I. 69–77

Most of the arguments for deleting 73 have been laid out by Dawe,Footnote ¹ and still hold good, despite subsequent arguments to the contrary.Footnote ² In particular:

1. (1) The emphatic statement that Greek is the language of Thebes (72–3) is gratuitous. That in 170 the gods are asked not to deliver Thebes to attackers who ‘speak an alien dialect’ (ἑτεροφώνωι) is beside the point: ἑτερόφωνος is not the same as βαρβαρόφωνος,Footnote ³ and nowhere in the play are the Argives envisaged as speakers of a foreign tongue,Footnote ⁴ for the obvious reason that this would fly in the face of common knowledge. It has been argued that the point of the emphasis on language is that the gods of Greece are expected to protect a Greek city;Footnote ⁵ but by the same token the Greek-speaking Argives are just as entitled to divine support as the Greek-speaking Thebans. There is also a linguistic difficulty:Footnote ⁶ when referring to human sounds, φθόγγος indicates timbre or intensity of sound rather than ‘language’ (see LSJ s.v.).

2. (2) As an alternative to φθόγγον χέουσαν, a supralineal note in the Laurentianus graecus 32.9 (M) offers (untypically for the manuscript tradition of Aeschylus)Footnote ⁷ the γρ-variant καὶ ὄλβον ῥέοντα, which is scriptio plena for κὤλβον ῥέοντα.Footnote ⁸ It seems unlikely that this variant represents a genuine piece of alternative tradition; rather, as Dawe saw, it is probably a misguided attempt to pad out line 72.Footnote ⁹ In particular, ‘crasis of καὶ + ο with words other than the article and relative is attested in A.’ only in frr. 205 and 361, the first of which is probably satyric.Footnote ¹⁰

3. (3) The locution δόμους ἐφεστίους (73) is unacceptable: houses are not ‘at the hearth’ (the opposite is true), nor can ἐφεστίους mean ‘equipped with hearths’ (despite Zuntz [n. 2])—if it could, one would have expected it to do so more often (in Ag. 851 read Karsten's δόμους ἐφέστιος).Footnote ¹¹ The phrase is merely a mechanical repetition of such tragic collocations as Soph. Trach. 262 ἐλθόντ’ ἐς δόμους ἐφέστιον or Eur. Med. 713 δόμοις ἐφέστιον (cf. also Aesch. Sept. 853; Supp. 365; Eum. 577, 669). In the above cases, however, ἐφέστιος never qualifies δόμος: its function is to emphasize that the hearth is a point fraught with special significance, particularly as a place where guests’ or suppliants’ rights are observed or infringed.Footnote ¹²

We may further note that Ἑλλάδος (72), which is usually taken as a genitive of separation from ἐκθαμνίσητε (‘do not eliminate my city from Greece’),Footnote ¹³ seems feeble without further qualification. The parallels adduced by Hutchinson (n. 8) contain more forceful formulations, with epithets adding emphasis: Aeschin. 3.133 Θῆβαι … ἐκ μέσης τῆς Ἑλλάδος ἀνήρπασται (cf. Din. 1.24); Thuc. 3.57.2 ἐκ παντὸς τοῦ Ἑλληνικοῦ … ἐξαλεῖψαι (sc. Πλάταιαν). It seems preferable to assume that one or more lines have dropped out after 72, and that Ἑλλάδος once belonged syntactically with those now-lost lines (see below).

Lines 74–5 are probably to be expunged as well, as was already suggested by Harberton.Footnote ¹⁴ The possibility of a life under foreign yoke, distressing as it is in itself, inevitably comes across as a feeble appendage after the harrowing threat of total annihilation conjured up in 71–2; as a result, the overall effect of 71–5 is anticlimactic. Likewise, γῆν τε καὶ Κάδμου πόλιν (74), coming so shortly after πόλιν (71), looks like a piece of weak repetitive padding, especially in view of the near-tautology γῆν … πόλιν.Footnote ¹⁵ Finally, while μήποτε σχεθεῖν (75) in itself could stand in lieu of a subjunctive in imperative sense (= μήποτε σχῆτε),Footnote ¹⁶ its coming after an imperative here (71–2 μή μοι … | ἐκθαμνίσητε) is syntactically jarring, hence Butler's emendation μὴ δότε σχεθεῖν.Footnote ¹⁷

Deleting 73–5 and positing a two-line lacuna after 72Footnote ¹⁸ leaves enough room for Eteocles’ prayer to have included, in the lacuna, a positive request on behalf of Thebes in addition to the negative one of 71–2 (again in two lines). In the lacuna (designated below as 72a–b), Eteocles will have asked, no doubt, for divine protection in the struggle against the Seven (cf. his request for ‘succour’, ἀλκή, in 76). For example:

For the hyperbolic Ἑλλάδος as a designation of the allied attackers, cf. the comparable use of Δαναοί in Eur. Phoen. 1145 τέκνα Δαναῶν, 1226 Δαναῶν ἀριστῆς (elsewhere in Eur. Δαναοί = Ἕλληνες, for example Hec. 113, 134, 138; Tro. 1122). For the scansion of ὄλεθρον (⏑—⏑), cf. Soph. OT 430. Presumably, once whatever stood in the lacuna was lost, lines 73–5 were devised as a stopgap solution, in order to make sense of Ἑλλάδος and to supply the missing positive part of the prayer that would lead up to γένεσθε δ’ ἀλκή (76). If the lacuna did contain a positive request, as argued above, then the MSS δ’ in 73, which would have had an acceptable adversative force if coming right after the negative content of 71–2,Footnote ¹⁹ must be changed into τ’.

II. 80–4

Line 83 is garbled beyond remedy. The paradosis is exhaustively discussed by West,Footnote ²⁰ who also mentions select emendations. Headlam's brilliant ἕλε δέ μ’ ἀσ<πίδων πάταγος …>Footnote ²¹ is based on the assumption that the copyist's eye jumped from ἀσπίδων to the similar-looking πεδι- below; however, for this use of ἕλε Headlam could only invoke the specious parallel of Aesch. Supp. 786 πατρὸ<ς> σκοπαὶ δέ μ’ εἷλον· οἴχομαι φόβωι, where the meaning (‘my father's lookout has trapped me’)Footnote ²² is quite different from what is required here (‘the clash of shields has overcome me with fear’).Footnote ²³

An undeservedly forgotten conjecture is Wilhelm Dindorf's εἷλε δ’ ἐμ<ὰς φρέ>νας <δέος>,Footnote ²⁴ which has the merit of neatly explaining the corruption as a result of the copyist's eye jumping from ἐμάς to φρένας because of the similarity of the word-endings. Dindorf's emendation may be expanded into, for example, ἕλε δ’ ἐμ<ὰς φρέ>νας <περίφοβον δέος> to yield a set of two identical dochmiacs (⏑ ⏕ — ⏑ —). For ἕλε … φρένας … δέος, cf. Hom. Il. 19.14 πάντας ἕλε τρόμος and Aesch. Supp. 379 φόβος μ’ ἔχει φρένας; for περίφοβον δέος, cf. Aesch. Supp. 736 περίφοβόν μ’ ἔχει τάρβος. The words of the chorus ‘I am seized with fear’ would be registering a personal reaction to the terrifying sights and sounds of the attack to complement the visual emphasis of line 81 (‘I am persuaded by the dust rising to the sky’). Wedged as it is between references to sights and sounds, ‘I am seized with fear’ may appear intrusive, but a comparable phenomenon occurs in 120–4, where φόβος δ’ ἀρείων ὅπλων is again sandwiched between references to sights and sounds.

III. 89

Both βοά and βοᾶι are transmitted, and both are equally possible. With the former, supply, for example, βοὰ <δ’ ὀρσίπους>, ‘a roar stirring the feet (of the chorus) to flight’; cf. Trag. Adesp. 245 Kannicht/Snell ὀρσίπους βοή. With the latter, taken as third person singular, supply, for example, βοᾶι <τραχὺ κῦμ’>, ‘the wave (of the approaching army) roars harshly’. For the image and phraseology, cf. Hom. Il. 14.394–5 οὔτε θαλάσσης κῦμα τόσον βοάαι ποτὶ χέρσον, | πόντοθεν ὀρνύμενον πνοιῆι Βορέω ἀλεγεινῆι. For hostile warriors as a billowing wave, cf. Eur. IT 316 κλύδωνα πολεμίων; Supp. 474–5 κλύδων | … δορός (~ Phoen. 859–60); Ion 60 πολέμιος κλύδων;Footnote ²⁵ already in the Babylonian Gilgamesh epic (I 34) the eponymous hero is, to enemy troops, ‘a violent floodwave that smashes a stone wall’.Footnote ²⁶ For τραχύς qualifying the roar of waves, cf. ?Aesch. PV 1048 κῦμα δὲ πόντου τραχεῖ ῥοθίωι. For the elision (κῦμ’) at the division between two dochmiacs, cf. (for example) Aesch. Sept. 164 ἄνασσ’ | Ὄγκα.

IV. 101–2

This vexed passage and its multiple problems have been most fully discussed by West.Footnote ²⁷ The paradosis is metrically unacceptable: πότ’ εἰ μὴ νῦν ἀμ- cannot be construed as a ‘dragged-out’ dochmiac (⏑— — — —),Footnote ²⁸ because there are no Aeschylean or Sophoclean parallels for the pattern × — — ×̅ | —, with word-end after a long second anceps (νῦν | ἀμ-).Footnote ²⁹ There are also linguistic difficulties. In trying to make sense of ἀμφὶ λιτὰν ἕξομεν, scholars at least as early as Thomas MagisterFootnote ³⁰ have taken it to mean ‘busy ourselves with supplication’; however, as West pointed out, ἀμφί τι ἔχειν in this sense ‘is an expression found several times in Xenophon and in later prose, but not so far as I know in verse’.Footnote ³¹ The further problem that λιτή in the singular never occurs in tragedy has been dealt effectively by Seidler, who emended into ἀμφὶ λίταν’ (sic; read λιτάν’, i.e. λιτανά) ἕξομεν;Footnote ³² however, although λιτάν’ must be considered certain, Seidler's emendation still leaves unresolved the problem of ἀμφί τι ἔχειν. Alternatively, it has been arguedFootnote ³³ that ἀμφὶ … ἕξομεν is tmesis for ἀμφέξομεν (from ἀμπέχω), with βρέτη supplied as its object from 99 (ἀκμάζει βρετέων ἔχεσθαι): ‘when, if not now, shall we put on (the gods’ statues) suppliant offerings of robes and garlands?’ But such tmesis of ἀμπέχω is unparalleled, conceivably because it would blur the distinction between ἀμπέχω and ἀμφί τι ἔχω.Footnote ³⁴

We must conclude that the paradosis is untenable. Brilliantly, West suggested that the key to the solution had been lying hidden in plain sight all along. He pointed out that a scholium (Σ 101a) in our earliest extant MS (Laur. gr. 32.9, or M) appears to preserve a paraphrase of the original now-lost reading:Footnote ³⁵

Cf. also Σ 101c:

The tell-tale ῥίψομεν … ἐπὶ γῆς (or πρὸς τὴν γῆν κάτωθεν ῥίψασαι), for which no corresponding locution survives in our text of the play, suggested to West that the scholiast had before him a text that read ἀμφὶ λιτὰν (or λιτανὰ) <βαλεῖν>. According to West, the scholiast, failing to realize the tmesis (= λιταν[ὰ] ἀμφιβαλεῖν), took instead βαλεῖν as a simplex (hence his ῥίψομεν) and ἀμφὶ λιταν(ά) as ‘busying ourselves with supplication’ (hence his τραπεῖσαι ἐπὶ λιτανείαν). With typical ingenuity, West suggested the following supplements:

The suggestion that Σ 101a (and Σ 101c) preserve traces of the original reading seems irresistible indeed. But West's emendation fails to account for the scholiast's (ῥίψομεν) ἐπὶ γῆς or πρὸς τὴν γῆν (ῥίψασαι). It is odd that the scholiast should specify that the robes and wreaths are to be thrown ‘on the ground’, since he was perfectly aware that such items were customarily put on the gods’ statues (cf. his ἐνέδυον γὰρ καὶ πέπλους τὰ ἀγάλματα). Surely, there must have been something in the original text to suggest the seemingly bizarre ἐπὶ γῆς / πρὸς τὴν γῆν? Here is one way of restoring such a text:

Construe πότ’ εἰ μὴ νῦν <ἀν>έξομεν λιταν(ὰ) ἀμφι<βόλωι γάνει> πέπλων καὶ στεφέων; (lit. ‘when, if not now, shall we offer a supplication by means of the delight of robes and wreaths thrown around [the gods’ statues]?’). For πότε εἰ μὴ νῦν with a future verb, cf. Ar. Vesp. 402 (πότε δ’ εἰ μὴ νῦν ἐπαρήξετέ μοι;) and Dem. 3.16 (πότε ἃ δεῖ πράξετε, εἰ μὴ νῦν;). For ἀμφίβολος = ‘thrown around’ as a garment, cf. Eur. Ion 1490 σπάργαν’ ἀμφίβολά σοι τάδ’ ἀνῆψα. For γάνος, ‘a thing to delight in’, used of offerings to the gods (and thus essentially a synonym of ἄγαλμα), cf. Aesch. Ag. 578–9 θεοῖς λάφυρα ταῦτα … | δόμοις ἐπασσάλευσαν ἀρχαῖον γάνος. For the periphrasis λιτάν’ <ἀν>έξομεν, cf. Soph. El. 635–6 ὅπως λυτηρίους | εὐχὰς ἀνάσχω δειμάτων.Footnote ³⁷ In the present passage, ἀνέχω would be particularly apt, since the raising of hands is a characteristic gesture of ritual supplication:Footnote ³⁸ cf. Eur. El. 592–4 ἄνεχε χέρας, ἄνεχε λόγον, ἵει λιτὰς | ἐς θεούς.

It now becomes easier to understand how the scholiast's ἐπὶ γῆς may have come about. Evidently, ἀμφιβόλωι γάνει was misread as ἀμφιβόλωι γᾶν εἰ, just as in Aesch. Ag. 1391–2 the paradosis διὸς νότω | γᾶν εἰ originated in an erroneous parsing of διοσδότωι | γάνει (Porson). It was such a misapprehension, I submit, that prompted the scholiast's ῥίψομεν … ἐπὶ γῆς, in a futile attempt to make sense of πέπλων … ἀμφιβόλωι γᾶν as an equivalent of, for example, πέπλα … βαλοῦμεν ἀμφὶ γῆν, with ἀμφιβόλωι possibly taken as a noun, ‘a thing thrown around’ (as in Eur. Tro. 537 κλωστοῦ δ’ ἀμφιβόλοις λίνοιο). It may even be that the meaningless εἰ λιτάν was ‘corrected’, mentally or in an actual manuscript, into εἰς λιτάν, hence the scholiast's τραπεῖσαι ἐπὶ λιτανείαν.

Finally, a word on the metre of 101. West's supplement <πότε> πότ’ εἰ μὴ νῦν produces a wholly legitimate ‘dragged-out’ dochmiac, which (albeit unparalleled in this parodos) occurs twelve times in Aeschylus, by Conomis's count.Footnote ³⁹ Łowiński'sFootnote ⁴⁰ πέπλων καὶ στεφέων εἰ μὴ νῦν πότ’ ἀμ- does produce a smoother set of dochmiacs, but the order εἰ μὴ νῦν πότε (rather than πότε εἰ μὴ νῦν) seems unnatural; cf. Ar. Vesp. 402, Dem. 3.16, cited above.Footnote ⁴¹

V. 108

In the second half of the line, the paradosis offers a very rare form of the dochmiac (⏑ ⏑⏑ ⏑⏑ — —), which would be unexceptional in itself,Footnote ⁴² but becomes problematic if taken (as I believe it must) in responsion with φιλόμαχον κράτος (⏑ ⏑⏑ — ⏑ —) in 129.Footnote ⁴³ Such freedom in dochmiac responsion—with a double resolution in 108 corresponding to a single resolution in 129 and, at the same time, a ‘drag-out’ in 108 corresponding to no drag in 129—is unparalleled in Aeschylus, as far as I can see, and otherwise extremely rare as well as attested only much later.Footnote ⁴⁴

In the app. crit. to his OCT, PageFootnote ⁴⁵ tentatively suggested emending into χθονὸς πάντες ἴτ’ ἴτ’ in order to remove what he saw as an illegitimate breuis in longo at the end of the dochmiac. However, as Hutchinson (n. 42) rightly remarks, the anadiplosis suggests that ἴτ’ ἴτε is a genuine imperative (‘come hither’) rather than an adverb of encouragement (‘go on’) with ἴδετε in the following line, and as such it guarantees the pause that is necessary for breuis in longo to occur at the end of 108.Footnote ⁴⁶ We may add that a pause is guaranteed at this place by the corresponding period-end created by the vocative in 129 σύ τ’ ὦ Διογενὲς … κράτος.Footnote ⁴⁷

Other notable emendations assume that πάντες has ousted a choicer word; hence Steusloff's ἴθ’ ἁθρόοι, or Headlam's ἴθ’ ἁλέες.Footnote ⁴⁸ As West remarked,Footnote ⁴⁹ these emendations have the merit of accounting for the aspirated ἴθ’ of a part of the paradosis. However, there may still be room for a less invasive procedure, which merely reshuffles the transmitted word order without introducing emendation. I suggest:

This produces — — ⏑⏑ ⏑ —, the same type of dochmiac as is found in 164 Ὄγκα πρὸ πόλεως; in both cases, the dochmiacs correspond to × ⏑⏑ — ⏑ — in the strophe or antistrophe.Footnote ⁵¹ For the period-end required for the breuis in longo (ἴτε̅), see above.

VI. 125–8, 146–9

At 147/8, στόνων ἀυτᾶς is probably interpolated. Apart from disturbing the metrical responsion with the strophe (126/7), the words cannot be syntactically accommodated in any meaningful way. Emendations are unconvincing: Wilamowitz'sFootnote ⁵² <κλῦθι δ’ ἀγα>στόνων ἀυτᾶς creates a feeble hysteron proteron with the preceding call to action; and Hutchinson's <ἀΐων> στόνων ἀυτᾶς entails a sequence of disagreeably redundant genitives (either στόνων or ἀυτᾶς should have been sufficient). WestFootnote ⁵³ suggested that στόνων ἀυτᾶς may be the remnant of a tragic(?) parallel cited by a scholiast in support of his interpretation of ἀπύουσαι (ἀυτοῦσαι Seidler) in 145 as στενάζουσαι; cf. Hom. Od. 11.383 στονόεσσαν ἀυτήν; Eur. Phoen. 1552 στενάχειν τάδ’, ἀυτεῖν. Alternatively, it may be that δαΐωι was wrongly parsed as δ’ ἀίω(ν), and the interpolated words were devised as a makeshift complement for ἀίων.

Further, σύ τ’ ὦ Λατογένει-|α κούρα is metrically problematic. The scansion ⏑ — — ⏑⏑ —, in which two shorts are substituted for the second single short of the normal dochmiac, is paralleled in Sept. 893/4 = 905, 935 = 949; Supp. 350 = 361.Footnote ⁵⁴ But none of these examples is in responsion to regular dochmiacs (here ⏑ — — ⏑ — in the strophe, which however is also problematic and in need of emendation, as argued below).Footnote ⁵⁵ There is also the further problem (pointed out by Parker, n. 55) that the line-initial bacchiac at 149 (–α κούρα), which is in synaphea with the preceding colon (i.e. in mid period), has word-end immediately after it, a phenomenon otherwise unparalleled. Emendation seems unavoidable: I suggest, exempli gratia, σύ θ’, ἁ Λατογενής | κόρα, σὸν τόξον (⏑ — — ⏑⏑ — | ⏑ — — — ⏑); cf. Eur. Ion 465 σὺ καὶ παῖς ἁ Λατογενής. The synaphea remains, but the word-end after the line-initial bacchiac is mitigated, because σόν here is a prepositive.Footnote ⁵⁶ For κόρα in tragic lyrics, cf. Aesch. Supp. 145; Soph. OT 508; Eur. Tro. 561. For the combination of a vocative (καὶ σὺ Λύκει’ ἄναξ) with a nominative (ἁ Λατογενής | κόρα) in addresses, cf. Aesch. Sept. 129–35 σύ τ’ ὦ Διογενὲς … κράτος | ῥυσίπολις γενοῦ | Παλλάς, ὅ θ’ ἵππιος … ἄναξ | … | ἐπίλυσιν φόβων, ἐπίλυσιν δίδου; Hom. Il. 3.276–7 Zεῦ πάτερ … | Ἠέλιός θ’; Ar. Nub. 264–5 ὦ δέσποτ’ ἄναξ, ἀμέτρητ’ Ἀήρ … | λαμπρός τ’ Αἰθήρ.Footnote ⁵⁷ For the substitution of the single short of the normal dochmiac by two shorts, see above (with n. 54).

The emendation forces us to reconsider the text of 127, which is suspect on independent grounds. πύλαις ἑβδόμαις can be made sense of only if taken as = πύλαις ἑπτά. This unique usage may perhaps be influenced by Hom. Il. 8.404 οὐδέ κεν ἐς δεκάτους … ἐνιαυτούς (‘not in the space of ten years’), where however ἐς δεκάτους ἐνιαυτούς seems to be a conflation of ἐς δέκα ἐνιαυτούς and δεκάτωι ἐνιαυτῶι.Footnote ⁵⁸ Still, it would perhaps have been easy for a later poet to assume that δέκατος simply stands for δέκα there; cf. Callim. Hymn 4.251 where ἑβδομάκις = ἑπτάκις;Footnote ⁵⁹ and the vague use of the ordinal has been comparedFootnote ⁶⁰ to Dante, Inferno 4.148 la sesta compagnia = ‘the company of six’, lit. ‘the sixth company’. When all is said and done, however, ἑβδόμαις cannot be exactly paralleled in Greek literature, and may be corrupt; contrast Sept. 714, where τάσδ’ ἐφ’ ἑβδόμαις πύλαις means, as it should, ‘this seventh gate’. No emendation readily suggests itself, but one might consider, for example, πύλαις ἑπταδύμοις,Footnote ⁶¹ or (more radically) ἑπταπύλοις στόμασιν. The former would yield perfect responsion with my emendation at 148:

VII. 158–65

In 159, the paradosis has been variously emended. As ἔρχεται demands an accusative expressing the target of the stones’ motion, Wilamowitz changed ἐπάλξεων into ἐπάλξεις.Footnote ⁶² But this is now contradicted by P.Oxy. 2179 (second century c.e.) ]ωνλιθασ[, which preserves traces of ἐπάλξε]ων. More importantly, Wilamowitz's emendation necessitates ἀκροβόλων to mean ‘skirmishers’, which seems inapposite: although slingers were of course used from an early age in siege warfare,Footnote ⁶³ it would be pedantically out of character for this chorus of young women to be concerned with which body of fighters the stones are coming from.Footnote ⁶⁴ Besides, it is wholly uncertain whether ἀκροβόλος as a military term can be dated anywhere near Aeschylus’ time: its earliest possible attestation is in IG v(1), 1426.10, a Messenian inscription of the early third century b.c.e.,Footnote ⁶⁵ where however Hiller's ἀκρ]οβόλοι is merely one possible supplement (another possibility would be, for example, λιθ]οβόλοι); otherwise, the term is attested only in lexicographers.

An alternative, suggested by Page,Footnote ⁶⁶ would be to take ἀκροβόλος δ’ (Ludwig) ἐπάλξεων λιθάς as meaning τὰ ἄκρα τῶν ἐπάλξεων βαλοῦσα λιθάς, ‘a shower of stones striking the top of the ramparts’.Footnote ⁶⁷ But the same meaning can be obtained with no change to the transmitted text, if we take (together with the ancient scholia and several modern editors)Footnote ⁶⁸ ἀκροβόλων as gen. plur. of ἀκρόβολος (pass.), ‘struck at the top’, with the recessive accentuation expressing, as usual, passive meaning (cf. λιθόβολος, ‘struck with stones’, as opposed to λιθοβόλος, ‘striking with stones’);Footnote ⁶⁹ cf. the use of ἀμφίβολος later in the play (Sept. 298–300) ἐπ’ ἀμφιβόλοισιν | ἰάπτουσι πολίταις | χερμάδ’ ὀκριόεσσαν.Footnote ⁷⁰

All in all, I see no reason to question the paradosis in ἀκροβόλων … ἐπάλξεων except only so far as to accept Butler's deletion of δ’:Footnote ⁷¹ connectives are used nowhere else in this purposely asyndetic passage (cf. 151, 153, 155, 161). The problem must lie with †ἔρχεται†: to explain ἔρχεται ἐπάλξεων along the lines of, for example, Hom. Il. 14.488 ὡρμήθη δ’ Ἀκάμαντος or 13.687 ἐπαΐσσοντα νεῶνFootnote ⁷² is to miss the important point that ἔρχομαι does not express the strongly purposeful motion of, for example, ὁρμάομαι, ἐπαΐσσω, ἰθύω, and the like; this is surely why ἔρχομαι with genitive of destination is otherwise unattested.Footnote ⁷³ In the present passage, it is surely just such a verb that ἔρχεται must have ousted: for instance, ἰθύει (cf. Hom. Il. 15.693 Ἕκτωρ ἴθυσε νεὸς κυανοπρώροιο) or ἵεται (cf. Hom. Od. 10.529 ἱέμενος ποταμοῖο ῥοάων).Footnote ⁷⁴

Line 162 (†καὶ Διόθεν†) as transmitted is unmetrical and meaningless. If 164–5 are sound, as indeed they seem to be,Footnote ⁷⁵ we must look for an emendation that discards καί in 162 (otherwise, σύ τε in 164 will be a syntactic impossibility) and provides either a vocative or an imperative to match ἄνασσ’ Ὄγκα or ἐπιρρύου respectively. Now, a vocative such as παῖ Διὸς (Westphal) ὅθεν (Van den Bergh)Footnote ⁷⁶ cannot stand on its own, without an accompanying imperative to match ἐπιρρύου below.Footnote ⁷⁷ It seems preferable to introduce an asyndetic imperative addressed to Apollo (160); for example, θὲς Διόθεν εὖ | πολεμόκραντον κτλ, ‘do arrange, by the will of Zeus, a good end of this battle, etc.’. Admittedly, the change from καί to θές may appear palaeographically implausible, but perhaps ΚΑΙ was devised as a stopgap once the sequence ΘϵϹ, in which each letter can be easily confused with one of the others, had been corrupted into some meaningless combination. For Διόθεν | … τέλος, cf. Aesch. Cho. 307–8 Διόθεν τῆιδε τελευτᾶν | ἧι τὸ δίκαιον μεταβαίνει; for Διόθεν expressing agency (‘by the will of Zeus’) rather than literal provenance (‘from Zeus’), cf. Aesch. Supp. 437 δίκαια Διόθεν κρατεῖ, ‘just dealings prevail by Zeus's will’;Footnote ⁷⁸ Pers. 101 θεόθεν γὰρ κατὰ Μοῖρ’ ἐκράτησεν, ‘by divine authority did Destiny prevail’; Cho. 39 θεόθεν ἔλακον, ‘they cried out with divine authority’.Footnote ⁷⁹ For θὲς … εὖ | … τέλος, cf. Soph. Trach. 26 τέλος δ’ ἔθηκε Zεὺς Ἀγώνιος καλῶς.Footnote ⁸⁰ For a deity asked to effect a prayer Διόθεν, ‘through the will of Zeus’, cf. Eur. Hel. 1504-5 (the Dioscuri) εὐαεῖς ἀνέμων | πέμποντες Διόθεν πνοάς. For the idea that the responsibility of arranging a favourable end lies ultimately with Zeus, cf. Aesch. Sept. 116 (ὦ Zεῦ πάτερ παντελές); Ag. 1485–7 (διαὶ Διὸς | παναιτίου πανεργέτα· | τί γὰρ βροτοῖς ἄνευ Διὸς τελεῖται;), 973 (Zεῦ τέλειε, τὰς ἐμὰς εὐχὰς τέλει); Supp. 525–6 (τελέων | τελειότατον κράτος, ὄλβιε Zεῦ), 823–4 (τί δ’ ἄνευ σέθεν θνατοῖσ<ι> τέλειόν ἐστιν;); Eum. 28 (τέλειον ὕψιστον Δία).