthians, the Cercyreans hint that a refusal would shame the Athenians: πολὺ δὲ ἐν πλέονι αἰτίᾳ ἡµεῖς µὴ πείσαντες ὑµᾶς ἕξοµεν. 162 There is mention of kinship with Aphrodite in 140–43, of sacrifices in 180–81. 163 ‘How do we find the right balance between those on the one hand who assert that Aeschylus is a difficult writer, whose style does not obey the logic of prose, so that 60 The metre: I follow the latest fashion and adopt Dale’s (1951, 21ff. [63 ff.]) ‘s–d’ notation throughout, on which see especially Sicking’s Griechische Verslehre (1993) and the comprehensive reviews of Slings (1996: laudatory) and West (1994: critical). This is useful for giving a concise indication of the general type of rhythm without necessarily having to attach a dubious label to each colon. I have put names on the more regular and well-known cola only. Following Slings (1996, 458–59), I note double-long segments as such, using ‘š’ instead of ‘n’, thus marking the affinity with the single-short and making it useful in notating syncopated iambics and other more ‘unwieldy’ metres.164 I have retained the convention of ^d for $$', which is useful for, among other things, indicating ἐπιπλοκή (see 87n.).165 emendation is to be practised as rarely as possible, and on the other hand those who assert that his style seems difficult only because his text is corrupt?’ (Garvie 2001, 3). Perhaps one day we shall have computer programs which are able to calculate the probability of corruption in any passage of a given author and propose the statistically most likely emendations: until then, we must to a degree trust our intuition, even (or in particular) when we are unable to give an a posteriore rationale. We should remember that intuition as such is not adverse to objective scholarship or ‘science’: I have quoted Popper in this matter elsewhere and need not repeat myself here (see Sandin 2001, 155, n. 36). 164 I do believe, pace Sicking (1993, 213) and Slings (l.c.), that syncopation is as useful and relevant a convention as resolution, contraction, and cholosis: see West (1994) 187–88, Diggle (1994b). In the listener-response perspective championed by Sicking and Slings, the rhythm may well be recognisable as akin to iambic with the proper intonation and/or musical accompaniment. Accordingly, it might be a good idea to mark the affinity of the double-long and the single-short segment: hence ‘š’. This may also appear within cola, for instance in syncopated iambics. Ionics, for example, may then be written ^dšdš etc. 165 There is, of course, a certain arbitrariness to the s–d notation. A rhythmical sequence will have a different notation depending on the context: e.g., the ‘ithyphallicon’ ('$'$'') will be s'š (= 2 ia sync) in the context of syncopated iambics, but ss× among ‘dactylo-epitrite’ or other more ‘open’ rhythms; similarly, the ‘lecythion’ ('$'$'$') may in one context be sss, in another s's. Also, some types of metre, for instance dochmiacs, yield curious results when analysed according to the s–d system. As for dochmiacs, the second ‘anceps’ must be analysed in terms of natural or dragged short to yield acceptable results; however, in this case the s–d notation clarifies the affinity with syncopated iambics (see below, 117–22 = 128–33n., n. 282). For the sake of consistency I have tried to set an s–d notation in every case. In my notes on the metre in the running commentary, I do reckon with defined metrical 61 The following symbols and abbreviations have been used: | Significant word-end (as defined by Maas 1962, 84) coinciding in strophe and antistrophe ¦ —“— including elision || Period-end indicated by brevis in longo or hiatus 166 ||| End of strophe $, + short syllable, brevis in longo ' long syllable %, &, ,, -, ), ( long syllable in strophe answered by short in antistrophe, and so on ! anceps or blunt close " short anceps in strophe as well as anti-strophe # long —“— s, ss, … '$', '$'$', … d, dd, … '$$', '$$'$$', … s's, d'd, … '$''$', '$$''$$', … ^ d truncated d-segment ($$' beginning a colon) d contracted d-segment (''') š syncopated s-segment ('') 167 s dragged s-segment (''') r s resolved s-segment ($$$') sr —“— ('$$$) r r s —“— ($$$$$) (r) s, etc resolved s-segment in strophe answering to unresolved in antistrophe, or vice versa ar aristophanean ch choriambic cr creticus dact dactylic dact-ep dactylo-epitrite δ dochmiac hem hemiepes cola as a reality; however, for a detailed critique of traditional, colon-based metrics, see Sicking (1993) and Cole (1988) with the review of the latter by Diggle (1990). 166 I have taken the ephymnia to be organic parts of their preceding stanzas. 167 On ‘drag’ or cholosis see Dale (1951) 23–24 (67–68). 62 ia io ith lec cat sync 4 dact, 2 ia, etc. + iambic ionic ithyphallicon lecythion catalectic syncopated dactylic tetrameter, iambic dimeter, etc. synapheia connecting different metra (ia + ch) or cola (ia + ar) To avoid more than necessary confusion, the symbols for anceps and blunt close (!, ", #) have been used only in the s–d notation and the symbols for long and short feet ($, ', +) only in the traditional one. The traditional notation is thus, apart from the colometric arrangement and the assumption of brevis in longo at period-end, purely descriptive, eschewing such notation as for instance * (which in some cantica denotes a contracted double-short), whereas the s–d notation contains certain measures of interpretation.168 Specific comments on the metre are given in the running commentary (in the commentary on the strophe, unless a certain textual problem in the antistrophe requires a metrical analysis). The metrical sequences thus commented on are marked with an asterisk (*). Footnotes indicate where my colometry differs from that of West. The first half of the ode is mainly dactylic (or choriambic), but iambic (singleshort) elements come to dominate in the latter half, being accompanied by a change in tone and content (see 112–75n.). 40 ~ 49 41–43 ~ 50–52 43–44 ~ 52–53 45–46 ~ 54–55 47–48 ~ 56–57 '$$'$$'| '''$$''$'''$'''$$'|* '$$'|$$'$$''|* '$|$'$$'$$'$$'$$|'$$''|* '$'|$$'$'''|||* dd (hem) d's#s#d (dact-ep) ddd! (4 dact cat) dddddd! (7 dact cat) sdss 168 The intention has been to minimise as far as possible the apparatus of notation and terminology. The chaotic state of this in modern metrical research is well brought out (and with stoical calm) by Danielewicz (1996, 9–32). 63 58 ~ 63 59 ~ 64 60–61 ~ 65–66 62 ~ 67 68–69 ~ 77–78 70–71 ~ 79–80 72 ~ 81 73 ~ 82 74 ~ 83 75 ~ 84 76 ~ 85 '$$''|$$''$$'| ''$$'$|'$'|* '$'$$''|$$|''$$'|?'$$|?''$$'|* ''$'¦$'$-||| '$$'|$$'$$'$$'$|$'+||* ''|'$$'|'$$'|$',|| $'$$'$''|* $'$¦'$'$'| ')'$$'|'$$'|* '$$'|$$'| '$|'$'$,||| 86 ~ 91 87 ~ 92 93 ~ 88 94 ~ 89 95 ~ 90 '''$$|'$$''|169 $$'$$|''''$|$''|* '''|$$'| '$'$'''$'| 170 $$'$''|||* 96 ~ 104 97–98 ~ 105–6 98–99 ~ 106–7 100 ~ 108 101 ~ 109 102–3 ~ 110–11 $''¦'$'| $'$''|$''|$'| $''$'$''|* '$$''$$'| '$$'&'$'|* '$$'$'$|''¦$$'|$''|||* "š's (2 ia sync) "s's's (3 ia sync) "š"s! d'd (2 ch) d!s (ch + ia) d"s'ds! (ch + ia + ar) 112 ~ 123 113 ~ 124 114–15 ~ 125–26 116 ~ 127 117 = 128 118–19 = 129–30 120 = 131 121 = 132 122 = 133 %'$$$$|$$$|$$$'|$'|* $$$¦$$$| $$$$'| * $'|$'|$'$'$'$'|171 '$$'$''|| '$''$|'$'|''|* ''$¦''|'|'|''|* '$$¦'$'|* 172 '$$$|$$$'| '$$'|$''||| !sr"rsr"s (3 ia) "rsr"rs (2 ia) "s"s"s (3 ia) ds! (ar) s's"s #s'š's ds (δ) srsrs ds! (ar) 134–35 ~ 144–45 $'$'|$'$'&¦$$$'|$'|$-|| 136–37 ~ 146–47 $'$'¦$'$¦'$'|'$|'$'|* 138–39 ~ 148–49 $''|'$'¦&'$'''| 169 d'd'd (3 ch) #dss sd'd'd'd'd #s's (2 ia) ddddd! (6 dact cat) dd'ds! "ds! "s"s (2 ia) dd'd dd (hem) sss (lec) ddd! (4 dact cat) ddd'd! dd (hem) ss#s ^ ds! ^ "s"s!rs"s (4 ia) "s"s"š!s (4 ia sync) "š's!s'š (4 ia sync) 86–87 ~ 91–92: εὖ θείη … παναληθῶς | ∆ιὸς ~ πίπτει δ’ ἀσφ- … ἐπὶ νώτῳ | κορυ-. 170 94–95 ~ 89–90: δάσκιοί τε … πόροι | κατιδεῖν ~ κἀν σκότῳ … τύχᾳ | µερόπεσσ-. 171 114–15 ~ 125–26: ἰή, ἰή … -πρεπῆ ~ ἰώ, ἰώ … πόνοι. 172 120–21 = 131–32: πόλλακι δ’ ἐµπίτνω | ξὺν. 64 140 ~ 150 141 = 151 142 = 152 143 = 153 '$'|$''| '$|''|$$|'+||173 ''|''|$+|| $$$$$$$'$'||| s"š (2 ia sync = ith) s'd! #š's (2 ia sync?) "rsr"s (2 ia) 154 ~ 168 155 ~ 169 156 ~ 170 157–58 ~ 171–72 159 ~ 173 160 ~ 174 161 ~ 175 '$¦'|$'-|| '$'$'|$'| ''$'|174 '$'$'$''$¦'$'$'| '$'$'$,|| '$'$',|| '$''|$'|$'$,||| ss! (ith) sss (lec) #s (ia) sss'sss (lec + lec) sss (lec) ss! (ith) s's!s (3 ia sync) 162 163 164 165 166–67 ''|''|$'|175 ''|''¦'$'|176 ''¦$'| $$'|'$$''|* 177 $$'$|''$$|'$|''||| #š's (2 ia sync?) #š#s (2 ia sync?) #s (ia) ^ d'd! (2 io) ^ ds'ds! 40–111. There is a significantly high frequency of epic words and wordforms in the first, dactylic, part of the choral ode (see above, 40–175n., and below, 112–75n., on the metre). Cf. 40 ἐπικεκλοµένα, 52 µνασαµένα, 60 Τηρεΐας, 63 ἐργοµένα, 67 δυσµάτορος, 68 τώς, 81 στυγόντες, 84 ἀρῆς, 90 µερόπεσσι, 101–3 αὐτόθεν κτἑ and also 44–46n., 83n.178 40–44. νῦν δ’ … βοὸς: cf. E. Ph. 676–81: on the two odes, see Willink (2002). 40. ἐπικεκλοµένα: the reduplicated forms in κεκλ- are aorists in Homer (pres. κέλοµαι), but might perhaps have been thought of as presents by the tragedians.179 An unequivocally present κέκλοµαι appears in Hellenistic times (A.R. 1.716 etc.). For the epicism, see 40–111n. above. 173 141–42 = 151–52: σπέρµα … µατρός | εὐνὰς. 156–57 ~ 170–71: τὸν γάϊον | τὸν ~ τὸν τᾶς βοός | παῖδ’. 175 162–63: ἆ Ζήν … ἰώ | µῆνις. 176 163–64: µῆνις … ἐκ θεῶν | κοννῶ. 177 165: γαµετᾶς οὐρανόνικον. 178 See further Sideras (1971) 109 (n. 57), 194, 210–11, 216, 244–45, 254, etc. There are epicisms, as noted by Sideras, in the latter half of the ode as well, but not as many and as conspicuous. 179 So Sideras (1971) 109, FJ–W. Cf. 591 κεκλοίµαν, S. OT 159 κεκλόµενος. 174 65 Turnebus’ emendation (-όµεναι M) is fairly certain in the light of the parallel construction, with the singular number, in the antistrophe (49 ἐπιλεξαµένα κτἑ). 41–44. ἀνθονόµον: The word-order of Mpc, ἶνίν τ’ ἀνθονόµου τᾶς προγόνου βοός, is impossible.180 Surely it would work only if ἀνθονόµου were not an attribute, but a designation, of Io: a personal name or an epithet so familiar as to have become a noun (Ἀνθονόµου, τᾶς προγόνου βοός). Interestingly enough, the priestess of Hera, which is the office held by Io before her metamorphosis (see 291–92n.), was given the title Ἄνθεια in Argos (Paus. 2.22.1; cf. FJ–W 43n.). But as the epithet ἀνθονόµος is not found outside this drama (also at 539), this has to be considered a less likely solution. Tucker’s ἀνθονόµον τᾶς (adopted by Murray) has been unjustly and summarily disregarded by the latest editors, but is now well defended by Willink (2002, 713).181 The argument against this emendation has been that such an epithet is irrelevant to Epaphus but fitting, almost traditional, for Io (see, e.g., FJ–W and Whittle 1964a). This argument makes a point but not, I believe, a very strong one (see Willink l.c.). Io’s native country (but not she herself) is called ἀνθονόµους ἐπωπάς in 539 and ποιονόµοις τόποις in 50. Also, flowers seem to be pictured as sprouting from her feet on an Attic hydria,182 and much later Severus the Sophist (i. 537 Walz) reports that violets (named after her: ἴον < Ἴω) grew at her feet (cf. also EM s.v. Εὔβοια, Suda s.v. Ἶσις). But if flowerbrowsing was traditionally connected with Io, we might as well say that Aeschylus, applying this epithet to her son—who, although perhaps metaphorically, is called πόρτις in 41, and again in 314—is simply being innovative. ἶνιν ἀνθονόµον would not by itself show (pace Whittle 1964a) that Epaphus is actually conceived in the form of a bull; one may for instance visualise a human Epaphus, being taught by his bovine mother to feed on flowers, which would be a likely scenario if violets actually did sprout at her feet. Aeschylus leaves it unclear whether Io is fully returned to human state before conceiving in 578 ff.; in 569–70 she is described as τὰν µὲν βοός, τὰν δ’ αὖ γυναικός. 180 For a (less than convincing) defence, see, however, Kraus (1957) 40 (who had adopted Porson’s ἀνθονοµούσας before in his 1948 edition). 181 The corruption of -νόµον into -νόµου could be due to an error when copying a minuscule source: see 110–11n., n. 276. 182 Beazley (1963) i. 579. They rather look like—and are thus described by Hoppin (1901) 335 ff.—‘four small bushes’, although painted in purple. 66 As for the appearance of Epaphus, Aeschylus may also be deliberately vague. It appears plausible that he, like Herodotus,183 would identify Epaphus with the Egyptian god Apis, who was definitely a bull; but he may well have considered the explicit image of a bovine king of Egypt unsuitable for an Athenian audience, preferring to leave the matter obscure.184 The Egyptian Apis is not mentioned by name in this drama (cf. on 117 = 128, 260–70). The balance between the two parts of the expression is, as Tucker observes, desirable: ἶνιν in the second part would seem to want at least one adjective to stand up against the formidable ∆ῖον πόρτιν ὑπερπόντιον τιµάορ’. ἶνιν ἀνθονόµον is the perfect way of expressing πόρτιν in other words (calf = flowerbrowsing son) as well as balancing not only the poetical rhetoric, but the two aspects of Epaphus’ heritage against each other, expressing a male–female polarity which reappears often throughout the Supplices: Zeus’s calf, the avenger—and the flower-browsing son of the cow. Being a paraphrase of πόρτιν, ἶνιν ἀνθονόµον at the same time offers a poetic contrast to ὑπερπόντιον τιµάορ’. Porson’s ἀνθονοµούσας has been the emendation of choice among a majority of prominent twentieth-century editors,185 but as Tucker observed, the lack of a definite article for the attributive participle is unacceptable. Whittle (1964a, 25) tried to refute Tucker, presenting a number of examples of what he claims to be parallels for attributive, ‘quasi-adjectival’ participles without a definite article. None of them is comparable to the present one: the reason for the lack of a definite article in Whittle’s examples is the simple fact that there is, unlike the present context, no definite aspect.186 West (W.SA) 183 2.38, 2.153, 3.27–28; cf. also Luc. Salt. 59, Ael. NA 11.10, the modern refs by Lloyd on Hdt. 2.38, and my 117 = 128n. 184 For the arguments of either side, see Tucker and Whittle (1964a) 24 ff. 185 For instance Wilamowitz, Friis Johansen, Page, FJ–W. 186 Cf. K–G i. 624, Anm. 5. Whittle compares Od. 4.446, 4.567, 8.222, 11.414, 12.70, 14.358, A. Th. 443, Supp. 779–80, Ag. 1234–36, S. Ant. 1146–47, Aj. 135, E. Tr. 1080, El. 771. But one does not look for a definite article in an expression like Ag. 1233–36 Σκύλλαν τινά (sic!) οἰκοῦσαν ἐν πέτραισι …, θυίουσαν Ἅιδου µητέρ’ ἄσπονδόν τ’ ἄρη … πνέουσαν, nor in S. Aj. 134–35 Τελαµώνιε παῖ, … Σαλαµῖνος ἔχων βάθρον, where ἔχων takes the vocative case. In S. Ant. 1146–47 ἰὼ πῦρ πνεόντων χοράγ’ ἄστρων … ἐπίσκοπε, ‘O! leader of stars breathing fire in the dance!’, ἄστρων is not, strictly, definite either, and it would not need an article if it were (see below). That only leaves a few of the Homeric examples (for more of which see also S.GG ii. 408), 67 defends Hartung’s ἀνθονόµου τὸν; I am inclined to agree with Lloyd-Jones (1993, 5) that the word-order hardly would favour this solution any more than that of the ms. 41–43(~50–52). The colon is one of the few examples of what appears to be dactylo-epitrite verse in Aeschylus.187 42. τιµάορ’: This noun is normally of the 2nd declension (τιµωρός, -άορος). Note that an audience familiar with the standard declension may well have conceived of the noun as a vocative, τιµάορε (cf. the similar inserted vocative, ὦ ∆ιὸς γένεθλον, in the Euripidean reminiscence cited above, 40–44n.). If this is an intended effect, we find that Aeschylean verbal ambiguity (cf. 8n., etc.) goes down even to the morphological level. τ’, deleted by Hermann, is better retained, pace FJ–W.188 Cf. D.GP 502 and see 60–62n. below. ‘Elmsleys canon’, defined by examples in Elmsley which are of little value, as the definite article proper is a rarity—according to some even non-existent—in Homer (see Russo on Od. 17.10, Monro 1891, 224–34, Chantraine 1963, 158–66). Even in Homer, however, a definite quality, if needed with the participle, is usually expressed by some other means, for instance a personal name: e.g., Od. 4.567 Ζεφύροιο λιγὺ πνείοντος. The same is true for words such as ἄστρα, θάλαττα which are in themselves ‘famous’ enough to confer a deﬁnite aspect without the article (see K–G i. 602–3). In our case, ἶνις ἀνθονοµούσας Ἰοῦς would have been theoretically acceptable as a Homerism. ἶνις ἀνθονοµούσας προγόνου βοός is ‘scarcely Greek’ (Tucker) and would, if at all possible, mean ‘son of an ancestral cow who is feeding on flowers’. 187 Contrast Pr. 526–60 and 887–907; and cf. Griffith (1977) 66–67. 188 FJ–W’s arguments against τ’ are feeble, except one which is inconsistent with the text they have themselves adopted. FJ–W refer to Parker (1966) 17 (cf. also 4–10), who argues that the word-end after a long anceps severs the ‘final choriamb’ in a way unparalleled in Aeschylus. Parker accepts Tucker’s ἀνθονόµον, which yields shared word-end in strophe and antistrophe after that word, hence suggesting end of colon and/or period: with Porson’s ἀνθονοµούσας, adopted by FJ–W, there will be no severed final choriamb, and Parker’s observation becomes irrelevant. As I favour Tucker’s emendation, however (see 41–44n.), this objection requires scrutiny. First, it is hard to tell how our instance of '|'$$'|| could be said to be ‘unparalleled’ in Aeschylus in any meaningful way, as the dactylo-epitrite metre (which Parker labels ‘trochaeo-choriambic’ [p. 17] or ‘choriambic admixture in iambo-trochaic context’ [p. 9]) is not found elsewhere, apart from parts of Pers. 852–906, which strophes are, however, mainly dactylic, like the present stanza. In the few existing cases in Aeschylus of similar metre, this kind of anceps (see Maas 1962, 40; Dale 1968, 179) is invariably long: here, in the middle of the s×s segment just before, and perhaps at 68 (ed. Medea) pp. 225–26 (241–43), and discussed further by Fraenkel on Ag. 1585, 1526, states that the Greek dual expressions in which two different interpersonal relationships of the same individual are described (for instance Pers. 151–52 µήτηρ βασιλέως, βασίλεια δ’ ἐµή, or S. Tr. 739–40 τὸν ἄνδρα τὸν σὸν ἴσθι, τὸν δ’ ἐµὸν λέγω πατέρα), have δέ as coupling particle (with µέν regularly withheld), or, in a few passages τε … τε. According to Elmsley (supported by Fraenkel), a single τε coupling the two elements is forbidden: ‘si τε non est in priore membro, non potest esse in posteriore, nisi hujus subjectum, ut vocant, diversum sit a subjecto prioris’. If we accept that such a canon may be formulated,189 we may observe an important difference in our case, which should make it exempt from the rule. The two relationships in question here (∆ῖον πόρτιν and ἶνιν … βοός) are actually of the same kind—the kind, that is, that exists between a child and a parent. Elmsley’s rule, if at all applicable, should be so only to instances where two specifically distinct relationships are coupled. δέ is naturally used where a contrast is implied, for instance between your daughter, but my sister 94~89 (q.v.). We may observe word-end after long anceps in the lyric choriambics / iambics in 109 of our drama (q.v.); furthermore, as Parker herself points out, the elision τ’ softens the impact of the word-end: ἶνίν τ’ ἀνθονόµον will be felt as a single entity. Parker’s article is an attempt to extend ‘Maas’s law’ to a more general prin-ciple concerning all metres of ‘serious Greek poetry’. Maas’s law reads (l.c., 34–35): ‘the following rule applies to several metres which contain the rhythm !'$'!: no word can end after long anceps, except at the caesura in the middle of the line’. He applies the rule to (1–2) the trochaic tetrameter and iambic trimeter of early iambo-graphers, tragedy and satyric drama (= Porson’s law), (3) the dactyloepitrites of Bacchylides (but not of Pindar and the tragedians!), (4–5) the trochaic tetrameter and dimeter of Alcm. fr. 1, (6) the catalectic trochaic tri- and pentameters of Call. fr. 202 and 399 respectively, (7) the end of the iambic tetrameter found in S. Ichn. (fr. 314) 298–328. As is obvious from several examples (S. OT 1090, E. Andr. 772, Hel. 1481), the law is not obeyed by Sophocles and Euripides in their dactyloepitrites, and Pindar is even more negligent. 189 There are, as suggested by FJ–W, perhaps rather too many examples of a single τε in similar passages in the mss. of our authors to justify it: at least Pi. O. 7.13–14 appears to have nothing to gain from an emendation to δέ (any more than it has been improved by the intensive punctuation bestowed by some editors, against which cf. Stinton 1977, 33 [317]): κατέβαν τὰν ποντίαν ὑµνέων παῖδ’ Ἀφροδίτας Ἀελίοιό τε νύµφαν Ῥόδον. On the other hand, a preparatory τ’ is easily supplied after Ἀφροδίτας. 69 (S. OC 322–23); whereas in the present case, where we have two different designations for the same child-parent relationship, there is no contrast to speak of: ‘calf of Zeus and (not “but”) flower-browsing son of the cow’. The same is also true in Pr. 137–40, which thus becomes irrelevant, pace FJ–W, as evidence against Elmsley’s canon (thus modified): Τηθύος ἔκγονα τοῦ περὶ πᾶσάν θ’ εἱλισσοµένου χθόνα … παῖδες πατρὸς Ὠκεανοῦ. That the τ’ is lacking in the scholiast’s paraphrase is irrelevant, as he leaves out the ∆ῖον πόρτιν part, thereby making it impossible to reproduce the τ’. 44–46. βοὸς … αἰών: the punctuation is of crucial importance. If the stop is put after ἔφαψιν, we get a very harsh anacoluthon, with no way of adjusting this word to the syntax of the previous sentence. The advantage, on the other hand, would be that ἔφαψιν becomes more closely connected to Ζηνός, as in 17–18 βοὸς ἐξ ἐπαφῆς κἀξ ἐπιπνοίας ∆ιὸς, of which 44–45 would be, as it were, a confused paraphrase (cf. W.SA). As for the following sentence, both manners of punctuation would be possible: if ἔφαψιν goes with Ζηνός, ἐπεκραίνετο takes the passive voice.190 That the fated αἰών was ‘fulfilled’ would presumably mean that Epaphus’ period of gestation was so (see FJ–W 46n.), resulting in his birth. With the other alternative, ἔφαψιν being taken with the subsequent passage, which I find preferable, we have a transitive middle ἐπεκραίνετο (as in Eu. 969) governing ἔφαψιν: the αἰών fulfilled the touch.191 There is still an anacoluthon, as the former clause lacks a finite verb, but this reading is nowhere near as difficult as the one yielded by the traditional punctuation. Besides the convincing arguments of Diggle (1982) who mentions, with Tucker, Pi. O. 2.10 αἰὼν δ’ ἔφεπε µόρσιµος,192 the single Ζηνός at the beginning of the verse would make an example (more apparent than Ζηνὸς ἔφαψιν) of the distinct Homeric stylistic feature of ‘progressive enjambment’ (see Kirk on the Iliad books 1–4, pp. 31 ff.), which is appropriate in dactylic 190 So the mss., Σ on 45–56 and, e.g., Wilamowitz, FJ–W, West. Willink (2002, 714) avoids the problem by reading ᾇ ’ξ ἐπιπνοίας together with Auratus’ ἐπωνυµίαν (on which see further below, 45–47n., n. 193): ‘…(calf/offspring) of the ancestral cow for whom, from the breathing-on of Zeus, the due time of birth aptly fulfilled the (god’s) “touching” (as) eponymous, and she brought forth “Epaphus”’. Both emendations are palaeographically easy, and the syntax becomes technically correct, but I do not find the style convincing. 191 Schütz (ed. 1794) followed by, e.g., Paley (ed. 1883), Murray, Diggle (1982). 192 Cf. also N. 1.46–47 ἀγχοµένοις δὲ χρόνος ψυχὰς ἀπέπνευσεν. 70 verse (so also εὐλόγως in the following: cf. Ag. 105, 124 and 86–87n. below) and also in accordance with the frequent Homerisms of the former half of the choral ode (see 40–111n.). Against West (W.SA), who maintains that the similar phrase found in 17–18 (see above) necessitates taking ἔφαψιν with Ζηνός here, one can argue that the proximity of Ζηνὸς will be sufficient to make his proprietorship of ἔφαψιν felt, almost as if Ζηνός were taken ἀπὸ κοινοῦ, not only with two nouns, but with two whole clauses. Syntactical ambiguity is common in the parodos and first choral ode: see on 15–18. 45–47. ἐπωνυµίᾳ means ‘naming’, ‘name-giving’, the point being that the µόρσιµος αἰών fulfilled Zeus’ touch εὐλόγως (‘significantly’, ‘meaningfully’), bringing forth a child named after it: ἔφαψιν – Ἔπαφος (cf. 252–53). ἐπωνυµίᾳ is emphasised by the postponement of δ’: it is in this very respect, the naming, that the touch is fulfilled εὐλόγως.193 On ‘significant’ names in Aeschylus, see Fraenkel on Ag. 682. This type of etymologising word-play is very common in Aeschylus, especially concerning proper names.194 The name Ἔπαφος (whose actual derivation is unknown) might in fact itself have been responsible for the myth that Epaphus was conceived by the ‘touch’ of Zeus (so Wilamowitz 1931, 246, n. 2)—a method that seems rather mundane in comparison with his other ways of producing offspring. 47–48(~56–57). On the dragged clausula sdss see Dale (1971) 9, Willink (1997) 298–300, Willink (2002) 717, n. 6. 48. Ἔπαφον δ’ ἐγέννασεν: δ’ is explanatory (cf. 4n.) of the statement in 45–47. The subject is uncertain and may refer to Zeus, Io, or even the ‘fated period’. In archaic and classical Greek, however, the active tense of γεννάω implies an active creation (‘beget’), such as women were not usually given credit for in the production of children (cf. Eu. 658 ff.). The verb appears, albeit rarely, as referring to the mother; it does not mean ‘give birth’, though, but rather ‘produce’, ‘generate’. Thus it should be translated at X. Lac. 1.3 (not, as LSJ, ‘bring forth, bear’), since the focus is on the proper way of feeding 193 Cf. Schweizer-Keller (1972) 25–26. Oberdick and Page adopt Schütz’s (ed. 1794) punctuation together with Auratus’ ἐπωνυµίαν (an adjective going with ἔφαψιν), which is detrimental. It does not make the postponement of δ’—which already has an exact parallel (direct object–dative adverbial–predicate) in Eu. 531 ἄλλ’ ἄλλᾳ δ’ ἐφορεύει—any easier. It also removes the emphasis on the word ἐπωνυµία. 194 See Kranz (1933) 83, 287–89; Schmid (1934) 297–98, and the further refs in FJ– W ad loc. 71 and caring for pregnant women, in order that they may produce the best possible offspring. The growth or gestation of the child is intended, not the moment of birth (cf. also S. Aj. 1077).195 To be sure, the subject of ἐγέννασεν may have appeared as ambiguous to Aeschylus’ contemporaries as it does to us. However, in the light of the usual sense of the verb, since Ζηνός is closer to the verb than προγόνου βοός, and since the very point of the sentence is the fulfilment of his touch, I am inclined (with Heath 1762) to favour Zeus as the intended begetter. FJ–W (46n.) argue that this sense is impossible because of an alleged parallel in 576–81: ‘the circumstances of [Epaphus’] uterine existence are detailed in strict chronological order in both 44–8 and 576–81: first the twofold impregnation […], secondly the gestation (46 ἐπεκραίνετο µόρσιµος αἰών ~ 580 λαβοῦσα δ’ ἕρµα ∆ῖον), and thirdly the birth (48 ἐγέννασεν ~ 581 γείνατο)’. But ἐπεκραίνετο … αἰών has no resemblance to λαβοῦσα δ’ ἕρµα, which is rather an exact parallel to Ἔπαφον δ’ ἐγέννασεν, but from the perspective of Io. Zeus begets Epaphus in Io’s womb with a touch: Io passively receives the ‘burden’ from him, in accordance with the contemporary view on conception.196 For the aorist in an explanatory δέ-clause referring to a time previous to the ‘explanandum’, cf. Il. 10.240. On the other hand, ἐγέννασεν could be said to be simultaneous to ἐπεκραίνετο: Zeus’s engendering action may be seen as extended in time throughout the gestation of Epaphus. Cf. 206 Ζεὺς γεννήτωρ. 49. ὅντ’: on the enjambment, see FJ–W. The credit for the certain emendation (δὲ γέννασ ἐ|όντ Mpc) is usually given to Porson, but West abstains from mentioning him in his apparatus criticus, seeing that the reading of Mac might have been δ’ ἐγέννασε | όντ’. ἐπιλεξαµένα: FJ–W argue, pace LSJ, that the middle voice always means either ‘select’ or ‘read’: they include ‘consider’ (the most frequent sense of the verb in the middle voice) as a specialised use of ‘select’: ‘[include] in one’s 195 The same goes for Arist. GA 716a22. In Pl. Lg. 930e γεννησάσῃ is interpolated from the use of the plural for ‘parents’, meaning simply ‘mother’. Cf. also the compound ὑψιγέννητος at Eu. 43 which, as Sommerstein ad loc. rightly observes, means ‘grown tall’—not, as LSJ, ‘born on high’. 196 Most critics have taken Io as the subject of ἐγέννασεν, but Wilamowitz (1914, 28, n. 1) suggested that the subject is left undetermined. West does not express an opinion but mentions Schmidt’s (1863, 233) conjecture ἐφίτυσεν, comparing 313, where this verb describes Zeus’s engendering of Epaphus, and also Hsch. s.v. 72 thoughts’. Most critics, however, as well as the scholium and Hsch. s.v. ἐπιλεξαµένη, have taken the verb as synonymous to ἐπικεκλοµένα here, something which ought not to be entirely impossible. It is not certain that the usage of the middle voice had yet cemented into the specific senses mentioned by FJ–W: moreover, Aeschylus tends to stretch and bend the common meanings of words, often discarding a conventional meaning in favour of a more suggestive one and sometimes taking the bare etymological elements of a compound as justification for a surprising sense: cf. 21 ἐγχειριδίοις taken in the unparalleled sense ‘thing held in hand’. On the method of mirroring similar words in strophe and antistrophe (ἐπιλεξαµένα – ἐπικεκλοµένα), see FJ–W ad loc. and below, 110–11n. 50–51. ποιονόµοις … τόποις: cf. above, 41–44n. 52–57. Similar promises to ‘tell the real story’ and to prove one’s case are usually found in the exordia of rhetorical speeches (see 40–175n.): cf., e.g., Lys. 1.5, 3.4, 7.3, 13.3–4, Antiphon Tetr. 1.1.3. 52–55. τά … ἐπιδείξω is a relative clause. Its correlate should be taken as the unexpressed subject of φανεῖται, with πιστὰ τεκµήρια as predicative: ‘(that) which I now shall show forth, will be seen as sure proof’.197 τε makes this a harsh anacoluthon, and it should possibly be emended (see below), unless one may understand τά τε as a form of ὅστε, a form of the relative which is used frequently by Aeschylus, for example just before in 49 (see D.GP 523– 24).198 As for the corruption in 54–55 (τά τ’ ἀνόµοια | οἶδ’ M), it would seem that τά τ’ has been mistakenly repeated from the previous verse, upsetting what originally followed. Hermann’s γαιονόµοισι δ’ is likely to have been what Aeschylus wrote.199 As well as being plausible for palaeographical reasons, 197 So the Greek paraphrase of Wilamowitz in his apparatus criticus (cf. Wilamowitz 1914, 28, n. 2), Untersteiner (ed. 1935), the translation of Friis Johansen and the notes of Verdenius (1985). 198 This possibility was pointed out to me by Dr. C. W. Willink, who observes that there is no parallel in sight (none in tragedy, and I suspect that none exists outside epic verse and Ionic verse and prose). He is certainly opposed to the reading. 199 Adopted by, e.g., Page and West. γαιονόµοισι δ’ is to be preferred to Dindorf’s (ed. 1857) γαιονόµοισιν (adopted by Wilamowitz). Apart from the former alternative’s being palaeographically easier, δέ adds an adversative force which fits the context. The lack of an exact parallel for δέ … περ is understandable: περ as such is a rarity in post-Homeric Greek (D.GP 481), as is δέ connecting participles (but cf. 369 73 the sense fits the context perfectly. γαιονόµοισι δ’ ἄελπτά περ ὄντα means that the evidence the Danaids will produce may be unexpected to the inhabitants of the land—a claim which is illustrated in one of the dramatic peripeteiai of the play, the scene where the young women persuade the king of Argos of their Argive ancestry (289 ff.), in which they come up with something quite unexpected: they are the offspring in the fifth generation of the sacred Io. The anacoluthon, which is FJ–W’s objection to reading τά as a relative, is severe, however. An easy and attractive emendation is τά γε (Professor Richard Janko).200 On γε (≈ δή) with relatives, see D.GP 123–24; cf. also, for ἀστοῖς δὲ … κοινώσας). The ‘proximaposition’ is not impossible in principle, however: the particles retain their independent forces and do not actually form a ‘particlecombination’. Parenthetic µηδέ περ + participle is found in, e.g., Ar. Ach. 223–24 µὴ γὰρ ἐγχάνοι ποτὲ µηδέ περ … ἐκφυγὼν Ἀχαρνέας. FJ–W admit the palaeographical plausibility of γαιονόµοισι δ’, and note 565 γᾶς … ἔννοµοι. What has bothered them and other critics is that compounds beginning γαιο- are unattested in classical Greek. On the other hand, compounds beginning γαια- and γαιη- do not exactly abound either, as observed by Sommerstein (1977, cf. W.SA): apart from the ogygian Homeric γαιήοχος there is not much to speak of. For other compounds with nouns of the α- and ο-stems as the first element, the conjunction may obviously be -ο- or -α-/-η-, regardless of the stem of the noun (cf. S.GG i. 438, and Debrunner 1917, 66: ‘sämtliche Stämme können als Vorderglieder ihren Ausgang durch ein -οerweitern oder verändern’). Sommerstein (1977) observes that ‘if Aeschylus can use ποιονόµος (50, Ag. 1169) when ποιηφάγος was an established form, then he can use γαιονόµος in the same stanza’. Cf. also θανατοφόρος (probably) in Ag. 1176, but θανατηφόρος in Ch. 369; αἱµατηφόρος in Th. 420 but αἱµατολοιχός in Ag. 1478; ξιφηφόρος but ξιφοδήλητος, etc. As for γαιο-, we find, in Hellenistic and later Greek, γαιοφάγος (Nic. Th. 784), γαιοδότης (Hdn. Epim. p. 209.13 Boissonade, Suda s.v., EM 223.17 which reads it in Call. fr. 43.64: v.l. γεωδαῖται in P.Oxy. 2080), γαιοειδής ([Ti.Locr.] p. 219.1 Marg), γαιοτραφής or -τρεφής (Synes. Hymn. 2.282), γαιογράφος (Hsch., Hdn. Orth. ii. 485 Lentz), γαιοµέτρης (Man. 4.210), γαιοφανής (Archig. ap. Orib. 8.2.4, Aët. Placit. 2.30.1 = Dox.Gr. p. 361), and possibly a compound beginning with γαιο- in Trag.adesp. 628a: !'$]ν̣ω̣νται δ’ Ἰλίαν γα̣ιο̣[$' (the metre seems to call for γεω-, unless the ο is long by position). 200 Less probable emendations would be τε τὰ or θ’ ἅτε—with τε coupling the participle clauses ὅντ’ ἐπιλεξαµένα and τῶν πρόσθε πόνων µνασαµένα. Such a postponement of τε is unparalleled (see D.GP 515 ff.). In Th. 7.84.4, where τε stands in fifth place (ἐς τὰ ἐπὶ θάτερά τε), this is a more regular postponement after preposition and nominal. 74 instance, S. Ph. 559 φράσον δ’ ἅπερ γ’ ἔλεξας, S. OT 342, Pl. Cra. 403c. Here it might also modify νῦν, promoting expectancy: ‘what I shall now show forth’ (on the position, cf. D.GP 149–50).201 On the other hand, the anacoluthon may not be impossible. It consists in a participial clause (µνασαµένα) and a finite one (φανεῖται, including a relative subordinate τά … ἐπιδείξω) coupled by τε. Verdenius (1985) observes that ‘transition from a participial construction to a finite verb is rather common in Greek poetry (cf. Bruhn Anh., § 191 [= Bruhn 1899])’. Cf. also Berti (1930) 238–53, who discusses Aeschylean nominativi pendentes in detail. One need only compare the previous strophe to find an example similar to the present one. While there may be no exact parallel to this particular leap from participle by τε to finite clause (including change of subject), there are several other similar anacolutha to be found among Berti’s and Bruhn’s (1899) examples.202 Many of them are themselves unparalleled, which indeed lies in the nature of the anacoluthon: if a syntactical aberrance is repeated often, it is not an aberrance at all, but part of accepted grammar. Dr. C. W. Willink notes the lack of a strong caesura in the dactylic heptameter, which he finds unacceptable, suggesting γαιονόµοις τάδ’ ἄελπτά περ ὄντα φανεῖται, with τάδ’ as a correlate to the previous relative τά (γε): ‘these (things), though unexpected, will appear (as) convincing proofs to inhabitants of the land’. I think the lack of a strong caesura may be acceptable in lyrical dactyls, however. Cf., for instance, Ag. 106–7~124–25, 129. Most critics take τά as the definite article, among them West (W.SA) who translates ‘I present these credentials now, and also later they will be made apparent to the local inhabitants, surprising as they may find them’. Apart from τὰ νῦν πιστὰ τεκµήρια (‘the present trustworthy credentials’) being an 201 It should not be taken as quippe quae (see D.GP 141–42), in which case it would have to refer back: ‘having recalled the former woes, those which I shall now show forth’: τά refers forward, to φανεῖται. 202 In S. El. 444–46 a relative and a finite clause with different subjects are linked by καί: (σκέψαι γὰρ εἰ … αὐτῇ δοκεῖ γέρα … δέχεσθαι …,) ὑφ’ ἧς θανὼν … ἐµασχαλίσθη κἀπὶ λουτροῖσιν κάρᾳ κηλῖδας ἐξέµαξεν. In OT 1199–1202 a participle and a finite verb are linked in a µέν … δέ complex: κατὰ µὲν φθίσας τὰν γαµψώνυχα παρθένον χρησµῳδόν, θανάτων δ’ ἐµᾷ χώρᾳ πύργος ἀνέστα. Participle and finite verb are coupled with τε … καί in Th. 4.100.1 προσέβαλον τῷ τειχίσµατι, ἄλλῳ τε τρόπῳ πειράσαντες καὶ µηχανὴν προσήγαγον. 75 improbable expression in the context,203 this necessitates taking φανεῖται as absolute, in a sense which is not attested. West translates this verb as ‘will be made apparent’ (adding ‘also later’ which has no equivalent in the text). But the passive of φαίνω simply means ‘be shown’, ‘appear’. The English ‘be apparent’ has another sense entirely, synonymous to ‘be evident’: it carries a cognitive significance which the Greek verb lacks.204 To have any meaning in the context, φανεῖται must take a predicative: ‘be shown as something’. Since ἄελπτα is already claimed by the participle, only πιστὰ remains. It is hard to believe that a spectator could have avoided understanding φανεῖται with πιστὰ τεκµήρια—at least if we take the doublelong in ἐπιδείξω as catalexis: τά … νῦν ἐπιδείξω (pause), πιστὰ τεκµήρια … φανεῖται.205 203 As rightly FJ–W. Thus τὰ γονέων Hermann, τὰ γένους Merkel (1858, 273), τοκέων Martin (1858, 16), defended by Liberman (1998): ‘I will show the trustworthy evidence of my birth …’, and also τάδε νῦν Page. These solutions, however, leave φανεῖται without its much-needed predicative (see below). 204 FJ–W argue that ‘be made manifest’ is ‘an acceptable meaning for φαίνεσθαι (cf. LSJ s.v. B. I. 3)’, but their reference offers no support for their claim: the paragraph in LSJ reads ‘of events, come about …; of sayings, be set forth, … S. Tr. 1, cf. OT 474 (lyr.), 848’. These passages are of little relevance to the present one (OT 848 has none, since φανέν there takes a predicative ὧδε). In all three instances the subject of φαίνεσθαι is ‘word’ (λόγος, φήµα and ἔπος), and, more precisely, ‘saying’: ‘old saying’ (S. Tr. 1) or ‘oracular saying’ (OT 474). The sense is not ‘be made manifest’ or ‘prove to be correct’, as we would need in our case, but simply ‘arise’, ‘come into existence’, or, as LSJ, ‘be set forth’. A better parallel for the sense ‘be made apparent’ would be Arist. EN 1175a29 (cf. LSJ s.v. B.II.2): διαφέρουσι δ’ αἱ τῆς διανοίας (sc. ἐνέργειαι) τῶν κατὰ τὰς αἰσθήσεις καὶ αὐταὶ ἀλλήλων κατ’ εἶδος […]. φανείη δ’ ἂν τοῦτο καὶ ἐκ τοῦ συνῳκειῶσθαι κτλ. Here, however, the predicative of φανείη is implicit in τοῦτο, referring back to an already stated proposition (διαφέρουσι): ‘this should appear from …’. ‘This’, i.e. τοῦτο, may ‘appear’, i.e. φαίνεσθαι (sc. εἶναι), in the sense ‘appear to be correct’, provided that τοῦτο is already defined as a complete proposition with a subject and a predicate. τὰ τεκµήρια φαίνεται, on the other hand, can mean no more than ‘the evidence appears’. If τά is taken as the definite article in our case, we have only the subject τά τε νῦν τεκµήρια that φανεῖται could refer back to: a literal translation would be ‘I will show the present trustworthy evidence; they will, albeit being unexpected, appear to the inhabitants of the land’. 205 Cf. the parallel adduced by Diggle (1982): E. HF 802–4 πιστόν µοι τὸ παλαιὸν ἤδη λέχος, ὦ Ζεῦ, σὸν ἐπ’ οὐκ ἐλπίδι φάνθη. 76 52. µνασαµένα: the rare middle aor. is probably an epicism (so FJ–W): cf. 40–111n. 53–54. ἐπιδείξω πιστὰ τεκµήρια: a commonplace in oratory (see above, 40–175n.): cf. D. 30.25 τῶν δ’ ἐπιδείξω µεγάλα τεκµήρια καὶ πίστεις ἱκανάς. More often, however, τεκµήριον takes an adverbial (instrumental) function with ἐπιδείξω, show by evidence (on the syntax here, see above on 52–55).206 In Aeschylus cf. also 271, Ag. 352, Eu. 447, 485, 662: on the latter passages see Kennedy (1963) 43. For τεκµήριον as a technical term, see also Arist. Rh. 1357a–b. 56–57. γνώσεται δὲ λόγου τις ἐν µάκει: Martin’s (1858) emendation (λόγουϲ M) is certain, and well defended by FJ–W. The expression appears to be a variation of a rhetorical stock phrase: cf. Isoc. Trapez. 19.3 (≈ Antiphon Caed.Her. 10) ὡς αὐτοὶ προϊόντος τοῦ λόγου γνώσεσθε, which is echoed in the scholium to our passage: ἐν µάκει] προϊόντος τοῦ λόγου. Cf. 52–57n. 58–67. A mythological digression, in which the Danaids compare themselves, somewhat farfetchedly (it may seem at first: see 63–64n.), to Procne, sister to Philomela and wife of Tereus. According to myth, Tereus raped or seduced Philomela and had her tongue cut out to ensure her silence, but she revealed the deed to her sister by means of a piece of embroidery. The sisters took revenge by killing Procne’s and Tereus’ son Itys, serving him to Tereus for dinner. As the deed was revealed and Tereus went after the sisters, all three were transformed into birds: Procne into a nightingale, Philomela into a swallow, and Tereus, in this version, into a sparrow hawk.207 59(~64). Bamberger’s (1839) ἐγγάϊος (ἔγγαιος M), on which see FJ–W, is necessary, together with Bothe’s (ed. 1805) deletion of οἰκτρὸν, to restore responsion between strophe and antistrophe—unless we read µὲν for νέον in 64 (q.v.). For the ×d start cf. 72~81. The colon (×dss and similar) is usually 206 Cf. D. 28.2 τεκµηρίοις µεγάλοις ἐπιδ. ὡς …, 29.22 ἐκ τοσούτων τεκµηρίων ἐπιδ. ὅτι…, Is. 10.6, Pl. Tht. 158c, etc. 207 In most extant versions (see Thompson 1936, 20; Frazer ed. [Apollod.] ii. 98 ff., and FJ–W for a fairly complete set of refs), Tereus is turned into a hoopoe, but the present one (also in Hyg. Fab. 45) is probably original: see Dunbar on Ar. Av. 15 for an interesting discussion. See also Hall (1989) 103 on the later ‘barbarization’ of Tereus: possibly it was Sophocles who first had the idea of making him out to be the king of Thrace. 77 called ‘enhoplian’ or ‘prosodiac’.208 A more common metric colon, iambic dimeter, of course becomes the result of the regular accentuation. See further 63–64n. 60. δοξάσει τιν’ ἀκούων: the real problem (pace most editors), is not the ἀκούων of the ms., but the τιϲ. Being an indefinite pronoun, it has to be just that—indefinite. A substantival, indefinite τις cannot, even in Aeschylus, refer back to a subject that is more definite in a previous (subjunctive) clause. τις πέλας οἰωνοπόλων ἐγγάϊος οἶκτον ἀΐων in 58–59, a quite well-defined entity, cannot subsequently simply be referred to as τις. If retained, τις in the apodosis refers to an indeterminate person other than the τις … ἐγγάϊος in the protasis. The labours invested in finding parallels consisting in a repetition of τις are misguided: the problem does not lie in the repetition per se, but in the fact that the first τις is not as undefined as the second, but forms part of a defined subject.209 FJ–W rightly observe that Ar. Ach. 569–71 comes close to making a parallel: εἴτε τις ἔστι ταξίαρχος ἢ στρατηγὸς ἢ | τειχοµάχας ἀνήρ, βοηθησάτω | τις ἀνύσας. But even here, the subjects of the two clauses are equally undefined, and/or strictly not referring to the same person: lit. ‘whether there is a taxiarch, or general, or wall-battling man at hand, may someone help, quickly!’. Unlike our passage, the protasis mentions a number 208 See Wifstrand (1965) 76–81, Dale (1968) 157–77, and FJ–W 64n. (ii. 62), who adduce metric parallels from Sophocles and Euripides. 209 West (1990, 9) compares the repeated indefinite pronouns in Ag. 662–63 and Eu. 545–49. In the first of these cases the latter τις is more defined than the former, serving to narrow down the subject and to add precision: τις ἐξέκλεψεν is expressed more exactly by θεός τις: ‘someone removed us … some god, not a human’ (see Fraenkel ad loc.). Similarly Eu. 508–14, where an undefined τις is later narrowed down to τις … πατήρ | ἢ τεκοῦσα, and E. Or. 1218–19 (adduced by Verdenius 1985), in which ξύµµαχός τις κτἑ defines a previous substantival τις further. In the latter of West’s examples, Eu. 545–49, the repetition of τις is a simple pleonasm, both τις occurring within the same clause, both being equally undeﬁned, ‘one’: πρὸς τάδε τις τοκέων σέβας εὖ προτίων καὶ ξενοτίµους ἐπιστροφὰς δωµάτων αἰδόµενός τις ἔστω. Similar pleonasms are found in E. Andr. 733–34 ἔστι … τις οὐ πρόσω Σπάρτης πόλις τις, X. Cyr. 1.6.11 (with further examples in K–G i. 665, Anm. 3): in all cases the latter, repeated τις is more or equally defined as the former, and not comparable to the present passage. (Verdenius 1985 also adduces E. Hec. 1178–79, in which the two τις refer to different people: εἴ τις γυναῖκας τῶν πρὶν εἴρηκεν κακῶς | ἢ νῦν λέγων ἔστιν τις.) 78 of possible subjects, and the τις in the apodosis, being indeﬁnite, conﬁrms that the list is not all-inclusive. ‘Let him come to our aid, anyone, and quickly!’, Henderson translates, allowing the indefinite ‘anyone’ to modify the toospecific ‘he’ on rendering the second τις.210 τις can never—as it invariably has been in Supp. 60—be translated simply by ‘he’, which, being a determinate pronoun, has almost the opposite meaning. ‘One’ sometimes suffices for the substantival, ‘a’ for the adjectival τις. In our passage, the subject has been defined as being a native seer (τις οἰωνοπόλων ἐγγάϊος), who is nearby (πέλας) and listening (ἀΐων). If τις in 60 is to stand, the chorus forgets about this whole description, saying that if a seer is listening, ‘anyone can seem to hear lament’. A retained τις must be translated ‘someone’, not—as in Tucker 51n., p. 202, Headlam, Wilamowitz (1914, 28), Smyth, Mazon, Rose—‘he’. However, Auratus’ τιν’ (adopted by, e.g., Murray, Page) is precisely what is needed. Quite contrary to the opinion of FJ–W, that ‘it is the voice of Tereus’ wife and no other voice, … that the augur will think he hears’, the vagueness and uncertainty that τιν’ confers fit the context perfectly (so Rose). Cf. Ag. 55 and 1142, the parallels adduced by Griffith (1986). δοξάζω, ‘think’, ‘believe’, may not be found elsewhere with a participle, but most other ‘cerebral’ verba sentiendi are, for instance µανθάνω, γιγνώσκω, νοµίζω, οἶδα, ἐπίσταµαι.211 Plato and other philosophers treat δοξάζω as an opposite to γιγνώσκω (‘presume’, ‘believe’, ‘take for granted’ vs. ‘know’), and it should be able to assume the same syntactical functions as that verb. 210 An even closer parallel would possibly be τινος in Pl. Phd. 87c, where however the text is difficult and disputed: ἐµοὶ γὰρ δοκεῖ ὁµοίως λέγεσθαι ταῦτα ὥσπερ ἄν τις περὶ ἀνθρώπου […] ἀποθανόντος λέγοι […] ὅτι οὐκ ἀπόλωλεν ὁ ἄνθρωπος ἀλλ’ ἔστι που σῶς, τεκµήριον δὲ παρέχοιτο θοιµάτιον ὃ ἠµπείχετο …, καὶ εἴ τις ἀπιστοίη [*Heindorf: ἀπιστῶν codd.] αὐτῷ, ἀνερωτῴη πότερον πολυχρονιώτερόν ἐστι τὸ γένος ἀνθρώπου ἢ ἱµατίου […], ἀποκριναµένου δή [v.l. δέ] τινος [secl. Burnet ed. 1900] ὅτι πολὺ τὸ τοῦ ἀνθρώπου, οἴοιτο ἀποδεδεῖχθαι ὅτι παντὸς ἄρα µᾶλλον ὅ γε ἄνθρωπος σῶς ἐστιν, ἐπειδὴ τό γε ὀλιγοχρονιώτερον οὐκ ἀπόλωλεν. Burnet (ed. 1911 ad loc.) is probably right to take τινος as intrusive and due to the corruption of ἀπιστοίη: he is certainly correct in assuming that the indefinite pronoun would mean that someone other than the original antagonist is answering. Strachan in the latest OCT adopts both Heindorf’s correction and Burnet’s seclusion of τινος. 211 LSJ s.vv., S.GG ii. 395–97, K–G ii. 48–52, 68–70. 79 The testimony of the ms. should in any case be trusted here: the participle conveys too good a sense to emend on such scanty lexicographical evidence as we have (δοξάζω is very rare with a simple infinitive). Rather than just ‘thinking that he hears’ something, the participle ἀκούων describes the state of listening in which the passer-by is placed by the hypnotic lament of Procne /the chorus. τιν’ removes any awkwardness. Cf. Young (1974), and Ag. 680 ἴσθι τἀληθῆ κλύων, 830 τὰ δ’ … µέµνηµαι κλύων,212 Pr. 824, S. OT 105. 60–62. ὄπα … ἀηδόνος: for the ‘maritonymic’ adjective Τηρεΐας see K–B ii. 294, Anm. 4. The cluster of genitives is confusing, however, and µήτιδος is obelised by Murray and Page. A large variety of solutions have been proposed (see n. 215 below). Personally, I think that it would be possible to take ὄπα τᾶς Τηρεΐας µήτιδος … ἀλόχου as ‘voice of the µῆτις of Tereus’ wife’, i.e., the voice of Procne’s thought, mind or skill213—the musical-poetical art is partly intended. For the word-order, cf. E. Supp. 628–29 τᾶς παλαιοµάτορος παιδογόνε πόριος Ἰνάχου. For µῆτις in this context, cf. Pi. N. 3.9: τᾶς ἀφθονίαν (sc. ἀοιδὰν) ὄπαζε µήτιος ἁµᾶς ἄπο, O. 1.8–9 ὁ πολύφατος ὕµνος ἀµφιβάλλεται σοφῶν µητίεσσι. The latter passage was adduced by Jurenka (1900, 184), of whose suggested solution the one presented here is a slightly modified version. Jurenka proposed that the grammatical function of µήτιδος is that of a periphrasis for Procne, such as he imagines Bacchylides’ (19.11) Κηΐα µέριµνα to be.214 Jebb’s note on that passage, however, would serve as well for Procne’s µῆτις: ‘µέριµνα is the musing, the fantasy, of the poet, — here half-personified’. Procne’s µῆτις could be said to be ‘half-personified’.215 212 κλυών Casaubon (680) and Wilamowitz (830). Cf. Th. 917–20 γόος … ἐκ φρενός, Ag. 546 πόλλ’ … ἐκ φρενός … ἀναστένειν. 214 Jurenka is oddly misunderstood by Dawe (1965), who claims that he ‘µητίδος [sic Dawe] “carmen” esse docet’. 215 The interpretations have otherwise varied greatly among editors and critics. Apart from conjectural solutions, there are five main groups: (1) Bücheler (1886) and Verdenius (1985) both suggested a solution related to the ones proposed by Jurenka and here, namely that µήτιδος = µητιοέσσης. Verdenius adduces Hes. Op. 191–92 κακῶν ῥεκτῆρα καὶ ὕβριν ἀνέρα τιµήσουσι, with the note from his own commentary on that passage, where he suggests that ὕβριν ἀνέρα is a contamination of the ‘parathesis’ of nouns found in expressions like ἄνδρας φύλακας, and the use of abstracts for concretes, e.g. in Od. 3.49 νεώτερός ἐστιν, ὁµηλικίη δ’ ἐµοί. This explanation is less than convincing, however, and without parallel. The text of Hesiod has been doubted: an easy solution might be ἄνδρες τιµήσουσι (ἄνερες αἰνήσουσι Evelyn-White 1915). (2) µήτιδος dependent on οἰκτρᾶς; ‘miserandae propter con213 80 οἰκτρᾶς may well go with µήτιδος: cf. Headlam, and the passage adduced by him, Od. 19.522–23 παῖδ’ ὀλοφυροµένη Ἴτυλον …, ὅν … κτεῖνε δι’ ἀφραδίας.216 The τε connecting the two parts of Procne’s designation is perfectly natural (see D.GP 502).217 The dual expression Τηρεΐας … ἀλόχου κιρκηλάτου τ’ ἀηδόνος is equivalent to 41–43 ∆ῖον πόρτιν … ἶνίν τ’ ἀνθονόµον. How- silium’ (Bothe ed. 1805 nolens, Mazon, Friis Johansen, Griffith 1986). It has been claimed that such a ‘causal’ genitive appears in exclamations only, which is not entirely true: cf. the examples of Smyth (1956, 335, § 1435), Tucker, and FJ–W. Tucker suggests that the definite article is a requisite in the case of a non-exclamation (e.g., Pl. Phd. 58e εὐδαίµων … ἐφαίνετο … τοῦ τρόπου); it is better to state that the gen.caus. is impermissible with attributive adjectives, but has to go with a predicative or an apposition: i.e., τλήµων σὺ τόλµης or Κρέουσα, τλήµων τόλµης, but not ἡ τόλµης τλήµων Κρέουσα. In our case the grammar might be more acceptable with de Pauw’s µήτιδας (defended by Sommerstein 1977). (3) Headlam takes µήτιδος οἰκτρᾶς as a gen.qual., dependent on ἀλόχου: ‘Tereus’ wife of lamentable counsel’. This is rather more Aeschylean, but also only possible as a predicative, except in the case of the genitive of measure (Smyth 1956, 317, § 1321). (4) Τηρεΐας µήτιδος periphrastic for Τηρέως, ‘cunning Tereus’ plaintive wife’ (FJ–W), in analogy with expressions like βίη Ἡρακληείη (K–G i. 280–81): so the scholiast and, e.g., Hermann, Paley, Whittle (1963, with argument); but Tereus’ µῆτις is irrelevant in the context, pace Whittle, and is also hardly general enough a quality to define his person as a whole (‘periphrasis subabsurda’ Weil). (5) Μῆτις (or Μητίς) as a proper name of Procne (first Welcker 1824, 503, then, e.g., Wilamowitz 1914, 28, n. 3, Fraenkel on Ag. 1526, FJ–W), but the name is not found elsewhere, and Griffith (1986) is right to call it ‘over-explicit and redundant in the context’. Of the various emendations which have been proposed (see above under (2) on de Pauw’s µήτιδας), Tucker’s ∆αυλίδος (‘woman from Daulis’) has been unjustly ignored. It is not so very difficult palaeographically as it may seem at first (δα > µ), and in the light of the passages he adduces (Th. 2.29, Plu. Conv. 727d, Catull. 65.14, Ov. Her. 15. 154), to which may be added Ar. fr. 936 PCG (ap. EM 250.8, Suda s.v., etc.), it deserves a mention in a critical apparatus. Burges’ (1810, 802) Ἀτθίδος is also attractive and has recently gained support from Liberman (1998), with a parallel from [Sen.] Herc.Oet. 199 fugit vultus Philomela suos | natumque sonat flebilis Atthis. 216 Not by mistake, pace the scholium on the passage, but by ‘senseless folly’. See Russo ad loc. who also comments on the variation with regard to the antagonists’ names in the Homeric and the Attic versions of the myth. 217 γ’ is suggested in Anon.Par. It is unnecessary, and with the reading κιρκηλάτου ἀηδόνος it results in awkward and unwelcome stress (pace Page, West). It is not comparable to any of the examples of epexegetic γε in D.GP 138–39. See, however, below on the particle in combination with a personal name. 81 ever, the text is not entirely certain. ἀηδόνος, being Turnebus’ emendation (ἀηδονῆς M), is the most attractive reading in my opinion, but the corruption is somewhat hard to account for. Possibly some scribe thought that a personal name was required; but it is still strange that he should then alter ἀηδόνος, as Ἀηδών is better attested as a personal name than Ἀηδονῆ.218 Perhaps the corruption involved some kind of confusion with the noun ἀηδονίς, or, as suggested by Whittle (1963, 250, n. 34), arose out of a faulty word-division (the Aldina has τᾶ ἠδονῆς, owing to a misunderstanding of the reading in Mc: see McCall 1985, 19). We cannot on the other hand entirely rule out the possibility that a personal name is what Aeschylus wrote. Palaeographically, this solution may seem more tenable than Turnebus’ conjecture in some respects: the paradosis ἀηδονῆς could be a (hyper-)Attification of Ἀηδονᾶς, which in turn is a contracted form of Ἀηδοναίας, a name which is actually attested: FJ–W and others note that it is written on a fifth-century kylix portraying the murder of Itys, where the murderess, i.e. Itys’ mother, is designated as ἀηδοναί<α>.219 Wilamowitz’ Ἀηδόνας may also be possible, although farther removed from the paradosis.220 However, either reading would require further emendation: a personal name in this position will make τ’ impossible. τε may connect non-equivalent designations or attributes (‘Zeus’ calf and Io’s son’, ‘wife of Tereus and hawk-chased nightingale’), but not personal names or pronouns with appositions or attributes (‘wife and hawk-chased Aedone’, ‘offspring and much-lamentable Iphigeneia’). The discussions of ‘epexegetic’ τε in D.GP 502 and by Fraenkel on Ag. 1526 are unsatisfying in this respect, the former accepting too much, the latter too little.221 Thus, with Ἀηδονᾶς or 218 For instance [Boeo] fr. 7, Σ Od. 19.518: both probably based on a misunderstanding of Od. 19.518, where ἀηδών should not be taken as a personal name. 219 Or ἀ[ ]ηδοναί<α> (ι suppl. Harrison 1887, 442; ϝ alii). There is actually no trace of the last letter of the name: see Harrison l.c. On the kylix see also, e.g., Mihailov (1955) 154–55 and Beazley (1963) i. 456 with refs. 220 In the critical apparatus: cf. Wilamowitz (1914) 28, n. 3 (Wilamowitz’ conjecture is not to be found, pace West, in his note on of E. HF 1022). Similarly, the spelling Ἀθήνη or -α alternates with -αία / -ᾶ, and, as noted by Wilamowitz l.c., the name Ἀλκυόνη is a variant of Ἀλκυών. 221 πολύκλαυτόν τ’ is impossible in Ag. 1526 and perhaps also ἥτε in Th. 501 (pace D.GP 501, 523): in both cases γε will confer the right nuance (suggested by Casaubon 82 Ἀηδόνας, the conjecture γ’ (Anon. Par.: see above, n. 217) would be necessary. On the whole, I think Turnebus’ emendation is the most satisfactory solution. 63–64. ἅ τ’ ἀπὸ χώρων ποταµῶν τ’ κτλ: The apparent irrelevance of these verses (see, e.g., W.SA) is due to the fact that the girls are not really talking about Procne, but about themselves. ‘Debarred from the lands and rivers she cries a strange lament over her old haunts’—the Danaids imagine Procne as an exile, like themselves. This is not a vital part of the myth; but it is, besides the fact that they are both being chased by fiancés/a husband, the only way in which Procne can offer a relevant parallel to the Danaids. As for the text, Hermann’s χλωρῶν … {τ’} is somewhat attractive, in the light of a number of parallels where this adjective is found in connection with the nightingale or with water.222 None of them is conspicuously close, however, and there is nothing intrinsically awkward about the combination ‘lands and rivers’, pace FJ–W:223 cf. 23 γῆ καὶ λευκὸν ὕδωρ, 1026–27 ποταµοὺς δ’ οἳ διὰ χώρας … πῶµα χέουσιν, Scylax in the sub-title of the Periplous: χῶραι καὶ λιµένες καὶ ποταµοί. The juxtaposition of the words is also common in Herodotus’ Egyptian geography (2.10, 2.13 bis, 2.14, 2.177); and if, as is likely, Aeschylus read about Egypt in, for example, Hecataeus (cf. 220–21n. with n. 381), he would certainly have encountered similar passages. Wecklein’s (ed. 1902) ἐπὶ (ἅ τ’ ἀπὸ Victorius, vulg.: ἃταπο Mac: -οπο Mpc) is detrimental, pace West (W.SA). We do not want a lyrical picture of Procne singing by her green rivers in the spring, notwithstanding Od. 19.518–20. The chase and especially the exile of Procne are the matters that produce significance and dramatic effect here. ‘The nightingale is not kept away from rivers’, in the former case, by Anon.Ald. and Anon.Barth. in the latter). γε ‘connects’ a personal name with another designation for the same person, and vice versa: see D.GP 139 for exact equivalents. On the other hand, τε is perfectly natural, pace Fraenkel on Ag. 1526, in passages like Supp. 42–43 and the present one (with Turnebus’ ἀηδόνος). 222 Od. 19.518 χλωρηῒς ἀηδών, S. OC 672–73 ἀηδὼν χλωραῖς ὑπὸ βάσσαις, E. Ph. 659–60 ῥέεθρα χλοερά, Hel. 349–50 χλωρόν Εὐρώταν (cf. ?Alcm. 10 ἄκουσα ταν ἀηδ[ον- …]| παρ’ Εὐρώτα): see FJ–W, W.SA, and Scheer (1914) 44, who first combined Hermann’s χλωρῶν (πετάλων {τ’}) with the paradosis ποταµῶν. 223 Nor does χώρων mean anything as specific as ‘fields’, as Wilamowitz argues in the apparatus: it simply refers to the native land of Procne. 83 asserts West, but the Danaids are kept away from the Nile, the River of rivers, and they project this predicament on Procne, who, they imagine, was chased away from her native land (as, incidentally, was Io). The passage is perfectly understood by Page: ‘cum non possit Procne ἀπὸ χώρων ποταµῶν τε generatim excludi, necessariam definitionem addit ἠθέων: a rure fluminibusque exclusa, de locis (illis) familiaribus lamentationem edit’. ἐργοµένα: Victorius’ εἰργοµένα is worth considering. The ms.’ ἐργ- is an epic/Ionic form, which does appear a few times in the mss. of Attic tragedians, but in Aeschylus elsewhere only at Ch. 446 ἄφερκτος. The high frequency of epicisms and Homeric reminiscences in the first half of the ode (see 40–111n.) might favour the paradosis, though. See also FJ–W. οἶκτον echoes the same word in the same place in the strophe (59), a common device in Aeschylean choral lyric (see FJ–W for other examples). Haecker’s (1861, 222–23) µέν (νέον mss.) has won much support among twentieth-century editors and critics.224 It is not even mentioned by West; and even if it may seem attractive, it is far from necessary. νέον is ambiguous, but arguably appropriate on several levels. The significance may approximate ‘unexpected, strange’ (LSJ II.2) and refer to the transformation of Procne (so Wecklein, ed. 1902), and perhaps at the same time, on a metaphorical level, either to the foreign tongue of the Danaids,225 or to a transformation on their part from Egyptian into Greek (see below). The novelty of their situation, the new and foreign land at which they have arrived, is an important aspect of the Procne-excursus. νέον could also refer to Procne’s exile as a new sorrow added to her former, and more well-known grief, the murder of her child. From the Danaids’ point of view, the slaying of her son is secondary, even unimportant (FJ–W); it cannot be left out completely, 224 For instance Wilamowitz, Murray (ed. 1955), Whittle (1963), Page, Diggle (1982), Griffith (1986). Incidentally, the adjective νέος seems to have been introduced by strange corruption in at least two other places in M: perhaps in 355 of this drama (q.v.) and in Eu. 490 καταστροαὶ †νέων, where I have elsewhere suggested δ’ ἐµῶν (Sandin 2002). Cf. also Ag. 1625, where *Wieseler’s µένων (νέον mss.) is adopted by, e.g., Page, Thomson, Denniston–Page with argument. 225 Foreigners in ancient Greek (and Latin) literature may conventionally refer to their own foreignness as if they were Greeks, even when at home: cf. 118 = 129 and several passages in Pers. where the Persians refer to themselves as βάρβαροι: cf. 187, 255, etc. 84 however, as it forms the central, best-known part of the myth. It probably also (so FJ–W 68–72n.) foreshadows the later deeds of the Danaids, the slaying of their husbands (cf. 6–10n.). Furthermore, an interesting possibility is that νέον οἶκτον may contain part of the same notions as Ἰαονίοισι νόµοισι in 69 (cf. Wecklein l.c., and see below), and that it could allude to a change in the music as well as to the change of country in general. Procne’s lament, like that of the Danaids, is conducted in a new style, adapted to the new environment, and perhaps also rendered in a new language. The two most compelling reasons for Haecker’s conjecture are that the metre becomes easier (iambic) and that we avoid the unparalleled ἐγγάϊος (ἔγγαιος M) in 59 (q.v.). The metric colon (on which see 59n.) is given adequate parallels by FJ–W 64n., however, and ἐγγάϊος is supported by Aeschylus’ use of γάϊος elsewhere.226 65. ξυντίθησι δὲ παιδὸς µόρον: ξυντίθησι may include the notion that Procne interweaves the lament over her child and that over her exile. 67. δυσµάτορος: apparently formed from the striking hapax δυσµήτηρ in Od. 23.97. 68. τὼς: this adverb, together with the dactylic rhythm, makes a virtual Homeric simile out of the Procne-excursus, although the introduction of Procne as the vision of a seer complicates the comparison in an un-Homeric way (cf. FJ–W). Cf. Sideras (1971) 96–97 and also 40–111n. above. 69. Ἰαονίοισι νόµοισι: West (W.SA), preceded by Cannatà Fera (1980), argues that this refers to an actual ‘Ionian scale’ in the song of the chorus here. This may well be the case (see above, 63–64n.); but the words may also refer to the new, Greek environment, contrasted to the Egyptian homeland of the Danaids. This does not mean that we should accentuate νοµοῖσι:227 just as in the previous passage, the song and music are in the focus of the lyrics. The idea may be that just as Procne cries a new sort of lament in her avianshaped exile, so the girls sing a new, Greek kind of song as they have reached Argos. So the scholium: <Ἰαονίοισι νόµοισι>] ἀντὶ τοῦ φωνῇ Ἑλληνικῇ.228 226 Reasonably certain examples are Th. 735 (see above, n. 136), Supp. 835; Supp. 155 (q.v.) and 826 are probable. 227 Whittle (1964a), FJ–W, supported by Diggle (1982) 134, n. 3. 228 Cf. Σ Ar. Ach. 106 πάντας τοὺς Ἕλληνας Ἰάονας ἐκάλουν οἱ βάρβαροι: so in the Persians, Ar. Ach. 104, 106 etc., and Hsch. s.v. Ἰαίνα. 85 71. ἁπαλὰν εἱλοθερῆ παρειάν: ‘soft’ and ‘sunburnt’ hint at foreign luxuriousness (cf. 235–36 and Hall 1989, 128). Musgrave’s εἱλοθερῆ is all but certain. FJ–W argue passionately for the paradosis Νειλοθερῆ, although omitting the strongest argument, stated by Paley and, among others, Whittle (1964a): Νειλοθερῆ could be interpreted as a contrast to the Ἰαονίοισι νόµοισι in 69: ‘I sing in a Greek manner, tearing my Egyptian cheeks’. However, the compound is hardly tolerable. FJ–W claim that it refers to the dark complexion of the Danaids, which is impossible: ‘coloured by the Nile-summer’ (Friis Johansen ed. 1970) or ‘summered by Nile’s sun’ (Smyth; cf. Mazon, Untersteiner ed. 1946) are not senses that can be read into the word, which does not by itself imply ‘sun’ or ‘colour’; and ‘summered by the Nile’ (FJ–W) is frankly nonsense. The sense ‘spending the summer by the Nile’, if at all possible (it cannot be inferred from S. Tr. 188 βουθερεῖ λειµῶνι, where it is the meadow that has cows in the summer, and not, obviously, the cows that are *λειµωνοθερεῖς), is not appropriate as an epithet pertaining to cheeks. -θερής does not by itself imply sunburn, and in combination with Νεῖλο- it should give either the sense ‘Nile-warm’ (i.e. ‘luke-warm’) or ‘Nile-harvested’ (so Σ), neither of which is appropriate here. Skin-colour is certainly the issue (cf. 154–55), and thus εἱλοθερῆ, ‘sun-heated’, is the obvious emendation. A metri gratia variant with -ο- (εἱληθερής is otherwise found in medical literature) is certainly admissible, as we may see from Aeschylean use of similar compounds elsewhere (see above, 52–55n., n. 199). 72. ἀπειρόδακρύν τε καρδίαν: because of their youth (cf. on 79–81) and perhaps also their nobility and ‘foreign luxuriousness’ (see above, 71n.). καρδίαν is disyllabic (‘kardyan’), as at 799. What happens to the accent is of little consequence, as the contraction only appears in lyrical passages where the pitch accent must give way to the music. 72(~81). West calls this colon ‘Hagesichorean’ after Alcm. fr. 1.57 (cf. West 1982, 30). The ×ds start is otherwise found in cola usually known as ‘enhoplians’ or ‘prosodiacs’ (so Dale 1971 ad loc. and FJ–W iii. 349): see 59(~64)n. 73. γοεδνὰ δ’ ἀνθεµίζοµαι: The simple verb is a hapax, but ἐπανθεµίζω in the active tense appears in S. Ichn. (fr. 314) 331.229 There the verb takes an internal acc. and obviously means, pace LSJ, ‘adorn as with flowers’, being 229 The pf. pass. διηνθεµισµένων is also read by conjecture (Kock i. 199) in Eub. fr. 104 PCG, meaning ‘with flowers throughout’. 86 thus identical in sense to ἐπανθίζω.230 An easy inference, then, would be that the simple verb ἀνθεµίζω is likewise a variant form of ἀνθίζω, which also, in classical Greek, means much the same thing as ἐπανθίζω: ‘deck with flowers’ (LSJ) or, in a metaphorical sense, ‘apply florid quality’ (colour, scent etc.). The passive voice is most common, meaning ‘be adorned’, ‘salved’, and the like. ἐπανθίζω appears several times in Aeschylus, and the aorist middle is found in Ag. 1459, meaning ‘adorn oneself’, apparently taking τελέαν as an internal acc.231 The same construction is possible for the present middle or passive ἀνθεµίζοµαι here, and is preferable to the vulgate (and scholiast) interpretation ‘cull’, ‘pluck’: the girls ‘adorn’ or ‘crown’ themselves with moaning.232 This sense is further supported by 115 ἰηλέµοισιν ἐµπρεπῆ (‘conspicuous by moaning’).233 74–76. δειµαίνουσ(α) … εἰ: ‘Fearing, lest not’ (K–G ii. 396–97). The words inbetween are a little more difficult. As for the reading of M, δειµαίνουσαφίλους (pc: φόλους ac) τᾶσ | δε φυγᾶς (or φυγὰς—one accent is written 230 χερ]οψάλακτός τις ὀµφὰ κατοιχνεῖ τόπου, πρεπτὰ … τόνου φάσµατ’ ἔγχωρ’ ἐπανθεµίζει: ‘A solemn voice, made by the hand that plucks the strings, goes forth over the land! Conspicuous … are the fantasies of sound that it scatters like flowers over the place!’ (Supplement and translation by Lloyd-Jones in the Loeb edition, my italics.) 231 The active verb is found in Th. 951–52 πολλοῖς ἐπανθίσαντες πόνοισι γενεάν, Ch. 150 κωκυτοῖς ἐπανθίζειν (κωκυτοὺς Paley ed. 1844, p. 16, on the present passage—but the emendation is not mentioned or printed in the 1845 issue of Choephoroi). ἐπανθίζω with an internal acc. is found in Luc. Hist.Conscr. 13. 232 The middle voice of the uncompounded ἀνθίζοµαι is otherwise unattested before the second century A.D., where it appears in App. BC 4.105, meaning ‘pluck’ (cf. Plu. Conv. 661 f., Clem.Al. Paed. 2.8.70.2). This is the interpretation of the scholium to our passage: γοεδνά· τῶν γόων τὸ ἄνθος ἀποδρέποµαι, and most editors and translators have adopted it. Even so, second-century and scholiast use of ἀνθίζοµαι has a limited value as evidence for the Aeschylean usage (of ἀνθεµίζοµαι). 233 Paley (ed. 1844), Wecklein (explicitly 1876, 334; inadequate notes in the editions), Rose and Sansone (1975, 34) have advocated the interpretation ‘blossom into grief’ (Sansone’s transl.), which is poetically effective and also supported by instances of bad things blossoming in Greek poetry. Sansone adduces the ‘hybris-trunk’ in 105 ff. and Ch. 1009 πάθος ἀνθεῖ, to which we may add Pers. 821 ὕβρις ἐξανθοῦσ’, Ag. 659 ἀνθοῦν … νεκροῖς and 743 δηξίθυµον ἔρωτος ἄνθος. Cf. also, e.g., S. Tr. 999, Ant. 960, fr. 786, E. IT 300, Pi. O. 13.23, N. 9.23, I. 4.18b, B. 15.57–59. In this case, however, an internal acc. is awkward and appears to be unparalleled, as are ἀνθίζω and compounds in the same sense as ἀνθέω. 87 over the other), there is a number of ways to elicit sense from it, two of which involve no more emendation than changes in word-division and accentuation. The first is to take φίλους at face value, meaning either ‘kin’ or ‘our own’ and denoting the Aegyptiads, or to take it as anticipatory of κηδεµών: ‘attractio pro ἀντὶ δειµαίνουσα εἴ τις φίλων ἐστὶ κηδεµών’ Meffert (1861, 8). This is awkward in many ways and unparalleled (see FJ–W), and I prefer a reading which has actually not been adopted by any editor so far: Me’s ἀφίλους τάσδε φυγὰς. This is probably an emendation by the scribe, not a traditional reading, but nevertheless it is a good one.234 The plural is perfectly acceptable; cf. 196, Eu. 424. Actually the reading might have a tiny traditional support in M, if the grave accent on φυγὰς is the original one and not a correction, which may be likely: if it were a correction we would perhaps have expected the accent on τᾶσδε to be corrected likewise.235 τάσδε φυγάς might approximate an accusative of extent (K–G i. 312–15): ‘fearing, on this friendless flight’ (rather than ‘fearing this friendless flight’). The accusative is found with δειµαίνειν in Pers. 600, where it is also uncertain whether it is to be taken as an adverbial or a direct object. 74(~83). On the metrical responsion, see 83n. 75. Ἀερίας: ‘the misty land’, a name for Egypt (as well as for other countries) according to late sources (see FJ–W). 78. εὖ: syntactically ambivalent, as FJ–W contend; it may be taken with κλύετ’ as well as ἰδόντες, although rhythm and sense may favour the former (Diggle 1982). Cf. on 15–18. 79–81. ἥβαν, in the sense of female sexual prime,236 and ὕβριν (cf. 31, 103, 234 The ‘scribe’ of this 16th-century ms. is the bishop Arsenius of Monembasia, also known as Aristobulos Apostolides (see FJ–W i. 68–69). The variants in Me are presented as conjectures in Arsenius’ name in West’s apparatus (see West p. xvii, W.SA 356–57). The emendation is repeated by Rogers (1894) and Rose. 235 FJ–W and West prefer Musgrave’s ἀφίλου. It is a possible alternative, but less economical. I do not think that the force of τᾶσδε φυγᾶς would be restricted to the εἰ-clause, however, ‘a helper on this flight’—rather, it might be taken as a genitive of place (K–G i. 384–85) with the participle: ‘fearing, on this friendless flight’. Another alternative is Enger’s (1854a, 392) δειµαίνουσα φίλος, which is preferred by, e.g., Wilamowitz, Page, Dawe (1972), Verdenius (1985). φίλους is retained by Wecklein in all eds and by Murray, and it is supported by Conacher (1996, 83, n. 19). 236 Cf. 663 ἥβας δ’ ἄνθος ἄδρεπτος, 997–98, Pers. 544. 88 etc.), masculine aggression, are on the level surface the two antagonistic motors of the drama.237 ἥβη is an ambiguous word: it means ‘youth’ but also ‘maturity’, in the sense of ‘manhood’ or ‘womanhood’. ἥβαν τέλεον … ἔχειν seems to mean ‘have full possession of (our) womanhood’,238 or, perhaps more to the point, ‘have possession of our womanhood so as to fulfil it’, as ἥβης τέλος elsewhere means ‘attainment of maturity’, i.e. ‘entrance into adulthood’.239 Through marriage, the Danaids would, unwillingly, enter into adulthood. On τέλεος and its connections with marriage, see FJ–W. West (W.SA), on the other hand, wants to take ἥβαν here as belonging to the Aegyptiads, ‘not allowing (anyone) to have ἥβη fulfilled (by marriage) in transgression of what is due’. This is perhaps possible, if we take ἥβη as meaning something like ‘youthful virility’, which would be ‘fulfilled’ through sexual intercourse (cf. the hints in 106–10). But the ἥβη of the girls has been stressed just before (70–72). Also, for a woman, marriage was the adequate rite of passage into adulthood, but hardly for a man (cf. Zeitlin 1990, 105 with n. 6). It is conceivable that the sense is ambiguous: ἥβη τέλεος refers to marriage or sexual intercourse; but the ἥβη may perhaps be either that of the man or that of the woman? ἥβαν appears to have been the reading of the scholium (ηβαι M240), who interprets it as τὴν τῶν Αἰγυπτιαδῶν. See FJ–W on the textual corruption. 81. στυγόντες: epic aorist II; contrast the echo in 528 ὕβριν εὖ στυγήσας. Just as in the case of 63 ἐργοµένα an easy emendation, Turnebus’ στυγοῦντες, would produce normal Attic; but there is even less cause to emend here. 82. πέλοιτ’ ἂν ἔνδικοι γάµοις: FJ–W, adopting Oberdick’s ἔνδικος γάµος, argue that the paradosis cannot mean ‘just towards our (or the institution of our) marriage’, since ἔνδικος etc. + dat. never takes this sense. But the meaning of the dative is presumably ‘in’, ‘by’, ‘through’, or ‘with regard to’ marriage: the sense is local or instrumental rather than proper dative. 237 But the contrast is to be problematised: see on 154–61 below. On the masculinefeminine polarity in the Supplices see also Zeitlin (1990). 238 ἥβη may also mean ‘genital parts’, regardless of sex, as often in medical and physiological literature (e.g. Arist. HA 518a18, GA 718a10), and perhaps some of that sense may be felt here as well as in 663 (cf. Dikt. 830 [fr. 47a col. II 32]). 239 Cf. Hes. fr. 30.31, E. Med. 920, AP 7.300 (Simon.), X. Cyr. 8.7.6.6, Thgn. 2.1326. 240 Several apparently conjectural variant readings are found in the apographs: ἡβαι Mb, ἧβαι Me, ἢ καὶ Ma, ἣ καὶ Mc, καὶ Md. 89 The scholium, ἐπὶ τοῖς νενοµισµένοις καὶ δόξασιν ἡµῖν, is mysterious. It would make sense if the person who wrote it read ἐνδίκοις γάµος: ‘marriage should come on lawful terms’ (cf. on 9–10).241 This passage may serve as a paedagogical example for those who wish to dispute that textual criticism is a legitimate area of scholarship (cf. the defence by W.SA 370–72). The exact reading of this particular verse is of crucial importance for what is one of the most controversial issues of the entire drama: are the Danaids adverse to marriage as such, or is it only this particular marriage they wish to escape? The former stance is taken by, for instance, Garvie (G.AS 221–22), who lists ten passages that allegedly support the idea of man-hating Danaids.242 FJ–W on the other hand take the latter position, discussing the problem at i. 29–33. Their conclusion (p. 32) is that ‘there is, in fact, not one passage in Supp. where the Danaids clearly express an attitude of general aversion to the institution of marriage, or to sexuality, or to the male sex as such’. Incidentally, as we saw, FJ–W adopt Oberdick’s conjecture, πέλοιτ’ ἂν ἔνδικος γάµος, in the text here: ‘then there might be a righteous marriage’. This also happens to be the only passage that carries any conviction among those they adduce as ‘unmistakable signs that [the Danaids’] attitude [towards marriage] is … positive’.243 I am not certain that the text supplies conclusive evidence either way, nor that Aeschylus necessarily thought that the Danaids’ agenda was as well-defined as critics have 241 Cf. Burges (1810, 803) who suggests that the scholiast’s reading was ἐνδίκοις γάµοις. Weil makes the same observation: ‘[scholiasten] vel vulgatam, vel ἐνδίκοις γάµοις habuisse puto’. He goes on to suggest <οὐ> πέλοιτ’ ἂν ἐκδίκοις γάµος, inaccurately reported by Wecklein (ed. 1885). ἐνδίκοις γάµος would, like Oberdick’s conjecture, result in an anacoluthon of the same type as the ones in 40–46, 50–55. On the other hand, ἡµῖν in the scholium might perhaps originate in a ms. reading ἐµοί (cj. by Rogers 1894: πέλοιτε σύνδικοι γ’ ἐµοῖς or ἐµοί). πέλοιτ’ ἂν ἔνδικοι γ’ ἐµοί (Griffith 1986) would be an easy enough correction. This would solve the alleged difficulty of the dative: the construction would be similar to that in S. Aj. 1363: ἄνδρας … Ἕλλησι πᾶσιν ἐνδίκους (where, however, δοκοῦντας has to be inferred from the context). There is nothing demonstrably wrong with the paradosis, however. 242 144–50, 392–93, 426, 528 ff., 643–45, 790, 798–99, 804–7, 818, 1017. See G.AS 215–23, and also Sommerstein (1995) 114–15 and n. 18, who follows Garvie on this particular issue (cf. also my Introduction, II 4). 243 Allegedly 79–82, 337, 978–79, 991–1009, 1031–33, 1052–53, 1062–67. 90 claimed. The immediate action and passion of the girls are what constitutes the drama, not any abstract mind-set, ideology, or opinion on their part. They do not want this marriage, that much is clear, and more than that we do not actually need to know about their state of mind (cf. 176–78n., n. 331, and the Introduction, II 4).244 83–85. The sense is equivalent to ἔστι δὲ καὶ φυγάσιν ἐκ πολέµου τειροµένοις ῥῦµα ἀρῆς· ὁ βωµός, δαιµόνων σέβας. Because of the Danaids’ concentration on the actual, concrete situation rather than on the abstract case, βωµός—the altar which is probably present on the stage (see 222–23n. and the Introduction, III 5)—intrudes prematurely in the gnome, and ῥῦµα becomes its apposition rather than the other way around. Thus another explanatory apposition is added, δαιµόνων σέβας. βωµός should not (pace FJ– W) be taken as ‘an altar’, but as an ideal abstract: The Altar. Taking (like the scholiast) ἐστι as a copulative—‘even for fugitives etc., the altar is a ward’—lessens the pathetic tone of the passage and makes the apposition δαιµόνων σέβας awkward. 83(~74). ἔστι δὲ κἀκ πολέµου: the first biceps of the hemiepes is contracted in the strophe (74 δειµαίνουσ’ ἀφίλους), something which Diggle (1982, 129) argues lacks sufficient parallel, thus supporting Enger’s (1854, 392) ἔστιν κἀκ.245 Diggle claims that 543~552 (ddd), where the same kind of responsion occurs (cf. also Eu. 1042~1046), is a dactylic tetrameter and not comparable to the hemiepes (dd), which is a questionable assertion. The hemiepes is often found in a dactylic context, and as it allows contraction of the double-short, it should be able to respond with contracted against uncontracted biceps. An exact parallel for this responsion appears in any case 244 However, I think that something may be said for Murray’s statement, although oldfashioned in tone (transl. 1952, p. 17; cited by G.AS 222): ‘a girl pressed to marry an unwelcome suitor usually says that she does not wish to marry at all’. Cf. also Gantz’s (1993, 204) reasonable opinion: ‘what the Danaids seem to fear most is a usurpation over which they have no control, and if this leads them to reject marriage altogether, we must remember that the impetuous … approach of their cousins constitutes their only experience in such matters’. 245 Enger’s conjecture seems to have been adopted only by Wecklein–Zomaridis, Wecklein (ed. 1902), and Page. We may note that Enger is partly anticipated by *Burney, who read ἔστιν δ’ ἐκ according to Wecklein (ed. 1885, ii. 98): the same conjecture also appears in Paley (ed. 1844). 91 to be found in 844~855 of this drama, and also in E. Ph. 797~815, neither of which appears to be corrupt.246 M’s πτολέµου is generally emended into the common πολ- from Ma and the scholium,247 the epic form being attested in tragedy only once (apart from in Νεοπτόλεµος and other personal names), metri gratia in E. Med. 643 ἀπτολέµους. It is worth remembering, however, that epic word-forms are fairly frequent in the first half of this choral ode (see 40–111n.). The hemiepes here and the dactylic rhythm of the strophe in general also lend some epic flavour (the strophe begins with a dactylic hexameter), so the epic form of the noun is not altogether impossible. 84. ἀρῆς: another epic word (see above): ‘bane’, ‘ruin’. ‘ionica flexio tenenda erat, ne ἀρά esse videretur’, Wilamowitz. 86–103. As noted by FJ–W 1n., Zeus is mentioned more times in the Supplices than in any other Aeschylean tragedy, and thus in any tragedy, apart from the Prometheus. He is doubly important here, not only being the supreme deity and the god of supplication, but also the γενέτωρ of the race of the Danaids. These verses, together with the ‘Hymn to Zeus’ in Ag. 160 ff., are perhaps the most intense expressions we have of Aeschylean piety. The focus is, even more explicitly than in the verses from Agamemnon, on the might of Zeus, in particular on his unlimited power to change human fortune, for better or for worse. Several scholars have noted that this passage moves Aeschylean religion away from pagan anthropomorphism towards monotheism and an abstract conception of the divine. An influence from Xenophanes has been suggested, 246 The former (Supp. 844) was obelised by Page, the latter (Ph. 815) by Diggle in their respective editions. Diggle (1982) also noted E. Med. 829~840, which, however, he claims is ‘possibly corrupt’. In any case, as a matter of principle, the hemiepes as such hardly occurs often enough to provide a statistical material that is sufficient to rule out ms. readings such as this one. We do find that responsion of two short against one long occur in all other rhythms involving double-shorts that will submit to contraction. Apart from dactylic, anapaestic, and dochmiac, it is also found in iambic, trochaic, ionic etc., where unresolved longa may answer to resolved. For this type of responsion in less regular cola, we may also compare E. Hipp. 738~ 748 and IT 1243~1268 (in both of which cases, however, Diggle accepts conjectures that produce exact responsion). 247 In the 19th century and later, πτολέµου is retained only in Porson, Paley (ed. 1844), the ultra-conservative Wellauer, and unreflectingly in some minor editions (Boissonade, Weise). 92 three major motifs of whose recur: monotheism, anti-anthropomorphism, and a stress on the ease with which God fulfils his desire.248 Other comparative material which has been stressed lately are the Near-Eastern sacred texts, in which the concept of the divine finds several expressions that are almost rendered verbatim in some Aeschylean passages. The parallels in the Supplices are discussed by West (1997) 557–66. See further the notes on 98–103, 100– 101, 169. 86–87, 93–95, 91–92, 88–90. There is little choice but to accept this disposition of the verses 86–95, first suggested by Westphal (1869, 158). Dawe (1964, 163) sums up the argument: ‘if there ever was a non sequitur it is ll. 93– 95, and that they should be introduced by γάρ only heightens the absurdity’.249 The case has been much discussed and questioned, and it goes against the principle for multiple transpositions set up in the Excursus (q.v.). It may be significant that the displacement concerns two symmetrically corresponding blocks of text in a strophe-antistrophe complex. One explanation may be that the verses were at some point colometrically arranged side by side instead of sequentially, the lower halves of the strophes perhaps being separated from the upper halves on account of the metrical and, possibly, the musical change of direction that occurs after the catalexis in 87~92.250 There is also the possibility that the lyrical passages may at some point have undergone a different type of transmission from the dialogue, perhaps being written down from memory with the aid of a musical score and/or a metrical chart.251 86–87. εὖ θείη ∆ιός, εἰ παναληθῶς ∆ιὸς ἵµερος· οὐκ εὐθήρατος ἐτύχθη: Wilamowitz’ (1914, 30, n. 1) emendation252 (εἰ θείη ∆ιὸς εὖ παν- M) is 248 On Xenophanes’ influence on Aeschylus, in particular regarding the concept of the divine in this and other passages, see Guthrie (1962) 374–75; Rösler (1970) 4–15 with refs at 10 (n. 25), 19–24; FJ–W 100–103n. with refs; Corbato (1995); Magini (1996); Sommerstein (1996) 378–79. 249 See FJ–W 88–90n., Friis Johansen (1966) for a detailed argument. Rash (1981, 214–16), following Booth (1974), argues, unconvincingly, for the traditional disposition. 250 An equivalent symmetrical displacement may have occurred in 905–10, where the lower halves of strophe and antistrophe have been exchanged. 251 See Dawe (1964) 161–64 on the hypothesis of an early oral transmission of the dramas of Aeschylus, and Dawe (1966, 1999) on possible strophic displacements of Pers. 93–100, Ag. 160–91, Ch. 434–38, 623–30. See also the Excursus. 252 After Hartung’s εὖ θείη θεός. εἰ δ’ ἄρ’ ἀληθῶς κτἑ. 93 ‘schöner, einfacher’ (Vürtheim), and rightly adopted by West.253 For a similar expression cf. 585 Ζηνός ἐστιν ἀληθῶς. οὐκ εὐθήρατος echoes, and contrasts with, εὖ θείη. On the quasi-adversative asyndeton, eased by the phonetic anaphora (which makes the reading virtually certain), cf. 181n. 87. ἵµερος: a rather surprising choice of word (see FJ–W) which emphasises the similarity between human and divine will, both being essentially a matter of emotion. Contrary to human desires, however, the divine wishes are not thwarted (see on 92, 98–103). 87(~92). The rising starts here and in 95~90 do not (necessarily) imply anapaests, but rather dactylic ἐπιπλοκή.254 93–95. δαῦλοι … δάσκιοί τε … ἄφραστοι: Fraenkel on Ag. 182 f. (i. 112, n. 1) compares, besides Supp. 1057–58 and Eu. 530, Hes. Op. 483–84: ἄλλοτε δ’ ἀλλοῖος Ζηνὸς νόος αἰγιόχοιο, ἀργαλέος δ’ ἄνδρεσσι καταθνητοῖσι νοῆσαι. Rather than the direct influence Fraenkel suspects, a common notion or proverb seems to account for the similarity: ‘the Lord works in mysterious ways’.255 93. δαῦλοι: on the accent, see Radt on fr. 27 (p. 146), Radt (1982). 92. κορυφᾷ is emphatically placed at the beginning of the verse.256 Standing in a polar relation to 87 ἵµερος, it signifies the ease with which the god carries out his purpose. A god may desire; but contrary to human desires, his is fulfilled—by a nod (see further on 98–103). κρανθῇ: ‘to [be] pronounce[d] and establish[ed] in binding and valid form with the guarantee of fulfilment in the future’, Fraenkel on Ag. 369, q.v. for a comprehensive discussion of Aeschylus’ use of κραίνειν. The predicative τέλειον is somewhat redundant (cf. below, 491n.). 253 FJ–W’s statement that ‘omission of the copula [i.e. εἶναι] in a conditional clause, except for εἰ δὲ µή, εἰ δ’ οὖν and the like, is a rarity (cf. KG i.41)’ makes one wonder how far one may trust their common assurances that no parallels are to be found to various phenomena. Cf. Th. 517, Ag. 160, Pr. 765, 978, S. Ph. 886, 1246, OT 896, Ant. 39, etc. 254 Cf. Cole (1988) 171, passim, Dale (1968) 40–41, Danielewicz (1996) 62–70. 255 Cf. also West (1997) 559–60 for similar expressions in Near Eastern literature. 256 Schmidt’s (1863, 228) ∆ιὸς εἰ κορυφᾷ is attractive for two reasons: the position of εἰ becomes normalised, and the position of ∆ιὸς becomes identical with the position of the same word in the strophe (∆ιὸς ἵµερος). Then again, κορυφᾷ benefits from the emphasised position at the beginning of the verse, and the chiastic responsion ∆ιὸς ἵµερος ~ κορυφᾷ ∆ιὸς is poetically effective. 94 88. παντᾷ: how to accentuate this adverb is a highly academic question, as Aeschylus did not write an accent of any kind and the music would not have allowed the pitch accent to be sounded anyway (cf. on 72 καρδίαν). However, convention requires that the regular prose accent be noted in editions. LSJ s.v. πάντῃ contend that the Doric form is accentuated παντᾷ, and the word is printed thus in, e.g., Vürtheim, Murray, Page and a slight majority of the 20th- and late 19th-century editions. M, however, accentuates πάντᾱι and is followed by, for instance, Wecklein (ed. 1902), Wilamowitz, FJ–W, and West. There is no discussion of the issue of accentuation in the latter camp and hardly any in the former. However, Vürtheim refers (with Dindorf in TGL vi. 169) to AB ii. 586 = A.D. Adv. p. 175 Schneider, where Apollonius states that the Doric accentuation was παντᾷ.257 In ‘authentic’ Doric, the adverb occurs twice in inscriptions of the third century or earlier, and five times in the Laconic passages of the Lysistrata. In Doric Kunstsprache, it is found in all three tragedians and in Aristophanes, Pindar, Bacchylides, and Theocritus.258 The sources may be of little or no use to us as regards the question of accentuation: the inscriptions lack accents, and medieval mss. and Roman papyri are hardly worth much as evidence for the ﬁfth-century authors’ non-expressed notions of accentuation. Nevertheless, a survey of the relevant critical editions reveals that the circumflex on the ultima on Doric παντᾷ is found in mss. of Pindar, Theocritus, and—once—at Lys. 1081, in the oldest ms. of Aristophanes (R, tenth century). The mss. of tragedians, on the other hand, always present paroxytone accentuation in this adverb.259 257 φάµεν γὰρ πάντῃ, ὅτι καὶ πάντως … οὐδαµῇ δέ, ὅτι καὶ οὐδαµῶς. διχῇ τε καὶ τριχῇ, ὅτι καὶ διχῶς καὶ τριχῶς. τούτῳ γὰρ τῷ λόγῳ καὶ ∆ωριεῖς παντᾷ φασιν, ὅτι καὶ τὸ ἐπίρρηµα παντῶς, καὶ ἀλλᾷ, ὅτι καὶ ἀλλῶς. The same claim is found in Hdn. Pros. cathol. i. 489 Lentz (cf. Σ Pi. P. 3.65), Σ Theoc. 8.41–44b–c. 258 Authentic Doric: Ar. Lys. 169, 180, 1013, 1081, 1096, Tab.Heracl. 141 (fourthcentury Laconic), GDI 4254.8 (third-century Rhodian). Kunstsprache: Pi. O. 1.116, 9.24, P. 1.96, 2.23, 4.171, 10.38, I. 1.41, B. 5.31, 9.48, 15.44, A. Pers. 282, Supp. 88, Eu. 255, 967, S. Tr. 647, E. Med. 853, Ion 205, Hipp. 563, Or. 1267, 1294, Ar. Av. 345, Theoc. 1.55, 8.41 ter, 15.6 bis, 21.53. Adverbial ἁπαντᾱι is also probable in E. Ph. 312 (see Mastronarde ad loc.). 259 The mss. of the tragedians usually exhibit πάντα, sometimes Attic πάντη or πάντηι. πάντᾱι is found for our passage and for Eu. 255 (Mpc). In the mss. of Aristophanes, apart from R mentioned above, we find πάντα or, in a few mss. of Av. 95 The issue becomes even more complicated seeing that several critics and editors of Euripides and Sophocles now argue that πάντᾱ, without the iota, is a correct form of the adverb.260 They do not touch upon the issue of accentuation but refer to S.GG i. 550, who states that -ᾱ and -η are instrumental case-endings in several adverbs, e.g. (Att.) λάθρᾱ, πῆ, ταύτη, (Dor.) κρυφᾶ, ταυτᾶ, hαµᾶ. The last two are found in authentic fifth-century Laconic, DGEE 12.4, 12.14, etc. The adverbs on -ᾳ and -ῃ are either dative– locative in origin or assimilated to this case, originally being instrumental (S.GG ii. 163). However, the dative-case adverbs often appear in Doric inscriptions; apart from the ones cited in n. 258 above, we find, for example, locative τᾶιδε in ﬁfth-century Megarian (DGEE 167a) and ἀλλᾱι in ﬁfthcentury Cretan (DGEE 179 VI 14, 37). Accordingly, the iota is found in all extant occurrences of πανται in early inscriptions,261 and the dative-case ending is also found in similar adverbs in Doric inscriptions more or less contemporary with Aeschylus. In our case, then, the choice is easy, seeing that the sense of the adverb is clearly locative, not instrumental, and (less importantly) that the iota is found in M.262 To return to the accentuation, S.GG (i. 384) does adhere to Apollonius’ doctrine, as do K–B (i. 326) and Thumb–Kieckers (1932, 76).263 The trag- 345, πάντη. The Bacchylides papyrus has παντᾱι everywhere, exhibiting an accent only at 15.44, where we read πάντᾱι. In Pindar the circumflex is found in about half of the mss., and sometimes the iota appears; likewise in Theocritus, as we may conclude from the silence of Gow, who prints παντᾷ everywhere but reports divergent ms. readings only at 21.53 (‘παντᾷ Iunt.Cal. πάντα X πάντα τε Tr.’). (The readings are gathered from Mommsen’s Pindar, Lloyd-Jones–Wilson’s Sophocles, Diggle’s Euripides, Henderson’s Lysistrata, Dunbar’s Birds, Gow’s Theocritus, and British Museum 1897.) 260 Barrett on E. Hipp. 563, followed by Stinton (1985, 419 [421]), Davies on S. Tr. 647. So also in the early editions of Aeschylus, and in TGL vi. 169 (s.v. πάντη). 261 παντᾶ is found in GDI 5200 I 9 (Sicily, prob. 1st century B.C.), in which inscription, however, the iota is lacking in all cases of the dat.sg. of the first and second declensions. 262 I shall not go into general principles regarding the iota in this and similar adverbs, except to suggest that the instrumental ending should perhaps be printed, pace Henderson p. xlviii, in several or all cases of παντᾶ in the Lysistrata, in the light of the contemporary Laconic inscription DGEE 12 discussed above (cf. Thumb– Kieckers 1932, 89). 263 If we had been looking at an instrumental-case adverb here, it should accordingly, following Schwyzer (S.GG) et al., be written παντᾶ: so Bothe (ed. 1830), fol- 96 edians may not have given the issue much thought: here (and of course in Pindar and Bacchylides) the Doric form of the adverb only occurs in lyrical passages, where the pitch accent is invalidated by the music. In the spoken (iambic) Laconic of Aristophanes, and in the Doric hexameters of Theocritus, the accentuation has relevance, however. Editors of these authors usually print the adverb (with or without the iota) with a circumflex, as do the editors of Pindar, where the mss. yield plenty of support for this accentuation. In my opinion, the best solution is to maintain consistency and accentuate παντᾷ (or -ᾶ) in all cases, including Bacchylides and the tragedians. 89. κἀν σκότῳ κελαινῷ ξὺν τύχᾳ: σκότῳ is flat without an attribute: ‘it blazes forth, even in the dark’. Conversely, τύχᾳ does not benefit from the ms. reading µελαίνᾳ: if we are already in the dark, what is the dramatic relevance of a sudden ‘black fortune’ that ‘blazes (!) forth’? The essence of τύχη is its inconstancy, changing good for bad and vice versa: in the phrase σὺν τύχῃ, however, it always elsewhere refers to good fortune, with or without an attribute.264 A contrast to the beginning of the next strophe is intended here: just as Zeus may bring good fortune in a desperate situation (89–80), which is what the Danaids wish for themselves, so may he on the other hand ‘hurl mortals to their ruin from high-towering hopes’ (96–98)—the fate desired by the girls for their suitors. κελαινῷ (Schmidt 1863, 229265) is thus as good as certain. For the corruption, cf. 785 with FJ–W’s note. ‘Black darkness’ in the context of changing fortune is also found, metaphorically, in Pi. fr. 108b (cf. also Pers. 301, Ag. 22– 23):266 θεῷ δὲ δυνατὸν µελαίνας ἐκ νυκτὸς ἀµίαντον ὄρσαι φάος, κελαινεφέϊ δὲ σκότει καλύψαι σέλας καθαρόν ἁµέρας. lowed by, e.g., Paley (ed. 1861), Dindorf (ed. 1869), Kirchhoff, Bassi. 264 For instance, Th. 472, Ch. 138, S. Ph. 775, Ar. Av. 1722, Sapph. fr. 20.4, Pi. P. 2.56, N. 4.7, 5.48, 6.24, I. 8.67, B. 5.52, 11.115, Hp. Loc.Hom.46.4. 265 Schmidt prints κελαινῷ in his lemma of 86–90, but then argues for the reading κλεαινᾷ. Tucker also suggested κελαίνῳ (sic) with further emendations. We may note that Bachvarova (2001, 51) translates (without comment) ‘even in dark gloom, with favor for mortal people’. 266 Part of verses 1–2 is cited on the title-page of Burges’ edition of the Supplices. 97 90. µερόπεσσι: on the epicism, see in general 40–111n. On the dative plural ending -εσσι in tragedy, see Diggle (1974) 22, n. 2 (117, n. 81). The etymology of the word is still unknown, and it is uncertain whether the popular one of Hesychius and other grammarians (µείροµαι + ὄπα, ‘have a voice allotted’) was known to, or appreciated by, Aeschylus and his contemporaries (the word also occurs at Ch. 1018, E. IT 1264). See further Russo on Od. 20. 49, with refs. 96–98. ἰάπτει … βροτούς: cf. 89n. 98–103. βίαν δ’ οὔτιν’ … πᾶν ἄπονον … ἥµεν’ … ἐξέπραξεν: the effortless omnipotence of the divine, admirably brought out in these verses (cf. also Eu. 650–51), also happens to be an important theme in the Greek art that was contemporaneous with Aeschylus. The tension between calm and strength, essential qualities of Greek gods, is best admired in the so-called severe style that dominated Greek sculpture of the early and middle fifth century. Apollo and Athena from the temple of Zeus at Olympia set the standard, emitting perfect calm in combination with supreme power. The cool and easy countenance of the young gods is effectively contrasted with the struggle and pain of lesser beings: Athena is set against a labouring Heracles, Apollo against a struggling Centaur.267 The calm and ease of the divine is also one of the motifs in respect of which an influence from Xenophanes on Aeschylus might have operated (see above, 86–103n.). The present passage has been compared to frr. 25–26 D–K: ἀλλ’ ἀπάνευθε πόνοιο νόου φρενὶ πάντα κρααίνει and αἰεὶ δ’ ἐν ταὐτῷ µίµνει κινούµενος οὐδέν | οὐδὲ µετέρχεσθαί µιν ἐπιπρέπει ἄλλοτε ἄλλῃ. The fragments do contain notable similarities in wording: ἀπάνευθε πόνοιο ~ ἄπονον : νόου φρενὶ ~ φρόνηµα : πάντα κρααίνει ~ ἐξέπραξεν ἔµπας : ἐν ταὐτῷ … κινούµενος οὐδέν ~ ἥµεν(α) … αὐτόθεν. (See also 101–3n. on the Homeric parallel for the last sentence.) 98–99(~106–7). On the (nameless) metric colon see Stinton (1975) 84–89 (114–19), Dale (1968) 95–96. 100–101. πᾶν ἄπονον δαιµονίων· ἥµεν’ ἄνω: FJ–W argue in favour of Bothe’s (ed. 1830) δαιµόνιον, which has not won much favour among later critics. Griffith (1986) claims that the metrical pause necessitated by the bre- 267 See Dörig (1987) on the Olympia master, whose surviving works are in the collections of the Olympia Archaeological Museum. 98 vis in longo is unwelcome. δαιµονίων is probably sound: however, pace Verdenius (1985) and Griffith (1986), the genitive is not possible to define exactly as either partitive or possessive, but contains both notions at the same time: ‘all is effortless of the Divine’. As for δαιµόνια, which is also the subject of ἥµεν’ ἄνω in 101, FJ–W claim that the sense ‘“divine beings” is unexemplified and inconceivable in tragedy’. This is too blunt a translation: one should understand ‘the Divine’ as an abstract, similar to the Euripidean τὸ δαιµόνιον.268 τὰ δαιµόνια include divine actions as well as divine will, mind, and being. Xenophanes may be partly responsible for the sophistication (see above, 86–103n., 98–103n.): cf. frr. 23–24 D–K εἷς θεός, … | οὔτι δέµας θνητοῖσιν ὁµοίιος οὐδὲ νόηµα and οὖλος ὁρᾷ, οὖλος δὲ νοεῖ, οὖλος δέ τ’ ἀκούει. As for the emendations, ἄπονον (de Pauw: ἄποινον M) is easily inferred from ἥµεν’ and from βίαν δ’ οὔτιν’ in the previous verse. πᾶν (Labbé: τὰν M) is certain. Wecklein’s (ed. 1885) ἥµεν’ is better, and certainly more economical, than the solution adopted by most modern editors (ἥµενος ὃν, Paley ed. 1844 and Haupt, respectively: ἥµενον ἄνω M). It is defended by Verdenius (1985), who adduces (with Paley ed. 1844) 597 οὔτινος ἄνωθεν ἡµένου σέβει κράτη and Pr. 312–13 ἀνωτέρω θακῶν … Ζεύς.269 101–3. ἥµεν’ … αὐτόθεν … ἑδράνων ἀφ’ ἁγνῶν: The sentence is recognisably adapted from an Homeric formula (cf. 40–111n.): cf. Od. 13.54–56 θεοῖσιν | … τοὶ οὐρανὸν εὐρὺν ἔχουσιν | αὐτόθεν ἐξ ἑδρέων, 21.420 αὐτόθεν ἐκ δίφροιο καθήµενος, Il. 19.77 αὐτόθεν ἐξ ἕδρης. ἀφ’ is a certain correction (Σ102, Anon.Span.: ἐφ’ M, Σ101b). 103(~111). On the colometry see 87(~92)n. 104–7. ὕβριν βρότειον οἵα νεάζει, πυθµήν … τεθαλώς: West’s (W.SA) Βήλειος οἵᾳ (οἵᾳ Schütz ed. 1794) has the advantage of clearing up the imagery as well as the syntax considerably. The ‘stock’ becomes recognisable as 268 Cf. E. Ba. 894 and Ph. 352 with the notes of Dodds and Mastronarde, respectively, and also fr. 152 Nauck, Alc. 1159 (= Andr. 1284). 269 Cf. also, e.g., Ag. 182–83, Ch. 165 (124a), fr. 159, Hes. Op. 8, S. Ant. 1072–73, fr. 684, fr. 941. 12, E. Hec. 791, Ar. Pax 854, Av. 843, Pl. Crat. 408c. If ἄνω is corrupt, a better reading than the one usually accepted would be ἥµεν’ ὃν φρόνηµα: the responsion is preserved as it is possible to read δια- as monosyllabic in the antistrophe (cf. 72, 799, Pers. 1007, 1038, West p. xxxiv). 99 the family stock of the Aegyptiads, sprouting through the hybris.270 Belus is the grandfather of the Danaids and the Aegyptiads (see 319), and West compares Ov. Her. 14.74, where Hypermestra addresses her bridegroom (Lynceus) Belide. However, attractive as this emendation may be, on principle we cannot take for granted that the paradosis is corrupt (apart from the unmetrical οἷα, easily emended to οἵα [Hermann]). ‘As a qualification of ὕβριν, βρότειον is otiose,’ according to West, ‘as there has been no thought of any hybris other than that of mortals’. Even so, βρότειον may not so much define what kind of ὕβρις we are dealing with as simply add pathos: ‘look at mortal insolence!’ The paradosis shows the hybris as a living, organic thing, growing, through the desire for marriage, into a stock. πυθµήν should be taken as predicative with νεάζει, the structure being equivalent to ἰδέσθω εἰς ὕβριν οἵα νεάζει, πυθµὴν γιγνοµένη.271 This interpretation is supported by several examples of hybris envisaged as an organic, blossoming entity (cited above, 73n., n. 233): cf. esp. the instances cited by FJ–W. 109–10. διάνοιαν µαινόλιν κέντρον … ἔχων: ἔχων is attracted to the masculine πυθµήν. ‘Aeschylus’ phrase is charged with ambiguity’ (FJ–W). The hybris goads the Aegyptiads with lustful thoughts—or is it the Danaids who are trying to escape the κέντρον ἄφυκτον? A Freudian nightmare. Cf. FJ–W for examples of κέντρον indicating sexual urges.272 109. καὶ διάνοιαν µαινόλιν: note word-end after long anceps in the final iambic here (cf. on 42, n. 188). Diggle (1982) notes S. OC 1055~1070 as the only tragic parallel, where the offending word-end occurs after a prepositive. Here perhaps the collision of nasals may offer a ‘bridge’ of sorts: -οιαµµαι-. 110–11. ἄτᾳ δ’ †ἀπάται µεταγνούϲ†: the corruption may be due to a kind of dittography or assimilation: ατα(ι) has been repeated from the previous line in 111, replacing the true ending of the word beginning ἀπ-. The corruption may go further, however: µεταγνούς is a word that is very hard to get any sense out of in the context. What can a ‘change of mind’ possibly have to do with anything? The verb appears once elsewhere in Aeschylus, Ag. 221, where it describes Agamemnon’s decision to sacrifice Iphigenia: τὸ παντότολµον φρονεῖν µετέγνω: ‘he altered his mind to be minded of the utmost 270 For the construction with the dative relative cf. S. Tr. 1044–45 ἔφριξα τάσδε συµφοράς … ἄνακτος, οἵαις … ἐλαύνεται. 271 Cf. Vürtheim, Friis Johansen (ed. 1970), and and Wecklein (ed. 1902). 272 On sexual innuendo in Aeschylus see also Sommerstein (1993), especially p. 61. 100 daring’, or words to that effect. But the Aegyptiads have not changed, or ‘altered’, their minds: their ‘decisive commitment to a fatal course of action’ (W.SA) has been taken for granted throughout the drama and cannot possibly be described as a ‘change of mind’. Moreover, the subject of the clause still appears to be the hybris-trunk, which has just been depicted as goading the Aegyptiads with ‘frenzied intention’. To say that the hybris has ‘by delusion [ἀπάτῃ] changed its mind so as to incur destruction [ἄταν]’ (FJ–W, adopting the reading of Mcac) is to translate Greek nonsense into English nonsense. The construction of µεταγιγνώσκω with a direct object, meaning something like ‘change one’s mind to (take on, infer, be committed to) something’, also appears to lack parallel.273 At the end of the corresponding strophe (103) we find ἀφ’ ἁγνῶν, which may supply a hint. µεταγνούς may conceal a word with the root ἁγν-, which would produce a ‘responsional parechesis’274 of the kind we find in, e.g., 750 ~757, where οὐλόφρονες in the strophe is echoed by περίφρονες in the antistrophe. The obvious choice would be ἀνάγνους,275 referring to the Aegyptiads: cf. 226–28, 751. A possible solution, then, would be ἄτᾳ δ’ ἀπατῶν ἀνάγνους: the hybris goads the suitors with διάνοιαν µαινόλιν, deceiving them (ἀπατῶν still attracted to πυθµήν), as it were, with ἄτη. Cf. Pers. 93– 98 δολόµητιν δ’ ἀπάταν θεοῦ τίς ἀνὴρ θνατὸς ἀλύξει; … παράγει βροτὸν εἰς ἀρκύστατ’ Ἄτα (W.SA: ἀρκύστατα codd.). The emendation is rather violent, and it may be hard to defend palaeographically: possibly the ν of -τῶν could have been read as µ, and we may note that the corruption ἀπατῶν > ἀπάταιµ might be an easy one if M was copied 273 With West’s (argument in W.SA) ἄταν δ’ ἀγαπᾶν (he compares Hes. Op. 58 ἑὸν κακὸν ἀµφαγαπῶντες), the construction becomes identical to that in Ag. 221; but the element of change inherent in µεταγνούς remains unexplained, and untranslated, by West: ‘having decided to embrace ruin’. Tucker’s ἄτας {δ’} ἀπάταν µεταλγοῦς removes the difficult participle but introduces an adjective which is not attested elsewhere (cf. 405–6n., n. 518). 274 On this manner of echoing words from the strophe in the antistrophe, see, for example, G.AS 42–43; FJ–W 49n.; Schmid (1934) 297, n. 2; Korzeniewski (1968) 162–70; Kühn (1905) passim; Rash (1981) passim. 275 ἀνάγνοις already suggested by Burges (1810, 803) in the quite passable ἄταις ἕπεταί µ’ ἀνάγνοις, inaccurately reported by Wecklein (ed. 1885). 101 from a minuscule source.276 αν turning into ετ is more difﬁcult. It might, however, be a further result of dittography: αταιδε from the previous may be repeated as (απ)αταιµε (suggested to me by Professor Richard Janko). 112–75. From here on, iambic (or single-short) movement dominates the ode. Rash (1981, 76) observes that this is answered by a transition from narrative to performative language: ‘there is little narrative of past events or philosophical reflection; rather, the women express their terror (strophic pair Z), appeal of divine assistance (strophic pair H), and threaten suicide’. This is true and accords with dactylic being proper to epic verse, whereas iambic is the natural metre for dramatic dialogue. See also 40–175n., on the content and thematic structure of the entire ode, and 40–111n. on the dactylic first half. 112–16. The emotional tone is stressed by the extreme resolution in the iambics (Rash 1981, 79) and leads to an outbreak with interjections in 115. 115–16. Though alive, the Danaids pay their own respects (τιµῶ) as if they were already dead (cf. e.g. Ch. 200, ‘Th.’ 1046, S. Ant. 904). Cf. on 123– 24. ἐµπρεπῆ goes (pace Page, FJ–W) with µε: see Headlam (1904) and the parallels he adduces.277 On the intruding gloss after ἐµπρεπῆ in the mss. 276 Friis Johansen (ed. 1970, p. 33) claims that this is the case, a claim which is iterated in FJ–W. It appears to have been first suggested by Turyn (1943) 14–15. Garvie (ed. Ch., p. lv) claims that it is ‘generally accepted’ that M’s exemplar was a minuscule ms., which is saying too much: only a few years earlier Page (p. x) and Diggle (1982, 127 with n. 2) had disagreed, arguing that the evidence is too scant. West does not touch upon the issue in his accounts of M (ed. pp. iv–vi, W.SA 321–23), but argues elsewhere (W.SA 163) for a possible corruption in M (Supp. 879) owing to a misreading of a minuscule source. The issue may be settled for L, the Sophoclean part of the codex Mediceus (see the refs of Garvie l.c.); but this does not necessarily mean that the same is true for M, which was written by another scribe than L and probably stems from a different source. However, Garvie’s list of corruptions (in the Choephoroi), which appear to be due to the erroneous copying of a minuscule source, is persuasive. To Friis Johansen’s (l.c.) similiar list for Supp., which includes 14, 271, and the more dubious (see Diggle l.c.) 116, 324, and 386, we may add the present passage, 195 and 276 (qq.v.) as possible examples. The simple minuscule corruption ον > ου and vice versa also appears in 43, 444, 584, 1063. 277 Ch. 12, 18, S. El. 1187. FJ–W complain that these examples support the idea of a conspicuous mourner, not a conspicuous mourned, and would thus call for ἐµπρεπής (Meffert 1861, 11, adopted by West); but I think this is a little pedantic, seeing that the mourners and the mourned are the same persons here. 102 (θρεοµένη µέλη, deleted by Porson), see FJ–W ad loc. and ii. 394. FJ–W (i. 61) adduce the marginal reading of M, ζώσατο οἷσ µε τιµᾶι, as evidence for a minuscule exemplar of M:278 see 110–11n. with n. 276. 117–22 = 128–33. Refrains of various sizes appear from time to time in Greek verse, not least in Aeschylus,279 a fact that appears to be satirised in Ar. Ra. 1261–80 (see Radt ad loc., test. 120, with refs). A received technical term for at least some kinds of refrains is ephymnion.280 The religious origin of the 278 ῶ > ᾶι would be a misreading of a minuscule text. Not counting repetitions of interjections in identical places in the strophe and antistrophe, but including strophical repetitions of single verses consisting of more than one word, refrains are also found in 141–43 = 151–53, 889–92 = 899–902, Pers. 663 = 671, Th. 975–77 = 986–88, Ag. 121 = 139 = 159, 1489–96 = 1513–20, Eu. 328–33 = 341–46, 1043 = 1047, fr. 204b.6–8 = 15–17 and with variation of one word in Eu. 1035 = 1039. In 1072–73 = 1076–77 Cassandra repeats an entire strophe verbatim in the antistrophe, and in 1080–81 = 1085–86 the first two verses (out of three) of a strophe are repeated. Refrains also appear in E. Ion 125–27 = 141–43, Ba. 877–81 = 897–901, 992–96 = 1011–16. 280 Not all the Aeschylean refrains listed in the previous footnote have been taken as ephymnia proper by modern scholars, and indeed there is some confusion as to which strophes ought actually to receive this label. West and Page, for instance, do not seem to think that 889–92 = 899–902, consisting mainly of vocatives and exclamations, are ephymnia, nor Th. 975–77 = 986–88. On the other hand, West prints ‘ephymn. 1’ beside the text of Ag. 1455–61 and ‘ephymn. 3’ at Ag. 1538–50, neither of which is repeated later. He thus contradicts his own definition of the term (West 1982, 80): ‘a refrain in which words as well as music are repeated’. On the latter passages West apparently follows Fraenkel on Ag. 1407–1576 (iii. 660–62), who speaks of 1455–61 and 1538–50 as ‘ephymnia’. Earlier, however, on Ag. 121, Fraenkel had defined ἐφύµνια (using Greek letters) as ‘refrains’. On any consistently maintained definition, Ag. 1455–61 and 1538–50 ought to be mesodes, as well as Ch. 807–11, 942–45, 961–64, Eu. 354–59 and 372–76, and 162–67 of the present drama, q.v. (Ch. 789–93 may be answered by an antistrophe in 827–30, rather than both of them being independent mesodes). Cf. G.AS 43–44, FJ–W on 117–22 and my notes on 141–43 and 162–67 below. An adequate working definition of ‘ephymnion’, at least for the purpose of consistent notation in the margins of critical editions, might be ‘identically worded refrain repeated after strophe and antistrophe’. There ought to be no reason to distinguish between refrains consisting of one line only and longer ones, still less between refrains that appear in the first choral ode of the Supplices and those appearing elsewhere. On ephymnia see further, e.g., Goebbel (1858), Schwarz (1897), Kranz (1933) 128–33, 199, and below on 141–43 = 151–53, 162–67. 279 103 drama is never more apparent than in the ephymnia, which practically always consist in invocations of gods.281 The metre of the present ephymnion is very irregular, and the s–d notation given in the metrical chart above is only a suggestion as to the general structure. In this context, the rhythm would presumably be received as a continuation of the single-short movement, although heavily syncopated: i.e. iambic. The molossi ending 117 = 128 and 118–19 = 129–30 could then suggest choliambic rhythm (syncopated trimeters), although this has not been recognised as a lyric metre. Cf., however, the irregular cholosis in 47–48~56–57. 120 = 131 have generally been taken as dochmiac (Dale 1968, 116; Dale 1971; Diggle 1982; West).282 117 = 128. ἱλεῶµαι: Schultze’s (1932) emendation is an otherwise unattested variant form (present indicative) of ἱλάοµαι, but apparently formed in a regular manner by quantitative metathesis from *ἱληϝοῦµαι (cf. ἵλεως).283 ἱλεοῦµαι (Wilamowitz) would be in greater accordance with attested usage,284 but is further removed from the paradosis ἱλέωµαι. On the other hand, Aeschylus would probably not have distinguished in spelling between -ωµαι and -ουµαι (Kirchhoff 1887, 95–96). µὲν contrasts with 118 = 129 καρβᾶνα δ’, q.v. The polarity is, as in 69–71, that between Greek and foreign: the Danaids’ attempt at appeasing the Argive land (Ἀπίαν βοῦνιν) is contrasted to their barbarian speech.285 281 See FJ–W on 117–22 = 128–33, Kranz (1933) 128–33, Kraus (1957) 15. In nontechnical ancient literature ἐφύµνιον usually means, in analogy with ἐφυµνεῖν, ‘sung invocation’ rather than ‘refrain’: see Sandin (2002–3) 181–84. 282 On the affinity of dochmiac to syncopated iambic, see Conomis (1964) 46–48 and Dale (1968) 107–11. Dale also discusses the emotional tone proper to these metra (110–14). The affinity is clarified by Dale’s s–d notation, if we are prepared to adopt this for dochimacs (see above on 40–175): ×šs is the ‘basic’ form of dochmiac, which thus comes across as a ‘compressed’ iambic dimeter (common forms of which are e.g. ×š's, ×š×s). See also on the mesode (162–67) below. 283 See Schultze (1932) 304, S.GG i. 245, FJ–W. 284 Pl. Lg. 804b and late prose (apart from the instances mentioned in LSJ s.v. ἱλάοµαι, e.g., Ael. NA 7.44 bis, Lib. Or. 64.96, Σ Il. 580a.2). We find the fut. ἱλεώσεσθαι in D.C. 78.34, aor. ἱλεώσασθαι in D.C. 59.27, 169.51, Ael. fr. 23.1, 47.4, etc. The form ἱλέοµαι (Turnebus) is unattested. 285 FJ–W advocate a contrast between ἱλεῶµαι µὲν and 120 = 131 πολλάκι δ’ ἐµπίτνω, taking the expressions to refer to words and gestures, respectively. The same is implied by West’s dashes before and after καρβᾶνα … κοννεῖς, which make this 104 Ἀπίαν βοῦνιν: two words of more than usually obscure derivation. The intended sense appears to be ‘Apian hill-country’, with ‘Apian’ meaning ‘Peloponnese’ (see below). βοῦνιν, whatever its true etymology (cf. 776),286 naturally suggests ‘cow’: another allusion to the origin of Io. To confuse matters further, there are (at least) two mythological characters by the name of Apis, both of whom are relevant as the presumed origin for the adjective Ἄπιος. As is evident from 260–70 (q.v.), Aeschylus (or at any rate Pelasgus, but it is nowhere hinted that he might be mistaken) derives Ἄπιος from the name of a son of Apollo. More famous, however, and certainly more relevant to the Danaids, is the ancient divine king of Egypt who took the form of a calf and who was, according to Herodotus and others, identical with Io’s son Epaphus (see on 41–44 with n. 183). This Apis is never mentioned by that name in the Supplices, though. Hall (1989, 170, n. 35) notes that the name Apia for the Peloponnese implies that Pelops had yet not arrived, hence also forming an example of Aeschylean ‘antiquarianism’ in the present drama: cf. 15n., 183n., 236–37n. 118–19 = 129–30. The ephymnion continues to abound in obscure words: καρβᾶνα (‘barbarian’) and κοννεῖς (‘know’) are both of unknown origin.287 The obscurity appears to be intentional here, representing the foreign speech (καρβᾶνα αὐδάν) of the Danaids. The feature of foreign language is not usually treated with any kind of consistency in Greek literature: poets mention it at times,288 but as a rule they ignore any difficulties of communication between foreigners and Greeks.289 Here the implication may be clause parenthetical. But this is an inferior solution, seeing that the contrast between Hellenic and foreign is not only much stronger here, but also a recurrent theme of the drama: cf. 234–327, 496–98, 719–20, 893–94, 914, 921–22, 952–53, and see G.AS 48–49. Bothe’s (ed. 1830) καρβᾶν’ αὐδὰν, defended by Diggle (1982), is thus detrimental. 286 The stem has survived in modern Greek with the meaning ‘mountain’: βουνίσιος, βουνό, etc. 287 See, however, van Windekens (1986) on the former, as well as the discussion by FJ–W. Boissonade’s conjecture for M’s εὐακοννεῖϲ (119) and εὐγακόννιϲ (130) seems certain: see Hsch. s.vv. κοννεῖν, κοννοῦσι. 288 For instance, Il. 2.803–6, 2.867, 4.437–38, Pers. 406, 635–36, Ag. 1050–51 (cf. 1060–63). 289 See Kranz (1933) 81–83, Thomson on Ag. 1059–61 (his 1043–45), FJ–W, Hall (1989) 17–22, 118–20, passim, on Aeschylus’ treatment of foreign languages. 105 that the Greek Earth can understand the Danaids, or rather perhaps may choose to listen to them, being the land of their origin. 120–22 = 131–33. The tearing of clothes is a recurrent theme in Aeschylus, usually proper to mourners290 and well in accordance with the imagery of death and funerals that appears in these strophes (see on 115–16, 123–24). Sommerstein (1977) ingeniously suggested that the tearing of veils may symbolise the rejection of marriage. 122 = 133. Σιδονίᾳ: The Phoenician city Saida was called Sidon by the Greeks. The workmanship of its artisans is celebrated in Homer,291 Il. 6.289– 91 explicitly referring to textiles. The reading (Mpc: σινδονία Mac) is certain: the adjective here continues the ‘barbarian-theme’ from the previous verses (cf. FJ–W). 123–24. θεοῖς δ’ … ἀπῇ: the scholiast, paraphrasing ὅπου δὲ θάνατος ἀπῇ, ἐκεῖ τῶν ἀνθρώπων εὐπραγούντων τιµαὶ τοῖς θεοῖς ἐπιτρέχουσιν, is, like the text, correct in the main (the ἀπῇ of the scholium is generally adopted for M’s ὅπηι). Owing to the difficulty of the word ἐναγέα in 123, many critics have refused to see the obvious and have, in my opinion, made things far too difficult for themselves. The passage can really only be interpreted as the scholium takes it: a pregnant expression of the common Greek sentiment with regard to the gods, do ut des, or here rather da ut dem. The Homeric heroes (see FJ–W for references) promise immense offerings to the gods in return for personal success. So also here: ‘For the gods … offerings [will, or are wont to] stream forth, if things turn out well, where death be apart.’ For the mention of sacrifices to the gods in a fearful situation, cf. also ?Lys. 2.39 ποῖαι δ’ οὐχ ἱκετεῖαι θεῶν ἐγένοντο ἢ θυσιῶν ἀναµνήσεις …;292 290 See IA s.v. λακίς, and Pers. 199, 538, 1030, 1060. Il. 6.289–91, 23.743, Od. 4.615–19 = 15.115–19, cf. 15.425 Σιδῶνος πολυχάλκου. See FJ–W, Richardson on Il. 23.740–49, Wace–Stubbings (1962) 542–43, Lorimer (1950) 64–67, 80. 292 See also Headlam ad loc., n. 2, who refers to Si. 17.27–28, the Greek version of which runs: ὑψίστῳ τίς αἰνέσει ἐν ᾅδου ἀντὶ ζώντων καὶ διδόντων ἀνθοµολόγησιν; ἀπὸ νεκροῦ ὡς µηδὲ ὄντος ἀπόλλυται ἐξοµολόγησις· ζῶν καὶ ὑγιὴς αἰνέσει τὸν κύριον. 291 106 As for ἐπίδροµ’, the great variety of meanings that words from this stem exhibit shows that its range is broad enough to be interpreted as required.293 Tucker compares Eu. 907 ἐπίρρυτος καρπός. An objection to the paradosis has been its alleged lack of connection to the previous strophe (FJ–W, Tucker). How could there possibly be any connection, one asks at first sight, seeing that 112–16 are almost entirely devoid of any substance to connect to (see ad loc.). ‘These are the pains of which I tell, heavy, shrill, [etc.], oh, conspicuous by moaning I honour myself with cries, yet alive.’ Only in the last sentence is there any suggestion of anything beyond the most general type of lament. ‘I honour myself with wailing’, the Danaids say, ‘but alive’, the point being that honours of this kind are usually paid to the dead (see ad loc., FJ–W). But these two words (ζῶσα … τιµῶ) do in fact constitute a thematic connection to the present passage, where they are matched by honour to the gods (τέλεα θεοῖς), where death be absent (ὁπόθι θάνατος ἀπῇ). Why talk of death, FJ–W ask (and Rose): there has been no mention of suicide before (there will be shortly, though; see 159–60). In case anyone should think that this is a valid objection, we will return to it shortly.294 Incidentally, a corresponding juxtaposition of sacrifice and suicide also appears in 450–65. The only substantial objection to the paradosis concerns the word ἐναγέα. The adjective ἐναγής usually means ‘polluted’ in classical Greek, and other words from this stem almost always denote offerings to dead people, for instance ἐνάγισµα, the gloss of the scholiast. One non-pejorative instance of ἐναγής may, however, occur in S. OT 656 (Creon has sworn a solemn oath that he is innocent; Oedipus presses him, and the chorus protests): 293 Some examples from LSJ: attacks, ships entering harbour, blood flowing, a ‘cord which runs along the upper edge of a net’ (X. Cyn. 6.9), horses racing, light and dark spreading. 294 FJ–W also object to the offering of sacrifice that ‘the general proposition is unsupported by any application of it to the Danaids (…), and it seems out of key with their mood and circumstances’. But the entire ode from 77 onwards consists in a prayer to the gods (see 40–175n. for an analysis). If such a mention of τέλεα is out of place here, I fail to see where it would be in place. 107 — φράζε δή· τί φῇς; — τὸν ἐναγῆ φίλον µήποτ’ ἐν αἰτίᾳ σὺν ἀφανεῖ λόγῳ <σ’> ἄτιµον βαλεῖν 656 ἀναγῆ *Musgrave φίλον plerique : φίλων N, P in lin., S, Suda µήποτέ σ’ *Nauck 657 λόγῳ <σ’> *Hermann : λόγον L : λόγων K : λόγων aut λόγῳ alii. Here ἐναγῆ, if indeed this is the correct reading, has been understood as ‘sworn’, ‘bound by oath’ or ‘liable to pollution because of an oath’.295 I find the Sophoclean passage perplexing, but it has been taken as a justification for such translations of ἐναγέα τέλεα in Supp. 123 as ‘sacrifices in satisfaction of vows’ (Smyth), ‘sacrificial rites in expiation’ (Tucker 106–7n.). On the other hand, Verdenius (1985) defends ἐναγέα in the general meaning of ‘solemn’. This seems to me to be the only possibility of defending the word in Aeschylus: simply to accept an in bonam partem meaning, just as the later word παναγής can mean ‘all-hallowed’ as well as ‘accursed’. So Wilamowitz (1914, 32): ‘die heiligen Steuern’. The words περαγής (‘holy’ prob. in Corinn. fr. 1 iii 47), εὐαγής, and δυσαγής might perhaps support this; cf. also the Mycenaean ti-mi-to a-ke-i,296 Hsch. α 407 ἄγεα· τεµένη, α 734–35, Phot. s.v. ἄγος, Suda ε 1086, DE s.v. ἄγος, van Windekens (1986) s.v. ἅζοµαι and, for instance, the English word ‘awful’. All this notwithstanding, Boissonade’s ἀναγέα, ‘untainted’, ‘pure’, is attractive. The corruption would be very easy if δὲ was written in scriptio plena (δὲ ἀναγέα).297 If the word ἀναγής is poorly attested in this sense, it nevertheless conveys too good a meaning in this context to be completely dismissed.298 With ἀναγέα, the passage becomes—like, for instance, 6 (see 6– 295 ‘The friend who has bound himself with a curse’, Jebb; ‘amico sacramento obstrictum’, Tucker 106–7n.; cf., e.g., Kugler (1905) 62, Dawe and Lloyd-Jones ad loc. 296 See Myc. 60, 176, 257 Ventris–Chadwick (and cf. p. 144): apparently the name of a place at Pylos. Professor Richard Janko called my attention to this instance of a non-pejorative ἄγος. 297 Cf. FJ–W iii. 398–99, and see my note on 296. 298 This is FJ–W’s objection to ἀναγέα: it is rarely attested in the meaning ‘pure’, ‘untainted’. This is true; but it is actually as rarely, and uncertainly, attested in any other meaning: the alleged instance in Herod. 2.70 is most likely ἐναγής, not ἀναγής, and Hsch. α 4222 is probably interpolated (so Schmidt): see Sandin (2002–3) 178– 108 10n.)—an ironic and ambiguous reference to the future killing of the Danaids’ husbands. The girls would be promising ‘for the gods untainted offerings—if things turn out well—streaming in, where death be apart’. What the gods will actually get—as the audience knows—is mass-pollution by the slaughter of 49 newlywed husbands. A similar hint, only more blunt, could perhaps be understood with ἐναγέα in bonam partem. 134–35. πλάτα … λινορραφής τε δόµος: A peculiar paraphrase for a ship. The naïvely elaborate description may perhaps be significant, as it would agree with the tradition that Danaus was the πρῶτος εὑρετής of the ship.299 That theme is not explored in this drama, but might perhaps have been noted in the Amymone, the invention of things being a stock motif in the satyrplay.300 Danaus describes himself as a ναύκληρος in 177, and Headlam (1898, 192) suggests that he may have been presented on stage in a skipper’s outfit, ornatus nauclericus, which is described in Plaut. Mil. 1177 and apparently worn by the faux merchant in S. Ph. 542–627 (see 128–29 with Jebb’s note). It appears far-fetched, at least to a landsman, to understand λινορραφής as referring to anything other than the sail. The scholium, however, supported by Tucker and FJ–W, takes it to mean some kind of packing of the hull (see FJ–W). But cf. Pr. 468 λινόπτερ’ … ναυτίλων ὀχήµατα, E. Hec. 1080–81 ναῦς ὅπως … λινόκροκον φᾶρος στέλλων, IT 410 νάϊον ὄχηµα λινοπόροις <σὺν> αὔραις.301 δορός is difficult: taken in its common meaning ‘ship’ it should refer to the same thing as δόµος earlier in the sentence, but the different syntactical functions of the two nouns make this awkward: ‘the flaxen-stitched house, keeping the sea out of the ship’, as if the house was not itself periphrastic for the ship. Friis Johansen (ap. Friis Johansen–Whittle 1975) suggested στέγον δόρυ as an explanatory apposition, which is neat, even a little too neat. So Griffith (1986) who, with Rose, argues for an attributive, ‘appositive’ (K–G i. 264–65), not a separative genitive: ‘the oars and the flax-stitched edifice of the boat, 81. There is little or no foundation for Schrevelius’ <µὴ> καθαρός in Hsch. s.v. ἀναγής (α 4227), adopted by Latte. 299 See Introduction, II 1, n. 21, for refs. On the wisdom and innovations of Danaus, see also 320n. 300 See Seaford’s Cyclops, pp. 36–37. For the motif of the πρῶτος εὑρετής in Aeschylus, cf. Theor. frr. 78a.1–22, 78c.49–60. 301 On πλευρὰς λινοζώστους in Tim. Pers. (fr. 15) 15, see FJ–W. 109 keeping out the sea’. The word order is against this, however. A better solution might be to take δορός in the broader sense of ‘timber’, ‘plank’, or ‘hull’: indeed one could argue that this is always the proper sense of δόρυ when used metaphorically for ‘ship’. The expression is still somewhat confused (‘the flaxen-stitched house, keeping the sea out of the hull’), but not impossible for Aeschylus.302 Cf. 186–87n. and also 15–18n. above. 136–37(~146–47). Stinton (1975, 89 ff. [119 ff.]) thinks that the sequence $'' $'$' (×š×s) is rare in tragedy and goes on to list about thirty possible examples, twenty-one of which he regards as ‘prima facie’. In the prima facie group there are fewer Aeschylean and Sophoclean than Euripidean examples (2:4:15), and Stinton argues that ‘ba. + ia. is barred in Aeschylus, Sophocles and early Euripides’ (l.c. 94 [126]). To begin with, however, the recognition of only two examples in Aeschylus makes for a rather mean count, seeing that the latest Oxford and Teubner editors (Page, West) accept yet another five.303 Secondly, with the Aeschylean count modified to, say, four, the proportions (4:4:15 = 1:1:3.75) are not strikingly different from those of plays preserved under the names of the respective tragedians (7:7:19 ≈ 1:1:2.7): if there is a statistical significance to the difference between the elder tragedians and Euripides, it is not great enough to justify emending the instances of the metrical sequence in the former. There is a caesura instead of the syncopated short syllable in strophe as well as antistrophe: πνοαῖς($)· | οὐδὲ ~ -σφαλές· | ($)παντὶ, and the peculiar metrical effect may well be intended.304 139. πατὴρ ὁ παντόπτας: cf. 86–103 with notes. 141–43 = 151–53. σπέρµα … ἐκφυγεῖν: pace FJ–W, the infinitives of these ephymnia appear to be syntactically dependent on the previous strophes, defining τελευτὰς πρευµενεῖς κτίσειεν and ἀδµῆτας ῥύσιος γενέσθω, respectively.305 If we are to nitpick about grammatical terms, 302 δόρει (or δορί) as an instrumental dative might make the expression easier but is hardly necessary. For the corruption, however, cf. 147 σθένει (σθενοσ Mac). 303 Th. 735–36~743–44, 767~773, Ch. 44–45~56–57, 81, Pr. 695. 304 Stinton suggests πνοαῖσιν here and ἅπαντι for παντὶ in the antistrophe (147), together with Heath’s (1762) ἀσφαλῶς. Both emendations are anticipated by Westphal (1869, 160). 305 FJ–W contend that all ephymnia in Aeschylus are ‘metrically and syntactically separated from the context’ (see their note on 117–22 = 128–33), but there is no foundation for such a claim. First, the Aeschylean ephymnia are too few and too different 110 ‘final-consecutive’ infinitive (without ὥστε) is preferable to ‘epexegetic’ (FJ– W).306 144–50. Artemis in invoked in her capacity as the goddess of maidenhood. She is not mentioned by name until 676, where she is invoked in another aspect, that of the goddess of childbirth; but she is subsequently called upon as ἁγνά at 1030. Artemis has an Egyptian alias, Bast: see 204–24n. with n. 359 on the Egyptian influence on the Danaids’ religion. Paus. 2.19.7 mentions that Danaus consecrated two wooden images of Zeus and Artemis at Argos. 144. θέλουσα … θέλουσαν: On the polyptoton see FJ–W ad loc. and 149n. 146. ἐνώπι(α) must primarily mean ‘countenance’, ‘face’, which is exactly what would be expected in an image where the goddess is ‘looking safely over’ her worshippers. So the scholium and also, for instance, Wecklein– Zomaridis, Mazon, FJ–W.307 Cf. LSJ s.v. ἐνωπή, ἐνώπιος, µετ-, ὑπώπιον. The meaning ‘temple’ has, contrary to vulgate opinion, little or no support from the extant appearances of the noun, which elsewhere only occurs in the Homeric formula ἐνώπια παµφανόωντα.308 In Homer, the word does not among themselves in other respects to make for any sort of reliable statistics, especially if one is prepared, as FJ–W apparently are, to make exceptions among them: ‘the repetition of one at the end of an epode, Ag. 159, is a special case’. Secondly, there is no apparent metrical independence in most cases, as is evident from FJ–W’s discussion on 117–22 (ii. 104): ‘metrically speaking, Aeschylean ephymnia are usually harmonized to a certain degree with their context’, the only independence being ‘the occurrence of cola which are rather abnormal within their category’. Thirdly, Eu. 328–33 = 341–46 and fr. 204b.6–8 = 15–17 are not syntactically independent insofar as they are connected to the previous passage by particles, δέ and δέ τοι, respectively. Just as in the case of stichomythia (see on 290–323, n. 463), we are justified in allowing a certain degree of artistic freedom as regards the ephymnia, as Kranz (1933, 131): ‘ein merkwürdiges Beispiel für die Freiheit dichterischer Anreihung des Refrains bieten Hik. 141 und 151: das Gebet an zwei verschiedene Götter … geht in denselben Wunschrefrain über, ohne gedanklich und sprachlich sich vom Vorhergehenden zu lösen’. 306 S.GG ii. 362–65 (especially 365), K–G ii. 3, 16–17; cf. also some of the examples ibid. pp. 7–8, 12–13. 307 To their parallels of ‘reverend countenances’ may be added X. Smp. 3.10.2 and, in malam partem, for example E. Alc. 773, Amphis fr. 13 PCG, Com.adesp. 1105.180 PCG (= Stratt. fr. 220.180 Austin). 308 Il. 8.435, 13.261, Od. 4.42, 22.121, and possibly Alc. 58.17, where FJ–W guess at the meaning ‘face’. Hesychius’ second entry of the word (ε 70) is probably derived from grammarians’ speculation on the Homeric passages. 111 mean ‘wall’, but refers to a particular sort of face, viz. ‘façade’: white plaster applied to a brick wall where it is in need of protection from the weather (Lorimer 1950, 428, n. 1), the characteristic of the archetypal Mediterranean house. Aeschylus simply uses the word in a more general sense (cf. on 21–22 ἐγχειρίδιον). Conversely, πρόσωπον can mean ‘façade’ (Pi. O. 6.3, P. 6.14). There may perhaps be a hint at a building, i.e. the safety of a temple in the case of Artemis; but this cannot be the primary meaning (and there is certainly no particular temple intended). Cf. 8n. on verbal amphiboly. ἀσφαλής: M’s ἀσφαλές is probably corrupt. To take it, with the scholium, with ἐπιδέτω (ἀσφαλῶς ἐπιδέτω µε) ought to be impossible on account of the distance between the words as well as of the metrical period-end after κόρα. Adverbial with ἔχουσα it is even more awkward: in the parallels adduced by Verdenius (1985), Il. 15.683–84 and Pi. P. 2.20 (cf. FJ–W on 146 ἐνώπια), ἀσφαλές is an internal accusative with an intransitive verb. We should accordingly choose between Heath’s (1762) ἀσφαλῶς, Sidgwick’s ἀσφαλῆ, and Young’s (1974) ἀσφαλής.309 The last is the most economical, and one might perhaps discern a stylistic device in the abundance of nominative attributes with the goddess here: the pairing of nominative participles with adjectives in emphatic positions in the sequence θέλουσα … ἁγνά, ἔχουσα … ἀσφαλής, †participle† … ἀδµήτα. 147. παντὶ σθένει: often used in official treaties of alliance (ἐπικουρεῖν παντὶ σθένει, etc.): see FJ–W and above, 40–175n. and n. 150. 148. †διωγµοῖσι δ’ ἀσφαλέασ†: if we are facing a dittography here of 146 ἀσφαλές, palaeography will be of little use in determining the probability of any conjectural reading. See FJ–W for a sound analysis of the textual corruption. Dittography may not be absolutely certain, as one could argue that a repeated form of ἀσφαλής would accord with the two other repetitions of adjectives/participles that appear in this strophe (144 θέλουσα … θέλουσαν, 149 ἀδµῆτας ἀδµήτα, q.v.). Thus Bücheler’s (1885) διωγµοῖς ἀσφαλέας would be a fairly attractive solution, with ἀσφαλέας taken, like ἀδµῆτας in 149 (q.v.), as a direct object of ῥύσιος γενέσθω: cf. 209 οἴκτιρε µὴ ἀπολωλότας with my note.310 309 Hermann’s Ἄρτεµις is less economical and over-explicit. See also K–G i. 296, S.GG ii. 73 and, e.g., Th. 289–90 τάρβος τὸν … λεών with Hutchinson’s note. 310 112 The dative διωγµοῖς then becomes very awkward, however, and from the context one would expect the same type of ‘dualistic’ prayer here as in 26–36 and 79–81 (cf. 89–90, 96–98): safety for the girls and aggression against their pursuers. Thus a nom. participle or adjective conveying the latter sense, taken with the dative διωγµοῖς (-ι δ’ is unmetrical), is a more attractive solution: see also above on 146 ἀσφαλής. Hermann’s ἀσχαλῶσ’ has been widely accepted; but FJ–W rightly observe that ἀσχαλᾶν + dat. elsewhere does not mean ‘be angry with someone’ but ‘be upset or vexed over something’, and vexation is not an emotion one would associate with gods (cf. 98–103n.), nor does it accord well with the adverbial παντὶ σθένει (‘upset with all her might’). It might be that Aeschylus has taken ἀσχαλῶσ’ in an eccentric meaning (cf. on 49 ἐπιλεξαµένα, 21 ἐγχειριδίοις); on the other hand any participle or adjective meaning ‘hostile to’ or the like, possibly beginning with α (or δ if the transmission of this letter is sound, as FJ–W suspect), could be right.311 διωγµοὺς διεφθορυῖ’ could explain the corruption -οῖσι δ’ but obviously calls for much special pleading. 149. ἀδµῆτας ἀδµήτα: whichever emendation we choose for M’s unmetrical ἀδµήτασ ἀδµήτα, it appears that the adjective must take either a different declination in each of the two instances (ἀδµῆτας ἀδµήτα, ἀδµῆτος ἀδµήτα312), or two generic endings in the first case and three in the second (ἄδµητος ἀδµήτᾳ, ἄδµητος ἀδµήτας313). The juxtaposition of thematic and athematic forms has parallels in Euripides in the formula δάκρυα δάκρυσι, however. FJ–W and West have adopted de Pauw’s ἀδµῆτος; but Westphal’s solution, the acc. pl. of the third declination, is easier.314 ἀδµῆτας ἀδµήτα answer chiastically to θέλουσα … θέλουσαν in 144. The change from singular 311 Thus for instance ἀντιβᾶσ’ (Jurenka 1900, cf. Pr. 234) and Wecklein’s ἀλκαθοῦσ’ (1893, 333), neither of which seems particularly attractive. 312 Westphal (1869, 160) and de Pauw, respectively. 313 de Pauw and Dindorf (1858, 498), respectively; however, the third-declination adjective always has three endings elsewhere, and the adjective cannot here be attracted to the following ῥύσιος, which is also an adjective. 314 So, e.g., Wecklein (ed. 1902), Wilamowitz, Verdenius (1985). For ῥύσιος with a direct object, see above on 148 and n. 310. ἀδµῆτος would be somewhat confusing: in the absence of pitch accents (cf. on 72), the audience might as well take the word as nom. ἄδµητος (and Aeschylus would not have differentiated the spelling of the forms). 113 (θέλουσαν … µ’) to plural is of no consequence at all, pace FJ–W: cf., apart from the instances they themselves adduce, 602–3 and 1008. 154–75. These lecythia and ithyphallics are distinct in mode from syncopated iambics (as labelled by Dale 1971): the double long (š) is absent, and the starts are falling throughout (except at 156 ~ 170). 154–61. This threat of suicide is the first clear indication of the undercurrent of violence and aggression in the minds of the Danaids. To threaten the supreme godhead with suicide if he does not grant a prayer seems like utter blasphemy to one raised in a Christian culture, and it would not be unproblematic to a pious heathen (such as Aeschylus). FJ–W suggest that the threat is an indication of ‘a kind of close “Homeric” relationship between men and gods’ in the Supplices, but this goes against what is said about Zeus earlier in the ode (86–103). Indeed Zeus has always, even in Homer, kept his distance from the humans, apart from the occasional mating. The threat is rather an indication of the hybris of the girls, giving an ironical twist to their recurring complaints about this fault in their suitors (31, 81, 104, 426, etc.). Suicide or the threat thereof is relatively often depicted as the woman’s prerogative in Greek antiquity. Hanging is the preferred feminine method, with 44 instances in myth and history of females hanging themselves, against 33 males.315 154–55. µελανθές … γένος: see Hall (1989) 139–43 on skin colour and masks in Greek drama, including this one. 155. ἡλιόκτυπον: the compound is hardly ‘audacious’ (FJ–W), as κτύπος and its compounds do not ‘invariably signify noise’ (pace FJ–W, Tucker, cf. Housman 1890, 106b), but are often found in the transferred sense ‘beat’.316 Cf. 71n. 156. γάϊον: Wellauer’s palaeographically easy but not altogether unproblematic emendation of M’s vox nihili ταιον. It may have been the reading of the scholiast: τὸν καταχθόνιον Ἅιδην (but see below). However, γάϊος, in 315 On suicide in antiquity, see van Hooff (1990) with useful statistical surveys at pp. 198–242; on suicide in tragedy, see Garrison (1985) with further refs at pp. 1–2, n. 1; on female suicide in antiquity, see van Hooff (1992). 316 See Verdenius (1985). The sun beats in biblical Hebrew (and Greek), too: Jn. 4.8, Is. 49.10, Ps. 121.6; and in English, perhaps by biblical influence: Spenser, The Shepheardes Calender, Aug. 47: The Sunnebeame so sore doth vs beate; T. S. Eliot, The Waste Land 1.22: A heap of broken images, where the sun beats. 114 the sense of (κατα-)χθόνιος seems to be unparalleled (cf. on 24–25 above). ἔγγαιος does appear in this sense in AP 7.480.7 (Leon.),317 and γῆ is often used as a synonym of χθών in tragedy (see FJ–W, Schuursma 1932, 72–73). On the other hand, some external evidence indicates that the reading may have been a word of the stem ζαγρ-. Thus Et.Gud. s.v. ζαγρεύς ≅ An.Ox. ii. 443 (derived from an article by the grammarian Seleucus according to Reitzenstein 1897, 172–73): Ζαγρεύς· ὁ µεγάλως ἀγρεύων, ὡς ‘πότνια Γῆ, Ζαγρεῦ τε θεῶν πανυπέρτατε πάντων’ ὁ τὴν Ἀλκµαιονίδα γράψας ἔφη [fr. 3]. τινὲς δὲ τὸν Ζαγρέα υἱὸν Ἅιδου φασίν, ὡς Αἰσχύλος ἐν Σισύφῳ ‘Ζαγρεῖ τε νῦν µοι καὶ πολυξένῳ χαίρειν’ [fr. 228]· ἐν δὲ Αἰγυπτίοις [fr. 5] οὕτως αὐτὸν τὸν Πλούτωνα καλεῖ ‘τὸν †ἀγραῖον, τὸν πολυξενώτατον, τὸν ∆ία τῶν κεκµηκότων’. 1 Ζαγρεῦ] αγρεῦ Et.Gud. d1, corr. d2 4 Αἰγυπτίοις] Αἰγύπτῳ Et.Gud. 5 post καλεῖ lacunam ci. Hermann (1846–47, 125 [181–82]), suppl. <ἐν δὲ Ἱκέτισι τὸν ∆ία> : <ὃν ἐν Ἱκέτισι> Wecklein : plura de verbo πολυξένῳ excidisse ci. Radt †ἀγραῖον] Ζαγρέα de Stefani : Ζαγραῖον Taplin (1977, 197) : γάϊον Hermann l.c. post Wellauer ad. Supp. 156 : ἄγριον Wilamowitz p. 379 : ζάγριον Schneidewin (1836) 6 ∆ία Welcker (1824, 557, n. 363) : Ζῆνα Hermann l.c. : διὰ Et. Gud. : δῖα An.Ox. As this witness318 is even more corrupt than the text of the Supplices, it is impossible to draw any certain conclusions. We ought probably to rule out the idea that virtually the same words (τὸν πολυξενώτατον κτἑ) appeared in another play in the trilogy, whether called Αἰγύπτιοι or Αἴγυπτος (cf. G.AS 189). The grammarian did quote the Supplices; but either (1) he got the title of the play confused (so Wilamowitz l.c.), or (2) he used Αἰγύπτιοι as the title for the whole trilogy (Welcker l.c.), or (3) there is a lacuna somewhere between the two quotations from Aeschylus (Hermann l.c., Radt). Only in the last case may we discard the passage as evidence for our text, i.e. if we, with Radt, assume that the comparison between the quotations concerns the word 317 Gow–Page ad loc. compare Plu. Pr.frig. 953a τὸ χθόνιον καὶ ἔγγαιον σκότος (= Erebus). 318 On which see, in particular, G.AS 188–89 and Radt pp. 125–26, 339 (on A. frr. 5, 228). 115 πολύξενος, not Ζαγρεύς, and that a discussion of the former word has fallen out. Then †ἀγραῖον may easily be taken as a misquotation or a corruption of γάϊον, irrelevant to the constitution of the text of the Supplices. The position of the adverb οὕτως and the fact that Seleucus’ article as a whole concerns the name Ζαγρεύς make this seem somewhat far-fetched, however. It is hard to avoid the conclusion that the grammarian did indeed read some word beginning with ζαγρ- in Supp. 156.319 Schneidewin’s (1836) ζάγριον is thus attractive, both in the witness and in our text.320 The adjective, although unparalleled,321 might agree with the scholium to our passage, τὸν καταχθόνιον ᾍδην, which has usually been interpreted as evidence that the scholiast read γάϊον. Hesychius glosses ζάγρη with βόθρος and λάπαθον: a hole in the ground. The epithet Ζάγριος might refer (by popular etymology) not to hunting but to the ritual of libation: βόθρος is a pit for this purpose in, for instance, Porph. Antr. 6, Phil.orac. 114, 118, 121 (read τοῖς δ’ ὑποχθονίοις for the ms. ἐπι-), and several times in Homer’s νέκυια (Od. 10.517, 11.25–95 passim). Thus ζάγριος would be virtually synonymous to χθόνιος.322 Hesychius s.v. equals Ζαγρεύς with χθόνιος ∆ιόνυσος, and Plutarch may actually connect the name with the element of earth (E.Delph. 388f–389a). GEW and DE dismiss ζάγρ- = *διάγρ- as a popular etymology, perhaps rightly (but pace Wilamowitz 1931, 250), but the early popular interpretation might not have been the ‘great hunter’ of the Et.Gud. but rather ‘through the ground’, i.e. formed from ἀγρός rather than ἄγρα,323 the epithet being a virtual synonym to χθόνιος. Cf. Οὐδαίῳ Ζανί in AP 14.123. 14 (Metrod.). Still, the evidence is inconclusive. ζάγριον is compelling in many ways, especially because of the witness of Seleucus, which is hard to explain away. On the whole, however, the lack of clear support for this adjective as well as the palaeographical easiness of Wellauer’s γάϊον should tip the balance in favour of the latter. †ἀγραῖον in the etymologies does imply that Seleucus 319 de Stefani observes that the corruption to ἀγραῖον would resemble that in Et. Gud. recension d, where the Ζ of Ζαγρεῦ has fallen out in the quotation from the Alcmaeonis. 320 For the corruption of ζ into τ, see FJ–W 194n. 321 Apart from what appears to be a curse ζάγριον (ζατρεῖον Meineke iv. 595) in Timostr. fr. 4 PCG (ap. Lex.Seg. Antatt. i. 98 Bekker = AB i. 98), and an unrelated (?) epithet of a mountain in Str. 11.12.4.26, 28. 322 Cf. also [Zonar.] s.v. ζαγρός· ἀνυπόδετος. 323 On the relationship between the two, see Chantraine (1956) 31–65. 116 read ζάγριον (-εα, -εον, or the like) in his text of the Supplices, but there is nothing that prevents this from being itself a corruption of γάϊον. 157. πολυξενώτατον is darkly ironic: nobody escapes the ‘hospitality’ of Hades. For parallels see FJ–W, who also observe that there is an ironic reference to the later invoked (627) Zeus Xenios. 158. Ζῆνα τῶν κεκµηκότων: ‘Zeus of the dead’ is Hades: cf. Il. 9. 457, Ag. 1386–87, etc. See also above on 24–25. 159–60. σὺν κλάδοις ἀρτάναις θανοῦσαι: by the juxtaposition of κλάδοις and ἀρτάναις the image is invoked of the Danaids using their nooses as suppliant boughs before Hades: cf. 21–22. As in that passage, the symbolic imagery hints that the supplication of the girls conceals violent, aggressive resolution. 161. µὴ τυχοῦσαι θεῶν Ὀλυµπίων: apparently a fusion of two common meanings of τυγχάνω: ‘succeed’ (LSJ s.v. B.I) and ‘obtain a thing from a person’ (B.II.c). Cf. E. Hipp. 328 with Barrett’s note. As often, the Olympians are contrasted to the chthonic gods (see on 24–25). Editors have adduced Turnus’ line in Verg. Aen. 7.312 flectere si nequeo superos Acheronta movebo: Turnus as well as the Danaids would turn to Hell where Heaven is unfavourably inclined.324 162–67. The stanza has been repeated after the antistrophe by a large majority of editors since this measure was first suggested by Canter.325 I cannot see any real justification for this, however. As FJ–W state, ‘whether this repetition accords with general principles of structure either in Aeschylus or in tragedy at large cannot be certainly inferred from the few and scattered lyric passages in and outside Aeschylus where ephymnia are unambiguously attested’ (175n.: see also my note on 117–22 = 128–33). In Ag. 1448–1566 and Eu. 321–96 we find mesodes as well as ephymnia within the same choral odes, and there is no reason why this should be impossible here. For positive reasons to follow the ms. (i.e. apart from mere conservatism) see G.AS 77, who observes that one feature of ring-composition326 in the ode is the parallel between the participles at the beginning (40–41) νῦν δ’ ἐπικεκλοµένα ∆ῖον 324 Embrace of Hell where Heaven is lost is probably best known from Milton (Paradise Lost 1.261–63): Here may we reign secure, and in my choice | To reign is worth ambition though in Hell: | Better to reign in Hell, then serve in Heav’n. 325 Notable exceptions: Porson, Hartung, Wecklein (ed. 1885). 326 See G.AS 74–78 with references at 74, n. 1. 117 πόρτιν and at the end (175) εὖ κλύοι (Zeus) καλούµενος. This parallel would be weakened by a repetition of the mesode after 175. In fact, 175 is a very effective conclusion to the entire choral ode, comprising as it does its very essence (as a prayer) in one sentence. We may also observe that 162–67 is different in tone and content from the two previous ephymnia: instead of an explicit supplication or cry for help, we find a brooding lament on the fate of Io (see ad locc.). See further G.AS 43–44 with refs. As in the case of the ephymnia, the mesode has irregular metre. It is notable that the double-long, ‘dactylic’ or ‘ionic’ rhythm recurs in 165–67. 162. ἆ Ζήν, Ἰοῦς, ἰώ: a sort of quasi-etymological word-play is suggested by the chiastic arrangement of names and interjections similar to each other,327 which makes Bamberger’s (1839) Ζάν attractive. In Homer ἆ always expresses pity (albeit sometimes condescending and hypocritical), invariably appearing in the formula ἆ δειλοί (δειλέ, etc.). Later the interjection conveys a broader range of emotions, but here pity appears to be intended with Io as the victim of Hera’s wrath.328 Thus punctuation with a comma after Ζήν is preferable, I think, to West’s colon. 163. µῆνις µάστειρ’: the wrath, in particular from Hera, that followed Io on her long flight. µάστειρα, the feminine form of µαστήρ, is only attested here. The scholiast’s paraphrase, ἡ παρὰ τῶν θεῶν µῆνις κατὰ Ἰοῦς †ὠδῆς [ἰωδής Hermann] ἐστὶ καὶ µαστιγωτική, either understands the adjective as derived from µάστιξ (cf. Homeric forms µάστι, µάστιν and µαστίω) or had something like µάστῑρ’ (Hermann with doubt) or µαστίκτειρ’ (Abresch 1763: cf. 466) in the text. 164. κοννῶ: see on 118–19 with n. 287. ἄγαν is Bamberger’s (1839) probable conjecture for the ms.’ ἄταν. The noun is also generally read in Ag. 131 (Hermann ap. Humboldt, p. 84; ἄτα codd.) and certainly attested in fr. 85 (Hesychius, glossing the dat.pl. with ζηλώσεσιν). The sense in malam par- 327 Wiel’s (1858) ἰῶ µῆνιν µαστῆρ’ (‘heal the inquisitorial wrath!’) is clever, but a direct request to Zeus here would hardly be in place between these two strophes, which contain direct accusations and threats against him. 328 See Sandin (2002) 149 for examples of compassionate ἆ after Homer. ἆ ἆ is a different case, expressing alarm, pain, or protest (to be distinguished from ἃ ἅ, laughter in E. Cyc. 157 and perhaps in Ba. 586, 596: see LSJ, Hsch. α 2). This may also be expressed by ἆ µή (on which see Barrett on E. Hipp. 503–4). 118 tem is found in Hdt. 6.61 and EM 8.49, too, whereas the word in Homer always means ‘amazement’, ‘wonder’. 165. The sense is fairly clear, and several conjectures have been put forward that would produce a sound text, the most expedient of which is perhaps Victorius’ simple γαµετᾶς οὐρανόνικον. The adjective, ‘heavenconquering’, is next to certain, and it may have been the reading of the scholium.329 The ‘οὐρανόνικον malice of a wife’ may not only refer to Hera having her way among the Olympians, but also contain a learned hint at the castration of Uranus by the design of his wife Gaia (Hes. Th. 159 ff.). The unqualified, ‘ambiguous’ (Paley) γαµετᾶς would thus be defensible as referring not only to Hera but to the malice and jealousy of (divine) wives in general. 166–67. χαλεποῦ γὰρ … χειµών: the sequitur is somewhat obscure: Weil’s δ’ is duly noted by West. The notes of Paley and Tucker might be useful: ‘the chorus speaks of Juno’s anger as a “breeze,” meaning that further troubles await them from this manifestation of it’ (Paley). 168–74. καὶ τότ’ οὐ … λιταῖσιν; Headlam (1892) was the first to see that the sentence must be formulated as a question. Rogers’ (1894) αὖ for οὐ in 168 was adopted by Page, but a positive statement in the indicative would be blunt; one would have expected the potential mode. 169. Porson’s ἐνέξεται for the ἐνεύξοµαι of the ms. is certain. Headlam’s (1892) ψόγοις is likely, as well. The ms. reading λόγοις seems weak and unAeschyleanly abstract (‘argumenta’ IA), although FJ–W argue that δικαίοις λόγοις would be understood as a forensic term by the audience: ‘then will not Zeus be liable to just pleas …?’. But the idea of Zeus being liable, i.e. standing trial, is absurd (pace FJ–W 154–61n.), and δικαίοις λόγοις is too general to convey this meaning. In legal contexts, ἐνέχοµαι is found with words like ἀρᾷ, ζηµίᾳ, αἰτίᾳ, ἐπιτιµίοις, νόµοις (LSJ). ψόγοις assumes an easy corruption (other examples are listed by FJ–W), and it might perhaps be supported by the echoes of Xenophanes established in 100–103 and other places in Aeschylus (see 86–103n. with the refs in n. 248). Xenophanes’ perhaps most famous couplet is fr. 11, where he speaks of Homer and Hesiod relating such deeds of the gods ὅσσα παρ’ ἀνθρώποισιν ὀνείδεα καὶ ψόγος ἐστίν. Unlike those (imaginary) stories of adultery and gluttony, would not this, the deserting of ﬁfty pious maidens, his own progeny, 329 τὴν τῆς Ἥρας τῆς ἐν ἀνδρὶ νικώσας πάντας τοὺς ἐν οὐρανῷ θεούς. 119 be a cause for δίκαιοι ψόγοι, ‘rightful censure’? It is exciting to imagine the passage as an allusion to a contemporary theological debate: can there be such a thing as justified censure of the gods?330 Euripides is likely to have had the passage from Xenophanes in mind at HF 1341–46 (so Lucas on Arist. Po. 1461a1). For rightful or legitimate censure, cf. also Pl. Smp. 182a ψόγον ἂν δικαίως φέροι. 172. γόνῳ: here ‘begetting’, ‘engendering’. This sense is not found elsewhere in tragedy, but is attested in classical prose. Portus’ γόνον may be worth considering. 175. καλούµενος: see 162–67n. 176–233. Danaus opens his mouth for the first time, having been standing in silence throughout the first choral ode. His daughters are given detailed instructions about how to act in the upcoming encounter with the Argives. The present scene has several similarities to 710–824: both are exceptionally long announcements of new arrivals on the stage, and both are dialogues between Danaus and his daughters about how to deal with a threat imposed by an approaching encounter (see Taplin 1977, 199–200). 176–78. φρονεῖν … φρονοῦντι … προµηθίαν are key-words in the mouth of Danaus, establishing his role as cautious adviser to his daughters. This is his only consistent, prominent character trait as well as his sole function in the drama. Character is to a very large degree dependent on dramatic function in Aeschylus, seldom consistent, rarely if ever with psychological depth.331 However, one cannot say that character is altogether non-existent as an independent factor: tradition, in the sense of the traditional, ‘ready-made’ 330 Gantz (1981) deals with the issue of divine morality in Aeschylus, taking the (reasonable) stand that it is the Danaids, not the gods, who are at fault here (p. 18, n. 6). The issue is also concisely addressed by Dover (2000), who notes a number of respectable people attributing moral blame to the gods while still actually believing in them and their actions: cf. Solon ap. Hdt. 1.32 and especially Thgn. 1.731–52. 331 The notion that (consistency of) character is subordinate to action in Greek drama was originally drawn up by T. von Wilamowitz (1917), who mainly concerned himself with Sophocles; Jones (1962) passim (see esp. pp. 18–20, 29–46, and 81–110 on the Oresteia) and Dawe (1963) developed the idea with a focus on Aeschylus. On Danaus in particular, see Lloyd-Jones (1964a) 368–69 (273–74). Easterling (1973) presents a more ﬂexible, less dogmatic view on presentation of character in Aeschylus, as does Lloyd-Jones (1964b) in his review of Jones (1962). See G.AS 132, nn. 1–4, for further bibliography. 120 characters of myth and the Homeric example, did offer a paradigm for a measure of independence in characterisation. In the case of Danaus, his wisdom and intelligence may be traditional: see the Introduction, II 1, and 134–35n., 320n., 496–99n. Garvie (G.AS 135–38) observes that Danaus, with the chorus taking on an unusually important dramatic role, is himself called on to perform some of the duties that normally fall to the lot of a chorus, not least the utterance of gnomes.332 The character of these conventional words of wisdom may also serve to delineate character: see Arist. l.c. (n. 302), and Lardinois (2000) on this use of gnomes in Homer. 177. ναυκλήρῳ: see 134–35n. 178–79. λαβών (Wordsworth 1832, 211) is preferable to the paradosis λαβεῖν, which would necessitate reading θ’ ἅµ’ (Heath 1762) in 179 and, contrary to the context, taking προµηθίαν … λαβεῖν to refer to the Danaids. Danaus has himself ‘taken precautions’ (see on 176–78 above), and the advice to his daughters is given after his own careful survey of the situation and the 332 See G.AS l.c. (with generous refs) for an analysis of Danaus’ role in the drama, and also Arist. Rh. 1394a–95b (2.21.1–16) on gnomes. To call these words of wisdom ‘platitudes’ (G.AS 137) and ‘not … profound’ (ibid. 138) is anachronistic and unfair: the Greeks were not, as we, hypersensitive to cliché. Sayings such as ‘the altar is greater protection than the fortlet’ (Supp. 190) and ‘bold talk behoves not the lowly’ (203) are not clichés to Aeschylus but simple, unchanging truths, such as constitute the κόσµος, the backdrop of order against which the δράµατα, the temporary outbursts of chaos, manifest themselves. Aristotle acknowledges the usefulness of common gnomes and encourages their use in rhetoric (Rh. 2.21.11): χρῆσθαι δὲ δεῖ καὶ ταῖς τεθρυληµέναις καὶ κοιναῖς γνώµαις, ἐὰν ὦσι χρήσιµοι· διὰ γὰρ τὸ εἶναι κοιναί, ὡς ὁµολογούντων πάντων, ὀρθῶς ἔχειν δοκοῦσιν. While acknowledging the comic effect of gnomes in rustic characters, he puts it down not to banality but to the propensity of self-important people to expatiate on things they know little about (2.21.9): ἁρµόττει δὲ γνωµολογεῖν ἡλικίᾳ µὲν πρεσβυτέροις, περὶ δὲ τούτων ὧν ἔµπειρός τίς ἐστιν, ὡς τὸ µὲν µὴ τηλικοῦτον ὄντα γνωµολογεῖν ἀπρεπές …, περὶ δ’ ὧν ἄπειρος, ἠλίθιον καὶ ἀπαίδευτον. σηµεῖον δ’ ἱκανόν· οἱ γὰρ ἀγροῖκοι µάλιστα γνωµοτύποι εἰσί. He states that one advantage of gnomes is their appeal to vulgar opinion (2.21.15): ἔχουσι … βοήθειαν µεγάλην µίαν µὲν δὴ διὰ τὴν φορικότητα τῶν ἀκροατῶν. Their greatest virtue, however, is not so Machiavellian: ἠθικοὺς γὰρ ποιεῖ τοὺς λόγους. […] αἱ δὲ γνῶµαι πᾶσαι τοῦτο ποιοῦσι διὰ τὸ ἀποφαίνεσθαι τὸν τὴν γνώµην λέγοντα καθόλου περὶ τῶν προαιρετῶν, ὥστ’ ἂν χρησταὶ ὦσιν αἱ γνῶµαι, καὶ χρηστοήθη φαίνεσθαι ποιοῦσι τὸν λέγοντα. This should hold water not only for rhetoric, but also for dramatic poetry: the purpose and effect of Danaus’ gnomes is to show him to be χρηστοήθης, a good man. 121 surroundings, the results of which are presented in 180–203. It is natural, after the stately introduction of himself in 176–78, that Danaus should do something of his own, rather than immediately begin handing out advice. See FJ– W, who also observe that the contrast between τἀπὶ χέρσου προµηθίαν λαβ. and ναυκλήρῳ is more effective if both the action and the epithet refer to the same person. 179. δελτουµένας: not on notepads, but on the writing-tablets of the mind. See FJ–W for classical parallels, and also West (1997) 561–62, who finds several Old Testament instances of ‘tablets of the heart’—not, however, with the original form of the noun δέλτος although this is a Semitic loanword, with the same origin as the letter δέλτα.333 180–81. ὁρῶ κόνιν: the asyndeton is common after a call for attention where a speech, long or short, is to follow (i.e., at the beginning of a speech). Cf., e.g., E. Supp. 518, Ar. Ach. 1000. Formally, it could perhaps be said to be ‘explicative’—of the reason, that is, for the call for attention (FJ–W 181n., ii. 147).334 σύριγγες οὐ σιγῶσιν: to call this asyndeton ‘explanatory’ is farfetched.335 Rather, its sense may be adversative (K–G ii. 342): the cloud is speechless, but the hubs do not keep silence. So the asyndeton at 86–87, by aid of the anaphora εὖ θείη — οὐκ εὐθήρατος.336 333 Cf. Hebr. daleth [‫]דּתל‬, ‘door’; also ‘column of text’ in Je. 36.23, and possibly ‘hinged writing-tablet’ in Hebr.Inscr. 1.004.3 Davies: see Clines–Elwolde (1995) 442–43, Diringer–Brock (1968) 42–43. 334 Some of the examples of ‘descriptive’ asyndeta in K–G ii. 340 are to be referred to this category: Il. 22.450–51, 294–95. 335 ‘…motivating the fact that the cloud, in spite of its speechlessness, announces the approach of an army’ (Verdenius 1990). This is awkward, making the present line semantically subordinate to the former. The visual and audible signs are treated as equally important indications of the approaching contingent. FJ–W defend Enger’s (1854a, 397) σιγῶσι δ’, claiming that asyndeton is impossible ‘because [it] cannot be either explicative … or inceptive …; nor can it fall into Denniston’s dubious category of “emotional” asyndeton’. Apart from the fact that these categories of asyndeton are not exhaustive, Denniston nowhere speaks of ‘emotional’ asyndeton as a separate category: the term he uses is ‘stylistic’. Although he mentions in passing (D.GP p. xlv) that ‘stylistic … asyndeton is used … for emotional effect’, it is clear from the examples in Denniston (1952), chapter 6, that emotion is inadequate as a defining quality of these asyndeta. Cf. the examples ibid. pp. 116–18, 121–23, and also K–G ii. 340–42. 336 A negation in the latter clause often seems to yield as it were a pseudo-adversative effect even when one cannot speak, as K–G l.c., of a ‘Gegensatz’. So in the cited pas- 122 183. καµπύλοις: another epicism. Rather than effecting a certain poetic style, the intention may be to lend an authentic ‘antique’ flavour to the action: cf. 15n., 117 = 128n., 236–37n. 184–85. τάχ’ ἂν … ὀπτῆρες εἶεν: a pregnant construction of the type exemplified in K–G i. 543 (§447.B.a). Cf. 189 πάγον προσίζειν. No exact parallel has been found, but K–G list four examples of pregnant expressions with verbs of observation337 and several others which are themselves unparalleled (cf. esp. Th. 4.57, Is. 5.46). A lacuna after 184 (O. Foss ap. Friis Johansen– Whittle 1975) is therefore unnecessary. For the sake of dramatic tension πρὸς ἡµᾶς should refer to the movement of the Argive party, pace Verdenius (1990) who suggests that the expression is equivalent to ὁρᾶν πρός τινα. The asyndeton is explanatory (K–G ii. 344–45). 186–87. τόνδ’ ἐπόρνυται στόλον: the masculine singular subject refers back to ὄχλον in 182.338 On this account, the scholiast’s reading of τόνδ’ … στόλον as an internal accusative of ἐπόρνυται339 is awkward, as στόλον, the internal acc., seems to refer to the same thing as ὄχλον, the implied subject. We found a similar problem in 134–35 λινορραφής … δόµος ἅλα στέγων δορός. The solution here may also be similar to the one proposed in that case: στόλον takes on a somewhat more abstract sense than usual, ‘expedition’ or ‘mission’ (FJ–W, cf. LSJ s.v. I.2.b) instead of simply ‘party’. If στόλος always refers to a group of people elsewhere in Aeschylus, it never means simply ‘party’, just as στάσις does not (see 12n.): the word refers to a party gathered for a specific purpose. στόλον here may still refer to much the same thing as ὄχλον in 182 but in a slightly narrower sense, which makes the sentence coherent, if a little awkward.340 sages, cf. D. 1.13 τοὺς µὲν ἐκβαλὼν τοὺς δὲ καταστήσας τῶν βασιλέων ἠσθένησε· πάλιν ῥᾴσας οὐκ ἐπὶ τὸ ῥᾳθυµεῖν ἀπέκλινεν, ἀλλ’ εὐθὺς …, Il. 2.24, 2.745, Pi. O. 8.45, I. 2.7 (with anticipatory γάρ?), B. 11.14, E. Cyc. 322. 337 Pl. Phdr. 268a, E. Hec. 1154, Il. 3.154, Od. 1.411, in all of which cases, however, the implicit movement is that of the object, not the observer. 338 So, e.g., Tucker, Untersteiner (ed. 1935), Verdenius (1990). 339 Cf. S. Ph. 1037–38 οὔποτ’ ἂν στόλον ἐπλεύσατ’ ἂν τόνδ’, X. Cyr. 8.6.20 ὥρµα … τὴν στρατείαν and the similar examples in K–G i. 307. 340 The conjectural attempts have been unsatisfactory, except possibly Kraus’s ἐπ’ ὄρνυται, if τόνδ’ … στόλον is understood to refer to the Danaids. For the sequence adj.–prep.–verb–noun, cf. h.Ven. 122 πολλὰ δ’ ἐπ’ ἤγαγεν [sc. µε] ἔργα. But the elided anastrophic preposition would be awkwardly ambiguous: see Friis Johansen– 123 189–90. This passage, among a number of others (79–85, 222–23, 345–46), seems to indicate that the altar is located rather close to the idols: i.e. at the rear rather than at the middle of the orchestra (see the Introduction, III 5). 189. πάγον: the cliff has not been mentioned before, and there is chaotic disagreement among scholars as to what, if anything, represents it on the stage. See the Introduction, III 4–5 for an assessment of the various theories: we will have to be content with the fact that an elevation of some kind is present. There are a number of statues or busts upon this elevation; Danaus may also have been standing on it the whole time since his entrance on stage (see 1–39n., 208 with note). τῶνδ’: if the demonstrative pronoun is attracted to θεῶν it is little cause for wonder, as the gods are important and the πάγος is not.341 Cf. 222 ἀνάκτων τῶνδε κοινοβωµίαν and also Verdenius (1990). ἀγωνίων θεῶν: simply ‘gods in assembly’, after the oldest meaning of ἀγών (LSJ s.v. I.1). See Fraenkel on Ag. 513, who suggests that the term as used of gods is Aeschylean in origin and influenced by the Homeric θεῖος ἀγών (Il. 7.298, 18.376, cf. Hes. Sc. 205). Possibly the Olympian pantheon and not just any group of gods assembled is specifically intended by ἀγώνιοι θεοί in Aeschylus (also at 242, 333, 355, Ag. 513): see on 204–24 below. 190 κρεῖσσον … σάκος: on the gnomic utterance see 176–78n. (n. 302). On the neuter predicative see K–G i. 58–60, Barrett on E. Hipp. 443–46. 192. ἀγάλµατ’: the word is used ‘etymologically’ in a general sense, meaning ‘glory’ or ‘adornment’ (cf. LSJ s.v. ἀγάλλω), in a context where one would at first expect the later, specific one: ‘statue’, ‘image’ of a god (cf. Th. 258, 265, Eu. 55). The intended effect may be one of archaizing. Cf. Od. 8.509, where a bull is a µέγ’ ἄγαλµα θεῶν, and also Eu. 920. Αἰδοίου ∆ιός: the epithet is not found elsewhere, but Zeus is of course in many ways αἰδοῖος, cf. S. OC 1267–68 ἔστι … Ζηνὶ σύνθακος θρόνων | Αἰδὼς Whittle (1975) 14. On Todt’s (1889) inelegant τῶνδ’ … στόλος see ibid. 13 = FJ–W 184–89n., ii. 149. The dative τῷδ’ … στόλῳ seems not to have been proposed: cf. 461, and note, for what it is worth, that the verb is always construed with a dative in Homer, the middle/passive voice occurring in Il. 21.324. Here the dative would be taken ἀπὸ κοινοῦ with ὠµῇ … ὀργῇ, and in my opinion with a rather intense dramatic effect. 341 Turnebus’ τόνδ’ is adopted by most editors, however, and vigorously defended by FJ–W. 124 ἐπ’ ἔργοις πᾶσι.342 The epithet should not be seen as transferred from the worshipper to the deity, but taken in a general sense, ‘Zeus of αἰδώς’ (cf. on 1, Ἀφίκτωρ). Cf. also Cairns (1993) 183–84. 194–95. αἰδοῖα … πρέπει: the girls are not urged to behave decorously, as αἰδοῖα might at first suggest: the context implies that the word is active in sense, ‘commanding reverence/pity’, i.e. ‘pitiful’ (LSJ s.v. I and II.3). The Danaids are to ‘play the part’, as it were, of miserable suppliants, with ‘pitiful’ and ‘plaintive’ words. This does not quite agree with Danaus’ subsequent guidelines for his daughters’ speech in 196 τορῶς λέγουσαι and 198 τὸ µὴ µάταιον (which refers to the speech; see ad loc.).343 Add to this that ἀµείβεσθαι is extremely rare with a double accusative,344 except of the type with a quasi-adverbial neuter pronoun (K–G i. 321–22, Anm. 4), and one may conclude that Paley (edd. 1844, 1879) had fair reasons to suspect 194, even if he did not state them. However, αἰδοῖα is echoed in 455 τέρµατ’ αἰδοίων λόγων, and ξένους ἀµείβεσθ’ seems to need some further modal qualification to which ὡς ἐπήλυδας πρέπει can refer. The verse is oddly corrupt, even for this drama. γοεδνὰ (Robortello, Turnebus) is certain for M’s vox nihili γοείδηα; but Geel’s (1830–31) ζαχρεῖ’, in the sense it is usually taken, ‘of sore need’ (LSJ), is neither ‘certain’ (FJ–W) nor a ‘corr.’ (Wilamowitz, Murray, West) of M’s τὰ χρέα: ζαχρεῖος is attested only once, as a neut.sg. adverb with a different meaning, in Theoc. 25.6.345 Portus’ κοὐκ ἀχρεῖ’, on the other hand, is paralleled in Pi. fr. 180 µὴ … ἀναρρῆξαι τὸν ἀχρεῖον λόγον. For the negation, cf. K–G ii. 180, 182–83. Note that the latter emendation may be rather easy palaeographically if M was copied from a minuscule source (see 110–11n., n. 276): similar corruptions of 342 Also, e.g., Pl. Prt. 329c. ἔλεγες γὰρ ὅτι ὁ Ζεὺς τὴν δικαιοσύνην καὶ τὴν αἰδῶ πέµψειε τοῖς ἀνθρώποις. 343 Cf. also Th. 181–286, where the over-emotional conduct of a female chorus is harshly censured. 344 Tucker notes Pi. P. 9.38–39, the sole extant example. de Pauw’s ξένοις is worth considering (cf. Eu. 442). 345 Giangrande’s (1979) interpretation of ζαχρεῖον in Theoc. 25.6 (supported by Chryssafis ad loc.) as an adverb synonymous to στερεῶς (formed as an opposite to the Homeric adverb ἀχρεῖον, ‘helpless’) is plausible. Cf. also Add.Et.Gud. s.v. ζαχρειῶν· καὶ ζάχρειον τὸ σφοδρόν. 125 ligatures into simple καί are possible in 276, 296, 504. Another possibility would be τὰ χρήστ’ (LSJ s.v. I.1).346 196. τάσδ’ ἀναιµάκτους φυγάς: cf. 6 οὔτιν’ ἐφ’ αἵµατι δηµηλασίαν, 6– 10n. 197–99. φθογγῇ δ’ ἑπέσθω: Aeschylus separates the abstract sense of the utterance from the sound of the vocal cords: the former ‘attends’ the latter (if the dative is taken as comitative with the verb, which seems likely). Bothe’s (ed. 1805) ἐπέστω is duly noted by West; to his parallels add Thgn. 1.85, 1.365, and cf. the similar corruption in Eu. 543: however, the paradosis might gain support from the somewhat similar expression in 523. There is, as noted by Tucker, a slight anacoluthon or a glide in the sense when we reach the prepositional phrases in 198–99. τὸ µὴ µάταιον ought to refer to the speech, as often in Greek and especially in Aeschylus, where the adjective is frequently used of speech that is indistinct and over-emotional owing to the influence of fear.347 However, the expressions ἐκ … προσώπων and especially ὄµµατος παρ’ ἡσύχου imply that the reference now is to some sort of general attitude, countenance, or ‘glance’.348 But the sense of µάταιος does not favour this: the adjective is awkward as referring to a facial expression or ‘attitude’. Possibly Danaus is still referring to his daughters’ speech: ὄµµατος παρ’ ἡσύχου may not mean, unlike E. Cret. III 14 (TrGFS p. 117; Pap. poet. fr. 11), ‘from (out of) a quiet eye’, but rather ‘from the general direction of’, ‘from beside’, or simply ‘beside a quiet eye’ (LSJ s.v. παρά A.I, III, K–G i. 509). So παρ’ ἀσπίδος in Th. 624 and Il. 4.468, and παρὰ µηροῦ in, for instance, Il. 1.190, Od. 9.300 strictly mean ‘(from) behind the shield’ and ‘from beside the thigh’. A parallel to the double prepositions ἐκ and παρά is found in a different context in Od. 15.58 ἀνστὰς ἐξ εὐνῆς, Ἑλένης πάρα καλλικόµοιο: ‘from out of the bed, from beside Helen’.349 346 Cf. S. Tr. 231, E. Supp. 296, Heracl. 555, El. 358, Ar. Av. 1449, and Men. fr. 806 PCG (Arsen. 7.98c) ἔστι δέ | γυνὴ λέγουσα χρησθ’ ὑπερβάλλων φόβος. 347 Cf. Th. 280, Ag. 1662, 1672, Ch. 846, Eu. 830. Others have taken it simply as ‘untrue’ (Wecklein ed. 1902, Vürtheim). 348 ‘and let your speech be accompanied … by no arrogance, and let no impudence proceed from gentle eyes’ (Friis Johansen); ‘let your speech be attended by no boldness, and let no froward glance proceed from countenances marked by a modest front’ (Smyth), ‘qu’aucune effronterie … ne se lise en votre regard pose’ (Mazon), ‘ciò che falso non è dal vostro volto’ (Untersteiner ed. 1946). 349 Cf. Il. 11.1 (= Od. 5.1), Od. 1.259, E. Ph. 1103, Th. 1.137.3. 126 παρά + gen.rei is no anomaly per se in Attic Greek: the rarity is the purely locative sense of the prepositional phrase, with the noun lacking an active function in the transaction, as it has in the expressions ‘learn from’ and ‘receive from’. One may learn and receive from ‘active things’, as it were, although naturally less often than from persons: e.g., Isoc. Antid. 223 παρὰ τῆς αὑτοῦ φύσεως ἐπίσταται, Antiphon Nov. 6 εἰδέναι παρὰ τῆς βασάνου.350 παρά + ‘pure’ gen. loc. does appear in Pindar and a few times in the drama, especially Euripides.351 †µετώπω σωφρονῶν†: by far the most attractive (if extensive) emendation is Dindorf’s (ed. 1857) σεσωφρονισµένων (cf. 724), which has been adopted in the text by West.352 200. ἐφολκός is passive, i.e. words would have to be dragged out of the girls (cf. LSJ s.v. ἐφέλκω I.4, II.1). 201. ἐπίφθονον γένος: editors point out that the Argives, like the Spartans, had a reputation for brevity of speech (cf. 273).353 It is unclear, however, why this should be connected with being ἐπίφθονος, which means ‘malicious’, here apparently ‘easily offended’. Perhaps something should be made of Hera’s special affinity with Argos (see on 291–92).354 Some (Hermann a.o.) find it strange that Danaus should have any knowledge of Argive national character, but the tradition (if meagre) does portray him as a knowledgeable man (see the Introduction and 320n., 134–35n.), so this need not be an unrealistic conceit. 203. On the gnome, cf. 176–78n. On Whittle’s (1968) transposition of 232– 33 to follow after this verse, see the note on these verses. 204–24. The Prayer: it is uncertain whether the Danaids move up on the πάγος (i.e., the rock, the stone wall, or the raised stage: see the Introduction, III 4–6), and whether they sit down, before the prayer. I am inclined to believe 350 Cf. also, e.g., Isoc. Paneg. 26, D. 18.308, 19.55, 25.81, Aeschin. 1.129.8. K–G i. 509. Cf., e.g., S. Ant. 966, 1123 (both corrupt), E. Ba. 118, HF 1127, 1222, Ion 1141, Rh. 384, Pi. O. 5.9, P. 4.103, N. 9.1, 11.36, fr. 30.2, Call. fr. 534, and the expressions παρὰ χειρός in E. Hyps. 58.7, B. 14.10 and πὰρ ποδός in Theogn. 1. 282, Pi. P. 3.60, 10.62. 352 On Porson’s (ed. Euripides 1802, p. xlix) less fortunate µετωποσωφρόνων, see FJ–W. 353 Pi. I. 6.58–59 with schol., S. frr. 64, 462, Cic. Brut. 50. 354 On the maliciousness of Hera, see Burkert (1985) 134; on her personality in general (especially in Homer), see Lindberg (1990). 351 127 that this is not the case (see on the constitution of the text below, and 208n.). Danaus probably also stands throughout the prayer. The prayer as such consists of a regular, symmetrical, stichomythia. With 207–9 remaining in their traditional places (see below), the stichomythia proper begins at 209, whereas the ‘profane’ dialogue before that is irregular (cf. Ch. 489–96).355 There is a symmetrical arrangement of the gods: Zeus is mentioned first, whereupon four gods are mentioned in order. On cues from Danaus, his daughters do reverence to each deity in turn. In three of the four cases, the prayer follows an identical pattern: on their father’s prompting, the Danaids utter a 3rd person optative or imperative pertaining to the god in question (on this type of prayer see Ziegler 1930, 19–26). In this manner Apollo (214–15), Poseidon (218–19), and Hermes (220–21) are venerated. Before Apollo, ‘Zeus’s bird’ is mentioned in a similar manner (212–13), probably taken for Helios (Re) by the Danaids (see my note ad loc.). Five male gods, then, are singled out as objects for the Danaids’ reverence. Connected with the gods in some way is an altar (κοινοβωµία, see the notes on 222–23, 345 and the Introduction, III 5). The original scenic arrangement is highly uncertain. However, the later text may suggest that each god as well as the altar is adorned with boughs: the altar is ‘crowned’ in 345 (cf. FJ–W 241–42n.), and it appears from 346, 354–55, and possibly 241–42, that the boughs are arranged so as to cast shadows on the gods. The following is pure speculation about the scenic arrangement, offered simply as a help to the reader in visualising the action. At any rate it has the advantage of agreeing with a conservative constitution of the text, which I think should be adopted in this passage (see below). Possibly on each cue from Danaus, one of his daughters climbs up towards the god in question and adorns his image with one of her suppliant boughs, while the others do reverence. The 355 In all extant examples of stichomythiae in Aeschylus, ‘short’ breaks of symmetry in the middle of the dialogue are avoided; i.e. there is no instance of a regular stichomythia (out of 27 examples of single-line stichomythiae of eight lines or more) where one of the speakers suddenly gets two consecutive lines, after which the single-line stichomythia recommences (with three lines or more of regular stichomythia on each side of the break). Any break of symmetry is either at the end or the beginning of the stichomythia and may thus be considered as not part of the stichomythia proper, or being so extensive as to make the recommencing stichomythia a new one entirely (four consecutive lines are spoken by the chorus in Ch. 770–73, before and after which there are short one-line stichomythiae involving the chorus and the nurse). 128 coryphaeus and perhaps one or several of her sisters adorn the altar in 222– 23, after which those who remain climb the ‘hill’, adorn the remaining gods (if there are any), and sit down as suppliants. See ‘The Text’ below, and also my notes on 241–42, 345, 351–53, 354–55. FJ–W on 209–23 argue that the number of images present is twelve, representing all the Twelve Olympians,356 which is possible but not provable. If only those gods that are mentioned are present on stage, and each god is adorned by one Danaid—except perhaps Helios, who has no idol of his own, but is represented by the eagle in Zeus’ hand (note that he is not, like the other gods, made the subject of an imperative or optative) and the altar by one or two Danaids, this leaves six of the girls without anything to do. These six may be the Danaids’ handmaidens, mentioned in 977 and perhaps in 954: they may not be carrying any boughs (cf. 1–39n.). If FJ–W are right, and all twelve Olympians are represented, there is no telling why these particular gods are selected for worship, nor why none of the female deities is named here.357 It is likely that Aeschylus lets the Danaids’ religious conduct reflect their Egyptian origin (see the separate notes on 212– 21). Herodotus 2.156 implies that Aeschylus was familiar with the Egyptian pantheon, stating that he (fr. 333) followed Egyptian tradition in making Artemis (Bast) the daughter of Demeter (Isis).358 FJ–W note that apart from Poseidon who, perhaps significantly, is not invoked by name, all gods mentioned here—and also Artemis, who was invoked in 144–50 (also at 676, 1030–31)—have an Egyptian alias.359 356 Cf. Paley 218n., Dale (1969) 263; for a different opinion Polacco (1983) 69–70. Artemis in particular is depicted elsewhere as a champion of the Danaids’ cause (see 144–50). Hera and Aphrodite are obviously adverse: cf. 165, 1032, etc. 358 Cf. also Paus. 8.37.6. Wilamowitz (1927, 287–88) also observed a striking resemblance between several passages from Clytaemnestra’s speech in Ag. 855–974 and an Egyptian hymn from the Middle Kingdom: see further Kranz (1933) 102, Hall (1989) 206. 359 Zeus–Amun (see 4n.); Apollo–Horus (Hdt. 2.156); Hermes–Thoth (Hdt. 2.67, 2.138, Aristox. fr. 23.6); Helios–Re (Hdt. 2.59, etc.; although the Egyptian name appears not to be known to him. In later Hellenistic cult Helios is identified with the Egyptian deity Serapis and with Zeus [C.Z. i. 187–90]); Artemis–Bast(et) (called Bubastis by Herodotus 2.37, 2.156 after the city associated with her). Note also Herodotus’ claim (2.4, through the mouths of Egyptian priests) that the Egyptians were the first to name the Twelve Gods, on which misunderstanding see Lloyd i. 92. 357 129 The text. Noting the change of speaker at 204, M thereafter fails to provide any information about the distribution of lines until 246. Accordingly, we are left to our own devices in determining who speaks which line in the first stichomythia of the drama. Critics have dealt with these lines differently, reordering, adding, and sometimes deleting lines after various fashions: there are almost as many versions as there are editors, and many more published in articles. For ease of reference for the user who does not have a conservative text of the present passage at hand, the traditional arrangement of the lines is given in n. 68 above (see the Translation, p. 25). This arrangement seems to me to be likelier than most or all of the conjectural versions that have been advanced. In fact, no apparent advantages adhere to any of the vulgate transpositions, except one: 210 may be moved to follow after 206 (Burges 1810, Hermann, most edd.). Thus ἴδοιτο δῆτα would follow directly upon Ζεὺς … ἴδοι, which accords with the common use of δῆτα in affirmations of this kind (D.GP 276–77): so at 215–16 συγγνοίη – συγγνοῖτο δῆτα. However, it is hardly an unshakeable rule of grammar that such a repetitive affirmation must follow directly upon the ‘affirmandum’: cf. Ar. Lys. 1242–45 —λαβὲ τὰ φυσατήρια, | ἵν’ ἐγὼ διποδιάξω τε κἀείσω καλόν | ἐς τὼς Ἀσαναίως … | — λαβὲ δῆτα τὰς φυσαλλίδας and Vesp. 172 (δῆτα in a question, but apparently repeating ἀποδόσθαι from 169). In 359 the same phrase, ἴδοιτο δῆτα, answers to εἴη three lines before, ignoring the message of the two lines inbetween (see ad loc.), also echoing ἴδε µε in the corresponding place in the strophe (348). See also Verdenius (1990). In the present case, 206–9 may be taken as a parenthesis of sorts: on Danaus’ cue, which consists of ignoring the Ζεὺς … ἴδοι in 205 and instead telling his daughters to ‘hurry up’, the Danaids take their appropriate places in order to pray (see above on the ‘Prayer’ section). When they subsequently address Zeus directly in a slightly more agitated fashion (ὦ Ζεῦ κτἑ), Danaus chooses to affirm their former utterance in 206 instead, which is more decorous (cf. Th. 78–286, especially 265–81). This would be easier still if we read σκοπῶν in 209 (Friis Johansen 1966; cf. 381, 402, 681), as ἴδοιτο δῆτα may answer ‘formally’ to this verb (cf. D.GP 277). In the light of the consequences of a transposition of 210, preservation of the paradosis is preferable to the commonly endorsed alternatives. The text as it stands presents no further prima facie difficulties (see the notes on the individual passages), whereas all the received transpositions will upset the text in such a way that further transpositions or lacunae become neces130 sary360—a conjectural measure that is exponentially less likely to be right than a single lacuna or transposition (see the Excursus). One alleged problem with the traditional arrangement is that 207–8 must mean that here the Danaids move up on the ‘mound’ beside Danaus and, it would seem, sit down (θρόνους ἔχειν). On the other hand, they are ordered ἑσµὸς ὡς πελειάδων ἵζεσθε at 223–24, and FJ–W claim, perhaps rightly, that the imperative must mean that they cannot have been sitting before that. At any rate it is very hard to imagine the prayer in 210–21 as executed by sitting worshippers, in particular if they are seated with their backs towards the gods.361 However, if we imagine, as suggested above, that the Danaids come forward one at the time and sit down by the gods, this will accommodate both 207–8 and 224: the ﬁrst girl walks up to Pelasgus in 208 (θέλοιµ’ ἂν ἤδη), puts a bough on Zeus in 209 and, having thus made proper reverence, sits down: similarly the others, one at the time, in 210–23, whereupon the remaining Danaids and the servants sit down at the request in 224, having adorned the altar. 360 Firstly, 207 µή νυν σχόλαζε cannot follow directly upon 210 without change of speaker. The change of tone from ceremonious prayer to mere parental impatience (‘hurry up!’) is incredible, especially in combination with asyndeton. Hermann’s solution, which has been adopted by several editors (e.g. Murray, West), was to let 207 and 208 change places (i.e. 206, 210, 208, 207, 209, 211); but this necessitates a lacuna after 211 in order to maintain the traditional type of stichomythia, with one line spoken in turn by each interlocutor. It is furthermore unlikely as the 1p. potential mode + ἤδη (208 θέλοιµ’ ἂν ἤδη) is commonly used in answer to requests or exhortations. Here it should certainly follow the imperative µή νυν σχόλαζε (cf. Th. 472, E. Or. 640, Ar. Eq. 40, Lys. 97). Oberdick instead let 211 move together with 210, which is better (so, e.g., Page); but the irregularity in the stichomythia that ensues at 208–9 (210–11 in Page’s edition), with two lines consecutively spoken by the coryphaeus, is awkward, and unparalleled in Aeschylus if a regular one-line exchange has already begun at 206 (see below, n. 362). Thus another lacuna becomes necessary after 208 (Kirchhoff, followed by Page)—and so on. 361 Prayers in antiquity generally appear to have been executed standing up; kneeling occurs, but seems to be associated with highly emotional prayer (such as that of the women of Thebes in Th. 87–180 and Ajax in S. Aj. 854–65). Possibly the context of supplication would admit this gesture in the present case; however, θρόνους ἔχειν means ‘sit’, not ‘kneel’. See Pulleyn (1997) 190 with refs, Burkert (1985) 75 with n. 19 (p. 376). 131 The possibility of course remains that the text is corrupt, and that some sort of disarrangement of verses has taken place. As for interference with the text, it is odd that apparently no critic after Bamberger has agreed that there is a far more economical and text-critically sound means to accommodate the problem with 210 than the wholesale re-shuffling, in combination with addition of single lines between 206 and 211, that has been current (see n. 360). Obviously the offending verses are 207–9. In the absence of these, few critics would seriously have considered re-arrangements of the text. Bamberger’s (1842) suggestion was the removal of these verses from their current place in the text, letting them follow after 233 instead: — κἀκεῖ δικάζει τἀµπλακήµαθ’ ὡς λόγος Ζεὺς ἄλλος ἐν καµοῦσιν ὑστάτας δίκας. σκοπεῖτε κἀµείβεσθε τόνδε τὸν τρόπον, ὅπως ἂν ὑµῖν πρᾶγος εὖ νικᾷ τόδε. µή νυν σχόλαζε, µηχανῆς δ’ ἔστω κράτος. — θέλοιµ’ ἂν ἤδη σοὶ πέλας θρόνους ἔχειν. ὦ Ζεῦ, σκοπῶν οἴκτιρε µὴ ἀπολωλότας. 230 233 207 208 209 207 fits well here, as µηχανῆς may refer to the entire plan for the girls’ conduct: better so than 232 τόνδε τὸν τρόπον, which ought to refer to something that has recently been said (on the problem of 232–33, see ad loc.). The absence of a connecting particle is natural with imperatives and similar concluding speeches (as already in 232; also, e.g., 289, 732, 1012, Th. 451, 480, 562).362 362 An alternative possibility would be a simple excision. The verses are not suspect from a stylistic viewpoint, however. An odd thing, presumably incidental, is that 207–9 and 210 are mutually exclusive, as it were: if 210 is removed instead of the previous three verses, 211–12 (both spoken by Danaus) follow naturally upon 209, maintaining a symmetrical arrangement of verses (which is always the case in Aeschylean stichomythiae). Two consecutive lines from the chorus would be followed by two from Danaus, whereupon a single-line stichomythia would ensue: — θέλοιµ’ ἂν ἤδη σοὶ πέλας θρόνους ἔχειν. ὦ Ζεῦ, σκοπῶν οἴκτιρε µὴ ἀπολωλότας. — κείνου θέλοντος εὖ τελευτήσει τάδε· καὶ Ζηνὸς ὄρνιν τόνδε νῦν κικλῄσκετε. 132 204. φρονούντως … φρονοῦντας: observe the similarity to 176 (ring-composition: cf. 162–67n.). The ‘generalising masculine plural’ (φρονοῦντας)363 has no parallel as referring to a group of women. The reason for this, however, probably is simply that groups of women are relatively seldom referred to in Greek literature. There is in any case no problem to speak of: this type of the masc.pl., a masculine plural attribute (adjective, participle, noun, possessive pronoun) implicitly designating a female (not further speciﬁed by name or with a personal pronoun), technically does not refer to a particular subject, whether singular or plural.364 Its purpose is to describe a situation in the abstract, as a general case, and the particular reference is implicit only. Here not ‘father, you speak prudently to us who are prudent’ but ‘… to people such as are prudent’. Cf., for example, Ch. 689 τοῖς κυρίοισι (implying Clytaemnestra), S. Tr. 1237 τοῖσιν ἐχθίστοισι (Iole), OC 148 σµικροῖς (Antigone), E. Med. 61 δεσπότας (nurse), Andr. 712 τίκτοντας ἄλλους (Andromache). The phenomenon has affinity with the gnome (see on 176–78) in its generalising of the concrete situation: the previous verse actually makes a borderline case, in which τοὺς ἥσσονας implicitly refers to the Danaids (cf. also E. Hipp. 358).365 205–6. φυλάξοµαι … µεµνῆσθαι: perhaps the expression should be seen as a ‘mixed construction’ (so Griffith 1986): cf. the Hesiodic366 (ἐν θυµῷ) ταῦτα (πάντα) φυλάσσεσθαι and, e.g., Pl. Smp. 200a τοῦτο … φύλαξον παρὰ σαυτῷ µεµνηµένος ὅτου, Lg. 783c φυλάξωµεν … τῇ µνήµῃ τὰ νυνδὴ 363 See K–G i. 18—who, however, restrict the usage to certain expressions only—and FJ–W ad loc. The case is to be distinguished from that where the masculine form of a participle actually signifies the feminine genus, on which see Fraenkel on Ag. 562, with refs, Barrett on E. Hipp. 1102–50, Langholf (1977). Yet another case is the selfreferential plural, on which see below, n. 365. 364 Cf. the similar use of the m.sg. in 245. 365 This ‘gnomic’ use of the 3pers. masc.pl. is essentially different from the use of the 1pers. masc.pl. referring to oneself alone: the latter is simply the ‘self-referential plural’ (more common in Latin), which always takes the masculine genus. K–G §§ 371.2 and 3 (i. 83–84) should be referred to the same category. In the case of singular women speaking of themselves in the plural, the fem.sg. of adjectives, participles, and pronouns may be used with verbs in the plural (E. Ion 1250–51 διωκόµεσθα … κρατηθεῖσ’, HF 858), alternatively the masc.pl. (A. Ch. 716–18 ἡµεῖς … κοινώσοµεν τε κοὐ σπανίζοντες … βουλευσόµεσθα, E. Andr. 355–58). The fem.pl. is never, to the best of my belief, used to designate a single person. The masc.sg. with a verb in the plural may be used in the case of a masculine subject (E. HF 1207–10). 366 Op. 263, 491, 561, 797–98 (see n. 371 below). 133 λεχθέντα. In any case, the construction is unexceptionable.367 The most common meaning of the verb is certainly ‘guard against’, from which follows that the construction is usually negative. There are, however, several instances of the verb where the sense is positive—‘take care that’, with according constructions. Thus the active voice is found with acc. + inf. in ?Pl. Epin. 982c φυλάττουσι τέλεον εἶναι τὸ … βεβουλευµένον and with non-negated ὅπως-clauses in Hdt. 4.190 φυλάσσοντες, … ὅκως µιν κατίσουσι, Pl. Ti. 90a φυλακτέον ὅπως ἂν ἔχωσιν τὰς κινήσεις πρὸς ἄλληλα συµµέτρους, Plu. Frat. am. 488b.368 Cf. also Ch. 579–80 φύλασσε τἀν οἴκῳ καλῶς, | ὅπως ἂν ἀρτίκολλα συµβαίνῃ τάδε. Just like the active voice, the middle voice of φυλάσσειν need not necessarily retain the negative sense of ‘guard against’, but may construe as a regular verb of purpose, thought, will, or action.369 LSJ s.v. φυλάττω B 9 and C II 3 present two examples of the middle voice with the non-negated final articular infinitive: Lxx Jo. 23.11 and [Ocell.] 4.13.370 Also, Hes. Op. 797–99 construes as the present passage, with a formally non-negated infinitive (albeit inherently negative in sense: ἀλεύασθαι).371 The reflexive middle voice here implies ‘guard with/for oneself’, and Turnebus’ φυλάξοµεν is unnecessary. 367 µεµνῆσθαι is attacked on grammatical grounds by FJ–W, who argue that the construction of φυλάττοµαι with a non-negated, non-articular, final-consecutive infinitive is unparalleled. The argument is presented in greater detail in Friis Johansen– Whittle (1975), where Friis Johansen suggests µὴ ἀµνηστεῖν. 368 The latter passage also presents a pendant acc. + inf.: φυλακτέον, ὅπως τὰ πράγµατα µάχηται καθ’ αὑτά, µηδὲν ἐκ φιλονικίας … προσθέντας ἀλλ’ ὥσπερ ἐπὶ ζυγοῦ τοῦ δικαίου τὴν ῥοπὴν κοινῶς ἀποθεωροῦντας καὶ τάχιστα ταῖς κρίσεσι καὶ ταῖς διαίταις τὴν ἀµφιλογίαν παραδιδόντας ἀποκαθῆραι. 369 K–G ii. 5–11, 28–29, 372–74. 370 Lxx Jo. 23.11 (gen.), presumably a Semitism: φυλάξασθε σφόδρα τοῦ ἀγαπᾶν κύριον τὸν θεὸν ὑµῶν, and [Ocell.] 4.13 (acc.): φυλάττεσθαι τὸ καθεστηκυίας τῆς διανοίας τὰς µίξεις γίνεσθαι. Friis Johansen–Whittle (1975, 14) argue that the article excludes the relevance of the latter passage as a parallel to our own; however, the articular infinitive with φυλάττεσθαι is also usually negative in sense, with or without the negation (K–G ii. 398, cf., e.g., Pl. Plt. 261e R. 424b Tht. 180a, Mx. 244e, X. Mem. 1.3.6, 4.2.5, Cyr. 3.1.27, 4.2.25, D. 18.258), and the rare occurrence of nonnegated instances thus ought to show that the same construction is possible, if rare, with the non-articular infinitive. 371 Hes. Op. 797–99 πεφύλαξο δὲ θύµῳ | τέτραδ’ ἀλεύασθαι φθίνοντός θ’ ἱσταµένου τε | ἄλγεα θυµοβορεῖν (ἄλγεσι West ad loc.); cf. S. Aj. 535 ἐγὼ ’φύλαξα 134 207–10. On the position of these verses, see on 204–24 above. 207. µηχανῆς δ’ ἔστω κράτος: cunning and plans are of little value unless there is strength to carry them out (cf. Paley, Mazon). The contrast is similar to the common one between λόγος and ἔργον (LSJ s.v. λόγος VI.1.c): cf. 241– 42n., n. 392. 208. σοὶ πέλας θρόνους ἔχειν: pace FJ–W (and the scholium), this does not mean that Danaus is already seated, only that the girls will move towards him (up onto the πάγος) and sit down beside him, whether he is sitting down or standing up. In fact, a seated Danaus giving instructions to his standing daughters amounts to a rather bad stage-direction, to say nothing of him sitting through the prayer he himself leads. It may well be that Danaus remains standing throughout the drama (on Danaus’ actions, see further 246–48n). θρόνους ἔχειν elsewhere means ‘have royal power’,372 and Wecklein (1893, 334) conjectured θάκους. FJ–W observe, however, that θρόνος in the general sense of ‘seat’ appears in 792. θέλοιµ’ ἂν: on the potential mode expressing resolve to act (K–G i. 233), see on 204–24, n. 360 above. 209. ἀπολωλότας: the predicative use of an oblique case of the pf. ptc. is apparently unparalleled (FJ–W): ‘have mercy on us (so that we are) not destroyed’ or, as Moorhouse (1948, 37): ‘pity us … not being, I pray, consigned to perdition’.373 On nominative participles as predicatives (with εἶναι, γίγνεσθαι, etc.), see also K–G i. 38–39. Friis Johansen’s (1966) σκοπῶν is palaeographically extremely easy, and the stem is associated with Zeus elsewhere in the drama (381, 402–3, 646–47). If the traditional arrangement of the stichomythia is kept, σκοπῶν may be answered by ἴδοιτο δῆτα in 210 (see 204–24n. above). 212–13. Ζηνὸς ὄρνιν … Ἡλίου: the emblem of one aspect of the Egyptian sun-god Re was the falcon, and the Danaids probably mistake Zeus’s eagle (referred to vaguely as a ‘bird’) for an emblem of the sun.374 This identiﬁcation τοῦτό γ’ ἀρκέσαι. (The Hesiodic example refutes the claim of Friis Johansen– Whittle 1975, 15, that φυλάττεσθαι + inf. is unattested before Aeschylus.) 372 S. Ant. 172, OC 425, 1354, E. HF 167, Ar. Ran. 769. 373 On the masculine gender, see 204n. Here it may possibly mean that Danaus is included in the reference (so FJ–W). 374 Cf. Morenz (1973) 129, 152, 178, Quirke (1992) 21 ff., C.Z. i. 341–42. On the eagle of Zeus in general, see C.Z. ii. 751–52, i 83–84, passim. Bamberger’s (1842) ἶνιν, rejected by himself and proposed again by Kiehl (1856), is thus unnecessary, creating more problems than it removes (see FJ–W). 135 fits remarkably well with the fusion that the Egyptian deities Amun and Re had undergone into Amun-Re (see 4n., n. 103, and cf. Pasquali 1924). The fact that the Danaids identify Zeus’s eagle with the Egyptian sun-falcon does not mean that they mistake the bird itself for a falcon: just as in the case of the other gods, none of whom is represented in the Egyptian fashion on stage (cf. 220 τοῖσιν Ἑλλήνων νόµοις), they allow for cultural differences in the outer manifestation of the deity. 214. φύγαδ’ ἀπ’ οὐρανοῦ θεόν: in all likelihood the servitude of Apollo to Admetus is intended (thus Plu. Def.orac. 417e–f, who also cites the verse in Exil. 607c).375 Another, slightly more elevated story about an ‘exile’ of Apollo’s is recorded by Alcaeus (fr. 307.1c, a prose paraphrase in Him. Or. 48.10–11), in which Zeus sends Apollo to Delphi, ἐκεῖθεν προφητεύσοντα δίκην καὶ θέµιν τοῖς Ἕλλησιν. Instead, however, he flies to the Hyperboreans with his swan-driven chariot, not to return until the next summer. Still, there is nothing to suggest that this ‘exile’ is involuntary. 218–19. Poseidon is not named, which perhaps reflects the fact that he has no indigenous Egyptian counterpart (see on 204–24 above).376 The Danaids seem to recognise him, however. Possibly τρίαιναν reveals him as a god of the sea, and thus as an appropriate deity to pay one’s respects to after a successful sea voyage (perhaps the first ever undertaken: cf. 134–35n.). West approves of Tucker’s critique of 220, to the effect that σηµεῖον θεοῦ is obvious and redundant information; but Danaus’ expression implies τόνδε τὸν θεόν, οὗ τὸ σηµεῖον τρίαινά ἐστι or τρίαιναν, σηµεῖον θεοῦ οὗ τὸ ὄνοµα οὐκ οἶδα. For the appositional phrase cf. 506. ὁρῶ may feel a little awkward in answer to a question. ὁρᾷς … θεοῦ; was suggested by Dawe (ap. Page). ὁρᾷς; with an answer beginning with ἀλλά is paralleled in S. Ph. 1255; cf. also E. Hec. 758–60. As support for the paradosis, cf. on the other hand Ch. 168. 375 Cf., e.g., E. Alc. 1–7 with Σ (= Pherecyd. fr. 76), S. fr. 851, Hes. fr. 54b, Call. Ap. 47–54 (who states that Apollo was in love with Admetus), [Apollod.] 3.10.4. 376 There was no god of the sea in the Old and Middle Kingdoms of Egypt; but a Semitic god, Yamm, was introduced in later myth: see Hornung (1983) 79 with n. 49. Hdt. 2.50 claims that the Egyptians learned of Poseidon from the Libyans. On Poseidon in North Africa, see Lloyd on Hdt. 2.50 (ii. 237–38). 136 220–21. Ἑρµῆς … κηρυκευέτω: Tucker ingeniously suggested that τοῖσιν Ἑλλήνων νόµοις refers to an ithyphallic representation of Hermes, the expression containing a sort of apology from Danaus to his chaste daughters for such crudity. This interpretation is accepted by Wilamowitz (in the apparatus) but rejected by FJ–W, who claim that the Olympian Hermes is never depicted this way. But would Aeschylus necessarily discriminate between the Olympian and the phallic Hermes in this context? And how is a depiction of the Olympian Hermes actually distinguished from any other type? S. Eitrem in RE viii. 764 ff. (s.v. ‘Hermes’) does not, pace FJ–W, provide much information on the matter. The classical, fully three-dimensional statues and realistic reliefs of Hermes seem not to have been ithyphallic; but the ithyphallic representation is the oldest, and it is the customary one in busts,377 which may well be how the gods on stage were represented. Certainly a bust the height of a man is a more stable thing to hang oneself from (cf. 455–65) than a realistic statue, especially a statue made out of wood, clay, or terracotta. Tucker notes a passage in Herodotus (2.51) which almost seems to be an illustration of his thesis: τοῦ δὲ Ἑρµέω τὰ ἀγάλµατα ὀρθὰ ἔχειν τὰ αἰδοῖα ποιεῦντες οὐκ ἀπ’ Αἰγυπτίων µεµαθήκασι [sc. οἱ Ἕλληνες], ἀλλ’ ἀπὸ Πελασγῶν. However, Herodotus—who may well have seen or even read Aeschylus’ Danaid trilogy (see Hdt. 2.156, the note on 204–24 above, and Radt on A. fr. 333)—nowhere connects the Pelasgians with the name of Pelasgus.378 Otherwise, Ἑλλήνων νόµοις might refer to the fact that Thoth, Hermes’ Egyptian counterpart, is usually depicted with the head of an ibis—i.e. that the Hellenic anthropomorphic Hermes is foreign to Egyptian custom. Herodotus does not explicitly mention Hermes’ Egyptian appearance, nor the name of Thoth; but he states (2.67) that the Egyptians bury ibises in Hermopolis. The Egyptian cult may well have been known to Aeschylus, if not the name of Thoth, which is not found in Greek literature before Plato (Phlb. 18b, Phdr. 274c–275b):379 Herodotus speaks of Heliopolis and once of Hermopolis in 377 See C. Scherer in Roscher i. 2391–94 (s.v. ‘Hermes’). On Herodotus’ tendency to make almost all Greek religion either Egyptian or Pelasgian in origin, see Lloyd i. 148–49. On herms, see id. on Hdt. 2.7, 2.51. 379 Plato does not mention Hermes in the context: he usually ignores all connections between Greek and Egyptian deities (cf. on 4–5, n. 105, Grg. 482b, Lg. 657a–b). In Ti. 21e he lets Critias report that Critias the elder once retold a story of Solon, in 378 137 Egypt as if the names were known to his readers.380 The names of these cities, as well as the notion that the Egyptians worshipped the same gods as the Greeks, were certainly presented in Hecataeus, probably even earlier (explicitly of Apollo in Hecat. fr. 305, cf. also frr. 300, 303–21, 324).381 West unearths and prints Kueck’s (1890, 13) κήρυξ for Ἑρµῆς, thereby making κηρυκευέτω in 221 easier to explain. κήρυξ would refer explicitly to the fact that a κηρύκειον, a herald’s staff, is represented with the god, whether in the hand of a three-dimensional statue or painted on a bust.382 That would correspond to the trident of Poseidon mentioned before. The corruption would easily have arisen from a gloss. There is no trace of a gloss on κήρυξ in the scholium, however, which clearly explains Ἑρµῆς ὅδ’ ἄλλος (the scholiast misunderstands ἄλλος; see below). Moreover, κηρυκευέτω is easily explicable from Hermes himself, with or without a staff: his well-known office as herald of the gods would be enough to make the Danaids use this verb in their prayer.383 We may also note the alliteration Ἑρµῆς ὅδ’ … Ἑλλήνων, followed by ἐλευθέροις … ἐσθλὰ in the subsequent verse. ὅδ’ ἄλλος: ‘this other one here’ (LSJ s.v. ἄλλος II.7–8). Vürtheim, Rose et al. are mistaken in adopting the scholiast’s explanation, ὡς τῶν Αἰγυπτίων ἄλλως αὐτὸν γραφόντων (see FJ–W). ἐλευθέροις … ἐσθλὰ κηρυκευέτω: the semantic weight of the imperative lies not on ἐσθλὰ κηρυκευέτω, but on ἐλευθέροις (as is indicated by the word- which the goddess Neit is said—by the Egyptian, Philathenian worshippers (ὡς ὁ ἐκείνων λόγος)—to be identical to Athena. One is almost tempted to see in this extreme caution a fear of controversy in religious matters, perhaps inspired by the execution of Socrates δαιµόνια καινὰ νοµίζοντα. As depicted by Plato, Socrates usually swear by Anubis, but never explicitly: the oath is νὴ τὸν κύνα, only in Grg. 482b explained as τὸν Αἰγυπτίων θεόν. 380 2.3 bis, 2.7 ter, 2.8 bis, 2.9, 2.59, 2.63, 2.67, 2.73. 381 On Greek Egyptology before Herodotus, see Lloyd i. 49–60, 116–39. On Hecataeus as a source for Aeschylean Egyptology and other ethnography and geography, see F. Jacoby in RE vii. 2680–81 (s.v. ‘Hekataios’ 3), Kranz (1933) 79–81, Powell (1935) 81–82, Lloyd i. 133, Hall (1989) 75–76, 171: cf. also 256–59n., 284–86n. 382 See C. Scherer in Roscher i. 2393–94 (s.v. ‘Hermes’). Σ218a states that Poseidon’s trident is represented in this way, ἐν γραφῇ. 383 Aeschylus could not possibly know, or be bothered to find out, that Thoth lacks this function in his Egyptian cult. 138 order): ‘let him announce good news to free people’, i.e. ‘let us be free as he announces good news’, or ‘now that we are free, let him …’. 222–23. κοινοβωµία is an abstract noun, ‘common-altarship’ (FJ–W), which could make one wonder whether the reference is actually to an altar present on stage, or whether it means something like ‘the collective object of reverence’ or ‘the collective that would share an altar’, especially as the demonstrative pronoun goes with ἀνάκτων, not κοινοβωµίαν. Portus’ τήνδε is more attractive here than Turnebus’ τόνδε in 189 (q.v., n. 341). However, -βωµία is too concrete an image to use in this case without an altar actually being present (cf. 83–85n.); moreover, πρύµναν in 345 (q.v.) probably refers to the altar. Perhaps the reference is to the altar of Dionysus in the middle of the orchestra, although this, if it existed at all (see the Introduction, III 5), would be separate from the gods, who did not stand in the centre of the orchestra: the juxtaposition of altar, gods, actors and chorus in the present scenes would be impossible. However, if the Danaids are to adorn the altar with boughs before sitting down (see 204–24n.), this may indicate that the altar is near the gods where they stand. 223–26. The bird-imagery from 60–67 is resuscitated, the Aegyptiads still figuring as κίρκοι (224, 62 κιρκηλάτου) while the Danaids now take the role of doves or pigeons.384 The dove seems to have a special affinity with the sanctuary, and especially with the oracle of Dodona (Thompson 1936, 229– 30; FJ–W). According to one story (Sil. Pun. 3. 678), two doves from Egyptian Thebes went out to found the oracles in Libya (see 4–5n.) and Dodona. 223–24. ἐν ἁγνῷ … ἵζεσθε: a parallel from Virgil has long been noted in Aen. 2.515–17: hic Hecuba et natae nequiquam altaria circum, | praecipites atra ceu tempestate columbae | condensae et divum amplexae simulacra sedebant. Note also the etymological connection (probably real, not only Aeschylean; see FJ–W) between ἑσµὸς and ἵζεσθε. Cf. 204–24n. 224–28. The bird-metaphor turns into an elaborate comparatio paratactica,385 in which the preying of bird upon bird is likened to the enmity (and, implicitly and less logically, the wooing) of kin and kin. 224–25. ὁµοπτέρων, ‘alike clad in feathers’ (see FJ–W) is answered by 384 Cf. Pr. 857 and Thompson (1936) 144–45, 227 for parallels with doves being followed by various birds of prey. 385 On this type of simile, see Friis Johansen (1959) 16–49 (21–26 on Aeschylus and the Supplices). 139 ὁµαίµων, ‘of the same blood’, in 225. On the oxymoron ἐχθρῶν ὁµαίµων see FJ–W. µιαινόντων γένος presumably refers to incest (cf. 37 λέκτρων ὧν Θέµις εἴργει). 226–28. ἁγνὸς here takes the sense ‘guiltless’ (= οὐκ ἐναγής: LSJ s.v. ἁγνός II.2), as distinct from the active holiness implied by the word in 223. On the alliteration, see on 227 below. 227. ἄκουσαν ἄκοντος: West p. xxx notes, with Threatte (1980–96, i. 503), IG i3 6 B 5 hακόσι[α], and accordingly puts a spiritus asper on the adjectives. A pronounced ‘h’ here does produce alliteration with ἁγνεύοι, ἁγνός and Ἅιδου in 226 and 228. On the polyptoton, see FJ–W 144n. 228–31. The idea of a final judgement in Hades (cf. 416) may, via the Orphic teachings, have been influenced by Egyptian religion (Burkert 1985, 198, 296 ff.). 229. µαταίων: ‘acts of worthlessness’, ‘acts of profanity’, i.e. not simply of stupidity (cf. on 197–99 above). LSJ s.v. II note that µάταιος often takes this meaning in Aeschylus. FJ–W note fr. 281a 17–19, in which µα]ταίοις stands opposite to δ[ι]καίοις. IA compares E. El. 1064. 231. Ζεὺς ἄλλος: Hades, cf. 158 Ζῆνα τῶν κεκµηκότων. 232–33. σκοπεῖτε … τόδε: the problem with these verses is that they appear to refer back to the advice given in 191–203, not to what has just been said, which would have been expected from τὸνδε τὸν τρόπον. This is nevertheless a necessary correction (Anon.Ald.) of M’s τόπον: ἀµείβοµαι in the local sense means ‘traverse’, and it is senseless in the context. κἀµείβεσθε can hardly refer to anything but the manner of speech recommended in 194– 203 (cf. 195 ξένους ἀµείβεσθ’), whereas σκοπεῖτε (‘consider’) appears to refer to the general advice handed out before that. The indignant lines 225– 31 may well be dictated by fatherly love, but they contain no advice which τόνδε … τρόπον could refer back to. In fact, 232–33 would fit perfectly after 203 (so Whittle 1968),386 and might not without some advantage be replaced 386 Hermann (1842, 179) on the other hand posited a lacuna after 231. (Whittle ap.) FJ–W would prefer one located between σκοπεῖτε and κἀµείβεσθε, which is farfetched. Nor does any of FJ–W’s arguments against Whittle’s (1968) transposition carry any weight: (1) the reversal of number from sg. in 200–202 to plural here is of no consequence in this drama, where the Danaids are referred to indiscriminately by singular or plural, whichever will best fit the verse (see 149n.); (2) µεµνῆσθαι in 205 (which according to FJ–W themselves [wrongly, see ad loc.] is corrupt), is not an ‘answering word’ to µέµνησο in 202, but simply a reminiscence, and a few extra 140 by 207–9 (see on 204–24 above). Whittle l.c. notes that ἀµείβεσθε would be an echo of that word in 195, just as the words ξένους, πρέπει, φυγάς, and θρασύ in 194–97 are all echoed in 202–3. 233. West (cf. p. xxxi) reads the enclitic, i.e. paroxytone form (see Barrett on E. Hipp., p. 424) of the pronoun, ὕµιν, which is not attested in Aeschylean mss. but deduced by Barrett (p. 425) as having been in common use in Attic. Barrett observed that the existence of paroxytone oblique forms of ἡµεῖς and ὑµεῖς is safely attested in Babrius, an author who always ends his iambic trimeters with a paroxytone word, in fifteen cases an oblique case of the 1pers. or 2pers. pl. pronoun.387 234–45. Enter Pelasgus, king of Argos. We can safely infer from 500 that a retinue is present with him on stage (Taplin 1977, 201); whether there is a chariot (cf. 181, 183) is impossible to determine for certain, though. Nothing actually speaks against it, seeing that chariot entries were rather common ‘in ganz alten und in ganz jungen Stücken’388 and that the Aeschylean ones are not particularly stressed by explicit announcements elsewhere.389 As for the size of the host that appears together with Pelasgus, Taplin (1977, 202–3) convincingly argues against the spectacle of multitudes (with sometimes over 200 persons present on stage) imagined by earlier critics, who believed that the chorus consisted of 50 Danaids. The number of coryphants may be half of that of the Danaids, i.e. perhaps six. 235. πέπλοισι βαρβάροισι: dress is always the safest sign of ethnicity in Greek drama, even more so than skin colour (cf. on 154–55): see Hall (1989) 136–38, and cf. on 122 = 133. 236. χλίοντα: pejorative, suggesting effeminacy, ‘oriental luxury’ (see Hall 1989, 128) and, sometimes, sex.390 verses in-between are of no consequence; (3) an imperative fits at least as well as a gnome at the end of a speech: cf. 204–24n., text for n. 362, and also 190–91 where a gnomic statement is followed immediately by an imperative. 387 1.9.9, 1.25.10, 1.26.11, 1.27.7, 1.33.11, 1.47.11, 1.58.9, 1.68.2, 1.90.4, 1.98.7, 2. 113.4, 2.119.8, 2.134.5, 2.134.15, 2.142.4. 388 Bodensteiner (1893) 707. However, Bodensteiner doubts whether a chariot was employed in the Supplices. 389 Taplin (1977) 77: see Pers. 155, Ag. 783, and Taplin (1977) 75–79 on the former passage. 390 Cf., e.g., Pers. 544 χλιδανῆς ἥβης τέρψιν, Supp. 1003, Ag. 1447, Pl. Smp 197d, Sapph. 60.8, which might perhaps read something like … ἄβᾳ χ]λιδάνᾳ πίθεισα (χ] suppl. Bechtel et al.: not Hunt, as claimed by Campbell ad loc.). The subject 141 236–37. Ἀργολίς … Ἑλλάδος τόπων: FJ–W complain about the logic (the Argolid is part of Hellas), but faulty logic is not by itself much ground for objection to an Aeschylean text. The expression is perhaps an attempt at antiquarianism; it is certainly an Homeric reminiscence: cf. Od. 1.344 καθ’ Ἑλλάδα καὶ µέσον Ἄργος (4.726, 4.816, 15.80). The Homeric usage of Ἑλλάς and Ἄργος was different from the classical one, in the Odyssey apparently meaning northern Greece and the Peloponnese, respectively.391 On Argos and the extent of Pelasgus’ reign, see 15n., 254–59 with notes. 238–39. οὔτε … ἡγητῶν: Hermann corrected οὐδὲ in 238, οὐδέ … τε being without parallel. Portus’ νόσφι θ’ is detrimental, however, as ἀπρόξενοί and νόσφιν ἡγητῶν in the second member are intimately connected: there are two, not three sides to this coin. Whereas κηρύκων, the official heralds, would be foreigners like the Danaids, πρόξενοι and ἡγηταί are both native helpers: the latter are ‘guides’ (see FJ–W), the former ‘patrons’ in general (cf. 419, LSJ s.v. I.2 and II). Thus, Pelasgus describes a polar predicament: the Danaids arrive (1) without proper announcement or official embassy, and (2) without friends on the inside to speak and cater for them (πρόξενοι, ἡγηταί). 241–42. κλάδοι … πρὸς θεοῖς: this is the first indication that the Danaids may have put suppliant boughs on the gods themselves (see 204–24n.). πρὸς with the dative may indeed mean that the boughs are wrapped somehow around, or placed on, the gods: cf. Pr. 4, 269, Ag. 996, fr. 210(?), LSJ s.v. B I 4. The matter is far from certain, however (see further 346–47n., 354–55n.). γε µὲν δή is always more or less adversative,392 here contrasting the Hellenic custom of supplication (see on 243) with the outlandish dress described above. παρ’ ὑµῶν (Auratus and Portus: παρ’ ὑµῖν M) has been adopted in would be erotic persuasion: ‘do not fight me … obey (the instincts of) your voluptuous youth’. The verse is probably !d'd'ds!: so Hunt ad loc. (p. 26); cf. West (1982) 32. Pace the latter, caesurae between the choriambs are presumably accidental: cf. e.g. frr. 58.10, 58.12, 62.7, 63.1–2, 63.7. 391 See S. West on Od. 1.344, Hoekstra on 15.80, Kirk on Il. 2.108, 2.529–30; and cf. also Str. 8.6.6. 392 Fraenkel on Ag. 887, D.GP 395. In Eu. 419 τιµάς γε µὲν δὴ the particle combination replaces a single δέ responding to γένος µὲν … κληδόνας τ’ in the previous line. The contrast is yet another variant of the λόγος–ἔργον dichotomy (see 207n.): the names of the Erinyes are known, soon their work will be. 142 recent Oxford and Teubner texts (Murray, Page, West), and it is indeed attractive. FJ–W argue that the emendation would exchange a common Attic idiom (κεῖσθαι παρά τινι) for a phrase hardly paralleled in Greek literature.393 The agent ‘by you’ is highly desirable in the context, however. If the expression becomes somewhat unusual, this is not due to any grammatical anomaly —the pregnant expression is perfectly regular (cf. above, 184–85n.)—but to unusual imagery. Also, the expression has fairly close parallels in Plato: cf. Smp. 197e ὁ παρ’ ἐµοῦ λόγος … τῷ θεῷ ἀνακείσθω, and also Lg. 926d τῷ δὲ ἡττηθέντι παρὰ τοῦ νοµοθέτου ψόγος καὶ ὄνειδος κείσθω. To defend the dative is certainly difficult: if we take πρὸς θεοῖς to mean ‘on the gods’, as argued above, the dative describing them as close to the Danaids is impossible from scenic considerations. If, on the other hand, the boughs lie ‘near’ the gods, the juxtaposition of two adverbials consisting of semantically almost identical prepositional phrases becomes very awkward without a conjunction. πρὸς θεοῖς <τ’> would then be an improvement. 243. συνοίσεται: ‘agree with’ (LSJ s.v. συµφέρω A.III.2, B.II). Cf. (with Headlam 1904) Call. Epigr. 5.6 ἔργῳ τοὔνοµα συµφέρεται, S. Aj. 431 τοὐµὸν ξυνοίσειν ὄνοµα τοῖς ἐµοῖς κακοῖς. The manner of supplication, as opposed to the dress and countenances, is in accordance with Greek custom. 244. τἄλλα is adverbial, ‘as regards the rest’ (see FJ–W), and does not go with πόλλ’. Martin’s (1858, 18) ἔτ’ (ἐπεικάσαι M) is likely on account of the caesura.394 245. παρόντι: perhaps the reference is to Danaus (see 246–48n.), and the sense is ‘he who stands beside you’ (LSJ s.v. I.2). Otherwise the reference is of the same implicit type (sc. τινί) as that of the generalising masculine plural discussed in 204n. For other examples of the masculine singular in this sense, see FJ–W. 246–48. εἴρηκας … ἀγόν; Paley’s (ed. 1883) suggestion that these verses are uttered by Danaus has been ignored by later critics and editors. It is not without merit, though. As for actual evidence, the attribution of the verses to 393 To FJ–W’s examples of κεῖσθαι παρά τινι, we may add S. Ichn. [fr. 314] 155, Ar. Pl. 742. 394 See Maas (1962) 66–67, West (1982) 82–83, and also FJ–W, Diggle (1982) and Kapsomenos (1983) for possible examples of iambic trimeters lacking a caesura in the second metron. 143 the chorus has no explicit support from the ms. tradition, which only marks the change of speaker with a paragraphus here.395 There are several advantages to letting Danaus utter the first words to the king. First, one would simply not expect a prince meeting the king of a foreign land to say nothing at all, allowing his daughters to do all the talking. Here lies the main problem of Danaus’ disappearance from the action in the following two hundred verses:396 as the leader of the supplicant host, he should at least say something. Then, perhaps, with the prerogative of the aged, he may delegate the business of negotiating to his industrious daughters (cf. Taplin 1977, 204 ff.). The initiative of questioning Pelasgus about his status would accord with Danaus’ function in the previous hundred verses, which has been to furnish his daughters with all the information they need to act properly. On the other hand, it is clearly the coryphaeus who answers Pelasgus in 274, and we hardly have any choice but to read the feminine participle ἔχουσα in 271 (q.v.), thus making Pelasgus address the chorus (or coryphaeus) directly there. At this point, we might speculate a little about a scenic movement designed to minimise the awkwardness of Danaus’ ‘disappearance’. If Danaus does speak 246–48, the scheme may have been the following: Danaus elicits this last piece of information for the benefit of the audience as well as his daughters, and then withdraws a bit—Pelasgus’ excursus on Apis (260–70), which may be directed at the girls (who have sung to the Ἀπία βοῦνις in 117, 129), gives him the opportunity to do so without insulting the king—climbing the hill (see 189n. and the Introduction, III 4) a little further and returning to his original business: watching the sea for his pursuers. Thus he lets his daughters do what he instructed them to do in 176–203: supplicating the king. This artifice would be much less convincing if the Danaids speak 246–48: we must then suppose that Danaus has withdrawn without a word during or just before Pelasgus’ first speech in 234–45. This would make him come across as something of a coward, or at least as very shy. The third option, to have Danaus remaining at the side of his daughters, saying nothing and being 395 As did the texts in Antiquity (see West 1973, 54–5). See Taplin (1977) 204 ff., G.AS 126–27 with refs on the problem of Danaus, which used to be taken as signs of the ‘immaturity’ of Aeschylus’ art and of the dif-ficulty of handling the novelty of a second actor. Both these assumptions were based on the notion that the Supplices is the oldest of Aeschylus’ preserved plays. 396 144 completely ignored the entire time they are talking to Pelasgus, is more awkward still. The argument that the awkwardness of Danaus’ (non-)behaviour is nothing but a ‘naturalistic prejudice’, and that characterisation always has to give in to dramatic effectiveness in Aeschylus,397 does not hold water: the awkwardness is a dramatic awkwardness, in that the most prominent persona is suddenly and inexplicably removed from the drama and yet remains on the stage, fully visible. It is of course impossible to determine exactly how Aeschylus disposed of Danaus when he was unwanted for dramatic purposes but had no rational cause for leaving the stage. It is more than likely, however, that he handled it in a way that did not cause too much disturbance to the audience’s sensibilities, seeing that the trilogy won the first prize (test. 70). At least part of the audience would be seasoned critics in these matters, as we see from Ar. Ra. 911–20, where Aeschylus is made fun of for having silent characters on stage.398 The difference between Achilles and Niobe on the one hand and Danaus on the other is that the former had strong, story-internal reasons for keeping their silence.399 246. ἀµφὶ κόσµον: ἀµφί + accusative in the abstract (non-local) sense usually implies a powerful emotional interest in the object: fight about or because of, cry over, grieve about, sing hymns to or about, etc. Elsewhere the neutral sense ‘speak about’, ‘tell about’ is found (periphrastic for ‘sing about’) only in the highly ceremonious context of the prooemia or invocations of the Homeric Hymns, e.g., 7.1–2 ἀµφὶ ∆ιώνυσον … µνήσοµαι.400 247. ἔτην: on this noun, see Rutherford (2001) 308, n. 8, with refs. 248. τηρὸν ἱερόρραβδον: both hapaces, the latter conjectural but certain (ἠτηρὸν ἡερου ῥάβδον M).401 249. πρὸς … ἐµοί: the verse is suspect. The imperatives at the beginning are out of place, seeing that a long speech from Pelasgus is to follow and that the Danaids are not given a chance to speak until 274. The nearest parallel 397 See 176–78n. with refs and cf. G.AS 127. In the Myrmonides, the Niobe and the Phryges: see Radt ad locc. (pp. 239–40, 265–66, 365). 399 See also 1–39n., text for n. 94, on Danaus’ silent activities on stage. 400 It was a stereotype of the dithyramb, where the formula went ἀµφὶ … ἄνακτα: see Allen–Halliday–Sikes ad loc. and cf. Σ Ar. Nub. 595, Terp. fr. 697, Cratin. fr. 72 PCG, Ar. fr. 62 PCG. 401 ἱερόραβδον Anon.Ald.: -ρρ- Tucker (ἱρο-) and Headlam ad loc., n. 6 (ἱερό-). For the necessity of duplicate -ρρ-, see FJ–W. 398 145 for such a thing seems to be Danaus’ imperative in 191 βᾶτε, which is not, however, in any way comparable: there the instructions to the Danaids go on, with a string of imperatives following immediately, not ending until 232–33 (see ad loc. and on 204–24 for the problems with the text). Here, a long speech from Pelasgus follows, concerning itself with Pelasgus, not the Danaids, and the imperatives are left without consequence. Moreover, the reference of πρὸς ταῦτ’ is very unclear: elsewhere the phrase always refers back, and it is always taken with the finite verb of the clause (see FJ–W). The reference here must somehow be to the request of Danaus or the Danaids in 246–48, which is incompatible with the imperatives.402 Instead one would expect the king to say something to signal his own willingness to answer: cf., e.g., E. Hipp. 697 ἔχω δὲ κἀγὼ πρὸς τάδ’, εἰ δέξῃ, λέγειν. The solution might be a lacuna—not before 249, as reluctantly suggested by FJ–W, but after πρὸς ταῦτ’. If a conjectural 249b begins with <σὺ δ᾿ αὖτ’> ἀµείβου, the corruption would be easily explainable by ‘parablepsy’, as would the imperatives in the new position, being uttered as a quid pro quo; αὖτ’ gives delayed effect to the imperatives ἀµείβου and λέγε: ‘I shall answer your question: you, in return…’. Thus, for instance, 249a 249b 402 πρὸς ταῦτ<α δείξω µὲν τὰ χρὴ τεκµήρια· σὺ δ’ αὖτ’> ἀµείβου καὶ λέγ’ εὐθαρσὴς ἐµοί. Paley, Wecklein (ed. 1902) a.o. take πρὸς ταῦτ’ = ‘for that matter’, ‘as for my rank’. Apart from FJ–W’s having shown that this is an unparalleled, and probably impossible, meaning of πρὸς ταῦτα, the result is nonsense: ‘as for my rank, answer me…’. The Danaids are not asked to expound on Pelasgus’ rank, but to disclose their own identity and business in Greece. Several emendations have been proposed, e.g. πρὸς πάντ’ (d’Arnaud 1728, 262), πρόσω τ’ (Friis Johansen ap. FJ–W), πάραυτ’ (Griffith 1986). None of these, however, takes into account the oddity of placing the imperatives at the beginning of the speech. Valckenaer on E. Ph. 1331 suggested a transposition of the line to follow 245, where it is completely out of place: moreover, the γάρ in 250 links it closely to this verse, and the conjectural attempts at removing this have been futile and deleterious (πάρειµ’ Burges 1811, 183; µέν Abbott 1850). Ercolani (2001) approves of Abbott’s µέν and suggests that 249 is spoken by the coryphaeus: however, the blunt imperatives are certainly out of place in the mouth of the suppliant (whether the coryphaeus or Danaus himself), and it is incredible that Pelasgus here should be told to ‘have courage’ (λέγ’ εὐθαρσής) by those who are entirely dependent on his good-will (Ercolani’s semantics in n. 34 are not helpful in this respect). 146 Cf., e.g., Eu. 226, 468, and also the reverse structure in 520–22: πρὸς ταῦτα µίµνε … ἐγὼ δὲ κτλ. Turnebus’ λέγ’ εὐθαρσὴς (λέγετ’ εὐθαρσεῖς M) has been adopted by most editors, including FJ–W, but Whittle’s (1961) εὖ θαρσοῦσ’ is demonstrably more idiomatic (see ibid., FJ–W), if somewhat more difficult palaeographically. Both the adjective and the expression εὖ θαρσεῖν are attested in Aeschylus (cf. 968, 1015, Th. 34, Ag. 930). 250–51. γηγενοῦς: Palaechthon was apparently one of the αὐτόχθονες, who were born of the earth itself: cf. Pl. Plt. 269b, Arist. GA 762b29. He was not the first inhabitant of Argos, however, as is evident from the following verses: the Danaids are descendants of Io, who lived in Argos five generations before. Πελασγός is Canter’s correction of M’s πελασγοῦ, which must be written down as a peculiarly absent-minded scribal error, perhaps influenced by the ending -οῦς of γηγενοῦς directly above in the previous verse. In several sources the king of Argos at the time of the Danaids’ arrival is named Gelanor.403 Aeschylus’ Pelasgus has little in common with the other mythological characters going under this name, except for lending his name to the Pelasgian race.404 252. εὐλόγως ἐπώνυµον: cf. 45–47. 253. Πελασγῶν … χθόνα: contemporary ‘archaeology’ held that the Pelasgians were an indigenous people in Greece and the eastern Mediterranean, but supplanted and/or assimilated by Dorians and others.405 Whether the Pelasgians were Greeks or barbarians seems to have been disputed: Homer ranges them among the Trojan allies (Il. 2.840–43), although the leaders have Greek names (see Kirk ad loc.); Herodotus’ guess (1.57) is that they did not speak Greek. Aeschylus apparently considers Pelasgus’ people as Greeks, 403 [Apollod.] 2.1.4, Paus. 2.16.1, 2.19.3–4, and, strangely, Plu. Pyrrh. 2.10, who has read the Supplices (cf. 214n.); also Geo.Sync. p. 178 Mosshammer. Σ Il. 1.42 quotes pseudo-Apollodorus on the matter almost verbatim, but changes the name of the king to Ἑλλάνωρ (cf. ΣD Il. 1.42 Heyne). 404 See P. Weizsäcker in Roscher iii. 1817–21 (s.v. ‘Pelasgos’). The most famous one seems to have been Arcadian, not Argive. Hesiod claimed that this Pelasgus was earth-born (fr. 160, ap. [Apollod.] 2.1.1, etc.), as did the epic poet Asius (fr. 8 PEG). 405 Cf. Hdt. 1.56–58, 6.137–140, Hecat. frr. 119, 127, Acus. fr. 11, Hellanic. fr. 4. 147 however.406 On Pelasgians and the view on them in classical Greece, see further Gomme on Th. 1.3.2. Homer once speaks of ‘the Pelasgic Argos’ (Il. 2.681), a phrase which often recurs in tragedy, especially in Euripides.407 However, according to Kirk ad loc. Homer referred not to the Peloponnesian Argos, but to ‘the region of the Sperkheios river and the Malian plain’ in south Thessaly. 254–55. Illogical: does Pelasgus rule over all the land (πᾶσαν αἶαν) of the river Strymon, or only over the western part? A look at the map suggests that the latter is intended: the Strymon, situated in western Thrace, makes a natural eastern border for Pelasgus’ pan-Greek (cf. 256–59n.) kingdom. One should probably not make too much of the deficient logic: Pelasgus simply exaggerates a bit, correcting himself at the last moment at the expense of consistent syntax (cf. 15–18n.). Note, however, that διά + gen. may denote extension along the side of something (LSJ s.v. A.I.4: ‘in Prose’), which would make the expression unexceptionable. 256–59. ὁρίζοµαι here means either ‘include within my borders’ or ‘have as outer border’. The latter is in any case the intended message: Pelasgus describes the northern border of his kingdom, from east to west. Obviously Pelasgus’ kingdom includes all of the Greek mainland. On Argos as the seat of power, see 15n. Friis Johansen (1966) makes a convincing case for transposing Περραιβῶν and Παιόνων in 256–57, as the latter people, according to all extant sources, lived by and in-between the rivers Axius and Strymon in Thrace and/or Macedon,408 whereas Mt. Pindus is situated in north-western Greece, near Dodona. The Perrhaebians, furthermore, are explicitly said to live περὶ ∆ωδώνην by Homer (Il. 2.749–50).409 West (W.SA) defends the ms. reading, 406 Cf. 220, 237, 243, 914. On Aeschylus’ Pelasgians, see also Kranz (1933) 79, Hall (1989) 171–72. 407 E.g., Ph. 256, Or. 692 (where see Willink), 1247, 1296, 1601. 408 Il. 2.848–50, 16.287–88, 21.154–58, Hdt. 5.1, 5.13, 5.98, etc., Th. 2.96.3–4, 2.98. 1–2, 2.99.4. The Paeonians are also placed in Thrace by Hecat. fr. 152, and more specifically in Chalcidice by Pi. Pae. 2 (fr. 52b) 61. Cf. also, e.g., [E.] Rh. 407–9, Jacoby on Hecat. fr. 150–57 FGrH (i. 346–47), Rutherford (2001) 270–71. 409 If Homer intended a Thessalian Dodona (so W. SA 135, comparing h.Ap. 218, pace Kirk on Il. 2.749–51), and not the one west of Mt. Pindus (in which the famous oracle was situated, cf. 223–26n.), this may have been lost on Aeschylus. Cf. also S. fr. 271 ῥεῖ γὰρ [sc. ὁ Ἴναχος] ἀπ’ ἄκρας Πίνδου Λάκµου τ’ ἀπὸ Περραιβῶν εἰς Ἀµφιλόχους καὶ Ἀκαρνᾶνας. 148 arguing, first, that the Perrhaebians in fifth-century literature are depicted as living north of Thessaly and north of the river Peneius, in the Tempe valley (Hdt. 7.128, 7.131, 7.173, Th. 4.78.5–6); this he also supposes to be the opinion of Aeschylus himself in the Perrhaebides (frr. 184–186a; see Radt ad loc., p. 300). Even so, that still puts them well to the south-west of the Paeonians. Secondly, West argues that Παιόνων πέλας means that the Paeonians are situated outside the borders of Pelasgus’ kingdom (so also Hall 1989, 171), and that ‘in order to give some general indication of the northern boundary Aeschylus chose to name a large barbarian nation’, with a rather dim conception of its location. However, the location of the Paeonians in classical times was apparently well known to most of Aeschylus’ famous contemporaries (see n. 408) as well as stated in Hecataeus (cf. 220–21n., n. 381). The ethnographic and geographical competence of Aeschylus is well documented,410 and the Paeonians, who fought on Aeschylus’ side in the Persian war (Hdt. 5.12–15, 7.185, etc.), would not be an obscure barbarian nation to him. In his account of the Persian war, Herodotus repeatedly mentions the Paeonians as living by the Strymon and in its vicinity.411 It is also unlikely that Aeschylus would overlook the Homeric evidence:412 the river Axius, mentioned three times by Homer as the home of the Paeonians in Heroic times, is by all accounts very far to the north-east of Mt. Pindus, on the opposite side of the Greek peninsula. The only way to understand 257 as transmitted is that the Paeonians are close to ‘Pindus and beyond’. This is unacceptable. Apart from Friis Johansen’s transposition, a solution was presented by Tucker in the conjecture Χαόνων for Παιόνων. The Chaonians, being situated ‘in the middle of Epirus’ according to Hecataeus (fr. 105), are indeed near to ‘Pindus and beyond’. However, the very large number of accounts of the Paeonians living by and in-between the Strymon and the Axius makes Friis Johansen’s transposition more attractive, as the Strymon is actually mentioned just before.413 259. ὑγρᾶς θαλάσσης: Chadwick (1996, 297) suggests ὑγρὸς, ‘wet’ being 410 See 3n., 279–90n., Kranz (1933) 79–80, Bacon (1961) 45–59, Hall (1989) 75–76 and also the refs in 220–21n., n. 381. 411 5.1, 5.13, 5.15–16, 5.98, 7.113, 7.124. 412 Cf., e.g., 15n., 63n., 122 = 133n., 236–37n. 413 Tucker over-ingeniously emended 254–55 to Ἅλιος ἔρχεται | Ἄκµων, taking the reference to be to the river Haliacmon north of the Tempe valley. 149 otiose as an epithet to θαλάσσης. However, it is a stock epithet of the sea, used in for instance Homer and Pindar, as noted by Chadwick himself.414 ὅρος ὑγρὸς θαλάσσης, ‘wet boundary of the sea’, will perhaps not be otiose but certainly stale and awkward. τῶνδε τἀπὶ τάδε: Canter’s certain emendation of the mss.’ τἄπειτα δὲ. The result is apparent redundancy: ‘I rule the land on the hither side (of the border)’. Perhaps τῶνδε is meant to refer to the mountains of Pindus and Dodona and the expression intended to clarify that these are indeed the outer border of the kingdom. 260–70. This excursus is somewhat uncalled for, and moreover confusing as it introduces an Apis who is distinct from the one sung about in 117 and 128 (q.v.). For speculation as to a possible scenic justification, see 246–48n. The notion that the Apis who gave his name to Apia (= Peloponnesus, see 117 = 128n.) was a son of Apollo is unique to Aeschylus (see Roscher i. 422, s.v. ‘Apis’ 5). In later sources this Apis is a cousin of Epaphus, being the son of Phoroneus and grandson of Inachus,415 neither of whom is mentioned by Aeschylus in the preserved tragedies and fragments. Apis’ identity as a physician son of Apollo, and the connection with snakes, is suggestive of Asclepius. 260–61. αὐτῆς … κέκληται: for the construction, see LSJ s.v. καλέω II.3.a. 265. τὰ δὴ: Turnebus’ correction of M’s τὰ δὲ. Dindorf’s (1873, 234b) ἃ δὴ has been adopted in the text by FJ–W, who observe that there is no certain example of a form of the relative pronoun on τ- in Aeschylus which is not used metri gratia. But τὰ δὴ is the better tradition in Ag. 342; note also the demonstrative τοί at Pers. 424, and see West p. xl. παλαίων αἱµάτων: these are unknown to us, but perhaps not to the scholiast who explains ὡς τῶν πολιτῶν αὐτοκτονησάντων (see FJ–W). 266. †µηνεῖται ἄκη†: FJ–W rightly observe that an apposition or predicative to τὰ δὴ416 requires an adjective or some other qualification. It appears likely that µην- is sound, being part of a nominal describing the wrath of the earth.417 One of the more attractive suggestions, in my opinion, is Weil’s 414 He suggests that the adjective originally meant ‘running as opposed to stagnant water’. 415 Rhian. fr. 13, Σ Il. 1.30, etc.: see Roscher l.c. 4. 416 ἄχη Martin (1858, 18; cf. Ch. 585–86), δάκη Turnebus, ἄγη Schwerdt (1863, 99; cf. Schwerdt 1886, 130). 417 Cf. Pl. Phdr. 244d–e: νόσων … καὶ πόνων …, ἃ δὴ παλαιῶν ἐκ µηνιµάτων 150 µήνιος τέκη. Burges’ (1811, 183) µῆνιν ἐνδακῆ is also worth mentioning, the abstract apposition not being out of place, as we can see from ξυνοικίαν in the following verse. The adjective ἐνδακής is not found elsewhere, but Burges compares Hsch. α 8405 αὐτοδακὴς µῆνις and reads ἐνδακεῖ· ἐµµανεῖ in Hsch. ε 2728 (ενδαγει mss.). Cf. also ε 2727: ἐνδακοῦσα· κατεσθίουσα. A middle participle would accommodate the η at the end of the verse. Margoliouth’s (1883, 19) µηχανωµένη is weak, however: µηχανή is a human contrivance, a direct opposite to nature’s produce, and the conjecture removes the root µην-. µηνιωµένη would fit, but µηνιάω is not attested earlier than the 1st century B.C. (D.H. Rh. 9.16, Lxx Si. 10.6, etc.) and the form µηνίω is found in Aeschylus, the middle voice in Eu. 101.418 268. ἄκη τοµαῖα καὶ λυτήρια: the phrase is formulaic and virtually untranslatable: my ‘knife and solvent’ is a rather desperate attempt at rendering some of the aspects of the expression. (ἐν-)τέµνω is commonly used with φάρµακον and the like (cf. Ag. 848–49), and, in extension, metaphorically with ἄκος and with ‘remedies’ in general.419 In Aeschylus, related expressions are found referring both to the cutting of herbs and, metaphorically, to ‘violent remedies’, i.e. killing, probably hinting at surgery.420 In the present case the notion of surgery is not very relevant, however. The coupling of τοµαῖα with λυτήρια makes it hard to take the latter as substantival, as FJ–W ποθὲν ἔν τισι τῶν γενῶν ἡ µανία ἐγγενοµένη καὶ προφητεύσασα, οἷς ἔδει ἀπαλλαγὴν ηὕρετο, καταφυγοῦσα πρὸς θεῶν εὐχὰς καὶ λατρείας, ὅθεν δὴ καθαρµῶν τε καὶ τελετῶν τυχοῦσα ἐξάντη ἐποίησε, and E. Ph. 931–35 δεῖ τόνδε θαλάµαις, οὗ δράκων ὁ γηγενὴς | ἐγένετο ∆ίρκης ναµάτων ἐπίσκοπος, | σφαγέντα φόνιον αἷµα γῆι δοῦναι χοάς, | Κάδµωι παλαιῶν Ἄρεος ἐκ µηνιµάτων, | ὃς γηγενεῖ δράκοντι τιµωρεῖ φόνον: ‘The use of [παλαιῶν … ἐκ µηνιµάτων] in Pl. [l.c.] is either a reminiscence of Tir.’s speech or evidence that the phrase was traditional in religious or oracular language connected with expiation’, Mastronarde ad loc. Probably the latter: cf. also Pl. Lg. 854b οἶστρος δέ σέ τις ἐµφυόµενος ἐκ παλαιῶν καὶ ἀκαθάρτων τοῖς ἀνθρώποις ἀδικηµάτων. 418 Other notable suggestions are, e.g., West’s γαῖ’ ἄµαχα µηνίµατα (cf. the previous footnote), µηνίοντ’ ἄχη Hadjistephanou (1991). Headlam’s (ap. Blaydes 1898) µηνίσασ’ ἄγει (µηνίσασ’ ἄχη already Martin 1858, 18) is incorrectly attributed to Blaydes in Dawe (1965) and in West’s apparatus criticus. µηνίω elsewhere takes gen.rei and dat.pers. 419 See LSJ s.v. τέµνω II.3, and cf. 807, Ag. 17, Ch. 539, Dikt. 779 (fr. 47a 15), E. Andr. 121. For λυτήρια, cf. also Eu. 645–46. 420 See Fraenkel and Thomson on Ag. 849 (the latter’s 837–41n.), Garvie on Ch. 539. 151 suggest (so in E. Melanipp.Sap. 17 ἄκη πόνων φράζουσα καὶ λυτήρια421). As attributes to ἄκη, the adjectives τοµαῖα and λυτήρια are disconnected, the one referring to the means, the other to the ends of the remedy: ‘cut and deliverant’. The same peculiarity is found, however, in Eu. 558–59: ἐν µέσᾳ δυσπαλεῖ τε δίνᾳ (cf. D.GP 501). Perhaps this could pass for a peculiar form of hendiadys.422 270. µνήµην … ἐν λιταῖς: Apis received ‘mention in prayers’ instead of a fee,423 i.e. he was honoured as a god or a hero: for the expression cf. E. Ba. 46 ἐν εὐχαῖς … οὐδαµοῦ µνείαν ἔχει and the formula from the end of several Homeric hymns, (καὶ σεῖο / ὑµέων τε) καὶ ἄλλης µνήσοµ’ ἀοιδῆς.424 In the light of these parallels, Rose’s suggestion that ἐν λιταῖς refers to Apis’ asking to receive honours (‘by his entreaties’) is hardly a possible interpretation. There is no evidence that a cult of the Argive Apis has ever existed, however (see further FJ–W 268–70n.). Kirchhoff’s τότ’ (ποτ’ ἀντίµισθον Turnebus, πονταντινεισθον M), adopted in the text by West, is attractive, agreeing well with the aetiological narrative: ‘it was then, that …’ (cf. Eu. 688, Pi. O. 6.70, 7.39). Turnebus’ ποτ’, being palaeographically easier, seems somehow too vague after the carefully narrated ‘aristeia’ of Apis. 271–73. Heimsoeth’s (1861, 420) ἔχουσα δ᾿ is the modern vulgate for M’s ἔχον δ’ ἂν in 271 (ἔχουσ’ ἂν Victorius, ἔχουσαν Mγρ). This would mean that the king is addressing the chorus, or coryphaeus, here. The coryphaeus is certainly answering to the request, and the king appears to be addressing the girls in 236–42, which makes this a plausible assumption (on Danaus’ participation, see 246–48n.). However, the corruption ἔχουσ > ἔχονδ is tenable only if the source copied is in minuscule lettering (see 110–11n., n. 276). *Klausen’s conjecture, ἔχων δ’ ἂν (ἔχων {δ’} ἂν Paley ed. 1855), postulates an easy phonological corruption of ω > ο which is common in M, but unparalleled in the case of such an easy reading as ἔχων.425 It would also 421 TrGFS p. 121, Pap.poet. fr. 14. See further FJ–W, Bollack and Fraenkel on Ag. 17, Lloyd-Jones (1978) 48–50 (321–23). 423 On the customary physician’s fee, see Thgn. 1.432–34, Pi. P. 3.50–57, Heraclit. fr. 58, and FJ–W for further post-Aeschylean references. 424 Also Isoc. Paneg. 43 εὐχάς … ποιησαµένους ἀναµνησθῆναι. 425 FJ–W iii. 372 list 107, 193 (σεµνὸς Mac), 204, 229, 366 (ἐµὸν Mac), 495, 625 and 958 (δεδωµάτοµαι Mac) as examples of this corruption in M. 422 152 mean that Pelasgus is addressing Danaus here, which would not be impossible if Paley’s (ed. 1883) attribution of 246–48 (q.v.) to him is correct. The tone of 273 is less than polite, however, and suggests that the king is addressing younger persons (contrast Theseus’ address to the suppliant Oedipus in S. OC 551 ff.). Moreover, the duplicate ἂν is out of place, as it would make Pelasgus’ statement hypothetical: see Fraenkel on Ag. 1048. 272. γένος τ’: on the position of τ’ see D.GP 518. 273. On Argive brevity of speech, see 201n. On the particle combination γε µὲν δή, see 241n. 274–75. γένος … σπέρµα τ’: on τ’ coupling two equivalent designations of the same thing, cf. 42, 62. Rules and categories pertaining to this usage are arranged by FJ–W ad loc. and 62n. Cf. also my notes on 42 and 60–62, D.GP 503. εὐτέκνου: as giving birth to Epaphus. 276. χὠς ταῦτ’ ἀληθῆ πάντα προσφύσω λόγον: ‘and that/how this is true, I shall fit in the entire evidence’, i.e., ‘I shall account in detail for the truth of this’. Or, if πάντα goes with ταῦτ’ ἀληθῆ, ‘I shall account for the truth of all this’. The matter may be deliberately vague; cf. especially 32 ξὺν ὄχῳ ταχυήρει, 78 εὖ, and see 15–18n. with further refs. Whittle’s (ap. Friis Johansen–Whittle 1975) λόγον (λόγων Mac, λόγωι Mpc) is easy and expedient together with Sommerstein’s (1977) χὠς (καὶ M).426 πάντα … λόγον is common in tragedy: cf., e.g., Pers. 246, Ag. 592, 599, Dikt. 785 (fr. 47a 21).427 Sommerstein backs up the subjunctive clause ὡς (ταῦτ’) ἀληθῆ (λέγω, etc.), a commonplace in forensic speeches (cf. 40–175n.), with twelve parallels from Lysias alone.428 The dative in the similar expression found in Ar. Nu. 372 τῷ νυνὶ λόγῳ εὖ προσέφυσας is irrelevant as a parallel:429 the Aristophanean expression 426 Whittle’s emendation assumes a phonological corruption (ο > ω) that is very common in M: FJ–W iii. 370 list 29 examples in Supp. only, with an exact parallel (λόγον > λόγων at the end of a trimeter) in 608. Sommerstein’s χὠς may imply yet another example of a corruption owing to a minuscule source (see 110–11n. with n. 276): for the corruption of the ligature, cf. 194–95, 296, 504. 427 Also Pr. 193, S. Tr. 484, Aj. 480, 734, OT 291, Ph. 1240, etc. 428 Apart from the orators, one may add the parallels of Hdt. 4.92 η, 8.77. 429 The conjectural alternatives of Page’s OCT and West’s Teubner preserve the dative, but at the cost of coherence and/or critical economy. Page’s τῷδ’ for M’s ταῦτ’ produces an expression he himself paraphrases as καὶ τῷδε λόγῳ λόγον παντελῶς ἀληθῆ προσφύσω: on the artificiality of this (ἀληθῆ πάντα = λόγον παντελῶς ἀληθῆ?), see FJ–W. West (argument in W.SA 137–38), adopting Sommerstein’ s 153 means ‘you attached it (the last argument) well to your present argument’, i.e. ‘your argumentation is consistent’, whereas in our case we have as yet no argument, no λόγος: this is what the Danaids are promising to ‘attach’ to their claim. λόγῳ has thus been taken to mean ‘claim’, ‘proposition’ here (see W.SA 138), which goes against the rationale of the word. λόγος as good as always denotes reason, explanation, narrative, discourse, etc. The Danaids’ ‘proposition’ is called µῦθος in 274, and something might indeed be made of a mythos–logos dichotomy here: to generalise, µῦθος is the ‘plain’ word, the story, λόγος the persuasive reason and argument. Here λόγος, argument, is ‘fitted’ to the µῦθος in order to prove the latter.430 FJ–W doubt that Mpc λόγωι (also in Md) could be an emendation by the diorthotes (see 8n., n. 104) but this is in fact rather likely, the emendation being elicited by what was felt (and is still felt by a majority of scholars) as a need for a dative with προσφύσω. The diorthotes elsewhere produces datives (rightly or mistakenly) by emendation in 111 bis, 122, 133, 147, 687 bis, 956 and 1041.431 However, there is no need for a dative. Parallels for προσφύω may be sought in the adverb προσφυῶς rather than in the concrete usages of the verb: προσφύειν λόγον appears to be a warped paraphrase of προσφυῶς λέγειν, ‘speak reasonably’ (Hdt. 1.27;432 cf. LSJ s.v. προσφυής II). 277–78. ἄπιστα µυθεῖσθ’ … ὅπως: cf. Ar. Pax 131–32 ἄπιστον εἶπας µῦθον … ὅπως κάκοσµον ζῷον ἦλθεν εἰς θεούς.433 278. γένος has a concrete sense, referring to the Danaids themselves, not χὠς, also prints Zakas’ (1890) πιστὰ for πάντα (‘and that this is true, I shall graft trustworthy guarantees on to what I have said’), suggesting an echo in Pelasgus’ ἄπιστα µυθεῖσθ’ in the subsequent verse. This echo (πιστὰ προσφύσω – ἄπιστα µυθεῖσθ’) is a banality compared to the ones West adduces as parallels (350–354, 375–376, 396–397, 437–438); moreover, the phrase ἄπιστα εἰπεῖν is a commonplace (see 277–78n.) and as such not very serviceable in a verbal echo of this type. 430 Cf., for instance, Pl. Smp. 189b λέγε ὡς δώσων λόγον: ‘speak only what you can defend’ (Lamb). 431 The dative, being obsolete in the Byzantine vernacular, would presumably—by the same psychological process as produces so-called ‘hyper-correct’ idiom in second-language acquirers—be extra tempting to a textual critic versed in Attic Greek. 432 Noted by Schwerdt ad loc. Cf. also, e.g., Ph. LA 3.161, Aet.mund. 54, Aristonic. Sign.Il. 9.17, Arr. Bithyn. fr. 64, D.H. Th. 5. 433 For ἄπιστα µυθεῖσθ’ cf. also E. IT 1293, Hel. 1520, El. 350, Pi. N. 9.33, Men. Sam. 545, Th. 6.33, X. Hier. 1.9, Pl. Thg. 130d, Demod. 385e, 386a, Lys. 3.24. 154 to the abstract ‘race’ or ‘family’, as is obvious from τόδ’ and the following lines. Wilamowitz and FJ–W suggest a conflation of two expressions (cf. my 15–18n.), including two different meanings of γένος: ‘ὅπως ὑµῖν Ἀργεῖον γένος (“lineage”) ἐστίν and ὅπως Ἀργεῖον τόδε γένος (“tribe”) ἐστίν’ (FJ– W). ‘Tribe’ is too narrow, however: the concept still means something like ‘race’, ‘breed’, or perhaps even more accurate, if rude in English, ‘batch’ (cf. 281 τοιοῦτον φυτόν). It denotes the Danaids as a collective sprung from the same source: ‘explain to me how this batch of yours is Argive’. ὕµιν is remarkable, as it might be said to refer to the same thing as τόδε γένος, i.e. the Danaids. Blaydes’s (1895) ὑµῶν is notable (‘consisting of’), but hardly necessary (cf. the similar problems in 134–35 and 186–87). Formally, the dative can hardly be adnominal; it has to go with the entire clause, as a ‘dative of interest’ or perhaps an ‘ethical’ dative: thus the enclitic form is preferable (cf. above, 233n.). The semantic effect does approximate the possessive, though.434 279–90. A competent ethnographical exposé, with Aeschylus perhaps showing off his knowledge a bit (cf. 256–59n.). Most of Pelasgus’ guesses involve a north-African origin on the part of the Danaids; their costumes probably suggested as much (cf. Hall 1989, 84, n. 127). Thus the tension is heightened by having Pelasgus making intelligent, plausible guesses, but lacking the vital information that would make him arrive at the truth. 279–81. Pelasgus’ two best guesses come first: Libya is Danaus’ grandmother (317) and the Nile valley his and his daughters’ native soil. 279. On the double comparative with µᾶλλον, see K–G i. 26. 282–83. κύπριος χαρακτήρ κτλ: the sense of these verses has long been discussed, without anyone having been able to arrive at a definite conclusion. The problems may be stated as follows:435 (1) The Cypriots are out of place in the enumeration of barbarian, mostly African, nations. Cyprus was colonised by Teucer, according to 434 Cf. K–G i. 421–23 (§ 423.18b, 18d), S.GG ii. 189–90. For a remarkable, if sound, example of adnominal ‘possessive’ dative, see A. Th. 926 (cf. my note in Sandin 2002, 149, n. 14). The use of dative for genitive is referred to by Lesb.Gramm. 8 as a σχῆµα Κολοφώνιον. 435 The metre led Wilamowitz to doubt Κύπριος, but the short υ (which is indeed natural; see LSJ s.v. Κύπρις) is paralleled in Pers. 891, and the initial anapaestic foot in Pers. 343 and Ag. 509. 155 myth,436 and it was seen as Greek; it is for instance included in Aeschylus’ list of Greek islands governed by Darius in Pers. 880–95.437 Of course, since the Supplices takes place several hundred years before the Trojan war, one might theoretically suppose that Aeschylus imagines an ancient, pre-Greek Cyprus; but this just seems too far-fetched and too demanding of the audience, who would see Cyprus as ‘the eastern limit of the Greek world’ in the words of Dodds on E. Ba. 402–16. (2) The imagery, apparently taken from handicraft, is obscure. In the context of handicraft, the words τύπος and χαρακτήρ usually denote impression, engraving, relief, and (especially) coinage; but why is the ‘Cyprian impression beaten’ by expressly ‘male artisans’? Critics have supposed (with the scholium καὶ γυναῖκες ἂν Κύπριαι ἀνδράσι µιγεῖσαι τέκοιεν καθ’ ὑµᾶς) that an image of a sexual nature is hidden behind the handicraft-metaphor, and that the male artisans are the fathers who produce their offspring in γυναικεῖοι τύποι, the mothers (cf. 48n.). This confuses the image beyond reason: in the context of image-making and handicraft, ‘female forms’ must refer to the Danaids, who are likened to artefacts of one kind or another: ‘a Cypriot χαρακτήρ in female moulds’. (3) The position of τ’ is suspect: the postponement after noun and attribute is rare, with 432, E. Tr. 1064, and Ar. Av. 257 as the only more or less certain examples in non-epic verse. All these occur in lyrical verse, with further mitigating circumstances: see 432n.438 Unlike the situation in these cases, τε here connects an entirely new sentence, which is too awkward. 436 Marm.Par. 26, Pi. N. 4.46–47, Isoc. Euag. 18, Clearch. fr. 19 (ap. Ath. 6.256b), Str. 14.6.3, Paus. 8.15.7, etc. Possibly Aeschylus’ Salaminiae (frr. 216–20) refers to this event, i.e. to the Cyprian city, not the island of Salamis. 437 Cf. also, for instance, Isoc. Paneg. 134; and see Hadjistyllis (1985) 517–19 for a review of the evidence in favour of a predominantly Greek population and culture on Cyprus in classical times. Cf. also Molyneux (1985) on Cyprus in Greek lyric poetry, and speculation as to the existence of a national Cypriot (lyric) literature. For a contrary argument, to the effect that Aeschylus indeed regarded the Cyprus of the time of the Supplices as ‘barbarian’, see Sommerstein (1977) 71 and Thomsen (1995) 33–34, who defend the present verses. Thomsen argues that ‘the Greeks’ attitude to the barbarians, with all its generalisations, stereotypes and blind spots, coloured their attitude to the Cypriots’. It is one thing to regard Cypriots as ‘barbaric’, however, and another one completely to suggest that they are barbarians and not Greek, which is the effect of Pelasgus’ words here. 438 See D.GP 517; of other examples, E. Alc. 818–19 are interpolated, yet others are due to conjecture (Tr. 1069, A. Ag. 229, S. fr. 859). 156 (4) εἰκὼς is awkwardly left without a dative. Common sense recommends Murray’s (ed. 1955) suggested interpretation, similis vobis, or perhaps a construction of ἐοικέναι resembling that of εἰκάζεσθαι, i.e. = Κύπριος εἰκασµένος χαρακτήρ, ‘what seems like a κύπριος χαρακτήρ’ (cf. Hdt. 3.28 and also 287–89n. below); but this appears to be without parallel, and πέπληκται is still awkward. FJ–W instead take εἰκὼς to refer to the agent τεκτόνων … ἀρσένων, ‘like to the male artisans’, which they take to mean the fathers (see under [1] above), but this interpretation comes across as less natural; moreover, the emphasis on a likeness between fathers and daughters is, as FJ–W themselves note, irrelevant and awkward in the presence of Danaus. To start with a new interpretation of the second crux, we may note that the juxtaposition of τύπος and χαρακτήρ is found in an interesting passage from Plutarch which deserves to be quoted here (Gen.Socr. 577f): <ἐπάνω δὲ> τοῦ µνήµατος <ἔκειτο> πίναξ χαλκοῦς ἔχων γράµµατα πολλὰ θαυµαστὸν ὡς παµπάλαια· γνῶναι γὰρ ἐξ αὑτῶν οὐδὲν παρεῖχε καίπερ ἐκφανέντα τοῦ χαλκοῦ καταπλυθέντος, ἀλλ᾿ ἴδιός τις ὁ τύπος καὶ βαρβαρικὸς τῶν χαρακτήρων ἐµφερέστατος Αἰγυπτίοις. The last clause translates something like ‘but the engraving was peculiar and foreign, of letters most alike to Egyptian ones’.439 The question arises: might the enigmatic Κύπριος χαρακτήρ refer to the Cypriot script, to the syllabary derived from Mycenaean Linear B? This was distinct from all archaic and classical Greek alphabets and would presumably seem incomprehensible and ‘barbarian’ to non-Cypriot Greeks. One is tempted to understand the phrase Κύπριος χαρακτήρ, ‘Cyprian letter(s)’, as an idiom denoting something that is foreign and incomprehensible, like the English ‘it is Greek to me’. A tentative translation: ‘(what seems like) a Cypriot script is engraved in female shapes by male craftsmen’. Still, a number of problems remain—all those listed under (2) to (4), in fact, as well as a few more. The above translation supposes that the verb πλήσσω is used here for engraving in stone or metal, presumably referring to 439 LSJ s.v. χαρακτήρ II.2 may be wrong to take τῶν χαρακτήρων as an objective genitive with τύπος: the function seems rather to be partitive, ‘of the various (types of) letters that exist’. 157 the striking of the hammer on the chisel. This appears to be unparalleled, as is the use of the verb in connection with any kind of craftsmanship (thus πεπλάσται Meffert 1861). Another problem is that χαρακτήρ in the sense of ‘symbol’, ‘letter’ (LSJ s.v. II.2), is not found before the second century B.C. Indeed, the juxtaposition of τύπος and χαρακτήρ is intrinsically suspect in this regard (see below). One other explanation that would make some sense of the verses was proposed by Hadjioannou (1975, 402–5),440 who suggested that κύπριος means not ‘from Cyprus’, but ‘of copper’. The image would then be lifted from coinage and refer to the dark complexion of the Danaids, as being similar to a face on a copper coin. The placing of such a comparison here would be less awkward if we, with Hadjistephanou (1990, see n. 442), delete τ’ and take the lines as a reference to the previous guesses at an African origin of the Danaids (‘explanatory asyndeton’, cf. K–G ii. 344–45): ‘you are more alike to Libyan women … and the Nile might feed such a plant: a copper coin-stamp is hammered in female-shaped relief by male craftsmen’. This makes good sense of the words χαρακτήρ and τύποις (LSJ s.v. τύπος II–IV; s.v. χαρακτήρ II.1.). There is also a parallel from Euripides, El. 558–69, where a face is likened to the χαρακτήρ of a silver coin: τί µ’ ἐσδέδορκεν ὥσπερ ἀργύρου σκοπῶν | λαµπρὸν χαρακτῆρ’; ἦ προσεικάζει µέ τῳ; The explanation is not flawless, however: besides the problems with εἰκώς and the awkward focus on the male gender of the artisans, the greatest objection is the scant and late evidence for this sense of κύπριος. Apart from Et.Gud. s.v. κύπρος· ἡ νῆσος καὶ χαλκός (χαλκούς mss.) and Gp. 10.64.4 κυπρίῳ ἥλῳ, the term appears to be found in two magical papyri only: PMag. 4.1847–48 of the fourth century A.D., and 7.466 dated to the third century A.D. (κυπρίνος). One may perhaps argue that if the reference is to a coin, the geographical epithet Κύπριος would naturally suggest copper, whether the adjective had yet assumed this sense in general or not: there was always copper in abundance in Cyprus.441 Apart from the attempts at solutions by conjecture, none of which actually solves very much, if anything,442 it has also been suggested that the reference 440 With enlarged argument in Hadjioannou (1985). See, e.g., Catling (1964) 7–8, 18–21. 442 κυπριοχάρακτός τ’ … | εἰκὼν or εἰκὼ (εἰκὼν already Murray ed. 1937) Friis Johansen (1966): ‘and of Cyprian stamp is the image impressed on your female forms 441 158 is to Cypriot art which some archaeologists claim to be distinct from that of mainland Greece, with a more Oriental or Egyptian character.443 Apart from this being a far from uncontroversial opinion, the comparison is far-fetched in itself and likely to have been incomprehensible to an Athenian audience. Whatever the actual sense of the verses may be, the most satisfactory solution with regard to the Supplices is, in my opinion, that of Friis Johansen– Whittle (1975): excision.444 Interpolation of passages quoted as parallels in the margin is not unique: Friis Johansen and Whittle l.c. list six possible examples from Aeschylus,445 one of which is certain: at Pers. 253 a verse from the Antigone (277) has been interpolated in a number of mss.; in others, as well as in a Byzantine paraphrase, it is quoted as a parallel (see West’s apparatus criticus). As for positive evidence of interpolation in our case, we may note that the juxtaposition of τύπος and χαρακτήρ with cognates is very common in the philosophical and scholarly discourse of the centuries surrounding the birth of Christ, but not found anywhere else in the Greek literature of the archaic and classical periods. The later instances refer to inscriptions, letters, and coinage, as well as to ‘types’ in various philosophical senses. The earliest (?) is SVF ii, fr. 749, containing a certain likeness to our passage in its use of mimetic art (painting) as a metaphor for human ‘types’.446 In the period between 200 B.C. and A.D. 200, the juxtaposition of various by male artisans’ (transl. Friis Johansen ed. 1970); εἰκὼς χαρακτήρ τ’ … Κυπρίοις Sommerstein (1977): ‘and a similar stamp is struck upon the dies of Cyprian womanhood by male artificers’; καὶ πρὸς χαρακτὴρ {τ’} Hadjistephanou (1990): ‘… and the Nile might foster such a stock (i.e. stock of such complexion) and, furthermore, the features too on women’s forms (i.e. on your faces) have been stamped similar [to those of Egyptians] by male craftsmen (i.e. the male authors of the race).’ We may note here that Κύπριοϲ is the diorthotes’ (see 8n., n. 114) correction of Mac κύπριϲ. 443 See, e.g., Myres–Richter (1899) 30, Casson (1937) 158, Gjerstad (1948) 356–61, 446–48, Srebrny (1950) 3–5, and the further refs in Hadjioannou (1985) 509–10. 444 Supported by Diggle (1982, 134, n. 3). 445 Pers. 253, Th. 601, Ag. 900 (on which see Fraenkel ii. 408, n. 4), Eu. 105, 286. Cf. also Fraenkel on Ag. 525 ff., 570–72, 836, 1290. 446 οἱ Στωϊκοί· ἀπὸ τοῦ σώµατος ὅλου καὶ τῆς ψυχῆς φέρεσθαι τὸ σπέρµα καὶ τῆς ὁµοιότητος ἀναπλάττεσθαι ἐκ τῶν αὐτῶν γενῶν τοὺς τύπους καὶ τοὺς χαρακτῆρας, ὥσπερ ἂν εἰ ζῳγράφον ἀπὸ τῶν ὁµοίων χρωµάτων εἰκόνα τοῦ βλεποµένου. 159 derivates of τύπος and χαρακτήρ is found 52 times in seventeen authors:447 before 200 B.C. it is not encountered once, except for our passage. Apart from this, one advantage of the excision is the improved symmetry in the geographical exposé: the three guesses at an African origin—Libya, Egypt, and ‘Indian tribes by the Ethiopians’—will then present an uninterrupted stretch from west to east, which accords with Aeschylus’ usual manner of presenting geographical matter (Bacon 1960, 46–47: cf. my 256–59n.). 284–86. Hecataeus and Scylax dealt with India,448 and Herodotus’ account of nomadic and camel-riding Indians in 3.98–106 may have been based on the former (see Jacoby on Hecat. fr. 295). Hecataeus would be a possible source for much of Aeschylus’ ethnographical material (see 220–21n., n. 381), including this on Indian nomadic tribes. In Herodotus’ account, nomadic tribes are mentioned at 3.98–99; camels, Ethiopians, and dark complexion, which may be the point of the comparison here, in 3.101. On the portrayal of Indian women in antiquity, see also Ruffing (2002). 284. Ἰνδάς is probably required by the feminine participles in 285 and 286 (Anon.Ald., ἰνδούϲ M), being an easier correction than Hartung’s Ἰνδῶν, which is printed by West.449 Robortello’s and Turnebus’ ἀκούω is a certain correction of M’s ἀκούων. ἱπποβάµοσιν (Turnebus: -οισιν M): i.e., as fast as horses (so Σ). One might speculate about a direct influence from Hecataeus: cf. Hdt. 3.102 αἱ γάρ σφι κάµηλοι ἵππων οὐχ ἥσσονες ἐς ταχυτῆτά εἰσι. Ptol.Ascal. p. 403 Heylbut (s.v. µεταµόρφωσις ≈ [Ammon.] Diff. 316 = Ph.Bybl. Div.verb. µ 116), Dionys.Scyt. fr. 8 FGrH (no. 32, i. 239, ap. D.S. 3.67.1), Ph. LA 3.16, 3.230, Sacr.Abel. 135, Det.pot. 83, Post.Cain. 99, 110, Immut. 121, Agr. 167, Plant. 18, Ebr. 90, 133, Heres. 181, 294, Somn. 1.129, 1.171, 2.17, Jos. 54, Decal. 101, Spec.leg. 1.30, 1.106, 1.325, 4.137, 4.146, Virt. 19, 52, Flacc. 144, Qu. Gen. 2.62, Philox.Gramm. fr. 18, Plu. Apophth.Lac. 214f, Mul.virt. 243c, Epict. Ench. 33.1, Sor. 2.32.2, Harp. 239 (s.v. παράσηµος ῥήτωρ), Heraclit. All. 65.2, Ph.Bybl. fr. 2 FGrH (no. 790, iii c (2) p. 812, ap. Eus. PE 1.10.36), Hdn. Mon.lex. ii. 908 Lentz, Poll. 3.86, 5.149, S.E. M. 7.408, 1.99, Hermog. Id. 2.10.148, Polyaen. 6. 52, Philostr. VA 8.31, Or. Cels. 6.31, Princ. 4.2.7, Jo. 10.24.140, Philoc. 1.14, 3.1, Sel. in Ps. PG xii. 1084, Schol. in Lc. PG xvii. 360. 448 Hecat. frr. 294, 296–99, cf. fr. 33, Scyl. frr. 3–5, 7. 449 The masculine case ending is not found elsewhere signifying a feminine, and Ἰνδή is attested in Ctes. fr. 45.19 (ap. Phot. Bibl. 72.46a), Callix. fr. 2 (ap. Ath. 5. 201a), etc. 447 160 285. ἀστραβιζούσας: the exact meaning of ἀστράβη and its cognates is rather obscure; and LSJ’s association with mules is misleading, being too narrow, as is evident from the present passage among others. The general idea of riding ἐπ’ ἀστράβης seems to be to travel on the back of an animal (any animal, but perhaps most often a mule) without taking the reins oneself, but with the aid of a ‘driver’, an ἀστραβηλάτης, who walks beside the animal and leads it (cf. Luc. Lex. 2). The manner of travel is thus distinct from riding on horseback as well as in a chariot.450 The noun ἀστράβη probably derives from the adjective ἀστραβής, referring to the relatively steady and ‘unshaking’ means of travelling (cf. DE s.v.), probably in contrast to the labour required when sitting astride a horse (ἀστράβη, ἀστραβίζω being equivalents to κέλης, κελητίζω). In literature the ἀστράβη is usually employed by women or by implicitly effeminate, often wealthy, men. It is often uncertain whether the noun refers to the animal itself or to a special kind of saddle or seat designed for the purpose. Much evidence from classical and Hellenistic times implies the former, pace LSJ, FJ–W et al. In the Attic of the classical period, only D. 21.133 seems to refer explicitly to a saddle or seat: ἐπ’ ἀστράβης δ’ ὀχούµενος ἀργυρᾶς τῆς ἐξ Εὐβοίας.451 450 Cf. Σ Pi. P. 5.10b, which has a list of πρῶτοι εὑρεταί of different ways of travelling with the aid of animals, enumerating four types: κέλητα καὶ χαλινὸν πρῶτος Βελλεροφόντης κατέζευξε, συνωρίδα Κάστωρ, ἅρµα Ἐριχθόνιος ὁ Ἀθηναῖος, ἀστράβην Ὀξύλος ὁ Αἰτωλός. 451 This reading only appears in one ms., S (in which the entire passage is obelised: see MacDowell’s ed. pp. 47–48), and in the testimonies of Men.Rh. ap. Σ D. 21.470a and Hellad. ap. Phot. Bibl. 533a. The other mss. have ἐξ Ἀργούρας τῆς Εὐβοίας or Ἀργούρας τῆς ἐξ Εὐβοίας, as do most of the testimonia: Ath. 11.481e, Hdn. Pros. cathol. i. 263 Lentz, Mon.lex. ii. 920 Lentz, Harp. s.v. ἀστράβη, Macrob. Sat. 5.21.8. This reading is the vulgate (δηµώδης) according to the scholium. For the city of Argura, cf. D. 21.132, 164, Ph.Bybl. fr. 27 FGrH (no. 790 iii C [2] p. 820, ap. St.Byz. s.v.). Editors are unanimous in adopting ἀργυρᾶς, however, which is the lectio difficilior. (MacDowell also deletes τῆς ἐξ Εὐβοίας, not even putting it in brackets, on the alleged authority of Menander Rhetor and Helladius [ll.cc.]; but neither testimony supplies evidence that these words were not in their texts of Demosthenes.) A ‘silver-plated ἀστράβη from Euboea’ is not as absurd as it may sound at first: the expression is paralleled in 158 τοῦ λευκοῦ ζεύγους τοῦ ἐκ Σικυῶνος, and it is known that Euboea did have metallurgic manufacture (A. Philippson in RE vi. 855, H. Kalcyk in Neue Pauly iv [s.v. ‘Euboia’]); indeed, ‘an Euboean talent’ seems to be a standard 161 The sense ‘saddle’ is also unambiguous in a papyrus from the third century B.C., PCair.Zen. 659.13, and in an interpolated passage in [Arist.] Col. post. 798a19. In most literary examples, the meaning ‘animal’ does seem to be preferable. In the oldest instance, Lys. 24.11–12, the antithesis, repeated twice, between ἐπ’ ἀστράβης ὀχεῖσθαι and ἐπὶ τοὺς ἀλλοτρίους or ᾐτηµένους ἵππους ἀναβαίνειν implies that ἀστράβη refers to the animal: a contrast between a ‘fine saddle’ and ‘someone else’s horse’ is nonsensical, pace Erbse (1979, 425). If ἀστράβη means ‘led mule’, the antithesis to a ‘borrowed horse’ is natural. Similarly, the proverb (Macar. 7.75) σοφόν γ’ ὁ βοῦς ἔφασκεν ἀστράβην ἰδών makes sense if the ἀστράβη is an animal whose job is to transport one person in a slow pace and a comfortable manner: the amount of work required would be considerably less than that of the ox, thus suggesting wisdom to the latter.452 Machon fr. 17.389 and 399 (ap. Ath. 13.582b–c) read κατέβαινε … ἐπ’ ἀστράβης | τὰ πάντ’ ἔχουσ’ ὀνάρια µεθ’ ἑαυτῆς τρία and τὰ γύναια ταυτὶ καταβαλῶ | σὺν τοῖς ὀναρίοις … καὶ ταῖς ἀστράβαις. In the first instance, τὰ πάντ’ (‘in all’) may suggest that the ἀστράβη is in fact counted as one of the ὀνάρια. On the latter passage, Gow notes that the ἀστράβαι are seen as ‘separate from the donkeys, not as saddles worn by them’. The expression ‘asses and astrabae’ will however ‘make sense’ if the one ἀστράβη present is in fact identical to one of the asses. measure of gold, silver, etc., even outside Euboea: cf., e.g., Hdt. 3.89, 3.95, Plb. 1.62.9, 15.18.7, etc., Posidon. fr. 19 (ap. Strab. 3.2.9). 452 σοφὸν ὁ βοῦς ἔφασκε δ’ κτἑ mss.: corr. von Leutsch. The proverb is edited as Com.adesp. 563 Kock (iii. 510, 754) and Iamb.adesp. 12 Diehl (iii. 75). The latter follows Crusius (1889, 459–60) in joining the verse (with the reading σοφῶς ὁ βοῦς κτἑ) with [Diogenian.] 7.9 οὐκ ἔστ’ ἐµὸν τὸ πρᾶγµα, πολλὰ χαιρέτω, adducing two Latin fragments, Com.pall.inc. 49 Ribbeck (ii. 144, ap. Cic. Att. 5.15, Amm.Marc. 16.5.9) clitellae bovi sunt impositae: plane non est nostrum onus, and Quint. Inst. 5.11.21 non nostrum inquit onus bos clitellas; the latter possibly corrupt. But, conjectural conservatism apart, (1) clitellae is not a luxurious saddle for human use, but a pack-saddle; (2) clitellas bovi imponere means to trust someone beyond his capacity, which is, together with the proverbial stupidity of the ox, irreconcilable with the latter speaking σοφῶς; (3) cf., with Lewis–Short (1879) s.v. bos 1, Hor. Epist. 1.14.43 optat ephippia bos, piger optat arare caballus. 162 The verb ἀστραβεύειν and the noun ἀστραβηλάτης453 also favour the interpretation ‘led animal’ rather than ‘saddle’. The former is wrongly translated ‘ride a mule’ by LSJ and other lexica, if we are to trust Pollux: τὸ δὲ ἀστραβηλάτου ῥῆµα ἀστραβεύειν Πλάτων … ἐν Ἑορταῖς: i.e., the verb refers to the driving, not the riding of the mule (or of the ἀστράβη). It is perhaps less than likely that the verb for the driving, and the noun for the driver, of an animal should both derive from the name of the animal’s saddle. The evidence is confusing, supporting different senses in different instances.454 Possibly ἀστράβη may refer to a unity, as it were, between animal and 453 The former in Pl.Com. fr. 38 PCG (ap. Poll. 7.186), the latter in Ar.Byz. Nom.aet. p. 276 Miller (ap. Eust. on Od. 9.220–22, i. 337 Stallbaum), Luc. Lex. 2, Poll. 7.186. 454 As for the rest of the literary instances from classical to Roman times, the context fails to give any clue as to whether a saddle or an animal is intended. Cf. Luc. Lex. 2 (bis), Alciphr. 4.18.17, Plaut. ii. 525 Leo (astraba is the title of a lost play). The ancient scholarly tradition is divided and confusing, several lexica, scholia, etc., offering both the ‘saddle’ and the ‘animal’ explanation. The latter is advocated in the oldest instance, by the grammarian Aristophanes, Nom.aet. p. 276 Miller: οἱ δὲ καταµόνας σωµατιοῦντες τῶν ἡµιόνων, ἤτοι οἱ φορταγωγοί, θηλυκῶς ἀστράβαι καλοῦνται, καὶ οἱ ἐλαύνοντες ἀστραβηλάται. Following Aristophanes in taking ἀστράβη as referring to the animal are Harp., Moer., Poll., Lex.Seg. Verb.util. p. 32b Boysen, Σ A. Supp. 285, Eust. i. 337 Stallbaum (whence the Aristophanean fragment). Most representatives of the ‘saddle’ or ‘seat’ explanation offer it as an alternative to the ‘animal’ one (or vice versa): thus Hsch., Phot., Hellad. ap. Phot. Bibl. 279.533a, Et.Gen., Suda, EM, An.Bachm. i. 154 (= Lex.Seg. Verb.util. i. 154 Bachmann), Σ Luc. Lex. 2. In some instances, the saddle explanation is the only one offered: Add.Et.Gud., Et.Sym., [Zonar.], AB i. 205 (= Lex.Seg. Gloss.rhet. i. 205 Bekker), Σ D. 21.133, 159, Σ Luc. Hist.conscr. 45, Σ Luc. Nav. 30. These are the most confused ones: the Lucianic scholia gloss ἐφίππιον, the Attic equestrian saddle (i.e. for a κέλης), with τὴν ἀστράβην φησὶν ἤτοι τὴν ἐφεστρίδα, ἣν νῦν σέλλαν φασί (so in the Nav., similarly in the Hist.conscr.). Furthermore, Hdn. Pros.cathol. i. 308 Lentz explains ἀστράβη as an εἶδος ἁµάξης. Cf. Probus p. 324 Hagen: carmen (sc. bucolicum) et astrabicon dictum est ex forma qua advecti fuerant qui illa cantaturi erant. sunt autem astrabae dicta παρὰ τὸ µὴ στρέφεσθαι (the last phrase recurring in Σ D. 21.133). In Heliod. ap. Orib. 49.4.34 ∆ηµοσθένους ἀστράβη is something entirely different: an εἶδος χειρουργικῆς ἐφαρµογῆς according to Dimitrakos (1933–50) s.v. Yet another sense, footpad, is found in the corpus glossariorum Latinorum (ii. 22.15, iv. 406.29, v. 591.17 Goetz). In medieval and modern Greek, finally, the sense appears to be different yet again: Tz. H. 9.847 explains ἀστράβη as ξύλον ὄρθιον τοῖς δίφροις τῶν ἁρµάτων εἰς ὃ ἐπικεκύφασιν οἱ 163 artifice—i.e., to the vehicle which becomes the result of fitting a mule, or some other animal, with a seat intended not for an autonomous horseman, but for led transport. This sense may then be the original and the narrower meaning ‘saddle’ secondary, or vice versa. 286. παρ’ Αἰθίοψιν: some ancients held that Africa and Asia were connected by a land bridge to the south, with the Indian Ocean as an inland sea.455 Ethiopia would thus be partly Asian—a notion perhaps influenced by Od. 1.22–24, where Homer divides the Ethiopians into a western and an eastern race.456 ἀστυγειτονουµένας: a hapax. FJ–W note that the middle voice of γειτονεῖν (or γειτνιᾶν) is not found elsewhere, nor a construction of either of these verbs with παρά + dat. However, I believe the compound verb will take a prepositional construction more naturally than that of a simple case.457 Burges’ ἀστυγειτονουµένην (-αν Pearson ap. Butler) hardly helps in either case, since the passive voice (if we, with FJ–W, take it as such) is just as unparalleled as the middle: even more so, one might argue, as the middle deponent of γειτονεύω is attested twice (Hp. Fract. 18, Art. 11). 287–89. On the dawning discrimination between myth and history among Aeschylus’ contemporaries, see Gomme on Th. 1.9.4 (i. 110). Seeing that the ‘sceptic’ Ephorus, who rejected the mythical period altogether (before the return of the Heraclidae), thought that the Amazons were a historical reality,458 it seems reasonable to assume that the same is true of Aeschylus.459 The ἡνίοχοι ἐλῶντες. Cf. Tz. Ep. 61, p. 91.15 Leone, Nicet.Acom. pp. 94, 109, 414 van Dieten, Leo Diac. p. 165 Hase, Zonar. Hist. i. 31 Dindorf, and Dimitrakos (1933–50) s.v. ἀστράβη 6–7. 455 See I. Gisinger in RE Suppl. iv. 558 (s.v. ‘Geographie’) and the note of Paley ad loc., and cf. Hdt. 3.114, 7.69–70. Some held that the Nile had its sources in India: cf. Gisinger l.c., Lloyd on Hdt. 2.28–34, [Arist.] fr. 1.4 FGrH (no. 646, iii C [1] p. 195), Str. 15.1.25–26, Arr. An. 6.1.2–4, Verg. G. 4.293. 456 See S. West ad loc., and also Hall (1989) 140–42 on the concept of Ethiopians in antiquity. Homer’s mythical Ethiopians have little in common with the historical ones, of whom Aeschylus speaks here. We find the former in Pr. 808–9, where the Ethiopians live πρὸς ἡλίου … πηγαῖς, at the eastern end of the world. 457 Cf. also Od. 5.489 ᾧ µὴ πάρα γείτονες ἄλλοι. 458 Ephor. test. 8 FGrH (no. 70, ii A p. 38, ap. D.S. 4.1.2); fr. 60 (ibid. p. 58, ap. Σ A.R. 1.1037). 459 On the myth see Hall (1989) 202, Kirk on Il. 3.187–89, who note that the theory 164 Amazons stand out, however, as being (if we accept excision or a non-ethnical reference of 282–83) the only non-African people mentioned by Pelasgus: they are usually described as living somewhere near the coast of the Black Sea (see FJ–W for refs). The comparison refers in the main not to the ethnicity but to the exclusively female sex of the fifty Danaids, perhaps even more specifically to the fifty grown women’s apparent lack of husbands (ἀνάνδρους).460 FJ–W observe that this may be yet another example of a hint at the future murder of the Danaids’ husbands (cf. 6–10n.): the Amazons were notorious man-killers (cf. Hdt. 4.110), indeed husband-killers according to one source (Ephor. fr. 60a). There were varying theories as to how the Amazons maintained their power over their men: they are said to have disjointed them (Hp. Art. 53) or blinded them (Xanth. fr. 11b) at birth. Ἀµάζονας … ᾔκασα ὑµᾶς: the construction of εἰκάζω with a double accusative is slightly anacoluthic (ὑµῖν Blaydes 1902). The intended sense is probably ‘I would have likened you to’, not ‘I would have guessed that you were’ (with εἶναι to be supplied), in accordance with the previous examples; but Pelasgus is forced to finish with the latter construction, having begun with an accusative (cf. Griffith 1986). Cf. also the sense ‘paint’, ‘represent’ of εἰκάζω (LSJ s.v. I): by way of the awkward construction the verb may attain a metaphorical quality, lit. ‘I would have painted you as the Amazons’ (cf. Hdt. 3.28 αἰετὸν εἰκασµένον). The attempts at tidying up grammar by extensive emendation are mistaken,461 since Aeschylus often modifies and distorts normal grammatical construction so as to enliven style and diction (cf. 15–18n., etc.). 287. κρεοβότους: the adjective (Scaliger: κρεοβρότους M) is attested in fr. 451l.17 and is thus preferable to the unattested κρεοβόρους (Anon.Ald.), even if κρεοβορεῖν is found in Roman times. The fact that the whole notion depends on the popular etymology Ἀµαζών < α priv. + µᾶζα (on the other popular etymology see 288n.), ‘without cake’, i.e. exclusively carnivorous, does not necessarily support the latter conjecture; ‘meat-herding’ of course implies meat-eating. of an origin in an actually existing matriarchy is now in disrepute. Cf. also Bremer (2000) on the Amazons in the imagination of male-dominated Athens. 460 Cf. also Melanipp. 757, cited in the Introduction II 1, n. 26. 461 ταῖς ἀνάνδροις … Ἀµάζοσιν Hartung; Friis Johansen (with argument in FJ–W) posited a lacuna after 287. 165 288. τοξοτευχεῖς: the traditional attire of Amazons. Combined with the popular etymology Ἀµαζών = ‘without breast’ (α priv. + µαζός), this led to the notion that they cut off their right breast in order to shoot better with a bow (see, e.g., Hdn. Pros.cathol. i. 28 Lentz, who also enumerates a large range of alternative explanations of the name). 289. διδαχθεὶς ἂν: Abresch (1763) supplied <δ’> and has been followed by most editors, but I wonder if the asyndeton would not be in place here, seeing that Pelasgus’ utterance has the imperative force of λέγοις ἄν and that τόδ’ not only refers forward to ὅπως, but connects the sentence closely to the former discussion. Cf. Ch. 105 and in stichomythia, e.g., Ag. 543 διδαχθεὶς τοῦδε δεσπόσω λόγου, Th. 261, Ch. 108, 167. Cf. also 323–24n. 290–323. Like the former stichomythia in 204–24, this one lacks indications of speakers in the mss. It takes the form of questions and answers, and it is clear that at least from 314 onward the questions belong to the king and the answers to the Danaids. The scholarly consensus has been that Pelasgus is questioning the Danaids on Argive lore at the beginning of the stichomythia as well. The reverse situation seems to me far more elegant and also more economical as regards the textual alterations necessitated: i.e., Tucker’s arrangement with the questions put by the Danaids at the beginning. This arrangement is adopted by Mazon, Smyth, Werner, Vílchez, and by Murray, who plausibly observed ‘ni fallor, coryphaeus historiam suam (274–5), quam rex incredibilem dixerat, ipsum affirmare cogit’. West (W.SA) argues that ‘the tentative tone of [295] µὴ καὶ λόγος τις …; is exactly that of [the Danaids’] initial gambit, [291–92] κλῃδοῦχον Ἥρας φασὶ …’. This appears irrefutable. Conversely, it is very awkward to let the king, after his firm, proud answer to the Danaids in 293, immediately begin an examination on their knowledge of Argive history, pertaining to exactly those questions that are relevant to the matter of their descent from Io and Zeus. He has just learned, reluctantly, by way of their comment on Io’s service as Hera’s priestess, that they have some insights: he reacts to the moderate tentativeness (φασί) of their statement with a proud affirmation ἦν ὡς µάλιστα. He cannot yet know what the Danaids are driving at when mentioning Hera, and he is unwilling to devote much patience to finding out: as is clear from 277–90, the king has (naturally) assumed a suspicious attitude towards the Danaids. He will not help them reach their goal, which is unknown to him, by asking exactly those leading questions about the union and lineage of Zeus and Io that would let them prove their descent. 166 To assume, with West, a lacuna after 296,462 only to allow the king to take over the questioning at this place, must be regarded as amounting to unsound textual criticism. 298 follows naturally on 296: the fact that Zeus’ adultery with Io was not hidden from Hera does imply (οὖν) a ‘quarrel’ (νείκη). On the whole, it is perfectly natural to let the Danaids continue their questioning in 298–306: the questions do not betray any lack of knowledge (pace W.SA), as is shown by the use of the particles οὖν, δῆτα, and especially οὐκοῦν (or οὔκουν) in 298, 300, 302, and 306. οὖν and δῆτα are ‘reasonable’ and ‘logic’ markers in questions: they invite the respondent to follow the current train of thought. οὐκοῦν as well as οὔκουν is used in leading questions, inviting assent from the respondent. Nothing of this would suit Pelasgus. The Danaids, on the other hand, eager to reach the final conclusion of their questions, have a clear motive to lead Pelasgus on: ‘— And did not Zeus again approach the well-horned cow? —So they say, in the guise of a cow-mounting bull. —Well then (δῆτα), what did the mighty wife of Zeus?’. The Danaids take an active part at first: not too active, to be sure, but not too passive either, as they are admonished by Danaus in 200–201: µὴ πρόλεσχος µηδ’ ἐφολκὸς ἐν λόγῳ. That scene (176–206), in which Danaus lectures his daughters on how they should behave in conversation with the Argive strangers, leads forward to the Danaids’ application of their strategy in regard to Pelasgus: not too bold a strategy, nor too meek, as they would have been if they had remained totally passive throughout the scene in 291–310. By posing humble questions (increasingly challenging in 300–306), the girls force the king himself to affirm their story. The case is put thus by Tucker p. 69: ‘it is obvious that all arrangements which necessitate transposition of verses or a large number of lacunae are little likely to be right.’ Several lacunae have to be assumed with the traditional distribution of questions and answers. A ‘traditional’ editor sensitive to style and diction, as Denys Page, finds himself obliged to assume no less than four lacunae in 298–324. FJ–W, who accept the awkward change of tone in 295 (see above) and give both this line and 293 to Pelasgus, nonetheless have to assume two lacunae (after 307 and 315) as well as a transposition of 309 and 310, and emendations of 309 (καὶ µήν for τοίγαρ) and 311 (τί γάρ for καὶ 462 So already Porson, which accounts for the lacuna being counted in the conventional numeration. 167 µήν).463 Wilamowitz assumes three lacunae, West three lacunae plus a transfer of 309 to follow after 311, and so on. The improbability of such measures 463 FJ–W present some arguments for the vulgate line distribution which are sophistic and unconvincing. (1) The respondent sometimes gives information not asked for (296, 301, 305), which allegedly serves to ‘demonstrate complete familiarity with the matter under examination’. Pelasgus would have no motive for this. But if we do not a priori take for granted that Pelasgus is the inquisitor, it is easy to see another motive: Pelasgus is becoming intrigued by the informed questions of the Danaids and reluctantly begins to take a certain delight in the conversation, hence providing more detailed answers. (2) The change of initiative in the middle of a stichomythia, i.e. the respondent becomes the inquisitor and vice versa, is allegedly unparalleled in Aeschylus. But FJ–W themselves provide an exception: Eu. 587–608. FJ–W argue that this stichomythia is ‘argumentative’ and not only ‘informative’, referring to Jens (1955) 26–27 (cf. also Gross 1905, 72–81). Reversal of direction in stichomythia, although not so as to present an exact parallel, also appears in Supp. 337 (on which see G.AS 124–25), Ag. 543, Eu. 427, S. Tr. 68 (see on argument 3 below). In the remarkable stichomythia in E. Ion 255–368 which, incidentally, deals with a theme similar to the present one, viz. the lineage of one of the protagonists (Ion), reversal appears in 308 and again in 331. In our case the reversal of direction will not be abrupt, if we read the text as Mazon, Smyth, and Murray would have it, disregarding Tucker’s awkward phrasings of 309 and 311 as questions. From 308 to 312 there will be a fourverses-long suspension of the questions and answers, after which Pelasgus takes over the initiative. What we witness in these lines is actually an ἀναγνώρισις (see ad loc.): Pelasgus finally understands that the Danaids are the lost daughters of the land. This puts the stichomythia near to the class of ‘recognition-scenes’ outlined by Gross (1905, 55–59). A change of initiative at some point is unavoidable in either case, as I argued, with W.SA, above. (3) The verb διδάσκω in 289, uttered by Pelasgus, διδαχθεὶς ἂν τόδ’ εἰδείην πλέον, ‘prepares the audience for immediate questions from him as the person who wants information’ according to FJ–W, who argue that the word is a regular marker of ‘inquisitive’ stichomythia and is used by the inquisitor only. But the meaning here is simply that the king wishes to know how the Danaids can vindicate their claim, not necessarily that he intends to question them. Cf. Eu. 431 ff., where the same word is not followed by any further questions, and Eu. 601 ff., where the word is followed by an uneven distribution of questions and answers. Cf. also the stichomythia in S. Tr. 64 ff., where the first question is put forward by Hyllus with δίδαξον, after which his mother Deianeira asks the questions for the rest of the stichomythia. On the whole, the argumentation of FJ–W in (2) and (3) assumes a set of strict rules for stichomythia for which there is no evidence. Rather, the convention as used by the poets of the classical era seems to provide a framework for all kinds of innovative dialogue. Take, for instance, the furious Eteocles vs. the timid chorus in Th. 245–63. Jens’s (1955, 7) notion that this is an 168 resulting in a sounder text is discussed in the Excursus. Almost all of them can be avoided without straining the Greek to mean something it cannot, or letting the protagonists talk nonsense. As in the case of the previous stichomythia, a conservative text (of 296–313, which is where I differ from West’s edition) is given in a foot-note to the translation (n. 74). There are no transpositions or deletions involved in the present version, and only one lacuna is assumed. Further arguments for this distribution of lines will be found in the notes on the particular passages. 291–315. The story of Io concentrated into 24 verses. The details, later ‘Überredungsstichomythie im typischen Stil’ is remarkable: this stichomythia, as Jens himself observes, is completely unique in that there is no argumentation whatsoever from the chorus; they even refuse to listen to what Eteocles says until 257. Then, suddenly having taken notice of him, they immediately obey his request as he repeats it. Compare this with the persuasive stichomythiae in Th. 712–19, Supp. 341– 46, Ag. 931–44, Ch. 908–30, where we find constant argumentation on both sides. Other creative and original examples in Aeschylus are the chorus refusing to understand Cassandra’s prophesies in Ag. 1245–55 and Clytaimnestra begging for her life before her son in Ch. 908–30. The classification of stichomythiae argued by Gross and Jens is, in practice, little more than those usages of the convention that are the most natural in a dramatic context; and although they may have developed from a few traditional forms, it is evident from the extant passages that a dramatic author is free to modify them as he sees fit. I am sure that if more had been preserved of Aeschylus’ dramatic output than the 6–8 % that we possess, we would have had to add a few more categories of stichomythia to his repertoire. This goes for the content of the stichomythiae; as regards the outer form, there is reason to believe that Aeschylus was stricter: see 204–24n. with n. 355 above. (4) The punctuation of 291– 92 as a question is, according to FJ–W, integral to the arrangement with the Danaids as active at the beginning of the stichomythia. This punctuation is dubious, according to FJ–W, who present no argumentation of their own for their view but refer to Tucker. Tucker’s only reason for phrasing these lines as a question is that he thought a positive statement would be too bold for the Danaids to use before the king of Argos. Apparently FJ–W do not agree with him in this, since they themselves, as well as most other editors, give these verses to the Danaids without phrasing them as a question. The statement is in fact moderate in tone: φασί, ‘they say’, is perfectly appropriate in a conversation with the king about his national history; it is even flattering, since it implies that the legends of Argos have reached outside the borders of Hellas. Murray (ed. 1955) does not consider a questionmark after 292 as integral to his version of the text, nor do West and Wilamowitz who, without phrasing 291–92 as a question, give the questions in 293 (Wilamowitz) and 296 (West) to the Danaids. 169 recounted poetically in 524–99, are very similar to those found in the Mesopotamian stories of the sexual union of the god Sin and the heifer Geme-Sin: see Bachvarova (2001) 53–58. 291–92 (Ch.). κλῃδοῦχον Ἥρας: the cult of Hera in Argos was famous, probably the greatest in all of Hellas.464 Hera is called Ἀργείη already in Homer (Il. 4.8 = 5.908), in Hesiod (Th. 11–12), and in the Phoronis (fr. 4 PEG) where her priestess is also called κλειδοῦχος. The great Argive Heraion was situated some distance to the north of the city, on a hill below Mt. Euboea, five km. from Mycenae.465 Whether Argos or Mycenae was the actual instigator of the cult is uncertain. It remains possible that Hera was a native palace goddess of Argos in Mycenaean times,466 although the current consensus among scholars rather favours an origin of the cult in the Geometric or Archaic period, under influence from the Heroic epos (Wright 1982, 199). On Hera’s priestess as a cow and on Hera and the cow in general, see C.Z. i. 441– 47, and cf. 41–44n. above. 293 (P.). πολλὴ: predicative: ‘mightily’ (LSJ s.v. 2.c.). Not, as FJ–W, ‘general’; Plu. Dem. 1.1 ὡς ὁ πολὺς κρατεῖ λόγος has a different meaning on account of the definite article: ‘the major story’, i.e. ‘most of the stories’. 296 (P.). κοὐ κρυπτά γ’: Portus for M’s καὶ κρυπτά: for the corruption cf. 194–95, 276, 504. Portus’ version is at least as easy as Hermann’s κἄκρυπτά, if the exemplar had καὶ οὐ in scriptio plena: οὐ would be removed for metrical reasons. On a more subjective note, the straightforward οὐ … γ’ accords better with Pelasgus’ earnest tone than the somewhat ironic κἄκρυπτά γ’. †παλλαγµάτων: West (W.SA) makes a good case for Butler’s παλαίσµατα (παλαισµάτων Stanley), adopting the reading ταῦτ’ ἄρ’ ἦν παλαίσµατα in his text. Zeus is indeed called παλαιστής by Cassandra in his capacity as her lover (Ag. 1206). This might seem a vulgar expression coming from the Dan- 464 See, e.g., Burkert (1985) 131, Roscher i. 2075–77, C.Z. i. 441–57 passim, iii. 65– 68, 566 n. 2, 1043–46. 465 On the Argive Heraion, see in particular Billon (1997) with refs, and, e.g., R. S. Mason in PECS p. 90, Wright (1982). 466 Nilsson (1967) 428. On the palace goddesses of Mycenaean times, see ibid. 345– 50. The one certain mention of Hera that appears in the Mycenaean corpus couples her with Zeus, both apparently minor deities at Pylos (see, e.g., Palmer 1963, 264), as the recipient of a golden cup. 170 aids (cf. Lloyd-Jones 1993); however, as noted by West, Pelasgus is the speaker here (see above, 290–323n.) and he is not afraid to use coarse language: cf. 301 βουθόρῳ. The scholiast’s gloss αἱ περιπλοκαὶ also seems to be in accordance with this reading:467 W.SA compares AP 5.259.5–7 (Paul.Sil.) κεἰ … ὁµιλήσασα παλαίστραις ταῦτα φέρεις, ὄλβου παντὸς ὑπερπέταται, ὅς σε περιπλέγδην ἔχε πήχεσιν.468 On the other hand, the sense of the stem παλλακ-, ‘concubinage’, fits the context. As noted by FJ–W 302n., παλλακίς is contrasted to ἄλοχος, wife, in Od. 14.202–3. *πάλλαγµα is hardly acceptable as such, however, presupposing *παλλάττω or *παλλάζω, neither of which verbs is attested, and neither of which seems a likely formation. Robortello suggested παλλακισµάτων (-κίσµατ’ ἦν Butler), but *παλλακίζω is likewise unattested, and the -ιζω formation does not seem congenial with the sense of the stem. A better option, then, is Dindorf’s (ed. 1841) παλλακευµάτων (-κεύµατ’ ἦν Hartung), regularly formed from the verb παλλακεύω which is attested in Herodotus and in several post-classical authors. The noun is unparalleled, but the tragedians are partial to forming new nomina actionis on -µα. Several remarkable Aeschylean hapaxes are noted by grammarians and lexicographers, constituting fragments; others are found in the extant plays.469 The genitive ending -ων may be preferable on the principle of conservatism.470 The partitive genitive seems unexceptionable: ‘of (all the various) 467 This has been taken as support for the noun ἐµπαλάγµατα (τἀµπαλάγµατα Hermann, -γµατ’ ἦν Wilamowitz), adopted by several editors, but the sense seems less than adequate in the context, the stem elsewhere denoting entanglement in the most concrete sense, and never in a sexual context. See FJ–W, W. SA. 468 Cf. also Ph. Det.pot. 41 οὐκέτ’ ἀπειρίᾳ σοφιστικῶν παλαισµάτων ὀκλάσοµεν, ἐξαναστάντες δὲ … τὰς ἐντέχνους αὐτῶν περιπλοκὰς εὐµαρῶς ἐκδυσόµεθα, Ach. Tat. 2.38.4 ἔξεστι … καὶ ἐν παλαίστρᾳ συµπεσεῖν καὶ φανερῶς περιχυθῆναι, καὶ οὐκ ἔχουσιν αἰσχύνην αἱ περιπλοκαί. 469 To mention only a few hapaces (some recurring in late authors) on -µα, we find Th. 278 ποίφυγµα (on which see Sandin 2001), Th. 523 εἴκασµα, Ag. 396 πρόστριµµα, Ag. 1284 ὑπτίασµα (also Pr. 1005), Ag. 1416 νόµευµα, fr. 79 σκώπευµα. 470 I very much doubt whether FJ–W’s observation that ‘omission of ἦν [or historic ἐστί] appears not to occur in stichomythia’ is statistically significant, or even true: cf., for instance, 926 and, not in stichomythia, 739. A theorem to the opposite effect might as well be formulated: the copula is omitted as a rule when the clause contains a predicative nominal in the nominative. Is there, in such a clause, any example in stichomythia where we do find explicit ἦν, ἦσαν, ἐστί or εἰσί? Outside of sticho- 171 embraces, these were not hidden from Hera’ (cf. S.GG ii. 116). The nominative case in the scholium may have been absentmindedly attracted to ταῦτα, or need not be based on a sound reading at all. If Stanley’s reading is correct, one may then suggest ταῦτα <µὲν> παλαισµάτων. An unanswered µέν is often found in Aeschylus after ταῦτα and τάδε (e.g. Pers. 1, Ch. 372, fr. 131), and it accords with the genitive as well as the stichomythia: cf. Eu. 589 ἓν µὲν τόδ’ ἤδη τῶν τριῶν παλαισµάτων. (297. There is no reason to suppose that a line has gone missing here, as Porson did; the conceit has nevertheless left its trace in the line-enumeration which is now conventional.) 300 (Ch.). Which is preferable, the ‘reasonable’ οὐκοῦν (M) or the ‘lively’ οὔκουν (Schütz ed. 1794)? D.GP 430 argues that the testimony of mss. is not to be overly trusted in cases where the choice stands between these two particles. He goes on to discuss the matter, but appears to me too eager to reach a clear-cut solution—viz. that interrogative οὔκουν belongs to drama and οὐκοῦν to prose, almost without exception. Here, if anywhere, the milder, reasonable οὐκοῦν ought to be in place, and it is retained by West. See also above, 290–323n. 301 (P.) argues for our case. I find it hard to believe that Aeschylus would put βουθόρῳ (‘vaccas iniens’, as LSJ and IA put it) in the mouth of those, as it will turn out, exceedingly chaste maidens. The Greeks may have been sexually uninhibited in many ways, but they were extremely protective regarding the conduct of noble virgins, who were hardly allowed to oversee the mating of cattle. 306–13. This is the turning point of the dialogue. Pelasgus is becoming more and more intrigued by the conversation, and he is slowly beginning to understand how the Danaids can claim to be of Argive descent. He now remembers that Io found rest in Egypt, a fact that was known to him before: see FJ–W 320n. and my 314–21n. 308–12. Pelasgus now understands what the Danaids are getting at. These lines present an ἀναγνώρισις of a familiar kind, the recognition of a lost relative. 308 (Ch.). οἶστρον: Aeschylus usually depicts the language of barbarians not by actual transcription of the foreign sounds, but by using Greek. Here mythia, we see that Pers. 513 ταῦτ’ ἔστ’ ἀληθῆ becomes conspicuous, and suspect (see Broadhead ad loc.), exactly because of the explicit copula. 172 he may be implying a direct translation from some expressive barbaric term (unless he simply means that Οἶστρος [personified, cf. Tucker] is the official name for Io’s tormentor among the Diaspora Argives). Cf. Hall (1989) 117–21 and especially 119–20 for the use of translation. 309 (P.). Pelasgus’ τοιγάρ may answer both to the statement in 308—that is, to the fact that the story is known in Egypt; for Io was driven far away (µακρῷ δρόµῳ) by the οἶστρος—and to his own statement in 307, etymologically picking up on βοηλάτην (τοιγάρ … ἤλασεν).471 The king is continuing to lose his reserve, being carried away by the story, thus ‘volunteering information’ (see n. 463 above). West understands M’s accentuation τοῖ (γάρ) as ‘illuc’ (see W.SA 142, West p. xl) which is attractive even without adopting his transposition (he lets 309 follow upon 311): it could refer to the Danaids’ mention of the Nile in the previous verse. However, the adverb is not found elsewhere in extant literature (but cf. Pers. 1002). 310 (Ch.). συγκόλλως ἐµοί: a triumphant note: all that Danaus has been forced to admit is in accordance with (‘glued to’) the Danaids’ claim of Argive descent. 311 (P.). καὶ µὴν: concessive (D.GP 353ff.), admitting to the Danaids’ triumph in 310, and implying that he finally accepts their claim (although he does not formally proclaim this until 325): ‘indeed, she came also to Canobus and to Memphis.’ Canobus is known from Hecataeus: see frr. 308–9 with Jacoby’s notes. Κάνωβον κἀπὶ Μέµφιν: on the single, postponed preposition, see Kiefner (1964) 27–29. (312. There is no need to, with Hermann, assume a lacuna here. Cf. 297n.) 313 (Ch.). The chorus replies to Pelasgus’ statement of Io’s Egyptian sojourn with a γε (‘yes’472), adding what Pelasgus already knows (see below, 314–21n.): that Zeus impregnated Io in Egypt. Ἐφάπτωρ: cf. 44–46n., 45– 47n. 314–21. From here on editors agree on the distribution of lines. Now, as the conversation turns towards the most important matter, the genealogy, Pelasgus takes over the questioning. Apparently he is familiar with the African-born descendants of Io and Zeus, as is suggested by the use of inferential οὖν in the questions and also by πάνσοφον in 320 (q.v.). He thus confirms what he has now guessed, that the Danaids are of Greek heritage. 471 472 For the latter interpretation, cf. D.GP 63 (§5). γε in answers (D.GP 130–31), here in combination with καί (cf. ibid. 157, 159). 173 314. ὁ ∆ῖος πόρτις … βοός: ‘Zeus’ calf by the cow’ (not the other way around). 315. ῥυσίων is difficult, and much debated. In order to validate the etymology of Epaphus, being ἀληθῶς ἐπώνυµος, the meaning of the noun should approximate ‘touch’ (cf. 45–47n.) or ‘seizure’ (cf. LSJ s.v. ἐφάπτοµαι II.1: ‘lay hold of’, II.1.d ‘lay (violent) hands upon’, and also the legal sense of the verb, ‘claim as one’s property’ [II.1.c]). The noun is elsewhere in this drama (412, 728) associated with the stem ἐφάπτ-,473 and presumably we are looking at a sort of ‘etymological simplicity’ like the one found in 21 ἐγχειριδίοις, 192 ἀγάλµατα. Here and elsewhere in Aeschylus (cf. 412, 424, 610, 728, Ag. 535, fr. 258) the sense of ‘reprisal’, ‘surety’ is not felt with the ῥυstem, and the noun apparently means ‘seizure’, ‘appropriation’ in general, reverting to the basic sense ‘draw’ (cf. DE s.v. ἐρύω). 316. Here, unlike in 297 and 312, it is hardly possible to deny that a lacuna is necessary (Stanley), in which Pelasgus would inquire after the offspring of Epaphus. Bothe (ed. 1830) supplied, exempli gratia, Ἐπάφου δὲ τίς ποτ’ ἐξεγεννήθη πατρός; 317. Λιβύη: the personified Libya is mentioned a few times in the tradition concerning the foundation of the Greek colony of Cyrene:474 otherwise, nothing much is known of her except her place in the genealogy (see FJ–W and O. Höfer in Roscher ii. 2035 ff. for further refs to classical literature). A marble relief in the British Museum from the second century A.D. depicts her as crowning the nymph Cyrene, who is wrestling with a lion: see Farnell (ed. Pindar) i, face 138. µέγιστον γῆς <$'> καρπουµένη: the country of Libya has traditionally been regarded as fertile: thus already in Od. 4.85–89 (see FJ–W for further refs). Two syllables are missing from the verse. If they belong after γῆς, as has usually been supposed, an ending on ϲ might account for the slip of the scribe: thus, e.g., θέρος (Dawe 1972). 318. τίν’ … ἄλλον: cf. Ch. 114. οὖν: see 314–21n. above. 319. Bῆλον: several mythical kings from the Near Orient go by this name in Greek mythology. The ultimate origin of them all is the Semitic god Baal 473 ‘ῥυσίων is to be treated as the verbal noun corresponding to ἐφάπτεσθαι’, Tucker. So also in S. OC 858–59 µεῖζον ἆρα ῥύσιον πόλει τάχα θήσεις· ἐφάψοµαι γὰρ οὐ ταύταιν µόναιν. Cf. Schuursma (1932) 116–17. 474 See J. M. Reynolds in OCD s.v. Libya, and cf. Pi. P. 4.14, 9.55, Paus. 10.15.6. 174 (Bel).475 In Greek genealogy Belus is usually named as the son of Poseidon and Libya, being the brother of Agenor. 320. τὸ πάνσοφον … ὄνοµα is probably right. If we accept the current vulgate τούτου (Portus: τοῦτό M), as convincingly argued by FJ–W, the transferred epithet from the dependent genitive is in fact regular: cf. K–G i. 263 (Anm. 2), S.GG ii. 180. FJ–W provide adequate parallels for similar adjectives with ὄνοµα in polite inquiries; usually, but not exclusively, καλόν, κλυτόν and the like, e.g. Pl. Thg. 122d τί καλὸν ὄνοµα τῷ νεανίσκῳ; (adduced already by Headlam 1898, 191). Apart from the examples of FJ–W and Headlam, cf. also, e.g., E. Ion 138 τὸ δ’ ὠφέλιµον ἐµοὶ πατέρος ὄνοµα λέγω Φοίβου τοῦ κατὰ ναόν.476 In our case the epithet may be relevant in two ways: presumably Danaus is already known to Pelasgus (so FJ–W) and has a reputation for wisdom (cf. 176–78n.). He is recorded as the inventor of shipbuilding (cf. 134–35n.) as well as of irrigation, and also as the introducer of the alphabet to Hellas (see Introduction II 1, pp. 6–7). However, there may also be a hint of ‘the allknowing name’ = ‘the name that will reveal all’, i.e. Danaus’ name and presence are the definite proof of and key to the Danaids’ alleged identity. 321. πεντηκοντόπαις: FJ–W have a point in arguing that the tradition gives better support for this form (Me: -κοστό- M) than for the vulgate -κοντά- (Heath 1762): to their examples of -ο- in similar compounds add Pi. fr. 93 (ap. Str. 13.4.6) πεντηκοντοκέφαλον, which form is retained by Maehler in the latest Teubner edition. In the present passage, West prints -άπαις with reference to his own note on Hes. Th. 312 where he argues, with Debrunner (1917, 69–70, §135 [n. 1]), that the -ο- forms are later; however, in his recent edition of the Iliad West retains πεντηκοντόγυον in 9.579. And indeed the choice of vowel in similar compounds seems next to arbitrary already in Homer: cf. Debrunner l.c. pp. 66–68 and my 52–55n., n. 199. 323. Αἴγυπτος. The name is attested once earlier than Aeschylus, in Phryn.Trag. fr. 1 (see the Introduction, p. 8). Obviously it serves the same purpose as that of Danaus, namely as an αἰτία for the ethnical denomination (ibid. p. 5). Aegyptus is not explicitly named as king of Egypt in this drama, 475 See, e.g., West (1997) 446; Bernhard in Roscher i. 778–79. τὸ *Musgrave: Diggle and Kovacs keep the mss.’ τὸν and accordingly adopt *Heath’s Φοῖβον τὸν later. Biehl ad loc. compares Plu. Rect.aud. 46f, but the parallel seems superficial. 476 175 but the audience may well have taken that for granted. See further the Introduction, chapter II. 323–24. εἰδὼς δ’: the ms. has δ’ (plus an extra δ’ after Αἴγυπτος, deleted by Turnebus), which is arguably more in place than in 289 (q.v.), although the construction is similar. Here the ‘imperative’ is not intimately connected with the former sentence: δέ is regularly connective, perhaps with an adversative note, suggesting that the interrogation may now be over and that it is time to move on to a different subject (cf. D.GP 167). ἀνστήσαις: the transitive ἀνίστηµι can mean ‘make suppliants rise and leave sanctuary’ (LSJ s.v. A III 3), FJ–W giving plenty of examples to supplement LSJ for this sense. This indeed fits the situation. M’s aorist participle ἀνστήσας is hard to fit into the syntax, however. As FJ–W argue, the ‘taking up’ of the Danaids still lies in the future. The diorthotes’ marginal variant (conjecture?) ἀντήσας is possible (‘act as if you have met an Argive host’), but still less attractive than the sense ‘do take us up’ which intuitively seems right, with ἀνστήσ- also being the lectio difficilior. With this construction Ἀργεῖον might be understood as predicative: ‘take up as Argive’ (cf. Hdt. 5.71 ἀνιστᾶσι … ὑπεγγύους). West retains the participle but does not defend the conservative reading. I suppose one would have to take the aorist participle as previous in relation to πράσσοις ἄν, but not to the present moment, and ὡς as going solely with Ἀργεῖον: ‘knowing about my origin, act, after having taken up the host as Argive’. This is hardly acceptable: the action now required from Danaus is precisely the ‘taking up’ of them, and this ought to be what πράσσοις refers to (so FJ–W). The sense requires a finite verb (ἀνστήσῃς Robortello,477 ἀνστήσεις Paley ed. 1844478), an infinitive (-στῆσαι Marckscheffel 1847, 184), or a future participle (-στήσων W. Headlam ap. FJ–W ad loc.479). FJ–W’s claim that ‘πράσσειν and its compounds are not combined with final ὡς (or ὅπως) in tragedy’ may be incorrect, the construction being retained and defended by Sommerstein in Eu. 769–71; in Ag. 364–66 the much477 Wrongly attributed to Victorius by Wecklein (ed. 1885) and FJ–W: Robortello does print ἀνϛήσης. 478 Thus anticipating Sidgwick (ed. 1900) who is credited with the emendation in W.SA 386, and Ellis (1893, 29) who gets the honour in Wecklein (1893, 335) and FJ– W. 479 From his unpublished ms. notes: see FJ–W i. 6, Whittle (1964a) 28, n. 2. 176 discussed clause ὅπως ἂν … σκήψειεν also seems to be dependent on πράξαντα in 360, although perhaps not exactly final in sense. Cf. also E. Cyc. 616–19. In any case, the extant examples of πράσσειν (ὅπ-)ὡς in tragedy are too few and too diverse in construction and sense to secure a statistically significant result. Final ὡς as such is frequent in Aeschylus (IA s.v. B II), and ὡς with the subjunctive does occur in tragedy with δρᾶν, being synonymous with πράσσειν (E. Cyc. 131, cf. S. OC 75 and 398–99, cited below). In Aeschylus, final ὡς with the subjunctive often appears as dependent on clauses consisting of prayers, entreaties, and commands, as is also the case here: cf. Th. 626–27 λιτάς | … τελεῖθ’, ὡς πόλις εὐτύχῃ, Ag. 1292–94 ἐπεύχοµαι δὲ καιρίας πληγῆς τυχεῖν, ὡς ἀσφάδᾳστος … ὄµµα συµβάλω, Ch. 735–77, 767, 770–72, 984. FJ–W argue for Marckscheffel’s infinitive and the consecutive clause ὡς ἀνστῆσαι, adducing the similar constructions occuring with (ἐκ-)πράττειν in Ag. 1380, Pers. 723, and Eu. 896. But it is doubtful whether these examples are comparable to the present one. In those passages the action of πράττω is previously defined, the conjunction ὡς referring back to a correlate (τόδ’ / τοῦτο (ἐκ-)πράττειν, ὥστε, and οὕτω πράττειν … ὡς), as is regular in the case of consecutive clauses (K–G ii. 501–2). The action of the governing clause is distinct from that of the dependent clause, so that naturally the sense becomes consecutive.480 In our case, the action inherent in πράσσοις ἂν is defined by, indeed identical to, the dependent clause and does not refer to a previous correlate: ‘act so as to take us up’. Accordingly, what we would expect is an object clause.481 These are construed as final clauses, regularly taking the future indicative or the subjunctive with or without ἄν.482 K–G ii. 372 claim that ὡς instead of ὅπως in such clauses appears only in Xenophon and Herodotus, but this may be pure chance. Besides, the line between an autonomous final clause and an object clause is not precise. In S. OC 399 ὡς apparently depends on δράσῃ in the previous verse, thus 480 The same is true for S. Ant. 303 ἐξέπραξαν ὡς δοῦναι δίκην, where ἐξέπραξαν refers back to an already defined action, the corruption referred to by Creon in the previous verses. 481 K–G ii. 372–77, Smyth (1956) 496–500. 482 Only very seldom, in anacoluthic constructions, does the infinitive appear (K–G ii. 377, Anm. 7). 177 presenting a parallel for an object or final clause with ὡς: (Creon comes!) —ὅπως τί δράσῃ …; … —ὥς σ’ ἄγχι γῆς στήσωσι. However, the infinitive can be defended: its strongest support may lie not in the regularly consecutive clauses adduced by FJ–W, but in the following observation by Smyth (1956, 509, §2267b): ‘a clause of intended result is often used where ὅπως might occur in an object clause after a verb of effort’. Indeed Smyth adduces a passage from Aeschylus, Eu. 82–83: µηχανὰς εὑρήσοµεν | ὥστ’ … σε τῶνδ’ ἀπαλλάξαι πόνων. Nevertheless, the consecutive clause is more natural in that case than in ours: µηχανὰς εὑρεῖν, ‘find means’, is a sufficiently autonomous action to be followed by a consecutive clause instead of an object clause; (τοιαύτας) µηχανὰς is implied: ‘means such as to relieve you’.483 The infinitive cannot be said to be impossible, but neither can the subjunctive. In fact, the grammatical construction of final and consecutive syntagmata is rather anarchical in Aeschylus (cf. above on 205–6), and we are at a complete loss as to the mode of the relevant verb in this case. It is conceivable that Aeschylus wrote the optative ἀνστήσαις, which palaeographically might be the most economic emendation: for the attraction of the modus, cf. Ar. Pax 412–13 βούλοιντ’ ἂν ἡµᾶς … ἐξολωλέναι, | ἵνα τὰς τελετὰς λάβοιεν αὐτοὶ, X. An. 3.1.18 ἆρ’ οὐκ ἂν … ἔλθοι ὡς ἡµᾶς … αἰκισάµενος … φόβον παράσχοι;484 325. Zakas’ (1890) µὲν is attractive as a supplement after δοκεῖτε; it is well defended by FJ–W and adopted in the text by West. On the other hand, Porson’s ἔµοιγε or Headlam’s (1904) ἐµοὶ <µὲν> are not as badly out of place as FJ–W argue; the stress on the king’s own person that these readings produce may be intended in relation to the people of the city, who also have a say in the matter of whether sanctuary is to be afforded to the Danaids (cf. 365– 69n., 397n.). 329. αἰόλ’ … πτερόν: a notable instance of mixed metaphor and ambiguous imagery. The abstract collective κακά, described as αἰόλα, hue483 Similarly, the examples given by K–G ii. 8–9, Anm. 6, seem like less obvious cases of objective clauses than the present one. A better example would perhaps be Supp. 773 φρόνει … ὡς ταρβοῦσα µὴ ἀµελεῖν θεῶν: here, however, ὡς is concessive and goes with the participle, and the construction is φρονεῖν + inf. In 622 we should read, with de Pauw (and West), ἔκραν’ … ὣς εἶναι τάδε. 484 Also, e.g., S. Aj. 1217–22, Tr. 953–57, K–G i. 256, S.GG ii. 326. 178 changing, metamorphosing, is concretisised in the next sentence and made manifest as a sort of exotic fauna of evils, with a different plumage in each location (οὐδαµοῦ). At the same time πτερόν suggests winged beasts that follow humans around, αἰόλα perhaps implying that they change hue as they move (cf. FJ–W ii. 263). 330–32. Difficult, and probably intentionally so: not even Pelasgus understands the full implications of the words (333). I think that LSJ, Whittle (1968), and FJ–W have a point in seeing κῆδος ἐγγενὲς as referring to the Danaids’ relation to their cousins, not to the Argives; there are also grounds for agreeing that this is the subject of κέλσειν and that it refers, in the context, to the Danaids themselves (cf. 38n. πατραδελφείαν). φυγήν is then an internal accusative or direct object of κέλσειν. So far we have ‘who would have thought that this … blood-related κῆδος would strand its unexpected flight in Argos?’ (on the sense of κῆδος, see below). τὸ πρίν appears awkward at first, as if the engagement or the enmity would somehow annul the blood relation. However, it does agree perfectly with the prefix µετ- in the next verse, which implies a change: τὸ πρίν ought to refer to the state of things prior to this change, which is brought on by ‘hatred of the marriage-beds’ (Turnebus’ ἔχθει [ἔχει M] is certain). †µετὰ πτοίουσαν† should then conceal a participle which describes the change undergone by the κῆδος, i.e. the Danaids or their relationship with their cousins. West suggests (W.SA) and prints µεταπτοηθὲν, which he translates as ‘fluttered into a change of location’.485 However, to take µετὰ as referring to a change of location produces no clear contrast to τὸ πρίν. Better, I think, in the light of the previous verses, to which ἐπεὶ intimately connects the present clause, would be µεταπτερωθὲν. Through loathing of marriage, the κῆδος, ‘grief’—which was ἐγγενὲς, ‘blood-related’, before— changes plumage as per 329 above, puts on wings, and flees to Hellas.486 The very meaning of the word κῆδος may be conceived as transformed by the change: from the mere ‘grief’ of being bullied by male cousins (the cousins 485 Before him, Blaydes (1895) had suggested µεταπτοιῆσαν which is palaeographically easier and gives much the same sense, although the syntax becomes next to impossible (the transitive verb taking φυγήν as object). 486 As in the case of West versus Blaydes (previous footnote), -ῶσαν would be a palaeographically easier version: the κῆδος changes wings on the flight (φυγήν) and puts it ashore in Argos. The logic now falters, but perhaps not fatally so? 179 were a nuisance even when not desiring marriage), the κῆδος turns into a dreaded ‘connexion by marriage’; or, from being the grief of a blood-related engagement, it turns into the grief of flight and exile. Cf. the play on the double meaning of the noun in Ag. 699–700, with the note of Denniston– Page, and perhaps a similar double-entendre in Il. 13.464 (see Janko ad loc., 463–37n.). 333. τί φῂς ἱκνεῖσθαι: ‘why do you say that you…’; ‘what do you mean is the reason for…’: the Danaids’ purpose is still unclear to Pelasgus. Cf. Ch. 778, S. Tr. 349, OT 655, and see further 335–40n. ἀγωνίων: see 189n. 334. λευκοστεφεῖς … κλάδους: see 21–22n. νεοδρέπτους: a requisite for supplication according to some sources (Σ E. Or. 383: see further FJ–W). This requisite probably extends to all usage of boughs or plants in a religious context: cf. S. Ant. 1201–2 (putting a body on a bed of fresh flowers), Theoc. 26.8 (Dionysian ritual νεοδρέπτων ἐπὶ βωµῶν), Gr.Naz. Or. 7.16.3 (dead pagans are honoured διὰ … στεµµάτων τε καὶ ἀνθέων νεοδρέπτων). 335–40. Griffith (1986) contends that ‘[337–39] are vital for understanding the play—but the text of 337 and interpretation of 338 and 339 are quite uncertain’. I agree: indeed, they are so uncertain that we may despair of ever fully understanding the play if these verses are a requisite. I think that a conservative approach to the constitution of the text (as tentatively by Page and West) is doomed to fail: 336–39 simply do not make any sense in the context. It is barely possible that 336–37 answers to 335, and the two following verses require an even greater amount of far-fetched interpretation to make any kind of sense. At least one lacuna of two verses or more has to be assumed. So first Wilamowitz (1914, 14), who suggested that lines have fallen out before as well as after 337. FJ–W accept the latter lacuna, which is also mentioned by West in the critical apparatus. It is impossible to see what Pelasgus thinks that he is talking about in 338. He has already stated (333–34) that he does not understand the purport of the Danaids’ speech (330–32), and their answer in 335 is not elucidating as such.487 Compare Pelasgus’ reluctance to understand the Danaids’ hints in 487 In this respect, Harberton’s (1903) deletion of 333–35 is not entirely without merit. The verses are unexceptionable in respect of general content and style, however. 180 457–65: in that passage he requires a plain statement; similarly, we should have an explicit mention and rejection of marriage to the Aegyptiads here, to which the discussion in 336–38 would relate. The statement that the Danaids would prefer not to be ‘slaves to Aegyptus’ race’ does not explain the matter, nevertheless Pelasgus’ remarks that follow (albeit obscure) suggest that he now fully understands. In particular, Pelasgus’ inquiry (336) about the reasons for the Danaids’ attitude is awkward: obviously nobody wants to be a slave.488 Thus a lacuna after 335 may be at least as likely as one or several around 337; for example: 335a ‘—But it is customary for women to subject themselves to their husbands’ 335b ‘—I’ll choose death before my cousins’ unclean bed’489 This does not remove all the problems: if 337–38 are interpreted, as is usually the case, as being about arranged marriages and/or the subjection of women to their husbands (σθένος … βροτοῖς in 338 referring to the benefits of procreation), then what is the meaning of 339 καὶ δυστυχούντων … ἀπαλλαγή? 337 is also textually and semantically uncertain. See the notes on the separate passages. 337. τίς δ’ ἂν φίλους †ὤνοιτο τοὺς κεκτηµένους: the vulgate solutions are incomprehensible to me, except possibly φιλοῦσ’ (Bamberger 1839) ὄνοιτο (ap. Robortello490), ‘who would object to their owners if liking them?’. 488 FJ–W’s explanation is unhelpful: ‘Pelasgus is confused by the apparent discrepancy between the Coryphaeus’ earlier expression εὐναίων γάµων (332n.) and her new term δµωΐς (335n.), which properly has no connection with marriage (cf. E. Fr. 132); he is accordingly uncertain whether it is a question of a union which is legitimate but for personal reasons odious, or of a wrongful servitude by inference involving concubinage’. There is no reason why Pelasgus should think that servitude or concubinage is wrong (µὴ θέµις) as such; and that servitude is unwelcome to a free woman is self-evident and already implicit in the Danaids’ use of the term δµωΐς in 335. Pelasgus’ question appears to relate to an entirely different assertion from the Danaids. 489 Such a supplement would imply that Pelasgus now fully understands the import of ἔχθει … εὐναίων γάµων in 332. But perhaps even two more verses are missing before this one: ‘Are you then to serve as concubines to your cousins? They will force us to assume the yoke of marital slavery’, or something to that effect. 490 Apparently by misprint: see McCall (1982). 181 More to the point, however, would be Portus’ οἴοιτο: ‘who would think their owners to be friends?’. This reading would, unlike the former, make some sense of the following verse (q.v.). The reading ὠνοῖτο (Turnebus and, e.g., Smyth, Mazon) introduces the completely irrelevant notion of a dowry. 338–39. If the verses are sound and if nothing has fallen out before 338, they appear to refer to the benefits of voluntary subjection, i.e. of regarding one’s masters or owners as friends (337 φίλους … τοὺς κεκτηµένους), in this particular case the women vis-à-vis their husbands. Such ‘happy slavery’ will increase the collective strength of humanity, removing conflicts that arise out of discontent. 339 is still enigmatic, however: what is meant by ‘easy riddance of the unfortunate’? Perhaps FJ–W are right: ‘a social system where the strong (men, husbands) have all the rights over the weak (women, wives) is advantageous not only to husbands, for the utilitarian reason advanced by Pelasgus (338), but also to rulers in Pelasgus’ present position, since it enables them to disregard the interests of the weak.’ The social awareness implied by such an interpretation seems rather modern, however. According to another interpretation the verse refers to divorce, i.e. to dissatisfied men getting rid of their wives; but that has little relevance to the context or to the Danaids’ argument, even if Gantz (1978, 282) suspects a double entendre referring to the wedding-night murders (see 6–10n.). 340. εὐσεβὴς. The inferential οὖν might suggest that the Danaids have mentioned something referring to σέβας earlier (in the verses that we presume are lost). FJ–W intimate that the adjective answers εὐµαρὴς in the previous verse, which seems feeble. 344. ἦν: West reads the subjunctive ᾖ, with argument in W.SA: ‘the reading ἦν makes it a reference to the existing case’. But the specific reference to the Danaids’ case appears rather more forceful than the generalising subjunctive: was Dike really present here from the start? 345. πρύµναν: the altar, or perhaps rather the entire sanctuary including both the altar and the gods (cf. on 222–23 and 346). The metaphor of the state as a ship often recurs in Aeschylus without particular reference to an altar, but to the leader of the state as being present at the stern of the ‘ship of the state’ (e.g., Th. 2).491 In this case we may perhaps imagine Pelasgus as stand- 491 See van Nes (1963) 101–2 on metaphoric πρύµναι in Aeschylus, and 71–92n. (87– 88n. on the present passage) on the metaphoric Ship of the State. 182 ing by the altar, below the Danaids, thus lending some concreteness to the metaphor. Cf. 351–52n. 346–47. Both these verses are uttered by Pelasgus (so already Arsenius, implicitly, in Me). Thus FJ–W and West. τάσδ’ ἕδρας κατασκίους: the seats may be either those of the gods (if the boughs are placed so as to cast shadows on them: see 204–24n., 241–42n.) or those of the Danaids, or both. W.SA 143–44 argues that the reference is to the Danaids alone, but I think the awe felt by Danaus would be more effectively illustrated by the towering, motionless, and silent statues of the deities than by the sitting women. While there is little reason for Danaus to shudder at the sight of the girls, the gods have the power to destroy the city, in particular Ζεὺς Ἱκέσιος: in adorning the image of Zeus (209–11) with suppliant boughs, the Danaids have made this aspect of the god manifest. See further 354–55n. γε µέντοι, like γε µὲν δή (see 241–42n.), is always, in accurate Attic, more or less adversative. Sometimes, as here, it opposes an implicit proposition: ‘to be sure’, German ‘doch’. Pelasgus grants a point to the Danaids in opposition to his subsisting reluctance to champion their cause.492 492 There are similar implicit adversative notes in the other examples in D.GP 413 of γε µέντοι ‘giving a partial ground for accepting a belief’ (§3)— as here, the ‘partial ground’, stands in opposition to an implicit assertion. In E. Hec. 600 the adversative tone is equivalent to English italics: ἆρ’ οἱ τεκόντες διαφέρουσιν ἢ τροφαί; ἔχει γε µέντοι καὶ τὸ θρεφθῆναι καλῶς δίδαξιν ἐσθλοῦ ‘there is (‘after all’, ‘doch’) instruction …’; in X. An. 3.1.27 the particle combination is ‘preparatory’, opposing, without much logic on Xenophon’s part, ἐπεὶ δ’ αὖ in 3.1.29 (‘You were present, surely […]. When, however …’ Brownson). In Xenophon, the adversative tone has often vanished as good as completely (D.GP 413); however, not in the examples given by Denniston: HG 6.5.39 βοηθῆσαι … ὑµῖν αὐτοῖς. […] συµφορώτερόν γε µεντἂν ὑµῖν αὐτοῖς βοηθήσαιτε ἐν ᾧ … ‘(I think you would) help yourself. […] To be sure, you should help yourself with (even) greater success in the case where…’; Cyr. 7.5.52 ἡγούµην … τότε σε … ἕξειν σχολήν. ὥς γε µέντοι ἦλθεν ἡ δεινὴ ἀγγελία … ἐγίγνωσκον ὅτι ταῦτα µέγιστα εἴη. εἰ δὲ ταῦτα καλῶς γένοιτο … ‘I thought that you would have time (for me) then. To be sure, when the dreadful message arrived, I realised that this was of outmost importance. But if that were to go well …’. Dettori (1986–87, 28) adduces a few alleged examples of non-adversative γε µέντοι outside of Xenophon, all of which actually carry a similar weak adversative tone: E. Alc. 725 (‘but nevertheless you shall die dishonoured’), Heracl. 1016 (‘Yet this is how things stand with me’ Kovacs), Ar. Eccl. 1008 (‘es mußt du doch’). Similarly S. Ant. 495 (see Jebb ad loc.), and Hdt. 7.103 (‘yet even of us not many but a few only’, Godley, who, however, adopts the v.l. γε µὲν). 183 348–437: Second Ode This ode takes the form of a dialogue, or debate, between the passionate women and the cautious king: a kommos or amoibaion.493 The ode is far from as dense and obscure as the previous one, rather ‘a lyric appeal a little more urgent than if spoken’ (Dale 1983 on 396~406), with few metaphors and similes (351–53 being the most conspicuous example). The Danaids’ lyrical impasses take the form of dochmiacs and single-shorts (cretics), as befits the performative (persuading) language, whereas Danaus answers in ordinary spoken iambic trimeters. On the relation of the last two strophic pairs to the previous strophes, see 418–37n. 348 ~ 359 349 ~ 360 350 ~ 361 351–52 ~ 362 352–53 ~ 363–64 $'$'$''$'| $$$'$'|$''$'| $$$'$$'|$$$$$$'|494 $$$'$|'$$$|'$''$$'|?*495 $''$$'$)?$|?''$$'$''|||*496 "s"šs (ia + δ) "ršs "šs (2δ) "ršd "ršd (2δ) "ršs"ršs'd (2δ + ch) "šd"(r)s'ds× (δ + ia + ar) 370 ~ 381 371 ~ 382 372 ~ 383 373 ~ 384 374 ~ 385 375 ~ 386 $'$'|$$|$'$'| $$$|$$$'| $'''$'$'$'| $'''$'$'$'| $'''$¦'&'|$'| '$$''|$'$''||| "s"ršs (ia + δ) "ršrs (δ) "š's"s (3ia sync) "š's"s (3ia sync) "š's!s (3ia sync) d'ss× (ch + ith) 392 ~ 402 393 ~ 403 394 ~ 404 395–96 ~ 405–6 '$$'|$''|$$'$'| &$$'$'|$''$'| %$$'$'|$)¦'$'| $'¦'$'¦$$$'|$''$$'$''|||*497 ds'ds (2δ) !ršs "šs (2δ) ×ršs "(r)šs (2δ) "šs"ršs'ds× (2δ + ar) 493 Arist. Po. 1452b: κοµµὸς δὲ θρῆνος κοινὸς χοροῦ καὶ ἀπὸ σκηνῆς. The term ἀµοιβαῖος is not Aristotelian, but Plu. Pomp. 48.7 may be relevant (Clodius poses insinuating questions to the mob): τίς ἐστιν αὐτοκράτωρ ἀκόλαστος; [κτἑ] οἱ δέ, ὥσπερ χορὸς εἰς ἀµοιβαῖα συγκεκροτηµένος, … ἐφ’ ἑκάστῳ µέγα βοῶντες ἀπεκρίναντο· Ποµπήϊος. 494 361: σὺ δὲ … µάθε <γε> γεραφρονῶν. 495 350–51 ~ 362: λυκοδίωκτον … ἀµ πέτραις ἠλιβάτοις | ἵν’ ἀλκᾷ ~ ποτιτρόπαιον … οὐ λιπερ<νής ποτ’ ἔσῃ | πρόσεισιν>. 496 352–53 ~ 363–64: ἵν’ ἀλκᾷ πίσυνος µέµυκε φράζουσα βοτῆρι µόχθους ~ <πρόσεισίν γε µάλ’ ἱ̄εροδόκα> θῶν λήµµατ’ ἀπ’ ἀνδρὸς ἁγνοῦ. 184 418 ~ 423 419–20 ~ 424–25 421 ~ 426 422 ~ 427 '$'¦'$'|)$'| '$''$''$$$'$'|498 '$$$|'$'| '$'($'||| s's'(r)s (3cr) s's'sr's (4cr) sr's (2cr) s'(r)s (2cr) 428–29 ~ 433–34 430 ~ 435 431 ~ 436 432 ~ 437 '$'|'$$$|'$'|* $$$'|$'|$'|($'| %('$'|* $$$'|$'$$$'|$'||| s'sr's (3cr) "ršs "šrs (2δ) ×(r)šs (δ) "ršs"ršs (2δ) 348–49. The entire vocabulary is Homeric: cf., for example, Il. 5.115 κλῦθί µευ αἰγιόχοιο ∆ιὸς τέκος Ἀτρυτώνη, 10.244 πρόφρων κραδίη. The Homeric language continues into the simile in 351–53, q.v. 348. Παλαίχθονος: see 250–51n. 351–53~362–64. It may not be a coincidence that the frequent doubleshorts make the dochmiacs of the first strophe appear almost like dactylic rhythm (rather than, as often, iambic: cf. 117–22 = 128–33n., n. 282): the simile in 351–53 is virtually Homeric (see below). For dactylic affinity with dochmiac, see West (1982) 112–13. A natural sense-pause occurs after ἠλιβάτοις, and a metrical pause here would seem natural: otherwise 351–53~362–64 will contain one entire period, which seems unlikely.499 Thus we should not, perhaps, analyse the metre as dochmiacs at every point (pace West, FJ–W, Dale 1983). An interesting result of assuming a pause after ἠλιβάτοις is that the remainder of the strophe may be analysed as two identical rhythmical phrases + blunt close: ×šds(r)×šds×. In another context, and apart from the resolution, these would be labelled glyconics (similarly the ending clausula of 396~406).500 A conventional analysis of the present colon would be δ + ia + ar. The colon 351–52~ 362 would be 2δ + ch, which is not a priori impossible: cf. Th. 916~927, 497 395–96 ~ 405–6: φυγάν … πρὸς θῶν ~ τί τῶνδ’ … ἔρξαι. 419–20 ~ 424–25 εὐσεβὴς … προδῷς ~ ῥυσιασθ- … χθονός. 499 The long period would not formally violate the limit of eight metra (Stinton 1977, 40 [327–28]; Maas 1962, 46); however, the irregular cola (dochmiacs with intermixed iambic elements, and aeolic clausula) ought to require a diaeresis inbetween, whether or not this would constitute a metrical pause (the lack of diaeresis is noted by Dale 1983, 11). 500 For glyconics in an iambo-dochmiac context, cf., e.g., the fourth ode of this drama (630–709), E. Tr. 308–40. 498 185 where a dochmiac is followed by choriambs. Dochmiacs in combination with iambics, cretics, or bacchii are also common cola in all tragedians (e.g., 348, Th. 107, 122); and as for the single choriamb at the end of an irregular colon (period?), cf. 41–43 ~ 50–52. 351–53. The simile is virtually Homeric, with epic imagery and vocabulary (πέτραις ἠλιβάτοις, ἀλκᾷ πίσυνος, µέµυκε, on which see Sideras 1971, 140, 192 and FJ–W ad locc.). However, the Homeric perspective is that of the predator rather than the prey: cf., e.g., Il. 3.23–26, 5.136–42, 5.554–57, 15.271–74, 16.352–56, 22.308–10, and see below on 352. The choice of the heifer instead of the lamb as the wolf ’s victim is unusual, and it accords with the images of Io presented before, the Danaids identifying with her (FJ–W 351n.). Cows are depicted as the prey of a lion in Il. 15.630–36. 351–52. πέτραις … ἀλκᾷ: it is tempting to take the image as alluding to the actual stage setting, an allusion that would be especially attractive if the elevation on which the Danaids stand is represented by a natural rock, as suggested by Hammond (1972): see the Introduction, III 4. FJ–W instead assume that the rock and the ‘protection’ in the simile correspond to the altar, as, they argue, does ἀλκή in 731, 832 and Eu. 258. But in 731 and in the passage from the Eumenides the reference appears to be more or less explicitly to statues of gods, not to an altar (cf. Supp. 725 τῶνδε µὴ ἀµελεῖν θεῶν, Eu. 259 περὶ βρέτει πλεχθεὶς θεᾶς). Here the reference must at least include the gods, thus according with the normal usage of ἀλκή which usually refers to protection given by, or strength residing in, persons. The gods have little in the way of a concrete counterpart in the simile, but this is not really necessary: as we saw in 63–68 (see 63–64n.), the Danaids concentrate more on themselves than on the poetic image they are conveying, so that the reality of their situation ‘spills over’, as it were, into the metaphor. 352. ἀλκᾷ πίσυνος: a similar phrase occurs in Homer in a typical predatorsimile, not unlike those listed in 351–53n.: Il. 5.299 βαῖνε λέων ὡς ἀλκὶ πεποιθώς. Aeschylus adapts the phrase so as to refer to the victim instead of the predator, and the sense of ἀλκᾷ is thus defensive: ‘protection’. 353. φράζουσα βοτῆρι µόχθους: a possible parallel for φράζειν in this context has been noted in Eup. fr. 1.2–3 PCG (iv. 303) †τί δὲ ἢν† [ἢν τ’ ἴδῃ Kock] λύκον κεκράξεται φράσει τε πρὸς τὸν αἰπόλον. 354–55. It seems almost inevitable to take the congregation (ὅµιλον) as that of the gods (pace W.SA), as the scholium does: ὁρῶ ὅµιλον θεῶν ἐστεµµένων. FJ–W argue that the notion that the gods are shaded by the suppli186 ants’ boughs is not indicated elsewhere in the text, but this may well be the implication of 346 and 241–42, q.v. (see also 204–24n.). In any case, the gods are the most important issue here: it is on them that Pelasgus’ gaze lingers, and it is before them he shudders (probably) in 346. It is they, not the Danaids, who have the power to take vengeance on the city. FJ–W and W.SA argue that 354–55 corresponds to 350 ἴδε µε τὰν ἱκέτιν, but whereas a confirmation from Pelasgus that he sees the Danaids sitting down is a banality, the grim statement about the gods is expressive and to the point: ‘—look at me! —I see the gods.’ That is: ‘it is not you who are the issue here, but the gods’.501 As for emendations, Bamberger’s (1839) νεύονθ’ is perhaps possible (adopted by Page and Murray); but it takes the edge off the laconic statement (‘I see the gods nodding approval’): cf. the objections of FJ–W. Musgrave’s ναίονθ’ is awkward. µένονθ’ might be possible (‘remain’ or perhaps ‘await’ a decision from Pelasgus); or, perhaps better, ἔνονθ’ (Professor Staffan Fogelmark), ‘be present’ (LSJ s.v. I 2), an easy corruption by metathesis: cf., for instance, 367 ἐκπονεῖν (ἐκπνοεῖν M). FJ–W adopt Harberton’s (1903) τόν τ’ for τῶνδ’ which is actually a rather elegant solution, allowing us to retain νέον θ’ at the beginning of the verse. Diggle (1982) argues that ‘the description of the Danaids as a ‘new’ throng is unbelievable’; however, as a contrast to the gods, who have been present for a very long time, it might not be out of place. It preserves the pertinent contrast between the (insignificant) Danaids and the (significant) gods: ‘I see, shaded by boughs, the new throng as well as that of the agonian gods’.502 The one thing that would favour the Danaids’ being the shaded party is Wecklein’s (1872, 83) conjecture ναύοντ’: according to Hesychius the verb 501 West (W.SA) has two more arguments for ὅµιλον referring to the Danaids, neither of which I find convincing: (1) the fact that the noun elsewhere in Aeschylus refers to mobile contingents is hardly relevant, seeing that groups of statues are not elsewhere given enough importance as to be mentioned repeatedly; (2) West argues that τάσδ’ ἕδρας κατασκίους in 345 refers to the Danaids, comparing S. OT 2–3, 656, and E. Or. 383 (cf. also Eu. 41), but he ignores the repeated occurrences of the expression in the present drama: 413 ἐν θεῶν ἕδραισιν, 423 ἐξ ἑδρᾶν πολυθέων, 493–94 θεῶν … ἕδρας, 501 θεῶν ἕδρας. ἕδρα also denotes the ‘seats’ of deities in Ag. 596, Eu. 11, 805, 855, 892. 502 For the incongruence of κατάσκιον, which would refer to both the Danaids and the gods, see S.GG ii. 604–5, K–G i. 80. 187 means ἱκετεύειν, and West (W.SA) observes that the verb is a contracted form of ναεύω, found in contemporary inscriptions from Gortyn (see LSJ with Supplement). Cf. 503n. 356. ἄνατον: ‘without ἄτη’, ambiguous: before the eyes of the gods (see the previous note), Pelasgus may want to wish the Danaids well, but the sense is predominantly active: ‘without harm to the city’. See my note on 359. This thought is repeated and deliberated upon in the next sentence. ἀστοξένων: a virtual hapax, with all other extant instances appearing in ancient lexica and scholarly treatises. The sense received in the ancient scholarly tradition may in fact be based on an interpretation of the present passage. One of the earliest instances, Poll. 3.60, reports that critics are divided as to the sense: ἀστόξενος δὲ κατὰ µέν τινας ὁ αὐτὸς τῷ ἰδιοξένῳ, κατὰ δέ τινας ὁ γένει µὲν ξένος τιµῇ δ’ ἀστός· ἐνίοις δὲ δοκεῖ ὁ φύσει µὲν ἀστὸς δόξῃ δὲ ξένος, ὡς ∆αναὸς Ἀργείοις. The last interpretation prevails in the later scholarly tradition.503 It is vindicated, it seems, by 618 in this drama, on which see the scholium. 359. ἴδοιτο δῆτ’ ἄνατον φυγάν: the Danaids choose to interpret Pelasgus’ utterance in 356 (q.v.) as well-wishing. ἴδοιτο also echoes the imperative ἴδε µε in the corresponding place in the strophe (350), perhaps with a hint of ‘if you won’t see me, may Justice look’. 360. Ἱκεσία Θέµις ∆ιὸς Κλαρίου: the person Themis Hikesia and the ‘suppliant justice’ of Zeus are the same thing: FJ–W rightly complain about the convention that requires upper-case letters in the case of personifications of abstract qualities. Themis is known elsewhere as Zeus’s consort, but it seems inevitable to understand her as his daughter here. The significance of Κλαρίου is not clear, but cf. Hsch. κ 2867 κλάρες· αἱ ἐπὶ ἐδάφους ἐσχάραι and κ 2870 κλάριοι· κλάδοι. Κλάριος is better known as an epithet of Apollo (of the sanctuary Κλᾶρος), but there appears to be no relation. Perhaps the reference here is to a judicial function, as ἑτερορρεπής in 403 (so Bothe ed. 1830 on the latter passage: see below, 402–4n., n. 515).504 503 With an interesting exception in Timaeus’ Lexicon Platonicum: ὁ µὴ ἔχων ἐν τοῖς πολίταις οἰκίαν ἰδίαν, a definition which recurs in Hsch. s.v. ἀστύξενοι· οἱ µὴ ἔχοντες ἐν τῇ πόλει τὴν οἰκείαν ἰδίαν. Ταραντῖνοι. 504 On Zeus Clarius see also C.Z. ii. 874 (n. 2), FJ–W, A. Adler in RE xi. 552 (s.v. ‘Klarios’). 188 361. παρ’ ὀψιγόνου µάθε †γεραφρόνων: ‘learn from the late-born, thinking old’ is the obvious sense, whatever the true reading may be of the apparently corrupt γεραφρόνων. For the topos, cf. Ch. 171 πῶς οὖν παλαιὰ παρὰ νεωτέρας µάθω; and the further instances cited by FJ–W. The present vulgate is γεραιόφρων (Burges 1811, 187), which is not attested elsewhere, although γηραιόφρων is found in Byzantine Greek (Tz. Ep. p. 81.2 Leone and, in a pejorative sense, MenRom i. 401); there is also παλαιόφρων in 593 and Eu. 838 = 871 (in both cases an attribute of gods). On the other hand, γεραφρονῶν would not actually be an impossible compound, a poetic alternative for γηρο- or γεραιο-. Cf. τερασκόπος (for τερατο-) in Ag. 978, the Homeric ὀνοµάκλυτος (Debrunner 1917, 65) and also the odd γερογνώµων in Apollon. Lex. p. 45.7 Bekker (s.v. ἀστεµφές).505 In Ar. Lys. 980 γερωχία, if sound, may be a compound formed directly on the stem γερα-, meaning ‘elder’s council’.506 Thus in our case one may consider µάθε <γε> γεραφρονῶν.507 γε would easily be removed by haplography,508 making this the most conservative emendation conceivable—together with Marckscheffel’s (1847, 170) γεραρὰ φρονῶν, which is also rather attractive: γεραρός always means ‘old’ in Aeschylus, and φρονέω with the neuter plural is common (LSJ s.v. II 2) in, for instance, Homer: e.g. Il. 1.542 κρυπτάδια φρονέοντα, 18.567 ἀταλὰ φρονέοντες, Od. 21.85 ἐφηµέρια φρονέοντες. Cf. also, e.g., Pers. 782 νέα φρονεῖ (?), fr. 399 ἐφήµερα φρονεῖ, Ar. Nu. 821 φρονεῖς ἀρχαιϊκά. In either case, there is hardly a concessive force to the adjective or participle, which would render the tone of the statement critical: ‘although old in mind, learn from the young’—but rather a causal force, and paradoxical sense: ‘since you are wise by age, learn (now) from the young’. The wisdom of the old is (naturally) proverbial in Greek: cf., e.g., E. fr. 291 Nauck, Men. Sent. 158, 164, 524 Jäkel. 362–64. Approximately nine syllables have fallen out (as noted first by Heath 1762), presumably before ἱερόδοκα in 363, which is probably more or 505 The word is absent from LSJ, Dimitrakos (1933–50), DGE, and TGL. Perhaps the reading should be γηρο-. 506 So Wackernagel (1916, 208, n. 15), who derives it from γεραοχία (cf. DE s.v. γέρων). 507 γε for metrical reasons following the verb instead of, more naturally, the adjective: cf. D.GP 149. 508 See FJ–W iii. 374 for other examples in M. 189 less sound. M’s οὖνπερ at the end of line 362 is certainly corrupt. Headlam (1893, 76) restored οὐ λιπερνής (with several variants on the same stem) from the scholium οὐ πτωχεύσεις, a conjecture which is palaeographically impeccable (ουλιπερ > ουνπερ) and almost certainly correct, despite the misgivings of FJ–W. As argued above (351–53~362–64n.), we need one pause in the supplemented passage that corresponds to one in the strophe. If there is indeed a metrical pause after ἠλιβάτοις in 352, Headlam’s οὐ λιπερ<νής ποτ’ ἔσει> (or ἔσῃ) will produce responsion, also accounting for the future tense in the reading of the scholium. Then, for instance, <πρόσεισίν γε µάλ’> ἱ̄εροδόκα (or ἱ̄ρο-) θεῶν λήµατ’ or, better, λήµµατ’ (Turnebus), in the sense ‘receivings’, ‘receipts’, which will govern the preposition ἀπ’ more satisfactorily than the adjective. The apparent redundance of the expression ἱεροδόκα θεῶν λήµµατ’ appears to be Aeschylean. On the metre, see 351–53~362–64n. It is true that ἱεροδόκα yields a harsh response to -ος µέµῡκε in the strophe; however, it becomes possible if we accept either epic lengthening of the iota (LSJ s.v. ἱερός V), or read ἱ̄ροδόκα (Heimsoeth 1861, 288, cf. Pers. 745) and scan θεῶν as monosyllabic. With monosyllabic θῶν and ἱ̄ερόδοκα we get the responsion -µέµῡκε ~ -ερόδοκα, which may have phonetic relevance (cf. 110–11n., text for n. 274). FJ–W and W.SA argue for one or more additional lacunae before or after θεῶν, but the odds are much higher (cf. my Excursus), and the integrity of the phrase ἱερόδοκα θεῶν λήµ(µ)ατ’ seems unexceptionable. 365–69. As often in Attic tragedy, at least in cases where the ancient king is portrayed as good and wise, the king’s rule is portrayed as more or less democratic (see FJ–W for refs). We ought not to assume that Aeschylus had a detailed conception of the constitution of Pelasgus’ Greece, however, and the discussion of FJ–W i. 28–29 appears somewhat pedantic. 368–69. οὐ … πάρος, ἀστοῖς δὲ … κοινώσας: ‘δέ links an adverb to a participal clause’ (D.GP p. xxxix, n. 3). Cf. 52–55n. 370–75. σύ τοι πόλις κτλ: an oblique reference to the absolute monarchy of Egypt, which the Danaids would be accustomed to? Cf. 373n. µονοψήφοισι, and also Bachvarova (2001) 51, with n. 6. Nowhere in the drama is the issue of the rulership in the Danaids’ homeland touched upon, but the audience may well have taken for granted that Aegyptus is the king (cf. 323n.). 370. δήµιον: FJ–W refer to Björck (1950) 171, 233, for retaining the Ionic η; West prints the Doric δάµιον (after Dindorf ed. 1869: cf. West p. xxvii). 190 Unlike Ch. 56 and Eu. 160, where the Doric α in this word is induced by phonological and etymological word-play (σέβας … ἀδάµατον … δι’ … φρενός … δαµίας, and δαΐου δαµίου), the low-intensity lyrical tone as well as the political jargon in the present passage (and also in 699 and Th. 177) favours the Attic vowel. 373. µονοψήφοισι: in outspoken opposition to the democratic practice of voting. Also at Pi. N. 10.6, where, with a slightly different sense, it denotes the dissentient ‘vote’ of Hypermestra’s dagger on the wedding night: µονόψαφον ἐν κολεῷ κατασχοῖσα ξίφος. 374–75. χρέος πᾶν ἐπικραίνεις: as FJ–W observe, there is a hint of a ‘debt’, or rather an ‘obligation’, ‘charge’: cf. 472. Ever so subtly the focus is shifted from the absolute power to the great responsibility of the sovereign: a point which is underscored by ἄγος φυλάσσου. 376. ἄγος … παλιγκότοις: utterances are ominous; that is, what you say may become true (cf. 512).509 This may be averted by formulaic utterances, for instance by wishing the mentioned evil upon one’s enemies, a common device (see FJ–W for refs). τοῖς ἐµοῖς: West prints Paley’s ἐµοὶ (ed. 1855), but the evil-averting formula loses some of its force with the adjective: instead of ‘curse upon my adversaries’, we are given ‘…upon those that are hostile against me’. A bold use of fairly common words, here the adjective παλίγκοτος as a noun meaning ‘adversary’ (so also Pi. N. 4.96) is an Aeschylean custom. Cf., for instance, 21 ἐγχειριδίοις, 192 ἀγάλµατ’, the subsequent note, and the parallels of FJ–W. 378. εὖφρον: apparently taken in a basic etymological sense (cf. the previous note), as 21 ἐγχειριδίοις, 315 ῥυσίων, 192 ἀγάλµατ’: i.e., not ‘nice’, but ‘well-minded’ = ‘prudent’, the adjective being used as an opposite to ἄφρων (Burges conjectured ἔµφρον). So elsewhere in Aeschylus: see IA s.v. II and Frankel on Ag. 806. LSJ s.v. III may well be right, however, pace Fraenkel l.c. (n. 1), to read piety and auspiciousness into the prefix εὖ (‘= εὔφηµος’). 380. καὶ τυχὴν ἑλεῖν functions as an explicative infinitive or as an entirely new clause (with φόβος understood): ‘fear (whether to) act or not to act; to seize opportunity’. Thus καὶ does not answer the previous τε’s. On τε … τε (and τε … καί) connecting mutually exclusive alternatives, see D.GP 515 and FJ–W ad loc. It has been noted that the passage probably influenced E. IA 509 On ominous utterances, see further I. Opelt in RAC vi. 947–64 (s.v. ‘Euphemismus’). 191 55–57: τὸ πρᾶγµα δ’ ἀπόρως εἶχε Τυνδάρεῳ πατρί, | δοῦναί τε µὴ δοῦναί τε, τῆς τύχης ὅπως | ἅψαιτ’ ἄριστα. 381. Zeus’s σκοπόν is treated as a separate deity in 646–50, possibly to be identified with the Ἀλάστωρ who is mentioned in 415–16 (so FJ–W). Here the phrase appears to refer to Zeus himself, as the scholium: τὸν ξένιον καὶ ἱκέσιον ∆ία. 383–84. οἳ … δίκας οὐ τυγχάνουσιν: FJ–W et al. argue that µή would have been expected instead of οὐ, being regular in ‘generic’ (FJ–W) relative clauses in Attic (but not in Homer). However, the relative clause here is not simply generic and abstract: οὐ makes it concrete, turning it into something that really happens. These people who are deprived of their rights are not hypothetical people in hypothetical situations, but consist of all those who are actually wronged in this way, now or in the future (cf. Griffith 1986). The negation pointedly states that the Danaids’ lawful rights are de facto in the process of being violated.510 385. Ἱκταίου: a hapax. Cf. 1n. 386. δυσπαραθέλκτοις παθόντος οἴκτοις: the latest editors (Page, FJ–W, West) are unanimous in adopting Schütz’s (comm. 1797) δυσπαράθελκτος, West giving the conjecture the status of a ‘corr.’ (M offers ῶ δυσπαρθέλκτοις with δυσπαρθενήτοις in the margin511). But this would mean that παθόντος οἴκτοις does not allude to the present state of the Danaids (despite e.g. 382 πολυπόνων, 353 µόχθους, and the entire lament in the first ode), but to the hypothetical suffering of one that is punished by the wrath of Zeus, who is hard to placate. I think the present vulgate is unacceptable. In fact there is (pace FJ–W) nothing wrong with the traditional dative -οις: cf. 433–36 below and also Pers. 807–8 σφιν κακῶν ὕψιστ’ ἐπαµµένει παθεῖν, both instances being cases of bad consequences that await (µένει) ill-doers or their children, 510 Cf. the examples in K–G ii. 192, Anm. 3. Paley suggests that the ῶ in the ms. may have its origin in a marginal conjecture -ῳ, i.e. someone suggesting that the dat.sg. δυσπαραθέλκτῳ should be read to avoid ambiguity due to the dat.pl. οἴκτοις. This is not impossible. We need not, pace FJ–W, expect that such a critic should prefer the acc.pl., seeing that a construction with µένει + dat. is perfectly normal and, moreover, occurs only 50 lines down in the text. However, it is also possible that it represents the attempt at clarification of one who read δυσπαράθελκτος, interpreting this as a nominative for a vocative (so FJ–W). 511 192 who take the dative case. µένει with the dative is also found in, e.g., Th. 902– 3, S. Ant. 563–64, E. fr. 733 Nauck.512 As for the alleged ambiguity of the dative plural, the context surely makes comprehension easier. The pronunciation may have helped the audience to identify the syntax immediately. In any case, Aeschylus can hardly be said to be very eager to avoid ambiguity of this kind: cf., for example, 60–62, 276 and the further cases of ambiguity discussed in 15–18n. ‘No light is shed upon the corruption by Σ’ (FJ–W), but at least it is certain that the scholiast did not read δυσπαράθελκτος (if not as a vocative: see n. 511) as he takes Zeus as the ally of the sufferer, not the inflicter of suffering: τοῖς θρήνοις τῶν πασχόντων συµµαχεῖ ὁ τοῦ ∆ιὸς χόλος. The aorist παθόντος is not a problem with the reading advocated here, pace FJ–W: Zeus’s wrath is a thing of the future which will show itself when the suffering of the Danaids has ended, or at least has ceased to be a concern of Pelasgus’. The aorist will be much more difficult if referring to the victim of Zeus’s wrath, in which case one would certainly expect the present tense: if the suffering is already over, what would be the point of Zeus not being appeasable? 387–91. The law alluded to is basically Athenian (see the Introduction, II 4, n. 35), and we have no knowledge of any hypothetical Egyptian counterpart. The Athenian law stated that the next of kin of fatherless girls had the right to marry them, and it would not be applicable in this case with Danaus being alive. FJ–W attractively suggest that Aeschylus’ knowledge of Egyptian endogamy would give him the idea of representing their law as even stricter in this respect. 389. West prints Mac τοῖς (‘oppose them’), which seems better and more idiomatic than τοῖσδ’ (Mpc: ‘oppose this’). The former is the lectio difficilior. 392–93. Denniston (1930, 213) advocates the supplement of γε (γενοίµαν γ’, κρατεσί γ’), since ‘probably always in Attic, where οὖν strengthens a negative, γε follows at a short interval’ (D.GP 422). This seems to me a questionable assertion. It is not true for οὔτε (µήτε) οὖν. Moreover, ποτε here imparts an ‘ancillary’ (cf. D.GP 418) note to οὖν and, together with the optative 512 We find µένειν with the accusative in a passage similar to the present one in S. Tr. 1239–40: ἀλλά τοι θεῶν ἀρὰ µενεῖ σ’ ἀπιστήσαντα τοῖς ἐµοῖς λόγοις. Smyth, Murray et al. thus adopt Burges’ (1811) δυσπαραθέλκτους. 193 mode, makes further limitation to its force unnecessary: this is a passionate wish, and γε would make it less so. 394–95. ὕπαστρον … φυγᾷ: intuitively, one expects the phrase to mean approximately ‘the stars are the limit for my flight’. This rather subtly opposes, and deliberately misunderstands Pelasgus’ assertion in 390, with ὕπαστρον being opposed to κατὰ νόµους τοὺς οἴκοθεν: ‘—you must φεύγειν according to the customs (law) of your homeland. —the sky is the limit for my φυγή’. The Danaids choose not to understand Pelasgus’ use of φεύγειν as a legal term. Tucker’s and Todt’s (1889) µῆκος is not without attraction, despite being the easier reading. The corruption might be explicable as influenced by ορ in the beginning of the next word. But µῆχαρ may well be right, extending the import of the phrase to concern not only space, but means: i.e., ‘I will employ every means under the stars to secure my flight’. In this case ὕπαστρον becomes more difficult, but perhaps not beyond what we may expect from Aeschylus. The dative (Victorius: φυγαί M) is more economical than Heath’s (1762) φυγάν and, despite FJ–W’s somewhat tangled argument to the contrary, perfectly all right semantically. 395–96(~405–6). On the aeolic clausula, see 351–53~362–64n. 395–96. Cf. 9–10n. As FJ–W note, ∆ίκαν stands in contrast to νόµους in 390: objective Right against mundane Law. τὸ πρὸς θεῶν: LSJ s.v. πρός A I 3–4. 397. κρῖµα: the long iota is not found in this noun in later verse (Orac.Sib. 8.298, Nonn. Jo. 9.176 etc.), and Wackernagel (1916, 76, n. 1) emended to κρεῖµα, on, I think, uncertain grounds. κρῖµα is found in the mss. of late prose works, for whatever that is worth.513 µή µ’: FJ–W’s reference to S.GG ii. 187 for ‘a degree of emphasis’ on the enclitic pronoun is criticised by Diggle (1982): ‘The phenomenon … illustrated [by S.GG ii. 187] is the quite separate one of an enclitic pronoun combined or contrasted with a following noun or non-enclitic pronoun.’ But surely the emphasis on µ’ in contrast to a following noun is what the present case is all about? Pelasgus contrasts himself with the δῆµος in the next line. The degree of emphasis on the pronoun is still debatable, and I believe West goes too far in adopting Tucker’s ’µ’ (as a ‘corr.’): as Diggle l.c. argues, this 513 Plu. Adv.Col. 1121e, Test.xii.patr. 12(T. Benj.).10.3, [Clem.Rom.] Ep.virg. 1.11.8, etc. 194 suggests that Pelasgus pleads personal incompetence (‘do not ask me to judge it’). The emphasis on µ’ comes a posteriori, as it were, with the introduction of the contrast.514 400–401. The phrase is, as noted by FJ–W, adapted from Il. 22.104–7: the xenophobic note is not in the original, however. 400. For µὴ λῷον (Schneidewin 1839, 153: καὶ µὴ τοῖον M: καὶ del. Turnebus), cf. Thgn. 690 ὅ τι µὴ λώϊον ᾖ τελέσαι, Philostr. VA 3.10.18, and cf., with West, Pers. 526 εἴ τι δὴ λῷον πέλοι. 402–4. ὁµαίµων is a partitive genitive referring to the Danaids and the Aegyptiads: ‘Zeus watches both parties of the kindred ones’. So ὅµαιµος elsewhere in the present drama (225, 474).515 This refers back to Pelasgus’ previous mention of the kinship of the Aegyptiads and the Danaids: Zeus takes account of both of the kindred parties, and he sees to right and wrong, not to legalistic niceties. The image of Zeus’s weighing in his scales is a little confused: does he dispense injustice to the bad and piety to the good? FJ–W explain ‘putting the ἄδικα of the wicked on their scale and the ὅσια of the law-abiding on theirs’, which perhaps approximates the sense intended. 514 Thus the rule of A.D. Constr. p. 170 Uhlig: Πᾶσα, φασίν, ἀντωνυµία συµπλεκοµένη ἐν τάσει ὀρθῇ ἐστι, ∆ιονυσίῳ ἐλάλησεν καὶ ἐµοί, ∆ιονύσιον τιµᾷ καὶ ἐµέ. εἰ γοῦν ἐκτὸς τοῦ συνδέσµου γένοιτο ἡ ἀντωνυµία, οὐ πάντως ὀρθοτονηθήσεται, ἐχαρίσατό σοι καὶ ∆ιονυσίῳ, ἐτίµησέ σε καὶ ∆ιονύσιον. The rule says (as demonstrated by Diggle 1982, n. 4) that the enclitic is possible before but not after the conjunction in the antonymy. 515 FJ–W, W.SA and others take ὁµαίµων as the nominative case, referring to ‘Zeus of Kinship’, the mention of him being ‘a warning that the Danaids’ claim on their Argive kin deserves respect’ (W.SA). But that kinship has not been mentioned since 325 ff. (probably not in 331, q.v.), whereas the more conspicuous kinship, that between the Danaids and their cousins, was referred to by Pelasgus only ten verses before, in an argument to the effect that the Aegyptiads have, being the next of kin, the legal right to marry the Danaids. This kinship, unlike that with the Argives, is based on Zeus as a common ancestor. Thus the Danaids’ unexplained reference to a ‘Zeus of Kinship’ would be very odd, and not at all supportive of their position. (So, e.g., Bothe ed. 1830: ‘neque … hic agitur Ζεὺς ὁµαίµων …, sed is, quo magis opus est Choro, Ζεὺς ἑτεροῤῥεπὴς [κλάριον dixit 331 (360)], qui … utriusque partis meritis pensitatis id, quod cuique debetur, appendit.’) In 652 the adjective may refer to the relation between the Argives and the Danaids; however, we should perhaps read ὁµαίµων there too, which would make γὰρ more understandable, referring back to 643–45: ‘they did not vote for the men … for of the kindred parties, they revere those who are suppliants of holy Zeus’. On 449 see ad loc. 195 The double accusative with ἐπισκοπεῖ may not be impossible, the n.pl. τάδε taking on a quasi-adverbial quality (cf. 194–95n.). But Schütz’s (comm. 1797) ἀµφοτέροις is attractive, agreeing with the later datives. ἑτερορρεπής as an epithet of Zeus appears to be active in this case (so FJ– W), ‘weighing each part’, instead of, as often later, passive: ‘inclining this way and that’. The classical image known from Egyptian mythology is the Psychostasia, the weighing of the heart of the deceased, usually heavy with sin, against a feather.516 405–6. It is possible that µεταλγεῖς is sound, the verb here meaning ‘hesitate’, ‘agonise’ (ὀκνεῖν), with µετ- signifying the conjunction or the simultaneity of the sensation with the action, instead of, as in E. Andr. 814, the posteriority.517 The preﬁx connects the agony with doing what is right: cf. E. Med. 996 µεταστένοµαι σὸν ἄλγος, Hec. 214. Cf. also the uncompounded verb at Ch. 1016 ἀλγῶ µὲν ἔργα: a very easy emendation here would be ἔτ’ ἀλγεῖς.518 406–9. The construction is mildly anacoluthous: δεῖ is construed with the genitive (cf. 417), but an acc. + inf. is also attached. It is not certainly defined whether the eye is that of the deep mind (Sansone 1975, 22–24) or of the diver (Liberman 1998). A quaint detail from divers’ lore is added: the diver must not drink too much wine!519 This may be alluded to in Anacr. 31: ἀρθεὶς δηὖτ’ ἀπὸ Λευκάδος πέτρης ἐς πολιὸν κῦµα κολυµβῶ µεθύων ἔρωτι: see FJ–W for further refs. On Greek diving, see also Auberger (1996) 48–56. 410. ἄνατα … πόλει: cf. 356–58. 412. ∆ῆρις ῥυσίων ἐφάψεται: on ῥυσίων see 315 (with n.). Here a hint of a ‘reprisal’ may be present with the noun: if Argos indulges the Danaids, Strife will claim hers in return. δῆρις is found personified in Emp. fr. 122, too. 516 See C. Seeber in LÄ iii. 249–50 (s.v. ‘Jenseitsgericht’), E. Martin-Pardey ibid. vi. 1084 (s.v. ‘Waage’). 517 For the soundness of the verb in Euripides, see Diggle (1981b) 94 (210–11). 518 ἔτ’ (ἀργεῖς) Musgrave: ἔτ’ ἄλγος Friis Johansen (ed. 1970). Sidgwick’s µεταλγὲς (‘bringing sorrow in its train’, LSJ s.v.) has been popular, adopted by Page and defended by Liberman here and at 111 (1998, 245–46). Seeing, however, that the focus of the present scene lies on Pelasgus’ agonising and uncertainty, the concrete stress on him that the verb in the 2nd person present conveys is attractive. Moreover, the adjective µεταλγής is not attested, and the formation as such, in the desired sense, is hardly more likely than that of µεταλγέω. 519 According to some sources the diver should not drink too much of any liquid: see Hp. Vict. 4.90, and further refs in FJ–W. 196 414–16. Ἀλάστορα: Zeus sometimes takes Ἀλάστωρ as an epithet (see C.Z. ii. 1098), but this cannot be the import here, as rightly argued by FJ–W. In Euripides the title of Alastor is often used as denoting vengeful deities or ‘demons’ from Hades, e.g. Med. 1059, Hipp. 820. So apparently also in Pers. 354, Ag. 1501, 1508,520 where it is implied that there is one Alastor for each misdeed (cf. Barrett on E. Hipp. 818–20). In E. Tr. 768, on the other hand, it is rather a personified abstract: πολλῶν δὲ πατέρων φηµί σ’ ἐκπεφυκέναι, Ἀλάστορος µὲν πρῶτον, εἶτα δὲ Φθόνου. 416. ἐν Ἅιδου: cf. 228–29. 418–37. An editorial problem is whether these strophes are to be regarded as a new ode (Friis Johansen, FJ–W) or as a continuation of the previous one (e.g., Page, Dale 1983, West). The question is not entirely academic: apart from the metrical affinity, the musical accompaniment would presumably have indicated if this was a new song or not. I follow the majority in labelling them as strophic pairs 4 and 5. 418–27. A comprehensive abstract of the Danaids’ position in easy metre, cretics with a few resolutions. The points made in the first antistrophe are elaborated on with more emotion and colourful imagery in the last pair of strophes (428–37). Jouanna (2002, 788–90) notes that the cretic metre presents a ‘manly’, ‘war-like’ ethos, comparing Eu. 328–33 (grave verses from the Erinyes), Ar. Ach. 665–75~692–701 (the Acharnians’ spirited invocation to the national muse), and an observation by Ephorus that Cretic rhythms are συντονώτατοι.521 The tone of 418–27 is certainly less emotional, more collected than in the previous and subsequent passages, while the imperatives at the beginning of each strophe preserve and focus on the sense of urgency. 419. εὐσεβὴς: cf. 340. 421. ἐκβολαῖς: ‘castings-out’, not quite in accordance with the Danaids’ earlier insistence on a voluntary exile (8–10). But unlike the variant of Md ἐµβολαῖς, which was earlier advocated as traditional and sound by Friis Johansen (1968, 363; opinion retracted in FJ–W), the word harmonises with the words ὀροµέναν and ἕκαθεν in the context. We should understand the ‘castings-out’ as describing the effect rather than the intention of the Aegyptiads’ actions. Thus we may counter FJ–W’s objection that ‘ἐκβολαί denoting one action of “casting out” is …unexampled’: the Aegyptiads did not commit 520 A radically different sense appears to be found in Eu. 236 (see Sommerstein ad loc.). 521 Ephor. fr. 149 FGrH (no. 70, ii A, p. 86, ap. Str. 10.4.16). 197 a single action of expulsion, but acted in a way that made expulsion inevitable. Or the Danaids may exaggerate a little, ‘like people who say they “have no option but to” when every one knows that they have’ (Dawe 1972). 422–23(~427, 418). de Pauw’s ὀρµέναν and Heimsoeth’s (1861, 287) πανθέων (ὀρο-, πολυ- M, vulg.) would produce exact responsion. The latter emendation also produces a type of verbal echo that is common in the choral odes of this drama (see 110–11n. with n. 274). However, the gods can hardly be described as ‘all’ gods, even if the twelve Olympians were indeed represented (see 204–24n.): 481–82 and 493–95 indicate that other gods exist in the city. The compound πανθε- is not found in classical Greek, except of a temple in Arist. Mir. 834a. In all extant examples it is formed from the adjective θεῖος. 424. ῥυσιασθεῖσαν: see 315n. 425. ὦ πᾶν κράτος ἔχων: see 370–75n. 427. κότον: that of the gods, especially Zeus (so Σ): cf. 347, 385, and 435– 36n. 428–32(~433–37). These verses are more urgent and emotionally charged than the previous ones. They expound the two last-mentioned themes: the dragging off of the Danaids (strophe) and the nature of Zeus’s wrath (antistrophe). The initial cretics of the strophe are succeeded by dochmiacs as the ode turns into a vivid depiction of the girls being led away like horses from the sanctuary. For the image, cf. Th. 326–29 τὰς δὲ κεχειρωµένας ἄγεσθαι … ἱππηδὸν πλοκάµων, περιρρηγνυµένων φαρέων, with Hutchinson’s note. 429. µή τι τλᾷς τὰν ἱκέτιν: the reading of the scholium, τλαίης, may not be a paraphrase but stem from a transmitted variant (FJ–W). It would be possible if τὰν were deleted. However, M’s reading τ’ ἀαΐσταν appears to be the result of a majuscule corruption of the present vulgate τλᾷς (Wellauer: τλῇς Turnebus): τλαισ > τααις. 431(~436). If the text, with the present vulgate ἱππᾰδὸν (*Voss: ἱππηδὸν pc M , -ών Mac), is sound, the responsion is rather irregular; but it is not impossible for dochmiacs, where free responsion of long against short ancipita and of resolved against unresolved syllables appears more often than in other metres. FJ–W provide several exact Aeschylean parallels, for instance, Ag. 1164~1175 (end of line). As for Voss’s conjecture, we may note with suspicion that the metri gratia variant -ᾰδον in adverbs which are regularly spelled -ηδον is not found outside epic and elegiac verse. In tragedy, however, the short alpha occurs in adverbs on -στᾰδόν and also in Ion Trag. 41b ἀµφᾰδόν. 198 Dindorf’s (ed. 1841) ἱππηδ(ά) is a clever alternative that produces closer responsion, but -ηδά as a variant of -ηδόν (or vice versa) is also not found in tragedy. Burges’ (1811) πλόκων for ἀµπύκων may be mentioned; but, notwithstanding the parallel from Septem contra Thebas cited above, πλόκων is (a) facilior lectio, and (b) removes the suggestive play on the horses’ ἀµπυκτῆρες (bridles: cf. Th. 461, Jebb on S. OC 1068–69). On ὁµοίαν in the antistrophe, see 436n. 432. πολυµίτων (Turnebus: -µήτων M) goes with πέπλων: ‘finely woven’, another hint at the luxuriousness of the Danaids’ oriental dress. It cannot refer to the headbands (despite Hsch. τ 1462522): the fine cloth of the dress has been repeatedly stressed before, in 121–22 = 132–33, cf. also 235–36. As for the postponed τ’, this is more acceptable in lyric verse than in dialogue (cf. 282–83n.): moreover, the colometry with probable colon-end after ἀµπύκων makes the structure clear and the postponement easier. Cf. E. Tr. 1064 and Ar. Av. 257, the only other certain examples in non-epic verse of this kind of postponement.523 434–36. The corruption in 435 does not obscure the general sense: whatever one does will have future implications for one’s family and estate. τάδε is the subject of µένει, which governs the dative παισὶ … καὶ δόµοις (see above, 386n.). As for †δρεικτίνειν (Mc: -τείνειν M), the latter part appears to be sound. Indeed τίνειν … θέµιν may be an Aeschylean thought, a self-styled figura etymologica. Of the hitherto suggested remedies of the corruption, Whittle’s (ap. Friis Johansen–Whittle 1975, cf. FJ–W) δεῖ (’κ)τίνειν is perhaps the most attractive: cf. the Σ paraphrase δίκαιόν ἐστιν ἀποδιδόναι ὁµοίαν δίκην. The intrusion of ρ into the commonplace δεῖ is somewhat hard to explain, however. A semantically bolder emendation would be τρὶς τίνειν. Cf. Orac.Sib. 2.304 τίσουσιν τρὶς τόσσον ὅσον κακὸν ἤλιτον ἔργον. The import here would be a different one: not punishment in the afterlife, but the 522 τριχαπτόν· τὸ βαµβύκινον ὕφασµα ὑπὲρ τῶν τριχῶν τῆς κεφαλῆς ἁπτόµενον, ἢ πολύµιτον. 523 See Fraenkel on Ag. 229–30; he does not, however, discuss the present passage. Of other examples, E. Alc. 818–19 is interpolated, Tr. 1069, A. Ag. 229, S. fr. 859 are due to conjecture. In the first two of these cases the conditions discussed here are present; the Sophoclean fragment is somewhat different with the conjectural τε (Headlam ap. Pearson) in third place after a prepositional phrase: φίλιπποι καὶ κερουλκοί, σὺν σάκει κωδωνοκρότῳ τε παλαισταί. 199 idea that the sins of the parents are accumulated and visited manifold upon either themselves or their offspring, a notion not unfamiliar to Aeschylus or to Greek thought, being the theme of the great tragic familial trilogies. The number three is highly significant in this context: three generations must pay for the sins of the ancestor (see Ch. 1065–76 with Garvie’s note). In Aeschylus’ Laïus – Oedipus – Septem we would have got to see the workings of such a curse in each generation, but only the finale has been preserved: see, however, Hutchinson’s Septem, pp. xxiii–xxx. The corruption may actually be easier to explain than in the case of δεῖ, as the ρ is accounted for: δ repeated from the previous verse, itacism, and the final σ misread as κ in κτείνειν. It will also make a little more sense of ὁµοίαν, which has not been satisfactorily explained. 438–54. Pelasgus’ final answer to the lyrical pleading of the Danaids is yet more non-committal politician-talk. Apparently the awkward metaphoric language was as incomprehensible to the ancient scribes as it is to us, resulting in some heavy corruption in 444–48. 438. ἐξοκέλλεται: the beginning of a difficult metaphorical passage. The subject is presumably ‘the present matter’, not (as FJ–W) ‘my reflections’, which would be too introspective. Cf. E. Alex. IV 3 πῶς οὖν ὀ̣[κέλ]λει ταῦτά γ’ ὥστ’ ἔχειν καλῶς;524 Middle and passive forms of ὀκέλλω and compounds are not found elsewhere. 440–41. Extremely difficult, as we do not know (1) what a στρέβλη actually is and (2) whether προσηγµένον, ‘neared to’, is sound. A ship has been bolted, i.e. built, that much is certain. Most commentators take στρέβλαι as either some sort of shipbuilding device (‘winch’, LSJ), or a means for transporting the ship on ground: ‘windlasses’, ‘ship-cables’ (FJ–W). The latter interpretation is attractive if one could take the phrase as meaning that the ship has been drawn towards the sea with the aid of στρέβλαι and is ready to set sail: ‘le vaisseau terminé à été mis à la mer et rien ne peut plus l’arrêter’ (Liberman 1998). The inherent sense ‘twisting’, ‘turning’ of the stem accords 524 So Snell (fr. 43.38), whose supplement appears better than Page’s (Pap.poet. fr. 9) <µ>ε[ταβα]λεῖ, in violation of ‘Lex Youtie’, an important guideline for papyrology and epigraphy: iuxta lacunam ne mutaveris (see Merkelbach 1980). Page followed the editor princeps Crönert in reading ε[, ignoring Snell’s (p. 1, n. 3) affirmation that ‘Die unrichtigen Lesungen Crönerts habe ich meist stillschweigend berichtigt’. The letter now appears to be illegible: see Diggle’s ed., TrGFS p. 84. 200 well with a parallel noted by Liberman, Moschio FGrH no. 575 (iii B p. 675, ap. Ath. 5.207b): ὡς δὲ περὶ τὸν καθελκυσµὸν αὐτοῦ [sc. τοῦ µέρους] τὸν εἰς τὴν θάλασσαν πολλὴ ζήτησις ἦν, Ἀρχιµήδης ὁ µηχανικὸς µόνος αὐτὸ κατήγαγε δι’ ὀλίγων σωµάτων. κατασκευάσας γὰρ ἕλικα τὸ τηλικοῦτον σκάφος εἰς τὴν θάλασσαν κατήγαγε. πρῶτος δ’ Ἀρχιµήδης εὗρε τὴν τῆς ἕλικος κατασκευήν. στρέβλαι appears to refer to much the same thing as Archimedes’ ἕλιξ, ‘screw-windlass’ (LSJ), even if this means that he cannot have been its inventor, being born two centuries after Aeschylus’ death. This image is hardly compatible with the previous notion that the issue has foundered, but that need not be an obstacle: the issue in 438 is not identical to the present ship. The fact that both the finite verb and the participle take the resultative perfect tense removes the alleged problem of the participle describing an action that would be subsequent to the bolting: the perfects do not express a temporal relation between the two actions. προσηγµένον is somewhat hard to take as absolute in the sense required: Liberman (1998) attractively suggests κατηγµένον with a reference to the passage cited. 442. καταστροφή: ‘halt’, ‘stop’: cf. Ag. 959, Pers. 787. 443–48. The passage is desperately corrupt, but the general sense is clear enough. Two lesser evils are deliberated upon: in 443–45 the loss of goods that may be replaced, and in 446–48 the verbal insult that may be verbally assuaged (µύθου µῦθος … θελκτήριος). These are contrasted to the spilling of kindred blood in 449 (where δ’ answers to µὲν in 443). Editors have tried to emend the corrupt mess with multiple transpositions of lines, usually following ‘Casaubon’525 and inverting the order of 444 and 445 as well as that of 447 and 448 (so e.g. Page, West, and Murray, oddly bracketing 448 after transposition). As I argue in the Excursus, discrete multiple transposition of lines is a very radical measure which should be used with the greatest caution, if at all, in editing. Indeed transposition per se is a less probable corruption than a lacuna or even an interpolation. In this case the vulgate transpositions do not produce a text of evident integrity but are a desperate remedy, at least in the case of 443–45. Friis Johansen’s relocation of 444 to follow 442 (ap. FJ–W) is hardly better: in fact it is no remedy at all, as FJ–W proceed to obelise the entire 444 after the transposition (similarly, Page obelises the same verse after adopting the transposition of ‘Casaubon’). There can be no 525 Marginalia in Cambridge Adv. b.3.3: however, according to Dawe (2001b), not by the hand of Casaubon. 201 reason to transpose 444 if we do not even know the sense and the approximately correct wording of the verse. After transposition Murray, followed by West, reads χρήµασιν … πορθουµένοις (*Voss) γένοιτ’ ἂν ἄλλα κτησίου ∆ιὸς χάριν | ἄτης γε µείζω, καὶ µετεµπλήσαι (Murray ed. 1937) γόµον. This seems artificial to me, as do the more conservative measures preferred by Young and Barigazzi.526 One might suggest ἅ τις for the impossible ἄτην (ἄτης Σ) in 444. This verse may simply be an elaboration on the looting and a description of the size of the stolen wealth. If, as FJ–W plausibly suspect, M’s γε µείζω stems from an intrusion into the text of the explanatory gloss γεµίζοντος (for ἐµπλήσας) in the scholium, the verse will be irremediable in the absence of further evidence (papyri). As a diagnostic (Maas 1958, 53–54) one could read, for instance, ἅ τις φέρει µέγιστον ἐµπλήσας γόµον. We thus assume (with e.g. Murray, Barigazzi l.c., West) that the better tradition is preserved by the scholium in ἄτης (ἄτην M) and γόµον (γόµου M), the former being an easy corruption of ἅ τις. But there is also quite possibly a lacuna after 444 (Dawe 1972): the dative χρήµασιν does appear to need a construction; and whereas the genitive χρηµάτων found in the margin of M is attractive, the corruption to the dative seems improbable. Voss’s πορθουµένοις is of no apparent help. As for 447–48, transposition is possible. The similar endings of the verses (-τήριος / -τήρια) may have led to one of them being inadvertently overlooked in the process of transcription, only later added in the margin—and then inserted at the wrong place. The transposition is hardly certain, however: one could as easily imagine a lacuna after 448, with the sense of something like ‘exchanging for soothing apologies’. 449. ὅµαιµον αἷµα: here the adjective seems to refer to the kinship of the Danaids, and especially of the Aegyptiads, with the Argives: a war between the latter two parties would cause kindred blood to be spilled, a dire pollution. (Cf. 402–4n.) 526 Young (1974) follows M, adopting only Scaliger’s γεµίζων, taking this and ἐµπλήσας as nominativi pendentes, and translating ‘On the one hand, with wealth from pillaged homes if I glut Ate, filling her up greatly with the cargo, —other (wealth) could accrue, thanks to Zeus of possessions’. Barigazzi (1983) is right to call this ‘eccessivo conservatismo’, but even moderate conservatism will be futile here, such as his ἄτης γε µείζω, καὶ µέγαν πλῆσαι γόµον (‘sia le richezze … possono diventare col favore di Zeus superiori alla perdita e tali da colmare un grande carico navale’). 202 450. δεῖ κάρτα θύειν καὶ πεσεῖν χρηστήρια: note the change of subject. The χρηστήρια are preliminary offerings before oracular consultation (LSJ s.v. II), not the other way around (consultation before sacrifice, as FJ–W). For this and for πεσεῖν in the context, cf. E. Ion 419: χρηστήριον πέπτωκε … πρὸ ναοῦ· βούλοµαι δ’ ἐν ἡµέρᾳ | τῇδ’ (αἰσία γάρ) θεοῦ λαβεῖν µαντεύµατα.527 χρηστήριον may also mean—and possibly does in this case—sacrifice in general, as in, for example, Pi. O. 6.70. κάρτα: ambiguous as to whether it defines δεῖ or θύειν, although the latter is perhaps more likely (see FJ–W). Cf. my notes on 22, 15–18. 452. παροίχοµαι: ‘I am lost’, ‘I am at a loss’. See Page on E. Med. 995: ‘the “not understanding” sense of “wander” is explained by ἄιδρις in [Supp.] 453 as here by οὐ κατειδώς 992.’ 455–67. The Danaids have the upper hand: their killing themselves on the holy precinct would mean unthinkable pollution and misery for the state of Argos. 455. αἰδοίων λόγων: cf. 194–95n. 456. ἤκουσα: the aorist is presumably a deliberate echo of this aspect of the verb in the Danaids’ imperative ἄκουσον in the previous verse, but the intended nuance is hard to understand. The tense is explained as ingressive by FJ–W: ‘have given ear’ (cf. LSJ s.v. I 3, II 1; K–G i. 163–65). 457. στρόφους: the exact sense is obscure, but if correct (Portus: στρόβους M), which is likely on account of the scholium and of Th. 871–72, the word apparently refers to a detail, presumably a cord or band of some sort (cf. Od. 13.438), which is characteristic of women’s clothing. So in Th. l.c., where the noun serves as the very definition of womankind: δυσαδελφόταται πασῶν ὁπόσαι στρόφον ἐσθῆσιν περιβάλλονται. As argued by FJ–W, it can hardly be synonymous to στρόφιον, ‘breast-band’, which would render the cited passage absurd—it has to be something worn outside the dress. 458. Marckscheffel’s (1847, 171) τάχ’ ἂν (τύχαν M) is certain. Page (in the apparatus criticus) argues that it is unsuited to the context, but one may perhaps discern some ironic detachment (so rightly FJ–W): ‘I suppose this would be appropriate for women…’. The ms.’ γυναικῶν … συµπρεπῆ is difficult, 527 Cf. also Σ A. Th. 230d: οὐ … µόνον χρηστήρια τὰ µαντεύµατα ἀλλὰ καὶ τὰ θύµατα. ἢ ὅτι θύοντες τοῖς θεοῖς καὶ αὐτῶν περί του δεόµενοι τὰς µαντείας δεχόµεθα. ἢ ὅτι µέρος µαντικῆς ἐστι καὶ τὸ διὰ σφαγίων τὰς µαντείας ποιεῖσθαι ὁρῶντας τὴν τοῦ ζῴου χολὴν καὶ τὸ ἦπαρ καὶ τὴν κύστην. 203 but perhaps not impossible, being construed in analogy with ἄξιος. Cf. E. HF 131 with the note of Wilamowitz and also FJ–W, who, however, adopt Hermann’s γυναικὶ. The genitive is retained by Page (with doubts), West, and Murray who refers to Wilamowitz l.c. 459–65. The situation is similar to that of Danaë in Dikt. (fr. 47a) 773–79: she is threatening to hang herself to prevent her being ravished by Satyrs: κνωδάλοις µε δώσετε … ἀγχόνην ἄρ’ ἅψοµαι … τεµοῦσα κωλυτήριον. In her case the threat is an emotional outburst rather than, as here, the result of cold calculation. 460. Elaborate periphrastic for ‘what?’, such as is a stock component of stichomythia. Cf., e.g., Th. 261, 713, Ag. 543, Ch. 117.528 γηρυθεῖσ’ ἔσῃ: not equivalent to a future perfect according to FJ–W (pace, e.g., Jebb on S. OT 1146, K–G i. 39). To be sure, the perfective aspect would be meaningless here. 462. Condescending: ‘then what will your device of girdles accomplish?’. 467. σαφέστερον: perhaps a redundant adverb (cf. Ch. 735, 767) rather than an adjective taken with an unexpressed object (σε or λόγον): ‘I clarified it more clearly.’ 468. The unmetrical verse is obviously corrupt, and unlikely to be emended by a palaeographically easy conjecture. I cannot see, however, why πολλαχῇ should be suspect, nor why it ‘must mean (in view of 469) not “on many sides” … but “in many ways”’ (FJ–W). The alliteration in 468–70 πολλαχῇ … (δυσ-)πάλαιστα … πράγµατα … πλῆθος … ποταµὸς … πέλαγος seems intentional, and makes it likely that the adverb is sound.529 There is the possibility that †καὶ µὴν† is simply intrusive (rather than just µὴν: see FJ–W). The tone does seem a little too reasonable and detached for Pelasgus to use in this situation. One would have expected something stronger, for example φεῦ (cf. Pers. 285, 739)—better, I think, than Paley’s ἦ (ed. 1883): we expect exclamation, not affirmation. γε would be regularly exclamatory, or rather perhaps explanatory of the emotional outburst: see D.GP 128, with 528 ‘A shepherd’s questioned mouth informed me that— | What? for I know not yet what you will say.’ (from A. E. Housman, Fragment of a Greek Tragedy). 529 On Aeschylean alliteration, see now Garvie (2002), who notes, however (p. 4) that alliterations on π are under suspicion, this consonant being especially frequent in the beginning of Greek words. 204 plenty of similar examples.530 φεῦ reinforces the above-mentioned alliteration. The corruption is hard to account for, however. West retains καὶ µὴν, adopting Sulzberger’s (1945, 139) lacuna (καὶ µὴν <… | καὶ> πολλαχῇ γε κτἑ). Anything spoken within the lacuna seems destined to be utterly redundant, unless perhaps, as suggested by West, the Danaids speak the verse beginning καὶ µὴν. 469–71. For the Waters of Evil cf., apart from FJ–W’s parallels, Th. 758– 61 κακῶν δ’ ὥσπερ θάλασσα κῦµ’ ἄγει· τὸ µὲν πίτνον, ἄλλο δ’ ἀείρει τρίχαλον, ὃ καὶ περὶ πρύµναν πόλεως καχλάζει. In that case ἀλκὰ δι’ ὀλίγου τείνει πύργος ἐν εὔρει. Here, there is no escape. 472–79. Pelasgus is finally convinced: the wrath of the godhead outweighs any secular considerations. There is no further discussion about the rightfulness of the Danaids’ cause: as suppliants in the holy precinct, they have a priori sanctity. The ultimate decision thus becomes Realpolitik, and not the result of moral considerations: the negative consequences for the state of each respective action tip the balance. Contrast to this the traditional (Athenian) propagandistic self-image as selfless champions of the righteous cause of suppliants: a stock component in patriotic speeches of the late fifth and fourth century.531 Contrast also Pelasgus’ chauvinistic and self-righteous dismissal of the Egyptian herald in 911–53: there necessity has already become virtue. See further 480–523n. below. 479 ἱκτῆρος: see 1n. 480–523. Regardless of the Danaids’ dubious means of persuasion, and the ultimately self-serving reasons for taking on their cause (472–79n.), Pelasgus 530 For instance, Pers. 739, E. Hel. 777, Andr. 184. See, e.g., Lys. 2 with Stevens’ notes on especially 7, 11, 17, 20, Kartes (2000) 21, 37–50, 155–56, Pl. Menex. 239a, 244e, passim, Isoc. Paneg. 28–40, 51–72, passim, D. 60.8–9, 11. Similarly Euripides’ propagandistic portrait of Theseus in the Supplices, on which see Collard i. 4–6, 24, 29 and his notes on 188–90 (with further parallels), 308–12, 577, etc. A rather more ‘realist’ view of Athenian imperialism is presented in the Athenians’ speech in Th. 1.73–76; a hostile view (with reluctant admiration) was given by the Cercyreans before them. As a contrast to the chauvinist speeches previously mentioned, see also Th. 1.70 with the notes of Hornblower. It is noteworthy that Pericles eschews any talk about Athenian altruism in the great Funeral Speech in Th. 2.35–46, whether because of his own or Thucydides’ distaste for hypocrisy. For a thorough discussion of the contrast between the speeches in Thucydides and in later Attic (and Atticism), see Strasburger (1958). 531 205 has now become their whole-hearted champion, and not only politically but morally. He even adopts the Danaids’ own biased language, speaking of the ὕβριν ἄρσενος στόλου in 487. Such he remains for the rest of the play, a fact which has been taken as an example of Aeschylus’ disregard of psychological realism (cf. 176–78n., and FJ–W i. 29, 478–79n.). The matter is not yet entirely settled: the king needs the approval of his people, whether formally or simply as a matter of prudence (cf. 365–69n.). For this purpose he and Danaus remove to the city, leaving the stage (in 504 and 523, respectively) to the Danaids for a lyrical interlude. 481. κλάδους τε τούτους: the anacoluthon, with τε left unanswered, is unparalleled in Aeschylus but perhaps not impossible: as Pelasgus elaborates the subjunctive clauses, briefly wallowing in self-pity (484–85), he forgets the original construction. But Auratus’ γε is also not impossible: the slight emphasis on the boughs is in order in directing Danaus’ attention towards them; and would it be entirely fanciful to detect aristocratic condescendence? (‘These boughs, take them…’). More extensive emendations worth mentioning are (1) Butler’s lacuna before this verse (adopted by West), (2) Weil’s αἶρ’ [αἶψ’ M] … βωµούς <τ’>: for the alternation between aorist and present imperatives, see Diggle (1981a) 62. 483. ἀφίξεως: see 1n. 484. ἀρχῆς γὰρ φιλαίτιος λεώς: a common sentiment in later Attic literature, and also one of which Aeschylus had seen instances in real life: Palladini (2001, 449–50) comments on this passage and compares it with examples of unjust cases of indictments and ostracism in the 490s to 470s, e.g., Militades in 489, Themistocles in 471. 486. καὶ γάρ does not mean ‘for in fact’ (pace FJ–W): καὶ links the formerly stated purport of the suggested action (expressed in two final clauses) with a new one (expressed with parataxis γάρ): ‘and also since…’ (D.GP 108, § I 1). There is thus nothing illogical in τάχ’, ‘perhaps’. Linwood’s (1843, 237) οἰκτίσας ἰδὼν (οἶκτος εἰσιδὼν M) stands a good chance of being right. 491. εὑρεθέντα … λαβεῖν (Porson: εὖρ’ ἐόντα Mac: εὖ ρέοντα Mpc) is not ‘languid’ (Tucker, approved by FJ–W): the pleonasm is equivalent to δίδωµι … φέρειν (LSJ s.v. φέρω A XI) or ἔχειν (cf. 80), and verbal redundancy as such is fairly common in Aeschylus (cf. my notes on 92, 364, 467). 492. ὀπάονας … ἐγχωρίων: the retinue that arrived with Pelasgus in 234 will now escort Danaus into town. The subsidiary, hitherto silent, chorus will reappear as Egyptians in 825. 206 496–99. Careful advice from Danaus, in accordance with his character and function in the drama (see 176–78n.). On the ethnical differences, cf. 154– 55n., 235n. and 497–98n. 497–98. Hall (1989, 173) notes that the implicit idea about the importance of climate and nurture for human physiology and temperament is found expressed in near-contemporaneous works: cf. Hdt. 3.12 and, especially, the Hippocratic De Aëre aquis et locis, where the influence of the climate on various nations is explained in detail. 498–99. Probably the idea is that outlandishness in combination with boldness (cf. 197, 203) produces fear in one’s neighbours, fear which leads to hate and violence, a well-known psychological process. Taking the fear as belonging to Danaus will produce a very awkward non sequitur. de Pauw’s φόνον would remove one of the logical steps in the process boldness–fear (hatred)–violence, and is therefore detrimental. καὶ δὴ appear to retain their separate senses, καὶ meaning ‘even’, stressing φίλον, and δὴ being regularly emphatic (‘indeed’). Cf. D.GP 250. The aorist ἔκταν’ is gnomic. 500. εὖ γὰρ ὁ ξένος λέγει: see 496–99n. 502–3. ναύτην: Pelasgus describes himself as a ναύκληρος in 177, and the feature of sea-voyaging is stressed in 134–35, q.v. The point here seems to be that the presence of a foreign seafarer would attract the attention of the townspeople. It has been suggested that the sense of the noun here is actually ‘suppliant’, derived from the verb ναύω or ναεύω: see 354–55n. and W.SA ad locc. Dawe (1972) objects that the audience could not possibly understand the homonym as anything other than ‘sailor’: however, the diphthong would perhaps be differently pronounced, with a long ᾱ: νά̄υτην. Wecklein (1872, 83– 84) suggested ναύστην or ναυστῆρ’, adducing Hsch. ν 149 ναυστῆρες· οἱ ἱκέται [*Lobeck: οἰκέται mss.]. 504. καὶ τεταγµένος κίοι: the optative is peculiar, but retained by most modern editors. κἂν might be an alternative, ‘and being ordered so, (I suppose) he should go’ (for the corruption cf. 194–95, 276, 296). Also Portus’ κίει, a very easy itacist corruption. Nevertheless a force of the bare optative akin to concessive or hortative subjunctive is possible.532 506. σηµεῖον πόνου: for the appositional phrase, cf. 218. 532 See K–G i. 229–30, S.GG ii. 322 (§ 4.3), Smyth (1956) 406 (§§1819-20). 207 508. λευρὸν … ἄλσος … ἐπιστρέφου: the Danaids are asked to move down from the πάγος, which is holy, and settle on the level ground of the precinct, which is less so (cf. 509 βέβηλον). There is an external dramatic purpose for this: the further action of the drama requires that the girls not be directly protected by the sanctity of the gods. Also, as noted by Jouanna (2002, 791), there is a scenic aspect: the chorus must move down into the orchestra to be able to perform the dance of the next odes (the kommos or amoibaion in 348– 437 was obviously performed without dancing). However, this measure taken by Pelasgus may also be construed as a precaution against unforeseen events: should the Aegyptiads arrive, he would not want the Danaids to have the opportunity to kill themselves before the gods in the manner previously described. On the other hand, nothing would prevent them from ascending to the πάγος once more when Pelasgus is gone, unless perhaps their leaving their boughs on the altar is somehow conceived as precluding this alternative for religious reasons. 509. βέβηλον ἄλσος: ‘some ἄλση were open to the public (βέβηλα), while others formed part of the sacred precinct proper’, Smyth in a footnote to the present passage. See Burkert (1985) 86 with refs in n. 30 (p. 381). 510–15. Cf. the dialogue between Eteocles and the frightened women of Thebes in Th. 245–63. In this case Danaus cannot afford to rage at the timidity of the women, but has to be content with sarcasm (510): ‘I will not deliver you to the ravage of birds’ (as if this was the most pressing danger). Cf. also 512n. 512. εὔφηµον … εὐφηµουµένῃ implies that Pelasgus was not only being sarcastic, but deliberately avoided mentioning the real danger in 510, in order not to be the bearer of ill omen (so FJ–W, cf. 376n.). The coryphaeus on the other hand was dangerously explicit in speaking of ‘those who are more loathsome than evil dragons’ (511); and Pelasgus may imply ‘as you were spoken to without ill portent, so should you speak yourself’. εὐφηµουµένῃ does not mean ‘be spoken well to’, which is far too weak a sense for a word which ordinarily means ‘speak without ill portent’ or ‘praise’, but retains its original sense. 513. φόβῳ φρενός: cf. 379. 514. †ἀεὶ δ’ ἀνάκτων†: Garvie’s (1973) λύειν is excellent,533 especially in 533 Garvie compares Th. 270. Friis Johansen (ed. 1970) first thought of a predicate infinitive in place of ἀεὶ δ’ (παύειν δ’). 208 the light of the subsequent verse(s). It might be better, palaeographically as well as semantically, if we delete δ’: λυειν > λυειδ > αιειδ.534 If the sense of the verse thus expresses Pelasgus’ willingness to offer consolation to the Danaids, not only does 515 become a perfectly natural answer, but the following speech from Pelasgus—the last before he leaves the stage—will also be in perfect accordance: he tries as best he can to assuage their fear. 516. ἐρηµώσει πατήρ: the expression seems unexceptionable, pace FJ– W (see LSJ s.v. III–IV). As the object is unexpressed, the sense becomes almost intransitive: ‘be gone.’ 517–23. It is difficult to understand how Turnebus’ πείσω in 518 (πιετω M, ἐπιέτω Md) has come to be so universally forgotten, hardly mentioned in any respectable edition of the twentieth century and ousted in favour of the likes of στείχω (Weil) and σπεύσω (Martin 1858). One may have to go back to Burges’ edition to find a discussion of the emendation: ‘vix et ne vix quidem Pelasgus dicere potuit, Πείσω―ὡς τιθῶ: potuit quidem, πείσω―ὡς τεθῇ (scil. Populus Argivus).’ To my mind, the former alternative is not so impossible, or even so awkward, as Burges thinks: if πείσω is taken with the previous sentence (517 λαοὺς … ἐγχωρίους), the ὡς ἂν-clause becomes independently consecutive: ‘assembling the host I shall persuade it, so that I may put the community in a favourable mood’. For the participum coniunctum in the present tense, cf. S. El. 778, X. Cyr. 1.4.22. στείχω is very far removed from the paradosis, which is true for all suggestions of verbs meaning ‘go’ except Wecklein’s (ed. 1885) πατῶ, which is indeed ‘semantically unsatisfactory’ (FJ–W). A verb which describes Pelasgus’ addressing the crowd is welcome here, making the structure persuasively simple: first the roles of Pelasgus and Danaus in the town, λαοὺς … πείσω … καὶ … διδάξω πατέρα … λέγειν, and then the Danaids contra Pelasgus πρὸς ταῦτα µίµνε … ἐγὼ δὲ ἐλεύσοµαι. This anticipates a third objection, viz. the meaningless repetition στείχω … ἐλεύσοµαι, ‘I will go … to persuade them … I will go to do this.’ Finally, we may note the repetition of the root in 523 πειθὼ and 527 πιθοῦ. 523. πειθὼ: see the previous note. ἕποιτο: cf. 197. τύχη πρακτήριος: in regular contrast to πειθὼ, words, we find as usual an expression of action. The admixture of luck gives an original touch to the old 534 Linwood’s (1847, 133) γυναικῶν accords with Aeschylus’ views (cf. Th. 182– 286), but the subsequent verse becomes somewhat of a non-sequitur. 209 cliché: in order to be successful this time, action in itself will not be enough; it has to be abetted by τύχη, here semi-personified (see FJ–W). Excursus: transpositions of lines The palaeographical and text-critical foundations for the editorial measure of multiple transpositions of non-consecutive lines do not seem to have come in for much consideration (one exception is R. D. Dawe: see below). In the present drama, four passages carry more or less broad editorial consensus as to the need for such emendation (see below). However, the critic should note that while it may sometimes be likely that one or a number of consecutive lines have been displaced in a given text, usually owing to their having for some reason been dislocated into the margin (see West 1973, 28), a corruption that involves several transpositions of discrete lines is much harder to explain. The principle that a textual corruption leads to further deterioration of the surrounding text cannot easily be applied to cases of transposition: one transposition will not by any likely process cause further transpositions to occur in the vicinity. Accordingly, in cases of single, unconnected lines of verse, several transpositions occurring within a given area will be exponentially more unlikely than one transposition. If there is, say, one chance in a hundred that a transposition should occur within a given sequence of verses (say, ten), the likelihood that two transpositions occur within the same sequence is one in ten thousand. Cf. Dawe (2001a, 122, cf. also 129) on A. Y. Campbell’s additions and transpositions in Ag. 929–72: ‘this arrangement is so complicated that even if it were, by some chance, right, it would be irresponsible to adopt it, because the mathematical odds against it are piled up in a way that would leave Ossa and Pelion looking like molehills.’ But so, I contend, would several if not all of the multiple transpositions in Dawe’s Sophocles. The chances that several transpositions of lines may have occurred in close proximity are simply not such that they should be allowed to encourage this measure of criticism—unless one is able to propose a plausible causal relation between the cases of transposition.535 535 One such relation would be scribal conjecture: a scribe confronted with lines in disorder would perhaps try to correct the mistake, instead ending up disordering them further. In this case, however, on account of the lectio difficilior-principle, one 210 The same principle is applicable to lacunae in combination with transpositions: there being no apparent causal relation between the two types of corruption, the probability that the two might occur in combination is exponentially lessened, and consequently extremely low.536 As for multiple lacunae, the case may not be quite as hopeless: material deterioration of the exemplar at some point in the tradition may be a plausible cause. This principle, or any principle of textual criticism, should not be adhered to slavishly. If the result is evident, it must be accepted. One exception to the rule, which I would set down as an actual case of discrete multiple displacement, is 86–95 of the present drama. The process of corruption is tentatively explained ad loc. In several other alleged cases, however, I would hesitate to introduce such a conjecture in the text, if the result is not absolutely evident.537 In the critical apparatus to Ant. 740–57, Dawe offers a tentative theory about the multiple transpositions: ‘ordinem codicum ex confuse histrionum memoria turbatum’; also, Dawe (1978, 111) on the same passage: ‘in so complex a case [the dislocation] must be attributed to a confused memory rather than scribal incompetence’. This would imply that the tradition does not ultimately stem from the author’s autograph, but from a written record of a re-enactment of the tragedy, where the actors (?) mixed up the order of their lines. To be sure, re-enactments with directors’ or actors’ interpolations have demonstrably had an impact on the tradition of several dramas, but would think it more likely that the scribe should correct the mistake than make it worse. 536 Cf. the so-called ‘Lex Youtie’: see above, 438n., n. 524. 537 In the latest Oxford and Teubner editions of the tragedians, several discrete (non-consecutive) lines in close proximity, or in combination with lacunae, are supposed to have been transposed in the following passages (I omit the names of the original authors of these transpositions and lacunae): Pers. 312–18 (Page, West), Th. 803–20 (Murray), 983–93 (Murray, West), Supp. 207–11 (Murray, Page, West), 294– 316 (West: Page only posits lacunae, albeit five of them), 444–48 (Murray, Page, West), 905–10 (West), Ag. 570–76 (West), Ch. 227–30 (Murray, Page), 237–43 (West), Eu. 367–80 (Murray, Page), 485–89 (West), S. Aj. 1066–70, El. 1047–53, 1205–10, OT 243–73, Ant. 740–57 (Dawe), E. Heracl. 683–91, El. 682–93, HF 1185– 88, Hel. 1226–30, Ba. 843–48 (Diggle). Lloyd-Jones–Wilson eschew multiple transpositions altogether in their Sophocles, adopting no more than three transpositions of single or consecutive lines in the entire text (Tr. 994–98, OC 189–99, 1028–33). 211 rather, one would think, so as to contaminate the original tradition with interpolations than to substitute it altogether.538 Apart from 86–95, three passages from the first half of the present drama have had multiple transpositions adopted in the text by several editors. I have attempted to elucidate them (ad locc.) in the light of the principles stated here. References Abbreviations of periodicals follow the standard of L’Année philologique. Abbreviations of journals and series not found there, or less well known, are listed below: Beitr.Alt. BSG Budé Fond.Hardt GGA HdA HThR IGmB JPh LCL NAC NGG OCT RAL RBPh RFIC SCI Teubner Tusculum VDI ZAlt 538 Beiträge zur Altertumskunde Berichte über die Verhandlungen der königl. sächsischen Gesellschaft der Wissenschaften zu Leipzig Collection des universités de France publiée sous le patronage de l’Association Guillaume Budé Entretiens pour l’étude de l’antiquité classique (Fondation Hardt) Göttingische gelehrte Anzeigen Handbuch der Altertumswissenschaft Harvard Theological Review Indogermanische Bibliothek The Journal of Philology Loeb Classical Library Numismatica e antichità classiche Nachrichten von der (königl.) Gesellschaft der Wissenschaften zu Göttingen, phil.-hist. Klasse Oxford Classical Texts Rendiconti della Classe di Scienze morali, storiche e filologiche dell’Accademia dei Lincei Revue belge de philologie et d’histoire Rivista di Filologia e d’Istruzione Classica Scripta classica Israelica Bibliotheca scriptorum Graecorum et Romanorum Teubneriana Tusculum-Bücherei Вестник древней истории (Revue d’histoire ancienne) Zeitschrift für die Alterthumswissenschaft See Hutchinson’s Septem, pp. xl–xlii. 212 I. Ancient Works Quotations and references usually follow the latest Teubner editions or the editions listed in the TLG Canon, third printed edition; or, for inscriptions, papyri, and some Byzantine lexicographers, the editions listed in LSJ, including the 1996 Supplement. In some cases the editions are explicitly noted and listed below under II 2, ‘editions, translations, and commentaries on other literature’. All fragments and testimonia of tragedy, except Euripides, are cited after the TrGF (see ibid.). Abbreviations of authors and texts, including inscriptions and papyri, adhere to LSJ and Lampe (in the case of Latin authors OCD), with the following additions and changes: A. Dan. Dikt. Epig. Myrm. Pr.pyr. Pr.sol. Salam. test. Theor. Thress. A.D. Constr. Add.Et.Gud. Antiphon Caed.Her. Nov. Tetr. Ar.Byz. Nom.aet. Aristonic. Sign.Il. Arr. Bithyn. Clem.Rom. Ep.virg. Com.pall.inc. Didasc. Dionys.Scyt. Et.Sym. E. Aeol. Alex. Bell. Tel. Aeschylus Danaïdes Diktyoulkoi Epigonoi Myrmidones Prometheus pyrkaieus Prometheus solutus Salaminiae testimonia Theoroi Thressae Apollonius Dyscolus De constructione Additamenta in Etymologicum Gudianum De caede Herodis In novercam Tetralogia Aristophanes grammaticus Nomina aetatum Aristonicus De signis Iliadis Flavius Arrianus Bithynica Clemens Romanus Epistulae de virginitate Comica pallata incerta Didascaliae Dionysius Scytobrachion Etymologicum Symeonis Euripides Aeolus Alexander Bellerophontes Telephus Eust.Macr. Hysm. Geo.Sync. Hdn. Mon.lex. Orth. Pros.cathol. Hebr.Inscr. Isid. Chron. Isoc. Antid. Busir. Euag. Paneg. Trapez. Leo Diac. Lex.Seg. Antatt. Verb.util. Gloss.rhet. Men. Sent. MenRom Myc. Nicet.Acom. Eustathius(-mathius?) Macrembolites Hysmine et Hysminias Georgius Syncellus Aelius Herodianus περὶ µονήρους λέξεως περὶ ὀρθογραφίας De prosodia catholica Ancient Hebrew Inscriptions Isidorus Chronica Isocrates Antidosis Busiris Euagoras Panegyricus Trapeziticus Leo Diaconus Lexica Segueriana Anonymus Antatticista Collectio verborum utilium e differentibus rhetoribus et sapientibus multis Glossae rhetoricae Menander Sententiae Menaea (see under II 2, ‘Editions [etc.]’) Documents in Mycenaean Greek (see Ventris– Chadwick under II 2, ‘Editions [etc.]’) Nicetas Acominates 213 Ps.Nonn. Schol.myth. Pseudo-Nonnus Scholia mythologica = sq. Comm.Gr.Naz.Or. In IV orationes Gregorii Nazianzeni commentarii Ph. Philo Judaeus Aet.mund. De aeternitate mundi Agr. De agricultura Decal. De decalogo Det.pot. Quod deterius potiori insidiari soleat Ebr. De ebrietate Flacc. In Flaccum Heres Quis rerum divinarum heres sit Immut. Quod Deus sit immutabilis Jos. De Josepho LA Legum allegoriarum libri III Plant. De plantatione Post.Cain. De posteritate Caini Qu.Gen. Quastiones in Genesim Sacr.Abel. De sacrificiis Abelis et Caini Somn. De somniis Spec.leg. De specialibus legibus Virt. De virtutibus Ph.Bybl. (H)eren(n)ius Philo Div.verb. De diversis verborum significationibus Phryn.Trag. Phrynichus Aeg. Aegyptii Phoen. Phoenissae Plu. Plutarchus Adv.Col. Adversus Colotem Aet.Rom. Apophth.Lac. Conv. Def.orac. E.Delph. Frat.am. Gen.Socr. Mul.virt. Pr.frig. Rect.aud. Porph. Phil.orac. S. Inach. test. Sen. Herc.Oet. Sept. Praecept. Σ ΣD Il. Σ rec. Σ Vat. D.T. Test.xii.patr. Tz. Ep. Zonar. Hist. Aetia Romana Apophthegmata Laconica Quaestiones convivales De defectu oraculorum De E apud Delphos De fraterno amore De genio Socratis Mulierum virtutes De primo frigido De recta ratione audiendi Porphyrius De philosophia ex oraculis Sophocles Inachus testimonia Seneca Hercules Oetaeus Septem Sapientes Praecepta Scholia vetera Scholia ‘D’ (Didymi) in Homeri Iliadem Scholia recentiora Commentaria in Dionysii Thracis Artem grammaticam: scholia Vaticana Testamenta xii patriarchum Joannes Tzetzes Epistulae Johannes Zonaras Epitome historiarum II. Modern Works Editions, translations and commentaries of/on ancient works are usually referred to by the last name of the editor (translator, commentator) and monographs and articles by last name and year of publication. A few works, including much-cited standard publications and lexica, are referred to by abbreviation, bracketed before the entry in the bibliography. References to conjectures usually follow the same rules as other references; in other words, the place where a conjecture was originally published is found in the bibliography if I have been able to track it down. This has proved impossible in some cases: an author who is not to be found in the bibliography is thus marked with an asterisk in the text, for instance ‘*Porson’. 214 1. The Supplices: editions, translations, and commentaries (separately or with other plays). [ALDINA] {Asulanus, F.} Αἰσχύλου τραγῳδίαι ἕξ ... Aeschyli tragoediae sex. Venice 1518. [Arsenius] Emendations by Arsenius or Aristoboulos Apostolidis, the scribe of the ms. Me (usually referred to simply as Me). Bassi, D. Eschilo: Le Supplici. Milan 1934. Boissonade, J. F. Αἰσχύλος: Aeschylus. i/ii, Paris 1825 (Poetarum Graecorum sylloge, 12). Bothe, F. H. Aeschyli dramata quae supersunt et deperditorum fragmenta. {i–ii,} Leipzig 1805. ——. Aeschyli Supplices. Leipzig 1830 (rep. in Bothe ed. 1831, pp. 113–84). ——. Aeschyli tragoediae. i/ii, Leipzig 1831 (Poetae scenici Graecorum, 9). Burges, G. Αἰσχύλου τραγῳδιοποιοῦ λείψανα: Aeschyli quae supersunt fabulae et fragmenta: Supplices. London 1821. Butler, S. (after Stanley, T.) Aeschyli tragoediae quae supersunt. ii/viii, Cambridge 1809. Canter, W. Αἰσχύλου τραγῳδίαι ζ´: Aeschyli tragoediae VII. Antwerp 1580. de Pauw, J. C. Aeschyli tragoediae superstites. i–ii, The Hague 1745. Dindorf, W. Αἰσχύλος: Aeschyli tragoediae superstites et deperditarum fragmenta. i–ii/iii, Oxford 1841. ——. Αἰσχύλου τραγῳδίαι: Aeschyli tragoediae. 3rd ed., Leipzig 1857 (prev. ed. 1827, -50; re-ed. -60, -65). ——. Poetarum scenicorum Graecorum ... fabulae superstites et perditarum fragmenta. 5th ed., Leipzig 1869 (the Aeschylus part was also published separately as Aeschyli fabulae superstites et perditarum fragmenta). [FJ–W] Friis Johansen, H., and Whittle, E. W. Aeschylus: The Suppliants. i–iii, Copenhagen 1980. Friis Johansen, H. Aeschylus: The Suppliants. Copenhagen 1970 (C&M Dissertationes, 7). Hartung, J. A. Aeschylos’ Werke. vii/viii, Die Danaiden. Leipzig 1854. Haupt, C. G. Aeschyli Supplices: Aeschylearum quaestionum specimen II. Leipzig 1829. Headlam, W. Aeschylus: The Suppliants. London 1900 (transl., rep. in id., The Plays of Aeschylus, London 1909). Hermann, G. Aeschyli tragoediae. i–ii, Leipzig 1852 (2nd ed. Berlin 1859). Kirchhoff, A. Aeschyli tragoediae. Berlin 1880. Kraus, W. Aischylos: Die Schutzsuchenden. Frankfurt a.M. {1948}. Mazon, P. Eschyle. i/ii, Les Suppliantes, Les Perses, Les Sept contre Thèbes, Prométhée enchainé. Paris 1920 (Budé). Merkel, R. Aeschyli quae supersunt in codice Laurentiano veterrimo. Oxford 1871 (diplomatic transcript). 215 Murray,* G. The Complete Plays of Aeschylus. London 1952 (transl.). ——. Aeschyli septem quae supersunt tragoediae. 2nd ed, Oxford 1955 (OCT, 1st ed. 1937). Oberdick, J. Die Schutzflehenden des Aeschylus. Berlin 1869. Page, D. Aeschyli septem quae supersunt tragoedias edidit D. Page. Oxford 1972 (OCT). Paley,† F. A. Αἰσχύλου Ἱκέτιδες: Aeschyli Supplices. Cambridge (etc.) 1844 (reissued as part of Aeschyli quae supersunt omnia, ii/ii, ibid. 1847). ——. The Tragedies of Aeschylus. 4th ed., London 1879 (prev. eds 1855, -61, -70). ——. Aeschyli fabulae Ἱκέτιδες Χοηφόροι. Cambridge 1883. {Porson, R.} Aeschyli tragoediae septem. i/ii, London (etc.) 1806.‡ Robortello, F. Αἰσχύλου τραγῳδίαι ἑπτά: Aeschyli tragoediae septem. Venice 1552. Rose, H. J. A Commentary on the Surviving Plays of Aeschylus. i/ii, Amsterdam 1957 (Verhandelingen der koninklijke nederlandse akademie van wetenschappen, afd. letterkunde, n.r. 64, fasc. 1). Scholefield, J. Αἰσχύλος: Aeschylus. Cambridge 1828. * If the edition or translation is not specified, the reference is to the 1955 edition. If the edition is not specified, the reference is to the 1879 edition. ‡ Two separate editions of Porson’s Aeschylus were printed before, in neither of which the editor received any credit whatsoever. The first, a folio volume of 357 pages, was issued by the Foulis press in Glasgow in 1795 and is described by Schütz (comm. 1797), who reports all of Porson’s emendations that appear in it. Having described the title-page and the general appearance of the book, Schütz goes on (pp. iv–v): ‘Editor, quem ferunt esse Cl. Porsonum, et nomen, et subsidia quibus usus est, et causas mutatae lectionis celavit; nihil enim aliud quam textum graecum exhibuit, ne tribus quidem verbis praefationis loco additis.’ The folio edition is also described in Cambridge Essays for 1857, 153–54 (excerpted on the front fly-leaf of Columbia University’s copy of the 1806 edition), where it is said to contain fewer corrections than the edition of 1806: so also according to Dindorf (ed. 1827, p. iii) and Wecklein (ed. 1885, ii, on the leaf following the title-page). Gaskell (1964, 388) is thus probably wrong to claim that the texts are identical: see, however, his pp. 386–87 on the folio edition. The ‘second’ Porson edition, in two volumes and including Stanley’s Latin translation, was printed by Foulis in Glasgow 1796, but it appears not to have been published at that time, except perhaps in a few copies (one example with a 1796 title-page which is at present on record in online library catalogues is in Cambridge, Univ. Lib. 7000.d.449–50). It was published in London and Oxford in 1806, i.e. as the edition listed here, with a cancel title-page ‘Glasguæ: excudebat Foulis, M,DCC,LXXXXIV … veneunt Londini, … Oxoniæ … M,DCCC, VI’. The former date is apparently a misprint for -XXXXVI (Gruys 1981, 340, n. 33). † 216 Schütz, C. G. Aeschyli tragoediae quae supersunt ac deperditarum fragmenta. iii/iii, Choephorae, Eumenides, Supplices. Halle 1794. ——. In Aeschyli tragoedias quae supersunt ac deperditarum fragmenta commentarius. iii/iii, In Choephoras, Eumenides et Supplices. Halle 1797. Schwerdt, F. I. Αἰσχύλου Ἱκέτιδες. i–ii, Berlin 1858. Sidgwick, A. Aeschyli tragoediae cum fabularum deperditarum fragmentis. Oxford {1900} (OCT, re-ed. 1902 with two minor changes). Smyth, H. W. Aeschylus. i/ii, Suppliant Maidens, Persians, Prometheus, Seven against Thebes. London (etc.) 1922 (LCL). Stanley, T. Αἰσχύλου τραγῳδίαι ἑπτά: Aeschyli tragoediae septem. London 1663. Tucker, T. G. Αἰσχύλου Ἱκέτιδες: The ‘Supplices’ of Aeschylus. London 1889. Turnebus, A. Αἰσχύλου Προµηθεὺς δεσµώτης, Ἑπτὰ ἐπὶ Θήβαις, Πέρσαι, Ἀγαµέµνων, Εὐµενίδες, Ἱκέτιδες. Paris 1552. Untersteiner, M. Eschilo: Le Supplici. Naples 1935. ——. Eschilo: Le tragedie. i/ii, Milan 1946. Vílchez, M. Esquilo: Tragedias. ii, Los Siete contra Tebas, Las Suplicantes. Madrid 1999. Vürtheim, J. Aischylos’ Schutzflehende. Amsterdam 1928. Wecklein, N. Aeschyli fabulae. i–ii, Berlin 1885 (an Auctarium to each volume was published in 1893: see below under ‘Editions, translations and commentaries on other ancient works’ and ‘Scholarly works other than editions, translations and commentaries’). ——. Äschylos: Die Schutzflehenden. Leipzig 1902. Wecklein, N., and Zomaridis, E. (Ζωµαρίδης, Ε.) Αἰσχύλου δράµατα σῳζόµενα καὶ ἀπολωλότων ἀποσπάσµατα. ii/iii, περιέχων Προµηθέα, Ἱκέτιδας καὶ ἀποσπασµάτια. Athens 1896 (Ζῳγραφεῖος Ἑλληνικὴς βιβλιοθήκη, 6). Weil, H. Aeschyli quae supersunt tragoediae. ii/ii, fasc. 3/4, Supplices. Giessen 1866. Weise, C. H. Αἰσχύλου τραγῳδίαι: Aeschyli tragoediae. Leipzig 1812. Wellauer, A. Aeschyli tragoediae. i/iv, Prometheum, Septem contra Thebas et Supplices continens. Leipzig 1823. Werner, O. Aeschylos: Tragödien und Fragmente. Munich 1959 (Tusculum). West, M. L. Aeschyli tragoediae cum incerti poetae Prometheo. Stuttgart (etc.) 1990 (Teubner, ed. corr. 1998; Aeschyli Supplices issued separately ibid. 1992). Wilamowitz-Moellendorff, U. von. Aeschyli tragoediae. Berlin 1914. 2. Editions, translations, and commentaries on other ancient works Allen, T. W., Halliday, W. R., and Sikes, E. E. The Homeric Hymns. 2nd ed., Oxford 1936. Austin, C. Comicorum Graecorum fragmenta in papyris reperta. Berlin 1973. Bachmann, L. Anecdota Graeca. i–ii, Leipzig 1828 (= An.Bachm. in LSJ). 217 Bagordo, A. Die antiken Traktate über das Drama. Stuttgart (etc.) 1998 (Beitr.Alt. 111). Bamberger, F. Aeschyli Choephori. Göttingen 1840. Barrett, W. S. Euripides: Hippolytos. Oxford 1964. Bechtel et al. ‘Graeca Halensis’ (= Bechtel, F.; Kern, O.; Praechter, K.; Robert, C.; Stern, E. von; Wilcken, U.; Wissowa, G.), Dikaiomata. Berlin 1913. Bekker, I. Anecdota Graeca. i–iii, Berlin 1814–21 (= AB in LSJ). ——. Apollonii Sophistae lexicon Homericum. Berlin 1833 (rep. Hildesheim 1967). Biehl, W. Euripides: Ion. Leipzig 1979 (Teubner). Boissonade, J. F. Herodiani partitiones. London 1819 (rep. Amsterdam 1963). Bollack, J. Agamemnon. i, part 1–2, Lille 1981 (Cahiers de philologie, 6–7). Boysen, C. Lexici Segueriani Συναγωγὴ λέξεων χρησίµων inscripti pars prima. Progr., Marburg 1891–92 (rep. in LGM pp. 12–38). Braswell, B. K. A Commentary on the Fourth Pythian Ode of Pindar. Berlin (etc.) 1988 (Texte und Kommentare, 14). Broadhead, H. D. The Persae of Aeschylus. Cambridge 1960. Brownson, C. A. Xenophon: Hellenica, Books VI & VII. Anabasis, Books I–III. London (etc.) 1921 (LCL). Burnet, J. Platonis opera. i/v, Oxford {1900} (OCT). ——. Plato’s Phaedo. Oxford 1911 (rep. 1949). Campbell, D. A. Greek Lyric. i/v, Sappho and Alcaeus. Cambridge, Mass. (etc.) 1982 (LCL, ed. corr. 1992). Chryssafis, G. A Textual and Stylistic Commentary on Theocritus’ Idyll XXV. Amsterdam 1981 (London Studies in Classical Philology, 1). Collard, C. Euripides: Supplices. i–ii, Groningen 1975. Crönert, W. ‘Griechische literarische Papyri aus Strassburg, Freiburg und Berlin’. NGG, 1922, fasc. 1, 1–46. Davies, G. I. Ancient Hebrew Inscriptions: Corpus and Concordance. Cambridge (etc.) 1991. Davies, M. Sophocles: Trachiniae. Oxford 1991. Dawe, R. D. Sophocles: Oedipus Rex. Cambridge 1982. ——. Sophoclis Aiax. 3rd ed., Stuttgart 1996 (Teubner). ——. Sophoclis Antigone. —“—. ——. Sophoclis Electra. —“—. ——. Sophoclis Oedipus Coloneus. —“—. ——. Sophoclis Oedipus rex. —“—. ——. Sophoclis Philoctetes. —“—. ——. Sophoclis Trachiniae. —“—. de Stefani, A. Etymologicum Gudianum quod vocatur. i–ii, Leipzig 1919–20. [DGEE] Dialectorum Graecarum exampla epigraphica potiora. Ed. E. Schwyzer, Leipzig 1923 (= Schwyzer in LSJ; rep. Hildesheim 1960). Diehl, E. Anthologia lyrica Graeca. i–iii, 3rd ed., Leipzig 1949–52 (Teubner). 218 van Dieten, J.-L. Nicetae Choniatae historia. i–ii, Berlin 1975 (Corpus fontium historiae Byzantinae, Series Berolinensis, 11). Diggle, J. Euripidis fabulae. i–iii, Oxford 1981–94 (OCT). ——. Tragicorum Graecorum fragmenta selecta = TrGFS. Dindorf, L. Ioannis Malalae Chronographia. Bonn 1831. ——. Ἰωάννου τοῦ Ζωνάρα ἐπιτοµὴ ἱστοριῶν: Ioannis Zonarae epitome historiarum. i–vi, Leipzig 1868–75 (Teubner). [D–K] Die Fragmente der Vorsokratiker. Ed. H. Diels and W. Kranz, i–iii, 6th ed., Berlin 1951–52 (rep. Zürich 1989–92). Dodds, E. R. Euripides: Bacchae. 2nd ed., Oxford 1960. Dover, K. Aristophanes: Frogs. Oxford 1993. [DOX.GR.] Doxographi Graeci. Ed. H. Diels, Berlin 1879 (rep. 1965). Dunbar, N. Aristophanes: Birds. Oxford 1995. Elmsley, P. Εὐριπίδου Μήδεια: Euripidis Medea. Oxford 1818 (2nd ed. Leipzig 1822: pages cited within parentheses). Farnell, L. R. The Works of Pindar. i–iii, London 1930–32 (vol. ii, Critical Commentary rep. Amsterdam 1965 as Critical Commentary to the Works of Pindar). [FGRH] Die Fragmente der griechischen Historiker. i–iii (various sub-vols), ed. F. Jacoby; iv, part A, ed. G. Schepens a.o., Berlin, Leiden 1923–. [FHG] Fragmenta historicorum Graecorum. i–v, fasc. 1, ed. C. Müller; v, fasc. 2, ed. V. Langlois, Paris 1828–83. Fowler, R. L. Early Greek Mythography. i/ii, Texts. Oxford 2000. Fraenkel, E. Aeschylus: Agamemnon. i–iii, Oxford 1950. Frazer, J. G. Apollodorus: The Library. i–ii, London (etc.) 1921 (LCL). Fromentin, V. Denys d’Halicarnasse: Antiquités romaines. i/iii, Paris 1998 (Budé). Garvie, A. F. Aeschylus: Choephori. Oxford 1986. Giannini, P. In Pindaro: Le Pitiche. Ed. P. A. Bernadini, E. Cingano, B. Gentili and P. Giannini, Verona 1995 (Giannini’s notes on Odes 4–6, 8–9). Godley, A. D. Herodotus. i–iv, London (etc.) 1920–25 (LCL). Goetz, G. Corpus glossariorum Latinorum. i–vii, Leipzig 1888–1903. Gomme, A. W. A Historical Commentary on Thucydides. i/v, Introduction and Commentary on Book I. Oxford 1945. Gow, A. S. F. Theocritus. i–ii, Cambridge 1950. ——. Machon: The Fragments. Cambridge 1965 (Cambridge Classical Texts and Commentaries, 1). Gow, A. S. F., and Page, D. L. The Greek Anthology: Hellenistic Epigrams. i–ii, Cambridge 1965. Griffith, M. Aeschylus: Prometheus Bound. Cambridge 1983. Hagen, H. Appendix Serviana. Leipzig 1902 (Servi Grammatici qui feruntur in Vergilii carmina commentarii, 3.2) Hall, E. Aeschylus: Persians. Warminster 1996. 219 Hase, K. B. Leonis Diaconi Caloënsis historiae libri decem. Bonn 1828 (Corpus scriptorum historiae Byzantinae, 11). Headlam, W. Herodas: The Mimes and Fragments. Ed. A. D. Knox, Cambridge 1922 (rep. 1966). Henderson, J. Aristophanes: Lysistrata. Oxford 1987. ——. Aristophanes. i–iv, Cambridge, Mass. (etc.) 1998–2002 (LCL). Herington, C. J. The older scholia on the Prometheus bound. Leiden 1972. Heylbut, H. ‘Ptolemaeus περὶ διαφορᾶς λέξεων’. Hermes 22 (1887), 388–410. Heyne, C. G. Ὁµήρου Ἰλιάς: Homeri Ilias. i–ii, Oxford 1834 (cited for Σ rec.). Hilgard, A. Scholia in Dionysii Thracis Artem grammaticam. iii/iii, Leipzig 1901 (Grammatici Graeci, 1.3). Hoekstra, A. In Heubeck, A., and Hoekstra, A., A Commentary on Homer’s Odyssey. ii/iii, Books IX–XVI. Oxford 1989 (Hoekstra’s notes on books 13–16). Hoffner, H. A. Hittite Myths. Atlanta 1990 (Society of Biblical Literature Writings from the Ancient World, 2 [transl.]). Hornblower, S. A Commentary on Thucydides. i/iii, Books I–III. Oxford 1991. Humboldt, W. von. Aischylos: Agamemnon. Leipzig 1816 (transl., cited for Hermann’s critical notes in the postscript). Hunt, A. S. ‘Sappho: Book iv’ (New Classical Fragments: 1787). The Oxyrhynchus Papyri, 15 (ed. B. P. Grenfell and A. S. Hunt), London 1908, pp. 26–46. Hutchinson, G. O. Aeschylus: Septem contra Thebas. Oxford 1985 (ed. corr. 1987). Jacoby, F. See FGrH above. Jäkel, S. Menandri sententiae. Leipzig 1964 (Teubner). Janko, R. The Iliad: A Commentary. iv/vi, Books 13–16. Cambridge (etc.) 1992 (general ed. G. S. Kirk). Jebb, R. C. Sophocles: The Plays and Fragments. i–vii, Cambridge 1887–1900. ——. Bacchylides: The Poems and Fragments. Cambridge 1905. Kirk, G. S. The Iliad: A Commentary. i /vi, Books 1–4. Cambridge (etc.) 1985. Kock, T. Comicorum Atticorum fragmenta. i–iii, Leipzig 1880–88. Kovacs, D. Euripides. i–vi, Cambridge, Mass. (etc.) 1994–2003 (LCL). Lamb, W. R. M. Plato: With an English Translation. v/xii, Lysis, Symposium, Gorgias. ‘Revised ed.’, London (etc.) 1932 (LCL). Latte, K. Hesychii Alexandrini lexicon. i–ii, Copenhagen 1953–66. Lentz, A. Herodiani Technici reliquiae. i–ii, Leipzig 1867–70 (Grammatici Graeci, 3.1–2). Leo, F. Plauti comoediae. i–ii, Berlin 1895–96. Leone, P. A.{L.} M. Ioannis Tzetzae epistulae. Leipzig 1972 (Teubner). von Leutsch, E. L. Corpus paroemiographorum Graecorum. ii/ii, Göttingen 1851 (rep. Hildesheim 1958). [LGM] Lexica Graeca minora. Ed. K. Latte and H. Erbse, Hildesheim 1965. Lloyd, A. B. Herodotus. Book II. i–iii, Leiden 1975–88 (Études préliminaires aux religions orientales dans l’empire romain, 43). 220 Lloyd-Jones, H. Sophocles. i–iii, Cambridge, Mass. (etc.) 1994–96 (LCL). Lloyd-Jones, H., and Wilson, N. G. Sophoclis fabulae. Oxford 1990 (OCT). Lobel, E. ‘Aeschylus: Various plays’ (New Classical Fragments: 2245–55). The Oxyrhynchus Papyri, 20 (ed. E. Lobel, C. P. Wegener and C. H. Roberts), London 1952, pp. 1–65. Lucas, D. W. Aristotle: Poetics. Oxford 1968. MacDowell, D. M. Demosthenes: Against Meidias. Oxford 1990. Maehler, H. Pindari Carmina cum fragmentis. ii/ii, Fragmenta. Indices. Leipzig 1989 (Teubner). Marg, W. ‘Περὶ φύσιος κόσµω καὶ ψυχᾶς’. In The Pythagorean Texts of the Hellenistic Period, ed. H. Thesleff, Åbo 1965 (Acta Academiae Aboensis, ser. A, 30.1), pp. 203–25. Mastronarde, D. J. Euripides: Phoenissae. Cambridge 1994 (Cambridge Classical Texts and Commentaries, 29). Meiggs, R., and Lewis, D. A Selection of Greek Historical Inscriptions to the End of the Fifth Century B.C. Oxford 1969 (rep. 1988). Meineke, A. Fragmenta poetarum comicorum Graecorum. i–v, Berlin 1839–57. ——. Stephan von Byzanz: Ethnika. Berlin 1849 (rep. Graz 1958). [MENROM] Μηναῖα τοῦ ὅλου ἐνιαυτοῦ. i–vi, Rome 1888–1901. Miller, E. C. B. ‘Aristophane de Byzance’. In id., Mélanges de littérature grecque, Paris 1868 (rep. Amsterdam 1965), 427–34 (rep. in LGM pp. 273–80). Mommsen, C. J. T. Pindari carmina. Berlin 1864. Mommsen, T. Chronica Minora: Saec. IV. V. VI. VII. ii/iii, Berlin 1894 (Monumenta Germaniae Historica, 11; rep. 1961). Mosshammer, A. A. Georgius Syncellus: Ecloga chronographica. Leipzig 1984 (Teubner). Müller, C. See FHG above. Nauck, A. Tragicorum Graecorum fragmenta. 2nd ed., Leipzig 1889 (rep. Hildesheim 1964). Otten, H. Eine althethitische Erzählung um die Stadt Zalpha. Wiesbaden 1973 (Studien zu den Bogazköy-Texten, 17). Page, D. L. Euripides: Medea. Oxford 1938. ——. Select Papyri = Pap.poet. Paley, F. A. Αἰσχύλου Χοηφόροι: Aeschyli Choephori. Cambridge 1845 (re-issued as part of Aeschyli quae supersunt omnia, i/ii, ibid. 1847). [PAP.POET.] Select Papyri. iii/iv, Literary Papyri: Poetry. Ed. D. L. Page, Cambridge, Mass. (etc.) 1941 (LCL). [PCG] Poetae comici Graeci. Ed. R. Kassel and C. Austin, i–viii, Berlin 1983–. Pearson, A. C. The fragments of Sophocles. Edited with additional notes from the papers of Sir R. C. Jebb and W. G. Headlam, i-iii, Cambridge 1917 (rep. in one vol. Amsterdam 1963). 221 [PEG] Poetarum epicorum Graecorum testimonia et fragmenta. i, ed. A. Bernabé, ed. corr., Stuttgart 1996 (Teubner, 1st ed. Leipzig 1987). [PG] Patrologiae cursus completus: Series Graeca. Ed. J.-P. Migne, i–clxi, Paris 1857–66. Porson, R. Euripidis tragoediae. 2nd ed., Leipzig 1802. Radt, S. Aeschylus. In TrGF iii (fragments and testimonia). ——. Sophocles. In TrGF iv (fragments and testimonia). Ribbeck, O. Scaenicae Romanorum poesis fragmenta. i–ii, 3rd ed., Leipzig 1897–98 (Teubner). Richardson, N. The Iliad: A Commentary. vi/vi, Books 21–24. Cambridge (etc.) 1993 (general ed. G. S. Kirk). Rowe, C. J. Plato: Phaedrus. Warminster 1986 (ed. corr. 1988). Russell, D. A. ‘Longinus’: On the Sublime. Oxford 1964. Russo, J. In Russo, J., Fernández-Galiano, M., and Heubeck, A. A Commentary on Homer’s Odyssey. iii/iii, Books XVII–XXIV. Oxford 1992 (Russo’s notes on books 17–20). Schmidt, M. Hesychii Alexandrini lexicon. 2nd ed. (‘ed. min.’), Jena 1867. Schneider, R. Apollonii Dyscoli quae supersunt. i/ii, fasc. 1/2, Apollonii scripta minora ... continens. Leipzig 1878 (Grammatici Graeci, 2.1.1). Seaford, R. A. S. Euripides: Cyclops. Oxford 1984 (ed. corr. 1988). Snell, B. Euripides Alexandros und andere straßburger Papyri mit Fragmenten griechischer Dichter. Berlin 1937 (Hermes Einzelschriften, 5). ——. Didascaliae. In TrGF i. 3–52. Sommerstein, A. H. Aeschylus: Eumenides. Cambridge (etc.) 1989. Stallbaum, G. Eustathii archiepiscopi Thessalonicensis commentarii ad Homeri Odysseam. i–ii, Leipzig 1825–26. Stevens, W. A. Select Orations of Lysias. Chicago 1876. Strachan, J. C. G. In Platonis opera, i, ed. E. A. Duke, W. F. Hicken, W. S. M. Nicoll, D. B. Robinson, J. C. G. Strachan, Oxford 1995 (OCT). [SVF] Stoicorum veterum fragmenta. i–iv, ed. H. von Arnim, Leipzig 1921–24. Thomson, G. The Oresteia of Aeschylus. i–ii, London 1938. [TRGF] Tragicorum Graecorum fragmenta. i, ed. B. Snell; ii, ed. R. Kannicht and B. Snell; iii–iv, ed. S. Radt, Göttingen 1977–99. [TRGFS] Tragicorum Graecorum fragmenta selecta. Ed. J. Diggle, Oxford 1998. Uhlig, G. Apollonii Dyscoli quae supersunt. ii/ii, Apollonii Dyscoli De constructione libri quattuor. Leipzig 1910 (Grammatici Graeci, 2.2). Valckenaer, L. C. Εὐριπίδου Φοίνισσαι: Euripidis tragoedia Phoenissae. Franeker 1755. van der Valk, M. Eustathii archiepiscopi Thessalonicensis commentarii ad Homeri Iliadem pertinentes. i–iv, Leiden 1971–87. Ventris, M., and Chadwick, J. Documents in Mycenaean Greek. 2nd ed., Cambridge 1973. 222 Verdenius, W. J. A Commentary on Hesiod, Works and Days, vv. 1–382. Leiden 1985 (Mnemosyne Suppl. 86). Walz, C. Rhetores Graeci. i–ix, Stuttgart 1832–36. Wecklein, N. Aeschyli fabulae: Voluminis I auctarium: fragmenta. Berlin 1893. West, M. L. Hesiod: Theogony. Oxford 1966. ——. Hesiod: Works & Days. Oxford 1978. ——. Homeri Ilias. i–ii, Stuttgart, Munich (etc.) 1998–2000 (Teubner). West, S. In Heubeck, A., West, S., and Hainsworth, J. B., A Commentary on Homer’s Odyssey. i/iii, Books I–VIII. Oxford 1988 (S. West’s notes on books 1–4). Wilamowitz-Moellendorff, U. von. Euripides: Herakles. i–ii, Berlin 1889. Willink, C. W. Euripides: Orestes. Oxford 1986. 3. Scholarly works other than editions, translations and commentaries Abbott, J. N. 1850. ‘On Aeschylus Suppl. 4 and 244–51’. The Classical Museum, 7, 309–11. Abresch, F. L. 1763. Animadversionum ad Aeschylum liber tertius. Zwolle. Ahrens, H. L. {1832}. De causis quibusdam Aeschyli nondum satis emendati commentatio. Progr. Ilfeld, Göttingen. Albinus, L. 2000. The House of Hades. Aarhus (Studies in Religion, 2). Alessandrì, S. 1995. ‘Dedica inedita a Hikesios da Metaponto’. Studi di Antichità (Lecce), 8, fasc. 2, 77–93. [Anon.Ald.] Anonymous conjectures in a copy of the Aldina, Cambridge University Library Adv. d.25.1 (= West’s ‘anon.3’: see W.SA 365). [Anon.Barth.] Anonymous conjectures in a copy of Victorius (British Library 11705 d.2), once owned by a certain Barthomier, otherwise unknown (= West’s ‘anon.2’; see West p. xxiv, W.SA 361, Mund-Dopchie 1984, 209). [Anon.Par.] Anonymous conjectures (probably including those of Auratus and Portus, cited without attribution) in the margins of one or several copies of Victorius* with provenance from Casaubon’s lectures in early 17th-century Paris (= West’s ‘anon.1’; see West pp. xii–xxiv, W.SA 358–64). [Anon.Span.] Anonymous conjectures reported by E. Spanheim from the margin of a copy of Stanley, lost after the second world war (= West’s ‘anon.4’). See MundDopchie (1984) 210 and n. 34, W.SA 359, n. 15 (the conjecture cited here is reported by Hermann). Anti, C. 1947. Teatri greci arcaici. Padua (Monografie di archeologia, 1). Anti, C., and Polacco, L. 1969. Nuove ricerche sui teatri greci arcaici. Padua (Pubblicazioni dell’Istituto di archeologia dell’università di Padova, 4). * British Library 11705 d.2; London National Art Library, Dyce 113.9 Q2; Bodleian Library, MS. Rawlinson G190; ibid., MS. Rawlinson G193; University Library of Cambridge Adv. b.3.3; ibid. Adv. c.25.5; Leiden, Bibliotheek der Rijksuniversiteit 756 D 22; ibid. 756 D 23. 223 Arnott, P. 1962. Greek Scenic Conventions in the Fifth Century B.C. Oxford. Ashby, C. 1988. ‘The Case for the Rectangular/Trapezoidal Orchestra’. Theatre Research International, 13, 1–20. —— 1991. ‘Where Was the Altar?’. Theatre Survey, 32, 3–21. —— 1999. Classical Greek Theatre. Iowa City. Auberger, J. 1996. ‘Quand la nage devint natation...’. Latomus, 55, 48–62. Auffarth, C. 1999. ‘Constructing the Identity of the Polis: The Danaides as “Ancestors”’. In Ancient Greek Hero Cult, ed. R. Hägg, Stockholm (Skrifter utgivna av Svenska Institutet i Athen, 8o, 16 [Proceedings of the Fifth International Seminar on Ancient Greek Cult, Göteborg University, 21–23 April 1995]), 39–48. [Auratus] Conjectures attributed to J. Auratus in the same sources as Anon.Par. (see above and W.SA 358–64). Bachvarova, M. R. 2001. ‘Successful Birth, Unsuccessful Marriage: Aeschylus’ Suppliants and Mesopotamian Birth Incantations’. Nin, 2, 49–90. Bacon, H. H. 1961. Barbarians in Greek Tragedy. New Haven. Bakewell, G. W. 1997. ‘Μετοκία in the Supplices of Aeschylus’. ClAnt 16, 209– 28. Bamberger, F. 1839. ‘Coniectaneorum in Aeschyli Supplices pars prior’. ZAlt 6, 878–85 (rep. in Bamberger 1856, 107–13). —— 1842. ‘Coniectaneorum in Aeschyli Supplices pars altera’. ZAlt 9, 693–712 (rep. in Bamberger 1856, 113–35). —— 1856. Opuscula philologica maximam partem Aeschylea. Ed. F. W. Schneidewin, Leipzig. Barigazzi, A. 1983. ‘Aesch. Suppl. 443 sgg.’. RFIC 111, 292–94. Beazley, J. D. 1963. Attic Red-Figure Vase-Painters. i–ii, 2nd ed., Oxford. Bees, R. 1995. ‘Die Skene in Aischylos’ “Persern”, “Sieben gegen Theben” und “Hiketiden”’. In Pöhlmann (1995), 73–106. Berti, M. 1930. ‘Anacoluti eschilei’. RAL 6, 231–74. Billon, M.-F. 1997. ‘Recherches archéologiques récentes à l’Héraion d’Argos’, and ‘Annexe: Propositions pour une restitution du temple archaïque’. In Héra: Images, espaces, cultes, Naples (Collection du centre Jean Bérard 15 [Actes du Colloque ... Lille, 29–30 novembre 1993]), 11–56, 57–81. Björck, G. 1950. Das Alpha impurum und die tragische Kunstsprache. Uppsala (Skrifter utgivna av K. Humanistiska Vetenskapssamfundet i Uppsala, 39, fasc. 1). Blaydes, F. H. M. 1895. Adversaria in Aeschylum. Halle. —— 1898. Adversaria in varios poetas Graecos et Latinos. Halle. —— 1902. Spicilegium tragicum observationes criticas in tragicos poetas Graecos continens. Halle. Bodensteiner, E. 1893. ‘Szenische Fragen über den Ort des Auftretens und Abgehens von Schauspielern und Chor im griechischen Drama’. JKPh Suppl. 19, 637–808. Bonner, C. 1900. ‘The Danaid Myth’. TAPhA 31, 27–36. —— 1902. ‘A Study of the Danaid Myth’. HSPh 13, 129–173. Booth, N. B. 1974. ‘Westphal’s Transposition in Aeschylus, Supplices 86–95’. CQ 224 24, 207–10. Bremer, J. M. 2000. ‘The Amazons in the Imagination of the Greeks’. AAntHung 40, 51–59. Brennan, C. J. 1893. ‘The Elision of the Dative Iota in Attic Tragedy’. CR 7, 17– 19. British Museum 1897. The Poems of Bacchylides: Facsimile of Papyrus DCCXXXIII in the British Museum. {London.} Bruhn, E. 1899. Anhang. In Sophocles, ed. F. W. Schneidewin and A. Nauck, viii/viii, Berlin. Buck, C. D. and Petersen, W. 1945. A Reverse Index of Greek Nouns and Adjectives. Chicago. Bücheler, F. 1885. ‘Aeschylus und der Parthenon’. RhM 40, 627–29. —— 1886. ‘Coniectanea’. RhM 41, 1–12. {Burges, G.} 1810. ‘Emendationes in Aeschyli Supplices’. CJ {2,} no. 4, 801–8. {——} 1811. ‘Emendationes in Aeschyli Supplices’. CJ {3,} nos. 5–6, 183–93, 414– 22. Burkert, W. 1982. ‘Enzyklopädie des Märchens’. Gnomon, 54, 713–19. —— 1985. Greek Religion. Cambridge, Mass. (revised transl. of id., Griechische Religion der archaischen und klassischen Epoche. Stuttgart, etc., 1977 = Die Religionen der Menschheit, 15). —— 1991. ‘Typen griechischer Mythen auf dem Hintergrund mykenischer und orientalischer Tradition’. In La Transizione dal Miceneo all’alto arcaismo: Dal palazzo alla città, ed. D. Musti a.o., Rome (Atti del Convegno Internazionale Roma, 14–19 marzo 1988), 527–38. Butterweck, G. 1974. ‘Das Dionysostheater’. Maske und Kothurn, 20, 105–47. Buxton, R. G. A. 1982. Persuasion in Greek Tragedy. Cambridge. Cairns, D. L. 1993. Aidos: The Psychology and Ethics of Honour and Shame in Ancient Greek Literature. Oxford. Cannatà Fera, M. 1980. ‘Ἰαονίοισι νόµοισι’. GIF 11, 189–93. [Casaubon] I. Casaubon’s marginalia in a copy of Victorius, University Library of Cambridge Adv. b.3.3 (see W.SA 362–64, Dawe 2001b). Casson, S. 1937. Ancient Cyprus. London. Castagnoli, F. 1972. Lavinium. i/ii, Rome. Catling, H. W. 1964. Cypriot Bronzework in the Mycenaean World. Oxford. Chadwick, J. 1996. Lexicographica Graeca. Oxford. Chantraine, P. 1956. Études sur le vocabulaire grec. Paris (Études et commentaires, 24). —— 1963. Grammaire homérique. ii/ii, 3rd ed., Paris (Collection de philologie classique 4). ——. Dictionnaire étymologique de la langue grecque = DE Clines, D. J. A. (ed.), and Elwolde, J. (exec. ed.) 1995. The Dictionary of Classical Hebrew. ii, ‫ו–ב‬. Sheffield. Cole, T. 1988. Epiploke. Cambridge, Mass. (etc.). 225 Conacher, D. J. 1996. Aeschylus: the Earlier Plays and Related Studies. Toronto (etc.). Conomis, N. C. 1964. ‘The Dochmiacs of Greek Drama’. Hermes, 92, 3–50. Corbato, C. 1995. ‘πᾶν ἄπονον δαιµονίων (Suppl. 99): Considerazioni sulla religiosità di Eschilo’. Paideia, 50, 199–206. [C.Z.] Cook, A. B. Zeus: A Study in Ancient Religion. i–iii, Cambridge 1914–40. Crusius, O. 1889. ‘Ad poetas Latinos exegetica’. RhM 44, 448–60. Dale, A. M. 1950. ‘Stasimon and Hyporcheme’. Eranos, 48, 14–20 (rep. in Dale 1969, 34–40: pages cited within parentheses). —— 1951. ‘The Metrical Units of Greek Lyric Verse. II–III’. CQ 1, 20–30, 119– 29 (rep. in Dale 1969, 61–97: pages cited within parentheses). —— 1968. The Lyric Metres of Greek Drama. 2nd ed., Cambridge. —— 1969. Collected Papers. Cambridge. —— 1971. Metrical Analyses of Tragic Choruses. Fasc. 1/3, Dactylo-Epitrite. London (BICS Suppl. 21, fasc. 1). —— 1983. Metrical Analyses of Tragic Choruses. Fasc. 3/3, Dochmiac–Iambic– Dactylic–Ionic. London (BICS Suppl. 21, fasc. 3). Danielewicz, J. 1996. The Metres of Greek Lyric Poetry. Bochum (Pomoerium Suppl., 1). d’Arnaud, G. 1728. Specimen animadversionum criticarum ad aliquos scriptores Graecos. Harlingen. Dawe, R. D. 1963. ‘Inconsistency of Plot and Character in Aeschylus’. PCPhS 9, 21–62. —— 1964. The Collation and Investigation of Manuscripts of Aeschylus. Cambridge. —— 1965. Repertory of Conjectures on Aeschylus. Leiden. —— 1966. ‘The place of the Hymn to Zeus in Aeschylus’ Agamemnon’. Eranos, 64, 1–21. —— 1972. [Review of Friis Johansen, ed. 1970]. Gnomon, 44, 121–27. —— 1978. Studies on the Text of Sophocles. iii/iii, Women of Trachis, Antigone, Philoctetes, Oedipus at Colonus. Leiden. —— 1999. ‘Strophic Displacement in Greek Tragedy’. Eranos, 97, 24–44. —— 2001a. ‘A. Y. Campbell: Method in Madness or Madness in Method?’. Lexis, 19, 119–30. —— 2001b. ‘Casaubon and Camb. Univ. Adv. B. 3. 3.’. Lexis, 19, 183–84. [DE] Chantraine, P. Dictionnaire étymologique de la langue grecque. i–v, Paris 1968–80. Debrunner, A. 1917. Griechische Wortbildungslehre. Heidelberg (IGmB, 2. Abteilung, Sprachwissenschaftliche Gymnasialbibliothek, 8). Denniston, J. D. 1930. ‘Notes on Greek Particles’. CR 44, 213–15. —— 1952. Greek Prose Style. Oxford. ——. The Greek Particles = D.GP. Dettori, E. 1986–87. ‘Note sulla interlocuzione in Eschilo’. MCr 21/22, 23–33. 226 [DGE] Diccionario Griego – Español. Ed. F. R. Adrados, E. Gangutia a.o., i–(vi), Madrid 1980–(2002). [D.GP] Denniston, J. D. The Greek Particles. 2nd ed., Oxford 1954. Diggle, J. 1974. ‘On the Heracles and Ion of Euripides’. PCPhS 20, 3–36 (rep. in Diggle 1994a, 90–136: the latter pages cited within parentheses). —— 1981a. Studies on the Text of Euripides. Oxford. —— 1981b. ‘On the Alcestis and Andromache of Euripides’. ICS 6, 82–101 (rep. in Diggle 1994a, 196–215: the latter pages are referred to within parentheses). —— 1982. [Review of FJ–W]. CR 32, 127–34. —— 1990. [Review of Cole 1988]. LCM 15.5, 74–79. —— 1994a. Euripidea: Collected Essays. Oxford. —— 1994b. [review of Sicking 1993]. CR 44, 324–25. Dimitrakos, D. 1933–50. (∆ηµήτρακος, ∆.) Μέγα λεξικὸν τῆς Ἑλληνικῆς γλώσσης. i–ix, Athens. Dindorf, W. 1858. ‘Aeschyleische Chorgesänge nach der mediceischen Handschrift nebst berichtigtem Texte’. Philologus, 13, 457–98. —— 1873. Lexicon Aeschyleum. i/ii, Α–Νεῖλος. Leipzig. Diringer, D., and Brock, S. P. 1968. ‘Words and Meanings in Early Hebrew Inscriptions’. In Words and Meanings: Essays presented to D. W. Thomas, ed. P. R. Ackroyd and B. Lindars, Cambridge, 39–45. Dobias-Lalou, C. 2001. ‘Ἀφίκτωρ, Eschyle, Suppliantes 1 et 241’. REG 114, 614–25. Dörig, J. 1987. The Olympia Master and his Collaborators. Leiden (etc.; Monumenta Graeca et Romana, 6). Dörpfeld, W., and Reisch, E. 1896. Das griechische Theater. Athens. Dover, K. 2000. ‘Are Gods Forgivable?’. In Double Standards in the Ancient and Medieval World, ed. K. Pollmann, Göttingen (Göttinger Forum für Altertumswissenschaft, Beihefte 1). Easterling, P. E. 1973. ‘Presentation of Character in Aeschylus’. G&R, 20, 3–19 (rep. in Greek Tragedy, ed. I. McAuslan and P. Walcot, Oxford 1993 [G&R Studies, 2], 12–28). —— 1997. ‘Form and Performance’. In id., ed., The Cambridge Companion to Greek Tragedy, Cambridge, 151–77. Ekroth, G. 2002. The Sacrificial Rituals of Greek Hero-Cults in the Archaic to the Early Hellenistic Periods. Liège (Kernos Suppl. 12; partly revised ver. of diss., Stockholm 1999). Ellis, R. 1893. ‘On Some Fragments of Aeschylus, and on the Supplices’. JPh 21, 25–36. Enger, R. 1854a. [Review of Hermann.] NJPhP 70, 361–410 (pp. 391–410 on the Supplices). —— 1854b. Observationes in locos quosdam Agamemnonis Aeschyleae. Progr., Ostrowo. Erbse, H. 1979. ‘Über die Midiana des Demosthenes’. In id., Ausgewählte Schriften zur klassischen Philologie, Berlin, 412–31. 227 Ercolani, A. 2001. ‘Aesch. Suppl. 249’. Lexis, 19, 159–69. Evelyn-White, H. G. 1915. ‘Hesiodea II’. CQ 9, 72–76. Fairbanks, A. 1900. ‘The Chthonic Gods of Greek Religion’. AJPh 21, 241–59. Flickinger, R. C. 1930. ‘The Theater of Aeschylus’. TAPhA 61, 80–110. Foucart, M. P. 1918. Le culte des Héros chez les grecs. Paris (Mémoires de l’Académie des Inscriptions et Belles-lettres, 42). Friis Johansen, H. 1959. General Reflection in Tragic Rhesis. Copenhagen. —— 1966. ‘Progymnasmata’. C&M 27, 39–64. —— 1968. ‘Codex Scurialensis t.i.15 and the Transmission of Aeschylus’ Suppliants’. GRBS 9, 359–83. Friis Johansen, H., and Whittle, E. W. 1975. ‘Textual Notes on Aeschylus: Supplices 1–503’. SO 50, 5–41. Gantz, T. 1978. ‘Love and Death in the Suppliants of Aischylos’. Phoenix, 32, 279–87. —— 1979. ‘The Aischylean Tetralogy: Prolegomena’. CJ 74, 289–304. —— 1980. ‘The Aischylean Tetralogy: Attested and Conjectured Groups’. AJPh 101, 133–64. —— 1981. ‘Divine Guilt in Aischylos’. CQ 31, 18–32. —— 1993. Early Greek Myth. Baltimore. Garrison, E. P. 1995. Groaning Tears. Leiden (etc.; Mnemosyne Suppl. 147). Garvie, A. F. 1970. ‘The Opening of the Choephori’. BICS 17, 79–91. —— 1973. [Review of Friis Johansen, ed. 1970]. CR 87, 21–22. —— 2001. ‘Aeschylus: When to Emend and When Not to Emend’. Lexis, 19, 1– 13. —— 2002. ‘Alliteration in Aeschylus’. Lexis, 20, 3–12. [G.AS] Garvie, A. F. Aeschylus’ Supplices: Play and Trilogy. Cambridge 1969. Gaskell, P. 1964. A Bibliography of The Foulis Press. London (The Soho Bibliographies, 14). Gebhard, E. 1974. ‘The Form of the Orchestra in the Early Greek Theater’. Hesperia, 43, 428–40. Geel, J. 1830–31. [review of Haupt]. Bibliotheca critica nova, 5, 57–100. [GEW] Frisk, H. Griechisches etymologisches Wörterbuch. i–iii, Heidelberg 1954– 72 (IGmB, 2. Reihe, Wörterbücher). Giangrande, G. 1979. ‘Trois endroits mal compris de Théocrite’. AC 48, 612–18. Gjerstad, E. 1948. The Swedish Cyprus Expedition. iv/iv, fasc. 2/3, The CyproGeometric, Cypro-Archaic and Cypro-Classical Periods. Stockholm. Goebbel, F. C. 1858. De Ephymniorum apud Graecos et Romanos rationibus. Diss., Göttingen. Gödde, S. 2000. Das Drama der Hikesie. Münster (Orbis antiquus, 35). Green, J. R. 1989. ‘Theatre Production: 1971–1986’. Lustrum, 31, 7–95, 273–78. —— 1995. ‘Theatre Production: 1987–1995’. Lustrum, 37, 7–202, 309–18. Griffin, J. 1998. ‘The Social Function of Attic Tragedy’. CQ 48, 39–61. Griffith, M. 1977. The Authenticity of ‘Prometheus Bound’. Cambridge. 228 —— 1986. ‘A New Edition of Aeschylus’ Suppliants’ [review of FJ–W]. Phoenix, 40, 323–40. Gross, A. 1905. Die Stichomythie in der griechischen Tragödie und Komödie. Berlin. Gruys, J. A. 1981. The Early Printed Editions (1518–1664) of Aeschylus. The Hague (Bibliotheca humanistica & reformatorica, 32). Guthrie, W. K. C. 1962. A History of Greek Philosophy. i/vi, Cambridge. Hadjioannou, K. 1975. (Χατζηιωάννου, Κ.) Ἡ ἀρχαία Κύπρος εἰς τὰς Ἑλληνικὰς πηγάς. iii/v, fasc. 1/2, Leukosia. —— 1985. ‘On the Interpretation of the κύπριος χαρακτήρ of Aeschylus’. In Papadopoullos–Hadjistyllis (1985), 509–13. Hadjistephanou, C. E. 1990. ‘Κύπριος χαρακτήρ in Aeschylus “Supplices” 282– 283’. Hermes, 118, 282–91. —— 1991. ‘Aeschylus Supplices 266’. Parnassos, 33, 161–63. Hadjistyllis, S. A. 1985. ‘“Κύπριος χαρακτήρ” (A. Supp. 282)’. In Papadopoullos–Hadjistyllis (1985), 515–20. Haecker, F. 1861. ‘Zur Hiketiden-Parodos’. ZG 15, 215–31. Haigh, A. E. 1907. The Attic Theatre. 3rd ed. (rev. by A. W. Pickard-Cambridge), Oxford. Hall, E. 1989. Inventing the Barbarian. Oxford. —— 1996. ‘When is a Myth Not a Myth? Bernal’s “Ancient Model”’. In Black Athena Revisited, ed. M. Lefkowitz and G. M. Rogers, Chapel Hill, 333–48 (revised version of article in Arethusa, 25, 1992, 181–201; rep. in Greeks and Barbarians, ed. T. Harrison, Edinburgh 2002, 133–52 [page-refs within parentheses]). Halliwell, S. 1997. ‘Between Public and Private’. In Pelling (1997), 121–41. Hammond, N. G. L. 1972. ‘The Conditions of Dramatic Production to the Death of Aeschylus’. GRBS 13, 387–450. —— 1988. ‘More on Conditions of Dramatic Production to the Death of Aeschylus’. GRBS 29, 5–33. {Harberton, J. S. P.} 1903. Notes and Emendations to Aeschylus, Sophocles, and Euripides: By a Graduate of Cambridge. London. Harrison, A. R. W. 1968. The Law of Athens. i/ii, The Family and Property. Oxford. Harrison, J. E. 1887. ‘Itys and Aedon: a Panaitios Cylix’. JHS 8, 439–45. —— 1922. Prolegomena to the Study of Greek Religion. 3rd ed., Cambridge (rep. Princeton 1991). Headlam, W. 1892. ‘Various Conjectures I’. JPh 20, 294–311. —— 1893. ‘Various Conjectures II’. JPh 21, 75–100. —— 1898. ‘Aeschylea’. CR 12, 189–93, 245–49. —— 1904. ‘Notes on Aeschylus’. CR 18, 241–43. Heath, B. 1762. Notae sive lectiones ad tragicorum Graecorum veterum Aeschyli, Sophoclis, Euripidis quae supersunt dramata deperditorumque relliquias. Oxford. Heimsoeth, F. 1861. Die Wiederherstellung der Dramen des Aeschylus. Bonn. 229 Hermann, G. 1820. De Aeschyli Danaidibus. Diss., Leipzig (rep. in Hermann 1827–77, ii. 319–36: pages cited within parentheses). —— 1827–77. Opuscula. i–viii, Leipzig (rep. Hildesheim 1970). —— 1842. [Review of Bamberger, Aeschyli Choephori]. [Wiener] Jahrbücher der Literatur, 100, 162–88 (the greater part, pp. 170–88, is a critique of Bamberger 1839 and 1842, with notes on Supplices). —— 1846–47. ‘Über einige Trilogien des Aeschylus’. BSG 1, 117–27 (rep. in Hermann 1827–77, viii. 173–84: pages cited within parentheses). van Hooff, A. J. L. 1990. From Autothanasia to Suicide. London. —— 1992. ‘Female Suicide’. Laverna, 3, 142–72. Hoppin, J. C. 1901. ‘Argos, Io, and the Prometheus of Aeschylus’. HSPh 12, 335– 45. Hornung, E. 1983. Conceptions of God in Ancient Egypt. London (first published as Der Eine und die Vielen, Darmstadt 1971). Hourmouziades, N. C. 1965 (Χουρµουζιάδης, Ν. Χ.) Production and Imagination in Euripides. Athens (Greek Society for Humanistic Studies, Ser. 2: Essays and Researches, 5). Housman, A. E. 1890. ‘Tucker’s Supplices of Aeschylus’ [review]. CR 4, 105–9 (rep. in The Classical Papers of A. E. Housman, ed. J. Diggle and F. R. D. Goodyear, Cambridge 1972, i/iii, 120–27). Hunter, R. L. 1979. ‘The Comic Chorus in the Fourth Century’. ZPE 36, 23–38. [IA] Italie, G. Index Aeschyleus. 2nd ed. (ed. S. L. Radt), Leiden 1964. James, A. W. 1970. Studies in the Language of Oppian of Cilicia. Amsterdam. Jens, W. 1955. Die Stichomythie in der frühen griechischen Tragödie. Munich (Zetemata, 11). Jones, J. 1962. On Aristotle and Greek Tragedy. London. Jouanna, J. 2002. ‘Le chant male des vierges: Eschyle, Suppliantes, v. 418–437’. REG 115, 783–91. Jurenka, H. 1900. ‘Ad Supplices Aeschyleam adversaria’. WS 22, 181–93. Kapsomenos, A. S. 1983. (Καψώµενος, Α. Σ.) ‘Ἰαµβικοί τρίµετροι χωρίς τοµή στον Αισχύλο’. EEThess 21, 149–61. Kartes, B. 2000. Der Epitaphios des Lysias. Diss., Saarbrücken. [K–B] Kühner, R. Ausführliche Grammatik der griechischen Sprache. 1. Teil, Elementar- und Formenlehre. Besorgt von F. Blass, i–ii, 3rd ed., Hannover (etc.) 1890–92. [K–G] Kühner, R. Ausführliche Grammatik der griechischen Sprache. 2. Teil, Satzlehre. Besorgt von B. Gerth, i–ii, 3rd ed., Hannover (etc.) 1898–1904. Kennedy, G. 1963. The Art of Persuasion in Greece. London (rep. Princeton 1974). Kirchhoff, A. 1887. Studien zur Geschichte des griechischen Alphabets. 4th ed., Gütersloh. Kiefner, G. 1964. Die Versparung. Wiesbaden (Klassisch-Philologische Studien, 25). Kiehl, E. J. 1856. ‘Aeschylos, Smeekelingen, 212’. Mnemosyne 5, p. 153. Korzeniewski, D. 1968. Griechische Metrik. Darmstadt. 230 Kranz, W. 1933. Stasimon. Berlin. Kraus, W. 1957. Strophengestaltung in der griechischen Tragödie. i, Aischylos und Sophokles. Vienna (SAWW 231.4). —— 1984. ‘Aischylos’ Danaidentrilogie’. In id., Aus allem Eines, Heidelberg, 85– 137. Kueck, E. 1890. Studia maxime critica in Aeschylum et scholia Aeschyli Medicea. Diss., Göttingen. Kühn, W. 1905. De vocum sonorumque in strophicis Aeschyli canticis aequabilitate. Diss., Halle. Kugler, L. 1905. De Sophoclis quae vocantur abusionibus. Diss., Göttingen. [LÄ] Lexicon der Ägyptologie. Begründet von W. Heck und E. Otto; Herausg. von W. Heck und W. Westendorf, i-vii, Wiesbaden 1975–92. [Labbé] C. Labbé (Labbaeus) in Bodleian Library, MS. Add. A 176f. 205r (see West p. lxvii, W.SA 360). Laistner, L. 1889. Das Rätsel der Sphinx: Grundzüge einer Mythengeschichte. i–ii, Berlin. Langholf, V. 1977. ‘Unmovierte Partizipien im Griechischen’. Hermes, 105, 290– 307. Lardinois, A. 2000. ‘Characterization through Gnomai in Homer’s Iliad’. Mnemosyne, 53, 641–61. Lefkowitz, M. R. 1981. The Lives of the Greek Poets. London. Lesky, A. 1971. Geschichte der griechischen Literatur. 3rd ed., Bern (etc.). Létoublon, F. 1980. ‘Le vocabulaire de la supplication en grec’. Lingua, 52, 325– 36. Lewis, C. T., and Short, C. 1879. A Latin Dictionary. Oxford (rep. 1955). Liberman, G. 1998. ‘Notes sur les Suppliantes d’Eschyle’. SemRom 1, 243–55. Lindberg, G. 1990. ‘Hera in Homer to Ancient and Modern Eyes’. In Greek and Latin Studies in Memory of Cajus Fabricius, ed. S.-T. Teodorsson, Gothenburg (Studia Graeca et Latina Gothoburgensia, 54), 65–80. Linwood, W. 1847. A Lexicon to Aeschylus. 2nd ed., London. Lloyd-Jones, H. 1956. ‘Zeus in Aeschylus’. JHS 76, 55–67 (rep. with added refs in Lloyd-Jones 1990, 238–61: pages cited within parentheses). —— 1964a. ‘The Supplices of Aeschylus’. AC 33, 356–74 (rep. with added refs in Lloyd-Jones 1990, 262–77: pages cited within parentheses). —— 1964b. [Review of Jones 1962]. Review of English Studies, 15, 221–24 (rep. in Lloyd-Jones 1990, 205–9: pages cited within parentheses). —— 1978. ‘Ten Notes on Aeschylus, Agamemnon’. In Dionysiaca ... Presented to Sir Denys Page on his Seventieth Birthday, ed. R. D. Dawe, J. Diggle, P. E. Easterling, Cambridge, 45–61 (rep. with added refs in Lloyd-Jones 1990, 318–34: pages cited within parentheses). —— 1985. ‘Pindar and the Afterlife’. In Pindare, ed. A. Hurst, Geneva (Fond. Hardt 31), 245–83 (rep. with added refs in Lloyd-Jones 1990, 80–109: pages cited within parentheses). —— 1990. Greek Epic, Lyric and Tragedy. Oxford. 231 Lloyd-Jones, H. 1993. [Review of West and of W.SA]. Gnomon, 65, 1–11. —— 1998. ‘Ritual and Tragedy’. In Ansichten griechischer Rituale: GeburtstagsSymposium für Walter Burkert, ed. F. Graf, Stuttgart (etc.), 271–95. Lorimer, H. L. 1950. Homer and the Monuments. London. [LSJ] Liddell, H. G., Scott, R., Stuart-Jones, H., a.o., A Greek-English Lexicon. 9th ed., with a revised supplement, Oxford 1996. Maas, P. 1958. Textual Criticism. Oxford (originally published in German as Textkritik, 3rd ed., Leipzig 1957). —— 1962. Greek Metre. Transl. (with additional notes) by H. Lloyd-Jones, Oxford. Magini, D. 1996. ‘Mente, intelligenza e realtà esterna: quattro variazioni eschilee su un tema senofaneo’. Acme, 49, 141–44. Manolopoulos, L. 1991. (Μανωλόπουλος, Λ.) Στάσις–Ἐπανάστασις– Νεωτερισµός–Κίνησις. Thessalonike. Marckscheffel, G. 1847. ‘De emendatione fabulae Aeschyliae quae Supplices inscribitur commentatio’. RhM 5, 161–215. Margoliouth, D. S. 1883. Studia scaenica. London. Martin, F. 1858. De Horatii carminum lib. II, 1. et I, 28., epistola ad Frid. Ritschelium. De aliquot locis Aeschyli Supplicum et Sophoclis tragoediarum. Progr., Posen. Masqueray, P. 1892. ‘Les systèmes anapestiques dans la tragédie grecque’. RPh 16, 117–36. McCall, M. 1982. ‘Robortello’s “Conjecture” at Aeschylus, Supplices 337’. CQ 32, 228–30. —— 1985. ‘The Source of the Aldine edition of Aeschylus’ Supplices’. BICS 32, 13–34. Meffert, F. 1861. Quaestiones criticae in Supplices Aeschyleas. Diss., Bratislava. Melchinger, S. 1974. Das Theater der Tragödie. Munich. {Merkel, R.} 1858. ‘Kritisches zu Aeschylus’ Supplices’. RhM 13, 268–83. Merkelbach, R. 1980. ‘Lex Youtie’. ZPE 38, 294. Meyer, E. 1892. Forschungen zur alten Geschichte. Halle (rep. 1966). Mihailov, G. 1955. (Михаилов, Г.) ‘La légende de Térée’. Годишник на Софийския Университет, филолог. фак.: Annuaire del’université de Sofia, fac. des lettres, 50, fasc. 2, 77–197. Molyneux, J. H. 1985. ‘Cyprus and Cypriot Myth in Ancient Greek Lyric Poetry’. In Papadopoullos–Hadjistyllis (1985), 499–507. Monro, D. B. 1891. A Grammar of the Homeric Dialect. 2nd ed., Oxford. Moorhouse, A. C. 1948. ‘On Negativing Greek Participles, where the Leading Verbs are of a Type to Require µή’. CQ 42, 35–40. Morenz, S. 1973. Egyptian Religion. Ithaka, NY (first published as Aegyptische Religion, Stuttgart 1960). Moretti, J.-C. 2000. ‘Le théâtre du sanctuaire de Dionysos Éleuthéreus à Athènes, au Ve s. av. J.-C.’. REG 113, 275–98. Müller, G. 1908. De Aeschyli Supplicum tempore atque indole. Diss., Halle. ——. 1966. ‘Textkritisches zu den Septem des Aischylos’. Hermes, 94, 257–66. 232 Mund-Dopchie, M. 1984. La survie d’Eschyle à la Renaissance. Louvain (Fonds René Draguet, Académie Royale de Belgique, Classe des lettres, 1). [Musgrave] S. Musgrave’s marginalia in British Library C.45.c.21–22 (see W.SA 361, n. 23). Myres, J. L. and Ohnefalsch-Richter, M. 1899. A Catalogue of the Cyprus Museum. Oxford. van Nes, D. 1963. Die maritime Bildersprache des Aischylos. Groningen. Nestle, W. 1930. Die Struktur des Eingangs in der attischen Tragödie. Stuttgart (Tübinger Beiträge zur Altertumswissenschaft, 10). Nilsson, M. P. 1967. Geschichte der griechischen Religion. i/ii, 3rd ed., Munich (HdA 5.2). [NP] Der neue Pauly. i–xii, Altertum. Ed. H. Cancik and H. Schneider, Stuttgart (etc.) 1996–2002. Noack, F. 1912/15. Σκηνὴ τραγική: Eine Studie über die scenischen Anlagen auf der Orchestra des Aischylos und der anderen Tragiker. Diss. Tübingen 1912, publ. ibid. 1915. Nock, A. D. 1944. ‘The Cult of Heroes’. HThR 37, 141–74 (rep. in id., Essays on Religion and the Ancient World, ed. Z. Stewart, Oxford 1972, i. 575–602). [OCD] The Oxford Classical Dictionary. Ed. S. Hornblower and A. Spawforth, 3rd ed., Oxford 1996. Palladini, L. P. 2001. ‘Traces of “Intellectualism” in Aeschylus’. Hermes, 129, 441–58. Palmer, L. R. 1963. The Interpretation of Mycenaean Greek Texts. Oxford. Papadopoullos, T., and Hadjistyllis, S. A. 1985. (Παπαδόπουλλος, Θ., Χατζηστύλλης, Σ. Α.) Ed., Πρακτικὰ τοῦ Β´ ∆ιεθνοῦς Κυπριολογικοῦ Συνεδρίου. i/iii, Leukosia. Parker, L. 1958. ‘Some Observations on the Incidence of Word-End in Anapaestic Paroemiacs and its Application to Textual Criticism’. CQ 8, 82–89. —— 1966. ‘Porson’s Law Extended’. CQ 16, 1–26. Pasquali, G. 1924. ‘Amonre nelle Supplici di Eschilo’. RFIC 2, 246–48. [PECS] The Princeton Encyclopedia of Classical Sites. Ed. R. Stillwell, Princeton 1976. Pelling, C. 1997. Ed., Greek Tragedy and the Historian. Oxford. Peppmüller, R. 1882. ‘Zu Hesiodos’. Philologus, 41, 1–10. Pickard-Cambridge, A. W. 1946. The Theatre of Dionysus in Athens. Oxford. Poe, J. P. 1989. ‘The Altar in the Fifth-Century Theater’. ClAnt 8, 116–39. Pöhlmann, E. 1981. ‘Die Proedrie des Dionysostheaters im 5. Jahrhundert und das Bühnenspiel der Klassik’. MH 38, 129–46. —— 1995. Ed., Studien zur Bühnendichtung und zum Theaterbau der Antike. Frankfurt a.M. (etc.; Studien zur klassischen Philologie, 93). —— 1997. ‘The Tragic Chorus and the Limits of Dramatic Convention in the Fifth Century B.C.’. In Scaenica Saravi-Varsovensia, ed. J. Axer and W. Görler, Warzaw, 1–10 (first published as ‘Der Chor der Tragödie an den Grenzen der Bühnen-konventionen des 5. Jahrhunderts’ in Pöhlmann 1995, 63–71). 233 Polacco, L. 1983. ‘Le Supplici, come le rappresentava Eschilo’. NAC 12, 65–89. —— 1990. Il teatro di Dioniso Eleutereo ad Atene. Rome (Monografie della scuola archeologica di Atene e delle missioni italiane in oriente, 4). [Portus] F. Portus’ marginalia in a copy of Victorius, Leiden University Library 756 D 22 (see W.SA 360). Also conjectures attributed to him in the same sources as those for Auratus and Anon.Par. (see above). Powell, J. E. 1935. ‘Notes on Herodotus’. CQ 29, 72–82. Pulleyn, S. 1997. Prayer in Greek Religion. Oxford. Quirke, S. 1992. Ancient Egyptian Religion. London. [RAC] Reallexikon für Antike und Christentum. i–xiii, ed. T. Klauser a.o., xiv–, ed. E. Dassman a.o., Suppl. i, ed. Klauser, Dassman a.o, Stuttgart 1941–. Radt, S. 1982. ‘δαῦλος / δαυλός’. Hermes, 110, 243–44 (rep. in Noch einmal zu...: Kleine Schriften von Stefan Radt zu seinem 75. Geburtstag, ed. A. Harder a.o., Leiden 2002 [Mnemosyne Suppl. 235], 293–95). Rash, J. N. 1981. Meter and Language in the Lyrics of the Suppliants of Aeschylus. New York. [RE] Paulys Real-Encyclopädie der classischen Altertumswissenschaft. Neue Bearbeitung, i–xxiv, i2–xv2, Suppl. i–xv, ed. G. Wissowa a.o., Stuttgart 1894– 1978. Rehm, R. 1988. ‘The Staging of Suppliant Plays’. GRBS 29, 263–307. Reitzenstein, R. 1897. Geschichte der griechischen Etymologika. Leipzig. Revermann, M. 1999. ‘The Shape of the Athenian Orchestra in the Fifth Century: Forgotten Evidence’. ZPE 128, 25–28. Rode, J. 1971. ‘Das Chorlied’. In Die Bauformen der griechischen Tragödie, ed. W. Jens, Munich (Poetica, Beiheft 6), 85–115. Rösler, W. 1970. Reflexe vorsokratichen Denkens bei Aischylos. Meisenheim (Beiträge zur klassischen Philologie, 37). —— 1992. ‘Danaos a propos des dangers de l’amour (Eschyle, Suppliantes, 991– 1013)’. Pallas, 38, 173–78. —— 1993. ‘Der Schluß der „Hiketiden“ und die Danaiden-Trilogie des Aischylos’. RhM 136, 1–22. Rogers, A. M. 1894. Emendations in Aeschylus. Baltimore. Rohweder, C. 1998. Macht und Gedeihen: Eine politische Interpretation der Hiketiden des Aischylos. Frankfurt am Main (etc.; diss. Freiburg 1991; Europäische Hochschulschriften, 3. Reihe: Geschichte und ihre Hilfswissenschaften, 772). [Roscher] Roscher, W. H., a. o., Ausführliches Lexikon der griechischen und römischen Mythologie. i–vi, Suppl. i–iv, Leipzig 1884–1937. Ruffing, K. 2002. ‘Einige Überlegungen zum Bild der indischen Frau in der antiken Literatur’. In Geschlechter, Frauen, Fremde Ethnien in Antiker Ethnographie: Theorie und Realität. Ed. C. Ulf and R. Rollinger, Innsbruck, 253– 68. Rupprecht, K. 1950. Einführung in die griechische Metrik. 3rd ed., Munich. Rutherford, I. 2001. Pindar’s Paeans. Oxford. 234 Sandin, P. 2001. ‘A Greek Delocutive Noun? Some Notes on ποίφυγµα and its Alleged Cognates’. Glotta, 77 [2003], 110–17. —— 2002. ‘Critical Notes on Aeschylus’. Eranos, 100, 146–60. —— 2002–3. ‘ἀναγής, ἐφύµνιον: Two Lexicographical Notes’. BICS 46, 177–87. Sansone, D. 1975. Aeschylean Metaphors for Intellectual Activity. Wiesbaden (Hermes Einzelschriften, 35). [Scaliger] J. J. Scaliger’s marginalia in a copy of Victorius, Leiden University Library 756 D 21 (see W.SA 359). Scheer, E. 1914. Studien zu den Dramen des Aeschylos. Leipzig. Schlesier, R. 1991–92. ‘Olympian versus Chthonian Religion’. SCI 11, 38–51. Schmid, W. 1934. Geschichte der griechischen Literatur, T. 1, Die klassische Periode der griechischen Literatur. i/v, Die griechische Literatur in der Zeit der attischen Hegemonie vor der Eingreifen der Sophistik. Munich (HdA 7.1.2). Schmidt, M. 1860. ‘Zu den Tragikern’. Philologus, 16, 161–63. —— 1863. ‘Ueber die Parodos der Hiketiden des Aeschylus’. Neues Schwezerisches Museum, 3, 219–34. Schneidewin, T. H. 1836. ‘In Aeschyli Supplicum vv. 145 sqq.’. RhM 4, 230–31. —— 1839. Coniectanea critica. Göttingen. Schoele, A. 1960. ‘µάκαρες θνητοί bei Hesiod’. AAntHung 8, 255–63. Schultze, W. 1932. ‘Zum griechischen Dual’. In Symbolae philologicae O. A. Danielsson octogenario dicatae, ed. A. Nelson, Uppsala, 299–304. Schuursma, J. A. 1932. De poetica vocabulorum abusione apud Aeschylum. Diss., Amsterdam. Schwarz, P. 1897. De ephymniorum apud Aeschylum usu. Diss., Halle. Schweizer-Keller, R. 1972. Von Umgang des Aischylos mit der Sprache: Interpretationen zu seinen Namensdeutungen. Diss. Zürich, Aarau. Scullion, S. 1994a. Three Studies in Athenian Dramaturgy. Stuttgart (Beitr.Alt. 25). —— 1994b. ‘Olympian and Chthonian’. ClAnt 13, 75–119. —— 2000. ‘Heroic and Chthonian Sacrifice : New Evidence from Selinous’. ZPE 132, 163–71. —— 2002a. ‘Tragic Dates’. CQ 52, 81–101. —— 2002b. ‘“Nothing to Do with Dionysus”: Tragedy Misconceived as Ritual’. CQ 52, 102–37. [S.GG] Schwyzer, E. Griechische Grammatik. i–ii/iv (vol. ii ed. A. Debrunner), Munich 1939–50 (HdA 2.1.1–2). Sicherl, M. 1986. ‘Die Tragik der Danaiden’. MH 43, 81–110. Sicking, C. M. J. 1993. Griechische Verslehre. Munich (HdA 2.4). Sideras, A. 1971. Aeschylus Homericus. Göttingen (Hypomnemata, 31). Silverman, D. P. 1991. ‘Divinity and Deities in Ancient Egypt’. In Religion in Ancient Egypt, ed. B. E. Shafer, Ithaca (etc.), 7–87. Slater, W. J. 2002–3. ‘Split-vision: Secondary Action in Greek Tragedy’. GRBS 43, 341–72. 235 Slings, S. R. 1996. [Review of Sicking 1993.] Mnemosyne, 49, 457–61. Smyth, H. W. 1956. Greek Grammar. 2nd ed. (rev. by G. M. Messing), Harvard. Sommerstein, A. H. 1977. ‘Notes on Aeschylus’ Suppliants’. BICS 24, 67–82. —— 1990–93. ‘Aesch. Suppl. 952s. (“Beer no drink for men”)’. MCr 25/28, 59– 64. —— 1995. ‘The Beginning and the End of Aeschylus’ Danaid Trilogy’. In Griechisch-römische Komödie und Tragödie, ed. B. Zimmermann, Stuttgart (Drama, 3), 111–34. —— 1996. Aeschylean Tragedy. Bari (le Rane, 15). —— 1997. ‘The Theatre Audience, the Demos, and the Suppliants of Aeschylus’. In Pelling (1997), 63–79. Srebrny, S. 1950. Critica et exegetica in Aeschylum. Toru (Towarzystwo naukowe w toruniu, prace wydziału filologiczno-filozoficznego, 2.1). Stanford, W. B. 1936. Greek Metaphor. Oxford. Stinton, T. C. W. 1975. ‘More Rare Verse-Forms’. BICS 22, 84–108 (rep. in Stinton 1990, 113–42: pages cited within parentheses). —— 1977. ‘Pause and Period in the Lyrics of Greek Tragedy’. CQ 27, 27–66 (rep. in Stinton 1990, 310–61: pages cited within parentheses) —— 1985. ‘Heracles’ Homecoming and Related Topics: The Second Stasimon of Sophocles’ Trachiniae’. Papers of the Liverpool Latin Seminar, 5, 403–32 (rep. in Stinton 1990, 402–29: pages cited within parentheses). —— 1990. Collected Papers on Greek Tragedy. Oxford. Stoessl, F. 1979. Die Hiketiden des Aischylos als geistesgeschichtliches und theatergeschichtliches Phänomen. Vienna (SAAW 356). Strasburger, H. 1958. ‘Thukydides und die politische Selbstdarstellung der Athener’. Hermes, 86, 17–40 (rep. with further refs in Thucydides, ed. H. Herter, Darmstadt 1968 [Wege der Forschung, 98], 498–530; again in Strasburger, H. Studien zur alten Geschichte, ii/ii, Hildesheim 1982 [Collectanea, 42], 676–708). Sulzberger, M. R. 1945. ‘Notes sur Eschyle’. RBPh 24, 139–44. Taplin, O. 1977. The Stagecraft of Aeschylus. Oxford. [TGL] Estienne, H. Θησαυρὸς τῆς Ἑλληνικῆς γλώσσης: Thesaurus Graecae linguae. i–viii, 3rd ed. (by C. G. Hase, W. Dindorf and L. Dindorf), Paris 1831– 65. Thompson, D’A. W. 1936. A Glossary of Greek Birds. 2nd ed., London. Thomsen, O. 1995. ‘Written Sources’. In Ancient Akamas, i/ii, ed. J. Jejfer, Aarhus, 26–40. Thomson, G. 1971. ‘The Suppliants of Aeschylus’. Eirene, 9, 25–30. —— 1973. Aeschylus and Athens. 4th ed., London (1st ed. 1941). Threatte, L. 1980–96. The Grammar of Attic Inscriptions. i–ii, Berlin. Thumb, A. and Kieckers, E. 1932. Handbuch der griechischen Dialekte. i/ii, 2nd ed., Heidelberg (IGmB, 1. Abtl., Sammlung indogermanischer Lehr- und Handbücher, 1. Reihe, Grammatiken; Bd.8). [TLG CANON] Berkowitz, L., and Squitier, K. A. Thesaurus Linguae Graecae Canon of Greek Authors and Works. 3rd ed., New York (etc.) 1990. 236 Todt, B. 1889. ‘Zur Erklärung und Kritik von Aeschylos’ Schutzflehenden’. Philologus, 48, 20–56. Travlos, J. 1971. Pictorial Dictionary of Ancient Athens. New York (etc.). Tronskij, I. M. 1957. (Тронский, И. М.) ‘Оксиринхская дидаскалия к тетралогии Эсхила о Данаидах’ (‘Didascalie oxyrhincienne à la tétralogie des Danaïdes d’ Eschyle’). VDI 60, 146–59. Turner, C. 2001. ‘Perverted Supplication and Other Inversions in Aeschylus’ Danaid trilogy’. CJ 97, 27–50. Turyn, A. 1943. The Manuscript Tradition of the Tragedies of Aeschylus. New York (Polish Institute Series, 2). Verdenius, W. J. 1985. ‘Notes on the Parodos of Aeschylus’ Suppliants’. Mnemosyne, 38, 281–306. —— 1990. ‘Danaus and his Daughters: Notes on Aeschylus’ Suppliants vv. 176– 233’. Mnemosyne, 43, 426–32. Wace, A. J. B. and Stubbings, F. H. 1962. A Companion to Homer. London. Wackernagel, J. 1916. Sprachliche Untersuchungen zu Homer. Göttingen (Forschungen zur griechischen und lateinischen Grammatik, 4). Wærn, I. 1951. Γῆς ὄστεα: The Kenning in Pre-Christian Greek Poetry. Uppsala. Wartelle, A. 1978. Bibliographie historique et critique d’Éschyle … 1518–1974. Paris. Watkins, C. 1989. ‘New Parameters in Historical Linguistics, Philology, and Culture History’. Language, 65, 783–99. —— 1995. How to Kill a Dragon: Aspects of Indo-European Poetics. New York (etc.). Wecklein, N. 1872. Studien zu Aeschylus. Berlin. —— 1876. ‘Die Aeschyleische literatur von 1859–1871. II’. Philologus, 34, 296– 369 (pp. 330–39 on Supplices). —— 1893. Aeschyli fabulae: Voluminis II auctarium: Appendix propagata. Berlin. Welcker, F. G. 1824. Die Aeschylische Trilogie Prometheus. Darmstadt. West, M. L. 1973. Textual Criticism and Editorial Technique. Stuttgart. —— 1979. ‘The Prometheus Trilogy’. JHS 99, 130–48. —— 1982. Greek Metre. Oxford. —— 1985. The Hesiodic Catalogue of Women. Oxford. —— 1990. ‘Colloquialism and Naïve Style in Aeschylus’. In Owls to Athens: Essays on Classical Subjects Presented to Sir Kenneth Dover, ed. E. M. Craik, Oxford, 3– 12. —— 1994. [Review of Sicking 1993.] GGA, 246, 183–97. —— 1997. The East Face of Helicon. Oxford. ——. Studies in Aeschylus = W.SA Westphal, R. 1869. Prolegomena zu Aeschylus Tragödien. Leipzig. Whittle, E. W. 1961. ‘Aeschylus, Supplices 249’. CQ 11, p. 9. —— 1963. ‘Notes on the text of Aeschylus, Supplices 58–67’. In Γέρας: Studies Presented to George Thomson on the Occasion of his 60th Birthday, ed. L. Varcl and R. F. Willetts, Prague (Graecolatina Pragensia, 2), 245–55. 237 Whittle, E. W. 1964a. ‘Two notes on Aeschylus, Supplices’. CQ 14, 24–31. —— 1964b. ‘An Ambiguity in Aeschylus, Supplices 315’. C&M 25, 1–7. —— 1968. ‘Textual Notes on Aeschylus’. C&M 29, 1–15. Wiel, J. 1858. Observationes in locos aliquot Aeschyleos. Diss., Bonn. Wifstrand, A. 1965. Grekisk metrik. 2nd ed., Lund. Wilamowitz-Moellendorff, T. von 1917. Die dramatische Technik des Sophokles. Berlin (Philologische Untersuchungen, 22). Wilamowitz-Moellendorff, U. von 1886. ‘Die Bühne des Aischylos’. Hermes, 21, 597–622 (rep in Wilamowitz 1935–72, i. 148–72). —— 1914. Aischylos: Interpretationen. Berlin. —— 1927. ‘Lesefrüchte’. Hermes, 62, 276–98 (rep. in Wilamowitz 1935–72, iv. 431–53). —— 1931. Der Glaube der Hellenen. i/ii, Berlin. —— 1935–72. Kleine Schriften. i–vii, Berlin. Wiles, D. 1997. Tragedy in Athens. Cambridge. Willink, C. W. 1997. ‘Sophocles, Electra 137–9’. CQ 47, 299–301. —— 2002. ‘The Invocations of Epaphus in Aeschylus, Supplices 40–57 and Euripides, Phoenissae 676–89’. Mnemosyne, 55, 711–19. van Windekens, A. J. 1986. Dictionnaire étymologique complémentaire de la langue grecque. Louvain. Winnington-Ingram, R. P. 1961. ‘The Danaid Trilogy of Aeschylus’. JHS 61, 141–52 (rep. with added refs in id., Studies in Aeschylus, Cambridge 1983, 55– 72). —— 1969. ‘Pindar’s Ninth Pythian Ode’. BICS 16, 9–15. Wordsworth, J. 1832. [Review of Scholefield.] The Philological Museum, 1, 209– 44. [W.SA] West, M. L. Studies in Aeschylus. Stuttgart 1990 (Beitr.Alt. 1). Young, D. C. C. 1974. ‘Conjectures and Interpretations in Aeschylus’ Supplices’. In Serta Turyniana: Studies ... in Honor of A. Turyn, ed. J. L. Heller, Urbana, 216–26. Zakas, A. J. 1890. (Ζάκας, Α. Ι.) Κριτικαὶ καὶ ἑρµηνευτικαὶ παρατηρήσεις εἰς Αἰσχύλον, Σοφοκλέα, Λυσίαν, Πλάτωνα, Λυκοῦργον και ∆ηµοσθένην. i, Αἰσχύλος. Athens. Zeitlin, F. I. 1990. ‘Patterns of Gender in Aeschylean Drama’. In Cabinet of the Muses: Essays ... in Honor of T. G. Rosenmeyer, ed. M. Griffith and D. J. Mastronarde, Atlanta, 103–15. Ziegler, K. 1930. De precationum apud Graecos formis quaestiones selectae. Diss., Breslau. 238 Indices The numbers refer to pages and footnotes (‘n.’). I. Index locorum * indicates that a new conjecture is proposed. A. Ag. 17: 151 n. 419 22–23: 97 105: 71 106–7: 75 124–25: 75 124: 71 129: 75 131: 118 160 ff.: 92 221: 100–101 314: 41 n. 107 342: 150 364–66: 176–77 452–54: 51 509: 155 n. 435 662–63: 78 n. 209 680: 80 699–700: 180 830: 80 848–49: 151 855–974: 129 n. 358 929–72: 210 1176: 74 (n. 199) 1234–36: 67 n. 186 1245–55: 169 (n. 463) 1292–94: 177 1343–71: 36 n. 88; 57 1386–87: 56 n. 148 1447: 141 n. 390 1448–1566: 117 1455–61: 103 n. 280 1459: 87 1478: 74 (n. 199) 1501: 197 1508: 197 1526: 83 (n. 221) 1538–50: 103 n. 280 1625: 84 n. 224 Ch. 56: 191 114: 46; 174 150: 87 n. 231 168: 136 306–509: 60 324 ff.: 55 n. 146 345–79: 55 n. 146 354–62: 55 n. 146 369: 74 (n. 199) 372: 172 394: 39 n. 98 400 ff.: 55 n. 146 446: 84 458: 46 476: 55 n. 146 483–85: 55 n. 146 489–96: 128 489: 55 n. 146 539: 151 n. 419 579–80: 134 689: 133 716–18: 133 n. 365 750–60: 57 770–73: 128 n. 355 783–806: 60 908–30: 169 (n. 463) 239 A. Ch. 936: 50 1016: 196 1065–76: 200 Eu. 43: 72 n. 195 82–83: 178 101: 151 160: 191 226: 147 258: 186 259: 186 289–91: 4 311: 46 321–96: 117 328–33 = 341–46: 111 (n. 305); 197 419: 142 n. 392 424: 88 427: 168 n. 463 431 ff.: 168 n. 463 468: 147 490: 84 n. 224 508–14: 78 n. 209 545–49: 78 n. 209 546: 50 558–59: 152 589: 172 601 ff.: 168 n. 463 645–46: 151 n. 419 650–51: 98 669–773: 4 688: 152 762–74: 4 769–71: 176 838(= 871): 189 920: 124 1042(~1046): 91 Pers. 1: 172 28: 57 n. 151 140–43: 14 151–52: 69 253: 159 301: 97 343: 155 n. 435 240 354: 197 424: 150 455: 46 513: 172 (n. 470) 526: 195 544: 141 n. 390; 88 n. 236 691: 55 n. 146 745: 190 782: 189 807–8: 192 852–906: 68 n. 188 880–95: 156 891: 155 n. 435 Pr. 137–40: 70 312–13: 99 468: 109 808–9: 164 n. 456 824: 80 Supp. 543(~552): 91 569–70: 66 576–81: 72 578 ff.: 66 580: 72 581: 72 585: 94 593: 189 597: 99 622: 178 n. 483 630–709: 185 n. 500 652: 195 n. 515* 658: 50 663: 88 n. 236; 89 n. 238 699: 191 710–824: 120 725: 186 731: 186 732: 132 773: 178 n. 483 807: 151 n. 419 825–910: 12 844(~855): 92, n. 246 905–10: 93 n. 250 911–53: 12–13 997–98: 88 n. 236 1003: 141 n. 390 1012: 132 1026–27: 83 Th. 87–180: 131 n. 361 108–81: 60 177: 191 181–286: 125 n. 343 245–63: 57; 168–69 n. 463; 208 265–81: 130 317: 46 326–29: 198 417–630: 60 420: 74 (n. 199) 451: 132 472: 131 n. 360 480: 132 501: 82 n. 221 562: 132 624: 126 626–27: 177 735: 50 n. 136; 85 n. 226 758–61: 205 831: 45 871–72: 203 951–52: 87 n. 231 fr. 44 (Dan.): 9 fr. 47a (Dikt. 773–79): 204 fr. 47a 15 (Dikt. 779): 151 n. 419 fr. 47a 32 (Dikt. 830): 89 n. 238 fr. 55 (Epig.): 53 fr. 78a.1–22 (Theor.): 109 n. 300 fr. 78c.49–60 (Theor.): 109 n. 300 fr. 85 (Thress.): 118 fr. 131 (Myrm.): 172 fr. 204b 6–8 = 15–17 (Pr.pyr.): 111 (n. 305) frr. 216–20 (Salam.): 156 n. 436 fr. 281a 17–19: 140 fr. 333: 129 fr. 379: 18–19 fr. 399: 189 fr. 488: 50 n. 136 Ach.Tat. 2.38.4: 171 n. 468 A.D. Adv. p. 175 Schneider: 95 Constr. p. 170 Uhlig: 195 n. 514 Alc. 58.17: 111 n. 308 307.1c: 136 Amphis fr. 13 PCG: 111 n. 307 Anacr. 31: 196 Anaximen.Lampsac. Rh. 15.3: 42 An.Ox. II. 443: 115 Antiphon Caed.Her. 10: 77 Nov. 6: 127 AP 5.259.5–7: 171 6.172: 47 7.480.7: 115 14.123.14: 116 Apollon. Lex. p. 45.7 Bekker (s.v. ἀστεµφές): 189 n. 505* Ar. Ach. 104: 85 n. 228 106: 85 n. 228 569–71: 78–79 665–75~692–701: 197 Av. 257: 156; 199 Eccl. 1008: 183 n. 492 Eq. 40: 131 n. 360 Lys. 97: 131 n. 360 980: 189 1081: 95 1242–45: 130 Nu. 372: 153–54 821: 189 Pax 131–32: 154 412–13: 178 731: 14 Ra. 911–20: 145 241 Ar. Ra. 1261–80: 103 Vesp. 172: 130 fr. 62 PCG: 145 n. 400 fr. 936 PCG: 81 (n. 215) Ar.Byz. Nom.aet. p. 276 Miller: 163 n. 454 Arist. EN 1175a29: 76 n. 204 GA 716a22: 72 n. 195 718a10: 89 n. 238 HA 518a18: 89 n. 238 Po. 1452b: 37 n. 90; 59 n. 159; 184 n. 493 Rh. 1394a–95b: 121 n. 332 [Arist.] Rh.Al. 1431b30: 42 B. 19.11: 80 Biblia sacra Is. 49.10: 114 n. 316 Je. 36.23 (Hebr): 122 n. 333 Jo. 23.11 (Greek): 134 n. 370 Jn. 4.8: 114 n. 316 Ps. 121.6: 114 n. 316 Si. 17.27–28 (Greek): 106 n. 292 Call. Epigr. 5.6: 143 Carm. conv. 894: 52 (n. 139) Com.adesp. 563 Kock: 162 n. 452 Com.adesp. 1105.180 PCG: 111 n. 307 Com.pall.inc. 49 Ribbeck: 162 n. 452 Corinn. fr. 1 iii 47: 108 Cratin. fr. 72 PCG: 145 n. 400 D. 1.13: 123 (n. 336) 21.133: 161 n. 451 28.2: 77 n. 206 29.22: 77 n. 206 30.25: 77 242 Danais fr. 1, PEG p. 122: 6; 8 DGEE 12.4: 96 12.14: 96 167a: 96 179 VI 14: 96 179 VI 37: 96 D.H. 1.64.5: 54 n. 143 [Diogenian.] 7.9: 162 n. 452 D.S. 11.71.4–6: 4 11.74.3–6: 4 E. Alc. 725: 183 n. 492 773: 111 n. 307 818–19: 156 n. 438 960: 46 1118: 42 n. 108 1159: 99 n. 268 Alex. IV 3 (fr. 43.38 Snell; TrGFS p. 84): 200 n. 524 Andr. 121: 151 n. 419 355–58: 133 n. 365 639: 46 712: 133 733–34: 78 n. 209 772: 69 (n. 188) 814: 196 1284: 99 n. 268 Ba. 46: 152 586: 118 n. 328 596: 118 n. 328 894: 99 n. 268 Cyc. 157: 118 n. 328 616–19: 177 El. 558–69: 158 Hec. 214: 196 600: 183 n. 492 1080–81: 109 1178–79: 78 n. 209 Hel. 1481: 69 (n. 188) Heracl. 1016: 183 n. 492 HF 131: 204 802–4: 76 n. 205 1207–10: 133 n. 365 1341–46: 120 1381: 43 n. 111 Hipp. 328: 117 358: 133 697: 146 738(~748): 92 n. 246 IA 1–48: 37 55–57: 191–92 Ion 138: 175 255–368: 168 n. 463 419: 203 1250–51: 133 n. 365 IT 410: 109 1243(~1268): 92 n. 246 Med. 61: 133 643: 92 829(~840): 92 n. 246 996: 196 Or. 640: 131 n. 360 692: 148 n. 407 1218–19: 78 n. 209 Ph. 187–90: 7 352: 99 n. 268 797~815: 92 n. 246 931–35: 151 (n. 417) Rh. 1–51: 59 n. 159 Supp. 628–29: 80 Tr. 308–40: 185 n. 500 768: 197 1064: 156; 199 fr. 21.5 Nauck (Aeol.): 42 n. 108 fr. 114 Nauck (Andromeda): 37; 44 n. 117 fr. 152 Nauck (Andromeda): 99 n. 268 fr. 291 Nauck (Bell.): 189 fr. 14.17 Pap.poet. = TrGFS p. 121 (Melanipp.sap.): 152 Et.Gud. s.v. Ζαγρεύς: 9 n. 28; 115 Eub. fr. 104 PCG: 86 n. 229 Eup. fr. 1.2–3 PCG: 186 Eus. Chron.Pasc. 162A: 3 Eust. ii. 129 Stallbaum: 39 n. 96 Eust.Macr. Hysm. 1.14.3: 56 (n. 147) Gp. 10.64.4: 158 Gr.Naz. Or. 7.16.3: 180 Hdn. Pros.cathol. i. 489 Lentz: 95 n. 257 Hdt. 1.27: 154 1.32: 120 n. 330 1.57: 147 2.4: 129 n. 359 2.38: 67 n. 183 2.50: 136 n. 376 2.51: 137 2.59: 129 n. 359 2.67: 137 2.153: 67 n. 183 2.156: 129; 137 3.12: 207 3.27–28: 67 n. 183 3.28: 157 3.98–99: 160 3.102: 160 4.190: 134 5.12–15: 149 5.71: 176 6.61: 119 7.103: 183 n. 492 7.185: 149 243 Hebr.Inscr. 1.004.3 Davies: 122 n. 333 Hecat. fr. 20: 7 n. 22 fr. 305: 138 Heniochus fr. 5 PCG: 14, n. 38 Heraclit. fr. 58: 152 n. 423 Herod. 2.70: 108 n. 298 Hes. Op. 141: 52 (n. 139) 166–73: 52 (n. 139) 191–92: 80 n. 215* 483–84: 94 797–99: 134 n. 371 Sc. 205: 124 Th. 140: 58 159 ff.: 119 845–46: 58 866: 40 fr. 128: 7 fr. 294 (Aegimus): 6, n. 19 Hom., hHom. Il. 1.542: 189 2.681: 148 2.749–50: 148 2.840–43: 147 4.468: 126 5.115: 185 5.299: 186 5.448: 46 6.289–91: 106 7.205: 46 7.298: 124 8.435: 111–112 n. 308 10.240: 72 10.244: 185 13.261: 111–112 n. 308 13.464: 180 18.376: 124 18.567: 189 244 19.77: 99 21.324: 124 (n. 340) 22.294–95: 122 n. 334 22.450–51: 122 n. 334 23.743: 106 n. 291 Od. 1.22–24: 164 1.344: 142 4.42: 111 n. 308 4.85–89: 174 4.567 : 68 (n. 186) 4.615–19: 106 n. 291 4.726: 142 4.816: 142 8.509: 124 10.517: 116 11.25–95: 116 13.54–56: 99 15.58: 126–27 15.80: 142 15.115–19: 106 n. 291 15.425: 106 n. 291 19.518–20: 83 19.522–23: 81 21.85: 189 21.420: 99 22.21: 111 n. 308 h.Ven. 122: 123 n. 340 h.Hom. 7.1–2: 145 Hor. Epist. 1.14.43: 162 n. 452 Hsch. α 4222: 108 n. 298 α 4227: 109 (n. 298) ε 2727: 151 ε 2728: 151 τ 1462: 199 n. 522 Hyg. Fab. 45: 77 n. 207 Iamb.adesp. 12 Diehl: 162 n. 452 Ibyc. 291: 52 (n. 139) IG i3 6 B 5: 140 i3 14.30–31 (= Meiggs–Lewis 40.30): 42 n. 109 iii 1.77: 39 n. 96 v 1.700: 39 n. 98 2.22.1: 66 Ion Trag. 41b: 198 Phryn. Eclog. 135: 17 n. 51 Isid. Chron. 174 Mommsen: 2 n. 5 Phryn.Trag. fr. 1 (Aeg.): 8; 175 frr. 8–12 (Phoen.): 15 fr. 8 (Phoen.): 37 Isoc. Antid. 223: 127 Paneg. 43: 152 n. 424 134: 156 n. 437 Trapez. 19.3 77 Luc. Lex. 2: 161 Lys. 2.39: 106 24.11–12: 162 Macar. 7.75: 162 Machon fr. 17.389: 162 17.399: 162 Marm.Par. A 56: 2, n. 5 Melanipp. fr. 1: 8, n. 26 Men. fr. 806 PCG: 126 n. 346 MenRom i. 401: 189 Moschio FGrH no. 575 (iii B p. 675): 201 Myc. 60 Ventris–Chadwick: 108 n. 296 176: 108 n. 296 257: 108 n. 296 Ps.Nonn. Schol.myth. 4.7: 11 n. 30 [Ocell.] 4.13: 134 n. 370 Ph. Det.pot. 41: 171 n. 468 Pherecyd. fr. 13 FHG: 8 Pi. I. 2.7: 123 (n. 336) N. 1.46–47: 70, n. 192 3.9: 80 4.96: 191 7.45–47: 52 (n. 139) 10.6: 191 O. 1.8–9: 80 2.10: 70 5.17: 56 6.3: 112 6.70: 152; 203 7.13–14: 69 n. 189* 7.39: 152 P. 3.50–57: 152 n. 423 4.159: 51 n. 138; 52 (n. 139) 5.93–101: 52 (n. 139) 6.14: 112 9.38–39: 125 n. 344 9.112–22: 7–8 fr. 30: 56 fr. 93: 175 fr. 108b: 97 fr. 180: 125 Pl. Cra. 403c: 75 Epin. 982c: 134 Orac.Sib. 2.304: 199 Grg. 482b: 137–38 n. 379 Ov. Her. 14.74: 100 Lg. 657a–b: 137 n. 379 717a–b: 51 n. 139 783c: 133–34 Paus. 2.19: 10; 11 n. 31; 111 245 Pl. Lg. 854b: 151 (n. 417) 926d: 143 930e: 72 n. 195 POxy 2256.2: 2, n. 6; 3 n. 7 POxy 2256.3: 1–2; 3 n. 7 Phd. 87c: 79 n. 210 Pratin.Lyr. fr. 708 = Pratin.Trag. fr. 3: 17 n. 51 Phdr. 244d–e: 150–51 n. 417 274c–275b: 137 274d-275d: 40 n. 105 Quint. Inst. 5.11.21: 162 n. 452 Phlb. 18b: 137 Prt. 329c: 125 n. 342 R. 392a: 52 (n. 139) Smp. 182a: 120 189b: 154 n. 430 197d: 141 n. 390 197e: 143 200a: 133 Thg. 122d: 175 Ti. 90a: 134 21e: 137–38 n. 379 Pl.Com. fr. 38 PCG: 163 n. 453 Plaut. Mil. 1177: 109 Plu. Cim. 8.8: 2, n.4, n. 5 Def.orac. 417e–f: 136 Dem. 1.1: 170 E.Delph. 388f–389a: 116 Exil. 607c: 116 Frat.am. 488b: 134 n. 368 Gen.Socr. 577f: 157 Pomp. 48.7: 184 n. 493 Pr.frig. 953a: 115 n. 317 PMag. 4.1847–48: 158 7.466: 158 Poll. 3.60: 188 Porph. Antr. 6: 116 Phil.orac. 114: 116 118: 116 121: 116* 246 S. Aj. 134–35: 67 n. 186 202: 54 n. 143 220: 57 n. 151 431: 143 535: 134–35 n. 371 854–65: 131 n. 361 1077: 72 1363: 90 n. 241 1416: 57 n. 151 Ant. 161: 57 n. 151 277: 159 303: 177 n. 480 495: 183 n. 492 864: 42–43 1146–47: 67 n. 186 1201–2: 180 El. 444–46: 75 n. 202 OC 148: 133 188: 44 n. 117 322–23: 69–70 399: 177–78 551 ff.: 153 858–59: 174 n. 473 1055(~1070): 100 1267–68: 124–25 1436: 42 n. 108 OT 105: 80 342: 75 474: 76 n. 204 656: 107–8, n. 295 848: 76 n. 204 1090: 69 (n. 188) 1199–1202: 75 n. 202 Ph. 542–627: 109 559: 75 1037–38: 123 n. 339 Tr. 1: 76 n. 204 64 ff.: 168 n. 463 188: 86 475: 42 n. 108 739–40: 69 1044–45: 100 n. 270 1237: 133 1239–40: 193 n. 512 1263: 57 n. 151, 59 fr. 271 (Inach.): 148 n. 409 fr. 314.331 (Ichn.): 86–87, n. 230 fr. 859: 199 n. 523 Sapph. 60.8: 141–42 n. 390* [Sen.] Herc.Oet. 199: 81 (n. 215) Severus i. 537 Walz: 66 Sim. 558: 52 (n. 139) Str. 11.12.4.26, 28: 116 n. 321 Stratt. fr. 220.180 Austin: 111n. 307 Suda s.v. θυµέλη: 17 n. 51 s.v. Σοφοκλῆς: 36 n. 88 SVF ii, fr. 749: 159 n. 446 Σ A. Th. 130b: 11 n. 30 Σ A. Th. 230d: 203 n. 527 Σ A. Supp. 37: 11 n. 30 Σ rec. A. Pers. 353: 11 n. 30 Σ Pi. O. 2.104b: 51 n. 138 Σ Pi. P. 5.10b: 161 n. 450 Terp. fr. 697: 145 n. 400 Th. 1.35.4: 60 n. 161 1.73–76: 205 n. 531 1.104: 4 1.126.12: 41 n. 106 2.35–46: 205 n. 531 4.100.1: 75 n. 202 Theoc. 25.6: 125 n. 345 26.8: 180 Thgn. 1.432–34: 152 n. 423 1.690: 195 1.731–52: 120 n. 330 Tim. Lex. s.v. ἀστόξενος: 188 n. 503 Timostr. fr. 4 PCG: 116 n. 321 Trag. adesp. 628: 74 (n. 199) Tz. Ep. p. 81.2 Leone: 189 H. 9.847: 163–64 n. 454 Verg. Aen. 2.515–17: 139 7.312: 117 X. An. 3.1.18: 178 3.1.27–29: 183 n. 492 Cyr. 1.6.11: 78 n. 209 7.5.52: 183 n. 492 8.6.20: 123 n. 339 HG 6.5.39: 183 n. 492 Lac. 1.3: 71–72 Smp. 3.10.2: 111 n. 307 Xenoph. fr. 11 D–K: 119 frr. 23–24 D–K: 99 frr. 25–26 D–K: 98 [Zonar.] s.v. ζαγρός: 116 n. 322 247 II. Index coniecturarum Aeschylearum This index lists conjectures on Aeschylus (including my own, cited without attribution) which are not mentioned in Wecklein (ed. 1885, Auctarium 1893), Dawe (1965), West, or W.SA; or which are incorrectly quoted or attributed in these sources. I owe to Professor Martin West my knowledge of the conjectures attributed to Auratus and Portus (see the Preface, p. vi). Nol(ens) = rejected by the author, or suggested by misprint; dub(itanter) = suggested without much confidence; del(evit) = the passage is deleted as an interpolation; [123] = mentioned on p. 123 in my commentary. Supplices 10 <καὶ ἄτιµον> [44]: ’πονοταζόµεναι (nol.) Whittle ap. FJ–W [45]; 25 χθονίους Portus [54]: χθονίας Auratus [54, n. 144]; 42 Ἀνθονόµου (nol.) [66]; 44–45 ᾇ ’ξ ἐπιπνοίας ... (ἐπωνυµίαν {δ’}) Willink (2002, 714) [70, n. 190]; 53 τε τὰ vel θ’ ἅτε (nol.) [74, n. 200]: τά γε Richard Janko [74]: (τά γε ...) γαιονόµοις τάδ’ ἄελπτά C. W. Willink [75]; 62 Ἀηδονᾶς (nol.) [82]; 82 ἐνδίκοις γάµος leg. Σ [90]: <οὐ> πέλοιτ’ ἂν ἐκδίκοις γάµος Weil [90, n. 241]; 89 κελαινῷ (nol.?) Schmidt (1863, 229) [97, n. 265]; 101 ἥµεν’ ὃν φρόνηµα (dub.) [99, n. 269]; 111 (ἄτᾳ δ’) ἀπατῶν ἀνάγνους [101]: (χὡ δι’ ἄνοιαν µαινόλιν ...) ἄταις ἕπεταί µ’ ἀνάγνοις Burges (1810, 803) [101, n. 275]; 135 δόρει vel δορί [110, n. 302]; 147 ἅπαντι Westphal (1869, 160) [110, n. 304]; 148 διωγµοὺς διεφθορυῖ’ (dub.) [113]; 172 γόνον Portus [120]; 187 τῷδ’ ... στόλῳ [124 (n. 340)]; 194 τὰ χρήστ’ [126]: κοὐκ ἀχρεῖ’ Portus [125]; 207–9 del. (nol.) [132, n. 362]; 210 del. (nol.) [ibid.]; 242 πρὸς θεοῖς <τ’> (dub.) [143]; 248 (ἱρό)ρρ(αβδον) Tucker [145, n. 401]: (ἱερό)ρρ(αβδον) dub. Headlam (ad loc., n. 6) [ibid.]; 249 πρὸς ταῦτ<α δείξω µὲν τὰ χρὴ τεκµήρια· | σὺ δ’ αὖτ’> ἀµείβου καὶ λέγ’ εὐθαρσὴς ἐµοί [146]: coryphaeo trib. Ercolani (2001) µὲν pro γὰρ (post Abbott) legens [146, n. 402]; 259 ὑγρὸς Chadwick (1996, 297) [149]; 266 µηνιωµένη (nol.) [151]: µηνίοντ’ ἄχη Hadjistephanou (1991) [151, n. 418]: (µηνίσασ’) ἄγει Headlam (ap. Blaydes 1898) [ibid.]; 281 καὶ πρὸς χαρακτὴρ {τ’} Hadjistephanou (1990) [159 (n. 442)]; 296 ταῦτα <µὲν> (παλαισµάτων) [172]; 324 ἀνστήσαις [178]; ἀνστήσῃς Robortello [176, n. 477]: ἀνστήσεις Paley (ed. 1844) [176]; 332 µεταπτερωθὲν [179]: µεταπτερῶσαν (dub.) [179, n. 486]; lacuna post 335 [181]; 337 οἴοιτο Portus [182]; 355 µένονθ’ [187]: ἔνονθ’ Staffan Fogelmark [ibid.]; 361 µάθε <γε> γεραφρονῶν [189]; 363 (οὐ λιπερ<νής ποτ’ ἔσῃ·) πρόσεισίν γε µάλ’> ἱ̄εροδόκα θῶν [190]; 405–6 (ἔτ’) ἀλγεῖς (dub.) [196]; 435 τρὶς τίνειν [199]; 441 κατηγµένον Liberman (1998) [201]; 444 ἅ τις φέρει µέγιστον ἐµπλήσας γόµον [202]; post 444 lacunam cj. Dawe 1972 [ibid.]; post 448 lacuna [ibid.]; 467 φεῦ· πολλαχῇ γε [204]; 481 γε Auratus [206]; 504 κἂν (dub.) [207]; 514 (λύειν){δ’} [209]; 652 ὁµαίµων [195, n. 515] Choephoroi 150 κωκυτοὺς Paley (ed. Supp. 1844, p. 16) [87, n. 231] Eumenides 490 καταστροφαὶ δ’ ἐµῶν (Sandin 2002, 155) [84, n. 224] 248 III. Index verborum α doricum 190–91 ἆ 118 ἄγαλµα 124 ἄγη 118 ἄγος 108 ἀγώνιος 124 ἄκων 59; 140 ἀλκή 186 Ἀµαζών 165–66 ἀµφί + acc. 145 ἀναγής 108–9 n. 298 ἀνέδην 47 ἀνθεµίζω 86–87 ἀνθίζω 86–87 ἀνθονόµος 66–67 ἀνίστηµι 176 ἀσεβής 44 ἀστόξενος 188 ἀστράβη -εύω -ίζω -ηλάτης 161–64 ἀσχαλάω 113 αὐτογενής 42–43 ἀφικετεία, -εύω 39–40 ἀφίκτωρ 38–40 βαρύτιµος 49–50 βοῦνις 105 γαιονόµος 73–74 n. 199 γε 74–75; 81 n. 217; 82–83 n. 221; 193–94 γε µὲν δή 142 γε µέντοι 183 n. 492 γεννάω 71–72 γένος 155 γερα-, γεραιο-, γήρα- 189 γεραφρονέω 189 γερογνώµων 189 γιγνώσκω 42 (pass.); 79 γόνος 120 δαιµόνιος 99 (n.pl.) δέ 69; 73 n. 199 δέ ... περ 73–74 n. 199 δέλτ- 122 δῆτα 130, 167 διὰ τὸ µή 11 n. 30 δίκη 45 -δικος 50 δοξαζω 79–80 δόρυ 109–10 ἐγχειρίδιον 48–49 εἴκω (pf.) 157 ἐναγής 106–8 ἐνώπιον 111–12 ἐπίδροµος 107 ἐπιλέγω (med.) 72–73 ἑτερορρεπής 196 εὐφηµ- 208 εὔφρων 191 ἐφύµνιον 103–4, nn. 280–81 ζαγρ- 115–16 ζαχρεῖος 125 Ζεὺς Αἰδοῖος 124–25 — Ἀλάστωρ 197 — Γεωργός 39 n. 98 — Ἱκέτης 39 n. 96 — Σωτήρ 55–56 ἥβη 88–89 ἡλιόκτυπος 114 -θερής 86 θυµέλη 17 n. 51 ἱλάοµαι 104 Ἰνδός 160 n. 449 καὶ γάρ 206 καὶ δή 207 κέλοµαι 65 κῆδος 179–80 κρεόβοτος 165 κρῖµα 194 κῦδος -αίνω -ιστος 46 κύπριος 155–58 λινορραφής 109 λόγος 153–54 249 µάταιος 126; 140 µε (encl.) 194–95 µέν ... δέ 41 n. 106 µένω + dat. 192–93; 199 µεταγιγνώσκω 100–101 µεταλγέω 196 µῆτις 80 µῦθος 154 ναεύω 187–88; 207 νεόδρεπτος 180 νέος 84–85 ὀκέλλω 200 ὀκρίβαντες 15 ὀνοτάζω 44 ὅστε (τά τε) 73 οὐ 192 οὐ (µή) ... οὖν 193–94 οὐκοῦν, οὔκουν 167, 172 οὖν ... ποτε 193–94 οὐρανόνικος 119 πάλαισµα 170–71 παλλακ- 171 παντᾷ 95–97 παρά + gen.loc. 126–27 πάροδος 59 n. 159 παροίχοµαι 203 πατραδελφεία 58 n. 158 πεντηκοντόπαις 175 250 πλήσσω 158 προσφυ- 154 ῥύσιον 174 σκοπός 192 στάσις 45–46 στόλος 123 στρέβλη 200–201 στρόφος 203 συµπρεπής + gen. 203–4 τά (rel.) 73; 150 τερασκόπος 189 τιµάωρ -ορος 68 τιµή 45, 50 -τιµος 49–50 τις 78–79 τυγχάνω 117 τύπος 156–60 τύχη 97 ὕβρις 88–89 φαίνω 76 n. 204 φεύγω 194 φυξανορία 43 φυλάσσω (med.) 133–34 χαρακτήρ 156–60 χθόνιος 50–55 χλίδ-, χλίω 141 χρηστήριον 203 ὡς (Conj.) 176–78 IV. Index parva nominum Apsephion: 2 n. 4 Archimedes: 201 Aristobulos Apostolides = Arsenius of Monembasia: 88 n. 234; 183 Campbell, A. Y.: 210 Cimon: 4 Critias: 137 n. 379 Dawe, R. D.: 210–11 ‘Diorthotes’ (anonymous corrector of the ms. M): 43 n. 114; 154; 176 Eliot, T. S.: 114 n. 316 Elmsley, Peter: 68–69 Fogelmark, Staffan: vi; 187 Friis Johansen, Holger: v Gorgias: 42 n. 110 Housman, A. E.: 204 n. 528 Janko, Richard: vi; 74; 102 Maas, Paul: 69 (n. 188) Militades: 206 Milton, John: 117 n. 324 Olympia master, the: 98 Pericleidas: 4 Pericles: 205 n. 531 Popper, Sir Karl: 61 (n. 163) Porson, Richard: 69 (n. 188); 216 n. ‡ Seleucus: 115–17 Socrates: 138 (n. 379) Solon: 137 n. 379 Spenser, Edmund: 114 n. 316 Themistocles: 206 West, Martin: v; vi; 248 Whittle, E. W.: v Willink, Sir Charles: vi; 73 n. 198; 75 Youtie, H. C.: 200 n. 524 251

RELATED PAPERS

RELATED TOPICS

Log In

Aeschylus’ Supplices: Introduction and Commentary on vv. 1–523

Aeschylus’ Supplices: Introduction and Commentary on vv. 1–523

Related Papers

RELATED PAPERS

RELATED TOPICS