Martin ————— Greek, Roman, and Byzantine Studies 50 (2010) 29–40 © 2010 Gunther Martin Three Deletions in Euripides’ Ion Gunther Martin HE TRIMETER SECTIONS in Euripides’ Ion, as scholars have noted, may contain more spurious text than has yet been identified.1 In what follows, I make three sug- gestions for deletion, partly anticipated long ago but rejected by the 20th-century editors, partly new (so far as I am aware). I. Lines 1398–1400 In the exodos, Creusa gives up her supplication after she has recognized the basket in which she had exposed her son. In the transmitted text she calls Ion her son and explains why she leaves her secure position at the altar:2 ΚΡ. τί δῆτα φάσµα τῶν ἀνελπίστων ὁρϱῶ; 1395 ΙΩΝ σίγα σύ· πῆµα κϰαὶ πάρϱοιθεν ἦσθά µοι.3 ΚΡ. οὐκϰ ἐν σιωπῇ τἀµά· µή µε νουθέτει. ὁρϱῶ γὰρϱ ἄγγος ᾧ ’ξέθηκϰ’ ἐγώ ποτε σέ γ’, ὦ τέκϰνον µοι, βρϱέφος ἔτ’ ὄντα νήπιον, Κέκϰρϱοπος ἐς ἄντρϱα κϰαὶ Μακϰρϱὰς πετρϱηρϱεφεῖς. 1400 1 M. D. Reeve, “Interpolation in Greek tragedy, III,” GRBS 14 (1973) 145–171, at 151; D. L. Page, Actors’ Interpolations in Greek tragedy. Studied with special reference to Euripides’ Iphigeneia in Aulis (Oxford 1934) 72, expresses sur- prise that there should be relatively few histrionic interpolations in the play. 2 The text I give is from Diggle’s OCT (1981). I cite the following commentaries of the play (partly in opera omnia) by authors’ names only: H. Grégoire (Paris 1923), K. H. Lee (Warminster 1997), A. S. Owen (Oxford 1939), U. v. Wilamowitz-Moellendorff (Berlin 1926). 3 The text of this line is highly problematic and probably needs cruces. However, the content of the first half is sufficiently clear from 1397, and the rest can be left as Diggle gives it without much impact on the present question. T 30 THREE DELETIONS IN EURIPIDES’ ION Greek, Roman, and Byzantine Studies 50 (2010) 29–40 λείψω δὲ βωµὸν τόνδε, κϰεἰ θανεῖν µε χρϱή. ΙΩΝ λάζυσθε τήνδε· θεοµανὴς γὰρϱ ἥλατο βωµοῦ λιποῦσα ξόανα· δεῖτε δ’ ὠλένας. CR. What unexpected sight is it that I am seeing? ION Be quiet. You have already been a pain to me before. CR. Silence is not what helps me now. Don’t lecture me. For I see the basket in which I once exposed none other than you, my dear child, when you were still a small infant, in the caves of Cecrops and the Makrai, roofed by rocks. I will leave this altar, even if I have to die. ION Arrest her. Driven by a god she has leapt away and left the statues of the altar. Bind her arms. 1396 σίγα L. Dindorf, πῆµα Broadhead, ἦσθα Musgrave, σιγᾶν σύ πολλὰ κϰαὶ πάρϱοιθεν οἶσθα µοι cod. 1398 ᾧ ’ξέθηκϰ’ Barnes, ὃ ’ξέθηκϰ’ cod. 1399 del. Cobet 1400 del. Paley Ion does not seem to realize what Creusa says in lines 1398– 1400. In his first reaction in 1402 he mentions only the fact that Creusa is now no longer protected by the altar; her clear statement regarding their relationship does not cause him to express any astonishment. This cannot be explained as Ion concentrating on the business at hand. For shortly after this section it becomes clear that he is not just ignoring Creusa’s claim to be his mother (be it part of θεοµανία or not) but un- aware of it:4 ΚΡ. σοῖς φίλοισιν εὑρϱίσκϰῃ φίλος. 1406 ΙΩΝ ἐγὼ φίλος σός; κϰᾆτά µ᾽᾿ ἔκϰτεινες λάθρϱᾳ; CR. You are found dear to those who are dear to you. ION I dear to you? And then you tried stealthily to kill me? Ion is surprised that Creusa calls herself “dear” (or “a friend”) to him and protests sharply—a strange contrast to his coolness when she uses the word τέκϰνον in 1399. Creusa explains the 4 That the lines are an aside and not heard by Ion (Lee, similarly W. Biehl, “Textprobleme in Euripides’ Ion. Bemerkungen zu Versumfang und Personenverteilung innerhalb der Sprechpartien,” Philologus 136 [1992] 14– 30, at 29) can be ruled out because of the emphatic ὦ τέκϰνον µοι and the declarative tone of 1397 and 1401. GUNTHER MARTIN 31 ————— Greek, Roman, and Byzantine Studies 50 (2010) 29–40 use of the word “dear” by stating her relation to him. Now Ion’s reaction is prompt and determined: ΚΡ. παῖς γ’, εἰ τόδ᾽᾿ ἐστὶ τοῖς τεκϰοῦσι φίλτατον. ΙΩΝ παῦσαι πλέκϰουσα – λήψοµαί σ᾽᾿ ἐγώ – πλοκϰάς. 1410 CR. Yes, since you are my child—I guess that is what is most dear to one’s parents. ION Stop weaving wiles—I’ll catch you out! Not only is the double declaration (1399, 1409) of Creusa’s motherhood awkward; the different reactions on Ion’s part, total disregard versus immediate suspicion, are psychologically unconvincing. If Ion has heard Creusa use the word τέκϰνον and the statement that she was the one who exposed him, the sardonic scepticism following φίλος is implausible, for that would be a natural claim by a mother. By contrast, no information or dramatic movement is lost if we delete the mention of Ion’s exposure in 1398–1400. In contrast to Ion, the audience has no problem following events on stage. Line 1395 shows that the first half of the mutual recognition has taken place and motivates Creusa’s action sufficiently: she has immediately realized that Ion is her son and so understands that silence is not the right response to the situation. Only Ion is puzzled by the sudden change in her behavior and attributes it to supernatural causes. The deletion of all three lines thus restores a dramatically effective and coherent text—without any loss of information or necessary elements of the plot. It has the additonal effect of removing lines that, although not obviously corrupt, have in- dividually given rise to various objections.5 5 For example, the transmitted text of 1398 (ὃ ἐξέθηκϰε) leaves σέ without a straightforward syntactical connection and has led to Cobet’s deletion. A possibility suggested by Biehl, Philologus 136 (1992) 29, is a dash after 1398: “das Gefäß—nein vielmehr dich selbst.” Similarly Grégoire translates: “Car je vois la corbeille où jadis j’exposai un enfant nouveau-né—ah! mon fils, c’était toi!,” producing a version by which the “correcting” γε (with σέ) precedes what it corrects (βρϱέφος). For this (not epexegetic!) use of γε in self- correction I find no parallel. Alternatives may be µᾶλλον δέ γε (e.g. Ar. 32 THREE DELETIONS IN EURIPIDES’ ION Greek, Roman, and Byzantine Studies 50 (2010) 29–40 It also solves a problem later in the text: Creusa says ἀνθέξοµαι / κϰαὶ τῆσδε κϰαὶ σοῦ τῶν τε σῶν κϰεκϰρϱυµµένων (1404– 1405).6 Most naturally, this is interpreted as a claim on the basket, the boy, and the recognition tokens, all of which be- longed to Creusa when she abandoned her child. However, the word used for the basket at its last mention was ἄγγος in 1398. So no matter whether Creusa refers back to ἄγγος or uses a deictic without a specific antecedent, we would expect τοῦδε.7 With 1398–1400 gone, the last reference to the basket is Ion’s ἀντίπηγος εὐκϰύκϰλου in 1391. The pronoun then has the gen- der we would expect. The deletion removes all the problems while leaving the action perfectly understandable. The interpolation should be regarded as one that makes the motive for Creusa’s sudden surrender explicit and heightens the pathos by reminding the audience of the exposure of the infant. II. Line 647 After the first (false) recognition of the play Xuthus invites his newly found (putative) son to accompany him from Delphi to Athens. Ion at first rejects that offer stating that he will face hostility from various sides if he goes to Athens, while the sanctuary of Apollo provides him with the opportunity to lead an ideal, peaceful life. Ion sums up his argument and asks ___ Vesp. 1485) or κϰαὶ σέ γε, cf. J. D. Denniston, The Greek Particles (Oxford 1954) 158. Diggle follows Barnes in substituting ᾧ, in parallel with 1413: σά γ’ ἔνδυθ’, οἷσί σ’ ἐξέθηκϰ’ ἐγώ ποτε. However, in 1398 the nature of the dative remains obscure, while 1413 can be explained, with Wilamowitz, as ἔνδυθ’, οἷς <ἔνδυτον>. 6 This is the text of the codex. Diggle adopts Tyrwhitt’s τῶν τ’ ἔσω, but this has no bearing on the problem as we agree upon the meaning of τῆσδε. 7 C. H. Whitman, “Two Passages in the Ion of Euripides,” CP 59 (1964) 257–259, tries to explain τῆσδε as the Pythia and translates “I will dispute the position, both with this woman and with you too, of those things” (γε for τε before σῶν). But he misinterprets his parallel Ar. Av. 1658, where the second genitive is possessive (if it is not replaced, as in recent editions, with a dative). The Pythia has left after κϰαὶ χαῖρϱ᾽᾿ in 1363, and it may be hard for an audience to identify her as the female or feminine referred to. GUNTHER MARTIN 33 ————— Greek, Roman, and Byzantine Studies 50 (2010) 29–40 Xuthus to allow him to stay in Delphi: ταῦτα συννοούµενος κϰρϱείσσω νοµίζω τἀνθάδ᾽᾿ ἢ τἀκϰεῖ, πάτερϱ. 645 ἔα δέ µ᾽᾿ αὐτοῦ ζῆν· ἴση γὰρϱ ἡ χάρϱις µεγάλοισι χαίρϱειν σµικϰρϱά θ᾽᾿ ἡδέως ἔχειν. Taking all these points into consideration I believe things here are better than in Athens, father. Let me live here: it is equally pleasurable to take delight in the big and to enjoy the small. 646 δέ µ᾽᾿ αὐτοῦ Badham, δ᾽᾿ ἐµαυτῶ cod. “Big” Athens with Xuthus’ riches and power is equated to the modesty of Ion’s current existence in Delphi.8 The contrast to sentiments expressed earlier in the same speech is striking: the life of power and wealth has been described as full of terror, the opposite of delight (630–632), whereas life in Delphi has been characterized in terms that let it appear as pure χάρϱις (633– 644). So the amount of pleasure to be expected in Athens is actually much smaller than in Delphi. David Kovacs has deleted lines 621–632, which includes the horrid description of life as a tyrant and the additional in- quietude that money brings with it. If we accept that deletion, there has been no reference to riches at all in this speech.9 In that case the contrast in the last line introduces a new ar- gument—the relative value money has for one’s happiness—in the briefest possible way, which does not allow it to gain per- suasive force. In either case ἴση χάρϱις taken as “equal delight/ pleasure” does not go well with κϰρϱείσσω νοµίζω τἀνθάδ’ ἢ τἀκϰεῖ, which also stresses the superiority of Delphi over Athens and makes the last line appear as undercutting Ion’s own argu- 8 For the sentiment cf. Men. PCG 843. 9 D. Kovacs, “Four Passages from Euripides’ Ion,” TAPA 109 (1979) 111– 124, at 116–124; the same suggestion is made independently by K. Alt, “Ion und die Tyrannis. Zu Euripides Ion V. 621–632,” in Ch.-F. Collatz (ed.), Dissertatiunculae criticae. Festschrift für Günther Christian Hansen (Würzburg 1998) 23–32. 34 THREE DELETIONS IN EURIPIDES’ ION Greek, Roman, and Byzantine Studies 50 (2010) 29–40 ments. So the last sentence should be rejected. However, the start of 646 is indispensable as conclusion and final appeal to Xuthus; that means that the second half of the line cannot be deleted without assuming a lacuna. But if the interpretation is changed, ἴση γὰρϱ ἡ χάρϱις can remain in the text: “the delight (χάρϱις) is the same” requires the explanatory infinitives, but “this is just as big a favor (χάρϱις)” does not. The sentence is then a close parallel to Hipp. 508, δευτέρϱα γὰρϱ ἡ χάρϱις: there a hierarchy of favors is established, whilst in the Ion it is denied. Line 647 is then a typical interpolated explanation, supposed to help the reader understand Euripides’ precise formulation but in reality altering the meaning.10 With 647 removed Ion appears as polite as at the start of the speech.11 He does not discredit the father’s invitation as un- welcome or against his interest. Instead, he emphasizes that Xuthus would not be ungenerous if he left his son behind. Xuthus believes he is doing something good to his son by taking him to Athens and liberating him from slavery, but Ion makes it clear that he (Ion) would appreciate the permission (ἔα) to stay in Delphi and regard it as a favor. The logic of γάρϱ in 646 thus becomes more stringent: it is not the case that Ion asks to stay “because” staying is as good as living in Athens. Instead, the particle refers to ἔα µ᾽᾿: allowing him to stay would be a favor; and that favor would be just as great as making him his son and giving him a home. III. Lines 612–620 Another passage earlier in the same speech may be worth considering as interpolation, as both the train of thought and 10 Compare, for example, the undoubtedly interpolated Or. 916, 1024, or, syntactically more similar, Hel. 764, deleted by Kirchhoff, who is followed by Dale and Diggle. 11 At the start Ion shows he is highly considerate towards Xuthus in his attempt not to offend his father while declining his invitation (587–589): ἐγὼ δὲ τὴν µὲν συµφορϱὰν ἀσπάζοµαι, / πατέρϱα σ’ ἀνευρϱών· ὧν δὲ γιγνώσκϰω, πάτερϱ, / ἄκϰουσον. GUNTHER MARTIN 35 ————— Greek, Roman, and Byzantine Studies 50 (2010) 29–40 the language are not beyond doubt. Lines 616–617 have long been disputed,12 but there may be a case for more extended scrutiny. Ion has mentioned that—being a bastard son of a non-autochthonous Athenian—he will be derided as a nobody by the public. As to his stepmother Creusa, she will hate him: ἐλθὼν δ᾽᾿ ἐς οἶκϰον ἀλλότρϱιον ἔπηλυς ὢν γυναῖκϰά θ᾽᾿ ὡς ἄτεκϰνον, ἣ κϰοινουµένη τῆς συµφορϱᾶς σοι πρϱόσθεν ἀπολαχοῦσα νῦν αὐτὴ κϰαθ’ αὑτὴν τὴν τύχην οἴσει πικϰρϱῶς, 610 πῶς [δ᾽᾿] οὐχ ὑπ’ αὐτῆς εἰκϰότως µισήσοµαι, ὅταν παρϱαστῶ σοὶ µὲν ἐγγύθεν ποδός, ἡ δ’ οὖσ’ ἄτεκϰνος τὰ σὰ φίλ’ εἰσορϱᾷ πικϰρϱῶς, κϰᾆτ’ ἢ πρϱοδοὺς σύ µ’ ἐς δάµαρϱτα σὴν βλέπῃς ἢ τἀµὰ τιµῶν δῶµα συγχέας ἔχῃς; 615 ὅσας σφαγὰς δὴ φαρϱµάκϰων <τε> θανασίµων γυναῖκϰες ηὗρϱον ἀνδρϱάσιν διαφθορϱάς. ἄλλως τε τὴν σὴν ἄλοχον οἰκϰτίρϱω, πάτερϱ, ἄπαιδα γηρϱάσκϰουσαν· οὐ γὰρϱ ἀξία πατέρϱων ἀπ’ ἐσθλῶν οὖσ’ ἀπαιδίᾳ νοσεῖν. 620 And when I come to the house of strangers, myself a foreigner, and to your childless wife, who—previously sharing her sorrow with you but now on her own with it— will feel bitterly the fate she must bear by herself, how will I not incur her hatred, and naturally so, when I stand near you, right by your foot, and she gives bitter looks to your dear son and when you then either abandon me, having regard to your wife instead, or honor me and thereby destroy your family? How many ways of murdering and death by fatal poison have women found for men! Besides, I pity your wife, father, when she ages in childlessness. She is undeserving, 12 For the main arguments see W. Kraus, “Textkritische Erwägungen zu Euripides’ Ion,” WS 102 (1989) 35–110, at 62. The suspicion and hostile attitude in it contrast strangely with the sympathy towards Creusa in the rest of the passage. 36 THREE DELETIONS IN EURIPIDES’ ION Greek, Roman, and Byzantine Studies 50 (2010) 29–40 because of her noble ancestry, to suffer from barrenness. 610 αὐτὴ Ald., αὐτὴν cod. 611 δ’ del. Canter 612–613 suspectos habuit Nauck 614–615 del. Nauck 616–617 del. L. Dindorf, <τε> add. Heath/Tyrwhitt 620 del. Nauck; ἀπαιδίᾳ L. Dindorf, ἀπαιδίαν cod. Before Xuthus discovered his son, Creusa did not have to cope with her childlessness all by herself. Instead she had her husband as a companion who halved her problem and suffer- ing by sharing them. With the intruder in Creusa’s house and Xuthus no longer in the same position she would be alone in her misery and (as a natural consequence) hate the person re- sponsible for it. Up through line 611 there are no difficulties. The text is syn- tactically complete and contains all the necessary information; the situation has been sufficiently described. Line 612 introduces another indication of time, parallel to ἐλθών in 608. This temporal clause, while seeming to repeat the idea of “entering into Creusa’s house,” takes that idea more literally than seems appropriate: ἐλθὼν δ᾽᾿ ἐς οἶκϰον ἀλλότρϱιον does not mean “to enter the space strangers dwell in,” but “to enter into strangers’ household and family.” This is in line with the emphasis on inheritance and racial purity in the rest of the speech and play and balances the thought that the Athenians will reject him as ignoble (592). This aspect of Creusa’s hatred is forgotten in the ὅταν-clause, which focuses solely on Creusa’s jealousy as she alone remains childless. The situation is nar- rowed down to her feelings when she sees Ion next to Xuthus.13 This restriction turns out to be unnecessary, as Creusa reacts strongly when she hears of Xuthus’ plans (cf. 864–865). The sequence of events is also peculiar: first Ion stands next to his father, then Xuthus decides to back either Ion or Creusa, and only then does Creusa hate. The disruption of the house 13 One might even wish to take the aorist in 612 as the description of a single action: the moment when Ion is introduced to Creusa. But at this moment in the play, Ion must assume that he is declared Xuthus’ son even before he goes to Athens (that changes only in 654). So the situation envisaged would even contradict the situation in the play. GUNTHER MARTIN 37 ————— Greek, Roman, and Byzantine Studies 50 (2010) 29–40 (being still part of the sub-clause) precedes Creusa’s hatred, even though it should be its consequence. What is more, Creusa’s hatred develops even in case Xuthus abandons Ion in her favor. In order to restore a more plausible sequence of thoughts we would need to bring lines 614–615 to the same syntactic level as 611, but the form ἔχῃς cannot be turned into a metrically fitting future tense. The formulation παρϱαστῶ σοὶ µὲν ἐγγύθεν ποδός is doubly unique in Euripides’ tragedies: the poet employs the periphrasis “your foot” for “you,”14 but the present usage does not seem to have parallels; instead we would expect an expression like ποδί σου. A second irregularity should prevent us from explaining the transmitted version as a free one-off formation by analogy: the use of ἐγγύθεν as a preposition15 is not tragic, not even classical. In archaic poetry I find four, partly doubtful, in- stances.16 In the other occurrences in fifth- and fourth-century literature, ἐγγύθεν is always used as an adverb, never as a prep- osition.17 The Hellenistic poets revive the construction (e.g. 14 Cf. Cyc. 6, Hipp. 661, Or. 1217, IA 627. 15 I see no possibility of a different construction: adverbial ἐγγύθεν (in analogy with ἐγγὺς παρϱεστώς in Alc. 1011) leaves ποδός without a syntactic connection. 16 Hom. Il. 11.723, Simonides Anth.Gr. 16.26, Solon fr.28 W., and Theognis 1.943. The four-line epigram of Simonides is said by D. L. Page, Further Greek Epigrams2 (Cambridge 1981) 189, to be of “indeterminable” date. There cannot be much more confidence about the authenticity of the later part of “Theognis’” first book. The fragment of Solon consists of a single line (Νείλου ἐπὶ πρϱοχοῇσι Κανωβίδος ἐγγύθεν ἀκϰτῆς), said by Plu- tarch to describe the poet’s voyage to Egypt. Which poem this is from is unknown (but cf. M. R. Lefkowitz, The Lives of the Greek Poets [Baltimore 1981] 44). 17 Of the instances of genitive in the proximity of ἐγγύθεν in tragedy, Aesch. Sept. 973 is corrupt, cf. G. O. Hutchinson, Aeschylus: Septem contra Thebas (Oxford 1985) ad loc.; M. L. West, Studies in Aeschylus (Stuttgart 1990) 123–124. In Eur. Ion 586 ἐγγύθεν is clearly an adverb. The same applies to Aesch. Cho. 852 εἴτ᾽᾿ αὐτὸς ἦν θνῄσκϰοντος ἐγγύθεν παρϱών, “standing near as [genitive absolute] he was dying”; the adverbial ἐγγύθεν describes παρϱών, cf. Soph. OT 1259. 38 THREE DELETIONS IN EURIPIDES’ ION Greek, Roman, and Byzantine Studies 50 (2010) 29–40 Theoc. Id. 7.112, Arat. Phaen. 181), perhaps regarding the single Homeric occurrence as an oddity worth copying.18 A later interpolator, using Euripidean diction such as ποῦς, would thus not feel that he is writing something that is not in line with poetic diction. In any case, the formulation should be regarded as dubious on two independent grounds. Deleting lines 612–617 would not only avoid these oddities and anomalies but yield additional advantages. Ion would not be uncourteous to Xuthus, as he is if the text is genuine: the periphrastic construction implies that Xuthus himself will cause the disruption of his family.19 This straightforward imputation against Xuthus—reaffirmed by πρϱοδούς—would be untypical for Ion, who deals with his father in a highly polite manner (see above on 644–646). In addition, Ion’s point is formulated much more forcefully if the sentence ends with the question rather than a long-winded and clumsy temporal clause. This change would also agree well with the deletion of lines 595–606 by David Kovacs (in addition to the ones mentioned above). His objections are based on considerations of context and linguistic irregularities and should be accepted.20 In that case, the sections in which Ion deals with the hostility of the autochthonous Athenian public and from Creusa move closer 18 This instance is the only one of this construction in Homer as opposed to eight for ἐγγύθεν + dative in Iliad and Odyssey (and one in the Hymn to Demeter) and eleven for ἐγγύθι + genitive. 19 W. J. Aerts, Periphrastica. An Investigation into the Use of εἶναι and ἔχειν as Auxiliaries or Pseudo-auxiliaries in Greek from Homer up to the Present Day (Am- sterdam 1965) 144. 20 The deletions have not met with the attention they deserve. Diggle does not seem to know them in his 1981 edition; Lee 225–226, 229–230, and K. Zacharia, Converging Truths. Euripides’ Ion and the Athenian Quest for Self- definition (Leiden/Boston 2003) 23–24, defend the passages on rather general grounds, showing what seems reconcilable with the drama rather than explaining the linguistic problems (e.g. τε and the infinitive in 598, λήψοµαι in 600, and the construction of ἐκϰνικϰᾶν in 629). These problems, however, as well as the inconsistency within the passage (Ion is envisaged first as a despised outsider, then as a democratic politician, and finally as a monarch) and with the rest of the scene need to be taken seriously. GUNTHER MARTIN 39 ————— Greek, Roman, and Byzantine Studies 50 (2010) 29–40 together (which tightens Ion’s argument). Moreover, πῶς οὐχ … µισήσοµαι in 611 then corresponds with κϰεκϰλήσοµαι at the end of the first section of the speech—and it may even be tempting to supplement <πῶς οὐ τὸ> µηδὲν κϰοὐδένων κϰε- κϰλήσοµαι in 594. In this way, both sections are of equal length and end on a sharp rhetorical question, further emphasized by homoeoteleuton: a very pointed formulation appropriate for the persuasive effect aimed at in the speech. If these lines are rejected, then 618–620 presumably have to go, too. The loss would not be great. The introduction with ἄλλως τε signals an addition: another reason why Ion would prefer not to go to Athens.21 But this is not what these lines give. Ion’s sympathy with Creusa has been expressed, albeit implicitly, in 608–610: she was unhappy before as a con- sequence of her barrenness, but now she is left alone with her calamity. Her husband seems to react to this passage in 657– 658: κϰαὶ γὰρϱ γυναῖκϰα τὴν ἐµὴν οὐ βούλοµαι / λυπεῖν ἄτεκϰνον οὖσαν αὐτὸς εὐτυχῶν. Lines 618–620, though more explicit, are not alluded to in a similar way. Ion has brought his point across by the earlier subtle remark. The idea that old age or Creusa’s noble birth may aggravate the problem is not recalled by Xuthus and seems irrelevant. Moreover, the connection of this passage with the next genuine one (633–646, if we follow Kovacs) is clearer without the three intervening lines 618–620. Ion has two arguments against living in Athens: public derision and private hatred. This is contrasted with the blessings of life in Delphi. If we retain the three lines, Ion’s description of his wonderful life in Apollo’s sanctuary (introduced in 633: ἃ δ’ ἐνθάδ’ εἶχον ἀγάθ’ ἄκϰουσόν µου, πάτερϱ) follows the somber mention of his pity for the aging and lonely Creusa. If we reject the lines, Ion’s blissful life at Delphi is directly juxtaposed with Creusa’s prospective hatred. Kovacs’ deletions taken together with the ones proposed here 21 For this use cf. Eur. IA 491, Soph. OT 1114. 40 THREE DELETIONS IN EURIPIDES’ ION Greek, Roman, and Byzantine Studies 50 (2010) 29–40 result in the speech being less than half as long as the version transmitted in the manuscripts. Features that have influenced many interpretations of the play are taken away, but Ion’s arguments are pruned to what contributes to his immediate purpose of persuasion. Ion himself is again the sensitive and thoughtful person devoted to Apollo and Delphi that we have encountered earlier in the play. The spirit of the argument is in harmony with his plea in lines 644–646: Ion does not say he will stay in Athens for the greater good of all, but because life in Delphi is so much preferable for him. That does not make him a disagreeable, self-centered person: he knows subtle ways of expressing his empathy with Creusa; and he presents his re- quest to Xuthus in a way that does not offend his benefactor, but assures him of his gratitude.22 November, 2009 Lady Margaret Hall Oxford OX2 6QA, U.K. gunther.martin@lmh.ox.ac.uk 22 I am indebted to Prof. Chris Collard and the anonymous readers, who gave generous advice even where they disagreed.