¹ Nisbet, R.G.M. and Hubbard, M., A Commentary on Horace Odes Book I (Oxford, 1970), 273–4Google Scholar note on 1.23 that the ode is modelled on Anacreon and ‘is not drawn directly from life’. Syndikus, H.P., Die Lyrik des Horaz. Eine Interpretation der Oden (Darmstadt, 2001 ³)Google Scholar, 1.229 associates the name ‘Chloe’ in 1.23 with other ‘fictional’ girls’ names in Horace and notes (229 n. 11) that in his choice of name Horace may have been influenced by Anacreon's use in his fawn-simile of the word νεοθηλής (PMG 408). However, it is possible that Chloe and other women like her in the Odes represent a kind of woman that Horace could have known, as Griffin, J., ‘Augustan poetry and the life of luxury’, JRS 66 (1976), 87–105, at 96–100, 102–4 arguesGoogle Scholar. Nisbet, R.G.M., ‘Pyrrha among roses: real life and poetic imagination in Augustan Rome’, JRS 77 (1987), 184–90, at 184Google Scholar, on the other hand, as a corrective to an overly literal reading, stresses that there are some unrealistic details in the poems Griffin discusses.

² Nisbet, R.G.M. and Rudd, N., A Commentary on Horace Odes Book III (Oxford, 2004), 117, 136, 317Google Scholar draw a parallel between the later Chloe odes and 1.23 only to comment on the meaning of the name, and at 3.26.9-10 comment: ‘Some editors compare 3.9.9 … ; but women in H's odes do not always have constant characteristics, and it is not certain that the earlier passage is meant to be remembered here.’ Rudd, N., Horace, Odes and Epodes (Cambridge, MA, 2004), 330 Google Scholar, in the index to his Loeb edition, under ‘Chloe’ separates ‘a girl loved by Horace’ in 1.23, 3.9 and 3.26 from ‘a woman in Oricus’ in 3.7. Quinn, K., Horace The Odes (Bristol, 1980), 262 Google Scholar similarly states at 3.9.9-12 that in that ode Chloe is called ‘Thracian’ ‘perhaps … to distinguish her from the Chloe of 3.7’. Davis, G., Polyhymnia: The Rhetoric of Horatian Lyric Discourse (Berkeley, 1991)Google Scholar, 253 n. 38 implies that the Chloes of 3.7 and 1.23, although sharing a significant name, are not the same girl.

³ Several scholars have linked two or more of the ‘Chloes’. Dettmer, H., Horace: A Study in Structure (Hildesheim, 1983), 198–9Google Scholar argues that 1.23 and 3.9 are linked by the fact that ‘in both Chloe is the object of the poet's affections’, and that, beyond a discernible narrative structure, ‘the thematic bond’ of the exclusus amator ‘among Odes 3.7, 3.9, and 3.26 lends further support to identifying Chloe as one and the same character’. Santirocco, M.S., Unity and Design in Horace's Odes (Chapel Hill, 1986), 144 Google Scholar connects the Chloe of 3.9 and 3.26. Watson, L., ‘Horace Odes 1.23 and 1.25: a thematic pairing?’, AUMLA 82 (1994), 67–84 Google Scholar, at 77 connects the Chloe and the Lydia of 1.23 and 1.25, respectively, to the women of the same names, ‘who, interestingly, recur as rivals for the affections of an unnamed amator at Odes 3.9.6-7’, although he hesitates (84 n. 75) to assume definitely that they are the same women from the previous odes. Doyen, C., ‘ Lydia, Glycera, Chloe: Analyse d'une triade feminine dans les Odes d'Horace’, LEC 72 (2004), 313–32, at 314–18Google Scholar argues that all the Chloes are the same woman through geographical references in the four odes in which she appears, all of which, he argues, can be connected to Thrace (the silua of 1.23, the fact that Gyges might be on his way to the Black Sea in 3.7, the adjectives Thressa and flaua in 3.9.9 and 3.9.19, and the reference to ‘Sithonian’, i.e. Thracian, snow in 3.26.10). Nadeau, Y., Erotica for Caesar Augustus. A Study of the Love-Poetry of Horace, Carmina, Books I to III (Brussels, 2008), 267, 268, 278, 404–5Google Scholar assumes all four mentions of a ‘Chloe’ to be references to the same person. Various other recurring women in the Odes have received similar attention. Dettmer (this note), 328–9, 432–45 considers Lydia, Glycera, Lalage and Lyde, in addition to Chloe, to be the same characters throughout the odes. Johnson, T.S., ‘Locking-in and locking-out Lydia: lyric form and power in Horace's C. I.25 and III.9’, CJ 99 (2003), 113–34, at 126Google Scholar takes the four instances of the name ‘Lydia’ to be linked. Doyen (this note), 314–23 argues that Lydia and Glycera, as well as Chloe, are the same women throughout the Odes and that the set of odes about these women form a coherent and structured set. Cf. Ligurinus in 4.1 and 4.10, who is taken to be the same youth (see e.g. Thomas, R. [ed.], Horace Odes Book IV and Carmen Saeculare [Cambridge, 2011], 214 Google Scholar).

⁴ D.R. Shackleton Bailey, Q. Horati Flacci Opera (Munich and Leipzig, 2001⁴) prints nam seu mobilibus ue<p>ris inhorruit | ad uentum foliis, accepting the conjectures of Gogau, Salmasius, and Bentley (uepris) and Muretus (ad uentum). On my translation of the MS reading, see n. 7 below.

⁵ LSJ s.v.

⁶ Ancona, R., Time and the Erotic in Horace's Odes (Durham, NC, 1994), 72 Google Scholar: ‘Chloe's name (Greek for “green bud” or “shoot”), introduced earlier in the poem, has prepared the reader to associate Chloe with spring.’

⁷ On the text, see n. 4 above. Commager, S., The Odes of Horace (Bloomington, 1962), 238 Google Scholar, defending the MS reading, points out that emending the text ‘to uepris … ad uentum ignores the Ode's controlling metaphor, which is a seasonal one’. My translation attempts to overcome one difficulty some scholars have felt with the stanza, that the arrival of spring does not cause leaves to rustle because leaves in early spring are not large enough to flutter: Bentley, R., Q. Horatius Flaccus (Cambridge, 1711)Google Scholar, ad loc. points out that there are not really leaves at all at the beginning of spring. In his review of Shackleton Bailey's Teubner, Nisbet, R.G.M., ‘A rival Teubner Horace’, CR 36 (1986), 227–34, at 229Google Scholar states that he is ‘reluctantly persuaded by Bentley's argument that the season is not the spring’. In the beginning of spring, however, the leaf-buds (and the plant shoots from the ground suggested by Chloe's name) are more bristly than fluttering, and it can truly be said that the arrival of the season causes the trees to bud with daily-changing leaves; therefore, with a slight alteration of the usual translations of inhorruit and mobilibus, we get ‘the arrival of spring bristles with changing foliage’. Ancona (n. 6), 72 remarks that through inhorreo (‘shudder, bristle, stand on end/become erect’) ‘the poet evokes the symptoms of (Chloe's) physical arousal’; Bannon, C.J., ‘Erotic brambles and the text of Horace Carmen 1.23.5-6’, CPh 88 (1993), 220–2Google Scholar also argues for the MS reading based on an erotic interpretation of this stanza. Renehan, R., ‘Shackleton Bailey and the editing of Latin poetry: a Latin classic’, CR 83 (1988), 311–28, at 321–2Google Scholar points out that ueris aduentus need not mean the very beginning of spring, but could indicate merely that spring has arrived. West, D., Horace Odes I: Carpe Diem (Oxford, 1995), 108–9Google Scholar and Syndikus (n. 1), 1.228-9, as well as Mayer, R., Horace Odes Book I (Cambridge, 2012)Google Scholar, ad loc. are moved by different arguments than the one I am opposing here to accept uepris … ad uentum.

⁸ A less usual word than Virgil's matura (iam matura uiro, plenus iam nubilibus annis, Aen. 7.53).

⁹ Cf. Fredricksmeyer, E.A., ‘Horace's Chloe (Odes 1.23): inamorata or victim?’, CJ 89 (1994), 251–9, at 254Google Scholar: ‘Because for the fawn the experience is strange … she is fearful of what in fact is altogether harmless.’

¹⁰ Arist. Hist. an. 578b12-18, Plin. HN 8.112; Renehan (n. 7), 322–3, like other scholars, takes Horace to mean a newly-born fawn and the reference to a fawn in spring to be a zoological mistake, a not uncommon occurrence in poets. In fact, however, the near or recent maturity of the fawn is another point of similarity with Chloe and makes the simile all the more fitting.

¹¹ As to whether the lacertae are specifically metaphorical for penes (see Nadeau, Y., ‘Aenigma, an eloquens structura? Hor. c. 1.23 (uitas inuleo)’, Latomus 46 [1987], 778–80, at 779–80Google Scholar; Ancona, R., ‘The subterfuge of reason: Horace Odes 1.23 and the construction of male desire’, Helios 16 [1989], 49–57, at 53Google Scholar; Bannon [n. 7], 221; Ronnick, M.V., ‘Green lizards in Horace: lacertae uirides in Odes 1.23’, Phoenix 47 [1993], 155–7, at 156CrossRefGoogle Scholar; Fredricksmeyer [n. 9], 251–9, 255–6; contra Mayer [n. 7], ad loc.), it matters not; Horace's point is that the love he offers is more like leaves and lizards than ferocious beasts. As to the seasonableness of lizards in the ode, the modern Lacerta viridis begins its hibernation in September and emerges in April (cf. Ronnick [this note], 156).

¹² E.g. Prop. 4.7.6; Ov. Am. 3.5.42, Pont. 1.7, 10.14.

¹³ The hospita (9) is the host's wife rather than a landlady: West, D., Horace Odes III: Dulce Periculum (Oxford, 2002), 74 Google Scholar; Nisbet and Rudd (n. 2), 117. Cf. the use of hospita at 1.15.2 to describe Helen's role as the wife of Paris’ host Menelaus at the time when Paris absconded with her (a situation not unlike the one Chloe here seems to desire).

¹⁴ Elsewhere in Latin poetry women originally located in Rome go abroad to follow a man: Gallus’ Lycoris is away with a soldier in the frozen north at Verg. Ecl. 10.22-3, 10.46-9; Propertius (1.8.20) imagines Cynthia, having set sail with another man, putting in at Oricus/Oricum. Beyond being a safe harbour just across the Adriatic from Italy, Oricus/Oricum, as a strategic military location, played a role in the civil war between Caesar and Pompey (Caes. BCiv. 3.11-12, 16, 34, 39).

¹⁵ Syndikus (n. 1), 2.94 points out that these dire exempla, given Chloe's and Gyges’ actual situation, are an exaggeration and a parody of the way elegy uses series of mythological exempla.

¹⁶ Nisbet and Rudd (n. 2), 137.

¹⁷ Cf. the lyre-playing and singing of women at 1.17.17-20, 3.28.9-16.

¹⁸ The anonymous referee for CQ suggests that this passage might be read as implying that Chloe is learned in Horace's dulcis modos, i.e. docta in the Odes’ literary doctrina. This might imply that she is to be identified with the previous ‘Chloes’. Also, Chloe's (temporary) success in love here, in contrast to her failure in 3.7, could also suggest that she has been paying attention to the themes of Horace's lyric poetry: in 3.7 her overly direct approach and threatening manner marked her as an inexperienced seductress, but perhaps she has now learned from the theme in the Odes of the attractiveness of the hard-to-get (i.e. coy: 1.9.23-4, 2.12.25-8) and the hard-to-keep (i.e. faithless: 2.8.5-8). Here in 3.9 Lydia is using to her advantage the fact that a potential lover often wants what he cannot have (note her insistence on her faithfulness to Calais at lines 14–16 and her praise of him at line 21). West, D., ‘Reading the metre in Horace, Odes 3.9’, in Harrison, S.J. (ed.), Homage to Horace (Oxford, 1995), 100–7, at 106 and (n. 13), 90–1Google Scholar points out that the metre and the word accentuation of the poem reveal that the line boasting about Chloe's musical ability is actually musically uninteresting in comparison with Lydia's stanzas; that is, they suggest that she is not so ‘skilled’. In fact, West (n. 13), 91 sees a ‘naïveté in the sound of line 10’, which would belie the words docta and sciens. That Chloe is not quite as skilled in love affairs as she could be is perhaps precisely Horace's point, as I shall argue below.

¹⁹ Shackleton Bailey (n. 4) prints Peerlkamp's reiecto for the MS reiectae, in which case Horace is the ‘rejected’ one and Lydiae is genitive. West (n. 18), 103 defends the MS reading on the grounds that the attitudes of Lydia and Horace throughout the poem suggest that Horace was originally the one who ended their relationship. As to whether reiectae Lydiae (keeping the MS reading) is genitive or dative, dative makes the best sense (cf. Nisbet and Rudd [n. 2], ad loc.), although some scholars have seen here a deliberate ambiguity: Kiessling, A. and Heinze, R., Q. Horatius Flaccus I. Oden und Epoden. Erklärt von Adolf Kiessling (Berlin, 1955 ⁸)Google Scholar, ad loc.; Wimmel, W., ‘Doppelsinnige Formulierung bei Horaz?’, Glotta 20 (1962), 119–43, at 124–7Google Scholar. Wimmel (this note), 125 points out that it is the man who is usually shut out from the house, not the woman; this reversal will be discussed below.

²⁰ Horace's question (quid si … ?, 17) is generally understood to be a proposal of reconciliation to Lydia: e.g. Wickham, E.C., The Works of Horace (Oxford, 1877), 1.222Google Scholar; Kiessling and Heinze (n. 19), ad loc.; Syndikus (n. 1), 2.104; Nisbet and Rudd (n. 2), ad loc. Cf. West (n. 18), 103: ‘If he is now offering to kick out Chloe, surely he is suggesting that he had once kicked out Lydia.’

²¹ Cf. Johnson, T.S., ‘Horace's elegiac criticism and the open-ended door (C. III.10)’, CJ 107 (2011), 165–88, at 174–5Google Scholar.

²² Jones, C.P., ‘ Tange Chloen semel arrogantem ’, HSPh 75 (1971), 81–3Google Scholar; Nisbet and Rudd (n. 2), 310–11.

²³ Wickham (n. 20), 243; Nisbet and Rudd (n. 2), 316.

²⁴ Capra, the constellation mentioned at 3.7.6, rises in early September (Col. Rust. 11.2.63).

²⁵ E.g. Verg. G. 1.73 (flaua farra), 96 (flaua Ceres), 316 (flauis aruis). Cf. Nisbet and Rudd (n. 2), 139: ‘As “Chloe” in Greek describes green vegetation, there seems to be an oxymoron with flaua, which suits ripe corn.’

²⁶ E.g. 3.25.9-11 Euhias | Hebrum prospiciens et niue candidam | Thracen; 1.25.11-12 Thracio bacchante … uento; cf. Ov. Am. 1.14.21 Threcia Bacche.

²⁷ Bacchus as a god of the autumn harvest: e.g. Verg. G. 2.4-8; Ov. Met. 8.273-4. Horace may evoke contrasting seasons by opposing harvest-time Chloe to Lydia, with whom he suggests he might be yoked (iugo cogit aeneo, 3.9.18) as if for early-spring ploughing (Verg. G. 1.43). Thrace and maenads also evoke the changing seasons in 1.25.11-12, where the increase of a bacchante-like ‘Thracian wind’ (Thracio bacchante magis … uento) is tied to the approach of winter and the eventual approach of Lydia's old age, the time when she will no longer be attractive to young men.

²⁸ Many thanks to Jenny Strauss Clay, Courtney Evans, A.J. Woodman and the anonymous referee for their suggestions and improvements on this article in its various stages of development.