¹ On Horace's alphabet games: Gowers, E., ‘The ends of the beginning’, in Houghton, L.B.T. and Wyke, M. (edd.), Perceptions of Horace: A Roman Poet and His Readers (Cambridge, 2009), 39–60, at 44–6Google Scholar; ead., Horace: Satires Book 1 (Cambridge, 2012)Google Scholar, on Sat. 1.1.25–6, 4.56–62.

² Fredericksen, E., ‘When enough is enough: an unnoticed telestich in Horace (Satires 1.4.14–18)’, CQ 68 (2018), 716–20CrossRefGoogle Scholar—an article I only saw as this piece went to press—persuasively argues for telestich SATIS, a brilliant example of Horatian wordplay in the Satires.

³ Thomas, R.F., Horace: Odes Book 4 and Carmen Saeculare (Cambridge, 2011)Google Scholar, ad loc., citing John Henderson per litteras.

⁴ The best acrostic found in Horace's Odes (= Carm.) so far is the imperative DISCE at Carm. 1.18.11–15 (Morgan, G., ‘Nullum, Vare … chance or choice in Odes 1.18?’, Philologus 137 [1993], 142–5CrossRefGoogle Scholar), picking up and modifying discernunt (Carm. 1.18.11) as a gamma-acrostic, and adding suitably didactic authority to Horace's revelations and reconcealments of Bacchic mysteries. (Does candideBassareu at Carm. 1.18.11 mimic Homer's unintentional ΛΕΥΚΗ at Il. 24.1–4, via Aratus’ ΛΕΠΤΗ, Phaen. 783–7?) Less persuasively, Adkin, N., ‘Quis est nam ludus in undis? (Virgil, Eclogue 9 39–43)’, Acta Classica 51 (2015), 43–58, at 49 nGoogle Scholar. 39 sees SAPIS at Carm. 2.10.9–13, reinforcing sapienter (2.10.22) but otherwise with little connection to the poem's contents.

⁵ Some pre-Horatian examples: Cicero cites Ennius’ Q. ENNIVS FECIT (not attested in surviving fragments) and the acrostics in the Sibylline Books, highlighting problems of intentionality (Diu. 2.111–12); Varro apparently commented on Sibylline acrostics to much the same effect (Dion. Hal. Ant. Rom. 4.62.6). Kronenberg, L.J., ‘The light side of the moon: a Lucretian acrostic (LUCE, 5.712–15) and its relationship to acrostics in Homer (LEUKĒ, Il. 24.1–5) and Aratus (LEPTĒ, Phaen. 783–87)’, CPh 114 (2019), 278–92Google Scholar finds an acrostic in Lucretius to match his interest in letterplay. On Latin acrostics generally: Damschen, G., ‘Das lateinische Akrostichon: neue Funde bei Ovid sowie Vergil, Grattius, Manilius und Silius Italicus’, Philologus 14 (2004), 88–115Google Scholar; on Greek: Luz, C., Technopaignia: Formspiele in der griechischen Dichtung (Leiden, 2010)CrossRefGoogle Scholar. A strong case for systematic allusive acrostic-poetics in Augustan poetry has been made by M. Robinson in two recent articles, ‘Arms and a mouse: approaching acrostics in Ovid and Vergil’, MD 82 (2019a), 23–73Google Scholar and ‘Looking edgeways: pursuing acrostics in Ovid and Vergil’, CQ 69 (2019b), 290–308CrossRefGoogle Scholar.

⁶ Gowers (n. 1 [2012]), 61.

⁷ The search for Virgilian acrostics continues unabated. I am particularly fond of MARS at Aen. 7.601–4, hesitantly proffered by Fowler, D.P., ‘An acrostic in Vergil (Aeneid 7.601–4)?’, CQ 33 (1983), 298CrossRefGoogle Scholar and now almost universally accepted. Also persuasively, Grishin, A., ‘Ludus in undis: an acrostic in Eclogue 9’, HSPh 104 (2008), 237–40Google Scholar finds VNDIS at Ecl. 9.34–8; Clauss, J.J., ‘An acrostic in Vergil (Eclogues 1.5–8): the chance that mimics choice’, Aevum(ant) 10 (1997), 267–87Google Scholar argues for FONS at Ecl. 1.5–8; and Brown, E.L., Numeri Virgiliani: Studies in Eclogues and Georgics (Brussels, 1963), 96–114Google Scholar first located the controversial signature Ma-Ve-Pu at G. 1.429–33—on which, see now Haslam, M., ‘Hidden signs: Aratus Diosemeiai 46ff., Vergil Georgics 1.424ff.’, HSPh 94 (1992), 199–204Google Scholar; Feeney, D. and Nelis, D., ‘Two Virgilian acrostics: certissima signa?’, CQ 55 (2005), 644–6CrossRefGoogle Scholar; and Somerville, T., ‘Note on a reversed acrostic in Vergil Georgics 1.429–33’, CPh 105 (2010), 202–9Google Scholar. The most level-headed account to date of Virgilian acrostic scholarship is given by Katz, J.T., ‘The Muse at play: an introduction’, in Kwapisz, J., Petrain, D. and Szymański, M. (edd.), The Muse at Play: Riddles and Wordplay in Greek and Latin poetry (Berlin, 2013), 1–30, at 4–10Google Scholar, and updated in id., ‘Another Vergilian signature in the Georgics?’, in Mitsis, P. and Ziogas, I. (edd.), Wordplay and Powerplay in Latin Poetry (Berlin, 2016), 69–85, at 72 nCrossRefGoogle Scholar. 9.

⁸ Both of these acrostics are listed in the scrapheap of ‘accidental acrostics’ piled up by Hilberg, I., ‘Ist die Ilias Latina von einem Italicus verfasst oder einem Italicus gewidmet?’, WS 21 (1899), 264–305, at 286 and 295Google Scholar, but neither has yet been rehabilitated in modern scholarship. Hilberg identifies other acrostics in the Satires, most of which do seem truly unintentional. I suspect, however, that it is no accident that the acrostic ADDI in the last four lines of Satires Book 1 (1.10.89–92) not only echoes the wickedly ironic ‘false closure’ of 1.1.121 (uerbum non amplius addam; cf. Gowers [n. 1 (2012)], 46) but also coincides with Horace's instructions to the puer to make a last-minute addition to the collection (i, puer, atque meo citus haec subscribe libello, 92); cf. Gowers (n. 1 [2012]), ad loc. MVS at 2.1.54–6 paradoxically encompasses large-scale lupus and bos (55) and anticipates the mures of 2.6.79–117—cf. Robinson (n. 5 [2019a]), on MVS at Ov. Met. 1.14–16—and both HIC (1.5.30–2) and NIL (2.5.69–71) are neat gamma-acrostics—but proving three-letter acrostics is a mug's game: cf. Hendry, M., ‘A Martial acronym in Ennius’, LCM 19 (1994), 7–8, at 7Google Scholar.

⁹ quidam notus mihi nomine tantum (3), an evasive identifier that Gowers, E., ‘Fragments of autobiography in Horace, Satires 1’, ClAnt 22 (2003), 55–91, at 59Google Scholar calls ‘the best joke of Satires 1’. Is the pest an anonymous acquaintance, or a wholly imaginary character? Oliensis, E., Horace and the Rhetoric of Authority (Cambridge, 1998), 36CrossRefGoogle Scholar takes him for Horace's ‘distorted double’; Henderson, J., ‘Be alert (your country needs lerts): Horace, Satires 1.9’, PCPhS 39 (1993), 67–93, at 70Google Scholar sees him as a version of his younger, more socially clumsy self, recalled from 1.3.63–6 (cf. Gowers [this note], 86 and 61–7, reading quo tendis? unde uenis? [1.9.62] as a biographical query). Ferriss-Hill, J.L., ‘A stroll with Lucilius: Horace, Satires 1.9 reconsidered’, AJPh 132 (2011), 429–55Google Scholar argues that the pest is Horace's satiric forefather Lucilius, and various other scholars maintain that he is one of Horace's poetic peers (listed by Ferriss-Hill [this note], 429 n. 3); none of these concrete identifications seems persuasive to me.

¹⁰ Hierarchical ‘patronage’ is subsumed under a veneer of ‘friendship’. Cf. Konstan, D., ‘Patrons and friends’, CPh 90 (1995), 328–42Google Scholar and Bowditch, P.L., Horace and the Gift Economy of Patronage (Berkeley, 2001), 19–26Google Scholar on the language of amicitia denoting affectionate personal friendship as well as more formal patronage structures. D. Armstrong, ‘Utility and affection in Epicurean friendship: Philodemus On the Gods 3, On Property Management, and Horace Sermones 2.6’, in Caston, R.R. and Kaster, R.A., Hope, Joy, and Affection in the Classical World (Oxford, 2016), 182–208, at 202–8CrossRefGoogle Scholar explores the Epicureanism of Horatian amicitia in the Satires.

¹¹ It is impossible to prove beyond doubt that an acrostic is deliberate. Commonly accepted factors include the length of the acrostic (usually four letters or more, following Hilberg [n. 8], 266), its relevance to the poetic context, and the presence of self-conscious instructions for the reader to ‘discover’ the acrostic. Trzaskoma, S.M., ‘Further possibilities regarding the acrostic at Aratus 783–7’, CQ 66 (2016), 785–90, at 786CrossRefGoogle Scholar notes ‘mild irony inasmuch as we are using the solution to elucidate the clues rather than the other way round’. Cf. Bing, P., ‘A pun on Aratus’ name in verse 2 of the Phainomena’, HSPh 93 (1990), 281–5, at 282Google Scholar; Clauss (n. 7), 267; Haslam (n. 7), 203 n. 12; Robinson (n. 5 [2019a]).

¹² notus (3), noris (7), notum (17) and add too ignosces (72); cf. Gowers (n. 1 [2012]), ad loc. On metapoetic signals or instructions for acrostic-reading, cf. (e.g.) Feeney and Nelis (n. 7); Danielewicz, J., ‘Vergil's certissima signa reinterpreted: the Aratean LEPTE-acrostic in Georgics 1’, Eos 100 (2013), 287–95Google Scholar; Trzaskoma (n. 11), Robinson (n. 5 [2019a]). Such signs commonly revolve around sight (e.g. σκέπτεο: Aratus, Phaen. 778; respicies: Verg. G. 1.425); Horace instead integrates his acrostic into the Satires’ obsession with recognition, comprehension, knowledge and notoriety, traced through the collection with notare as well as noscere: cf. Gowers (n. 1 [2012]), on Sat. 3.24.

¹³ membrum commonly denotes a small constituent part of a composite whole—in rhetoric and criticism, usually a clause or colon (TLL 8.645.19–30; cf. Horace's famous disiecti membra poetae, 1.4.62), but an extension to alphabet letters (along the lines of Lucretian elementa) is not unfeasible. Whether the question refers to physical or linguistic agility, the unspoken answer is, of course, Horace himself.

¹⁴ Cf., for example, Catull. 68.98, Prop. 2.24.35, Tib. 3.2.26, Ov. Met. 4.157. Tibullus’ use, like Horace's, retains metapoetic overtones: sic ego componi uersus in ossa uelim, followed by his projected epitaph (cf. littera 27, carmina 28), perhaps implies not only ‘to be buried, turned into bones’ but also ‘to be composed as a line of verse’ (cf., for example, Petron. Sat. 127.10).

¹⁵ Henderson (n. 9), 74 notes poetic terminology and a textual turn here: ‘his abjection of ille into another indirect lexis is at once seized on and dismantled (est tibi … te … est opus ~ haud mihi … composui)’. Welch, T.S., ‘Horace's journey through Arcadia’, TAPhA 138 (2008), 47–74, at 60–1Google Scholar compares Verg. Ecl. 1.23: sic paruis componere magna solebam. Oliensis (n. 9), 219 renders omnes composui as ‘I've versified them all’.

¹⁶ TLL 3.2117.36 s.v. II.A.1.b.

¹⁷ Gowers (n. 1 [2012]), ad loc. translates it as ‘finish me off’; Oliensis (n. 9), 219 as ‘kill me’ or ‘write me’—but the lack of an object in the Latin lends ambiguity to the command.

¹⁸ amicus amico: Plaut. Curc. 332, Mil. 660; Ter. Phorm. 562; Accius, Deiphobus fr. 4 Ribbeck; Petron. Sat. 43 and 44. Wills, J., Repetition in Latin Poetry: Figures of Allusion (Oxford, 1966), 205–6Google Scholar notes that this structure, illustrating ‘the reciprocal sense of polyptoton’, is as suited to the perversities of civil war as to friendship.

¹⁹ Reciprocity: cf., for example, Saller, R.P., Personal Patronage under the Early Empire (Cambridge, 1982), iCrossRefGoogle Scholar; Konstan (n. 10), 328. Henderson (n. 9), 80 offers the inverse: ‘you scratch my back, I'll knife yours’.

²⁰ So already Cicero: hoc scriptoris est, non furentis (Diu. 2.112).

²¹ Gowers (n. 1 [2012]), passim on the interplay between Horace's ‘rough exterior’ and concealed sophistication. Cf. Henderson (n. 9), 79; Gowers, E., ‘The restless companion: Horace, Satires 1 and 2’, in Freudenburg, K. (ed.), The Cambridge Companion to Roman Satire (Cambridge, 2005), 48–61, at 49CrossRefGoogle Scholar.

²² Cf. Gowers (n. 1 [2012]), 14 on satire between oral and written; Henderson (n. 9), 78–9 on Horace as storyteller. I wonder if the acrostic's distribution over both Horace's and the pest's speech brings about a sort of coerced reciprocity, forcing the sparring pair to collaborate on its production.

²³ Zetzel, J.E.G., ‘Dreaming about Quirinus: Horace's Satires and the development of Augustan poetry’, in Feeney, D. and Woodman, T. (edd.), Traditions and Contexts in the Poetry of Horace (Cambridge, 2002), 38–52CrossRefGoogle Scholar; Thomas, R.F., ‘Horace and Hellenistic poetry’, in Harrison, S., The Cambridge Companion to Horace (Cambridge, 2007), 50–62CrossRefGoogle Scholar; Tarrant, R., ‘Horace and Roman literary history’, in Harrison, S., The Cambridge Companion to Horace (Cambridge, 2007), 63–76, at 66–71; Gowers (n. 1 [2012]), 305–6Google Scholar.

²⁴ On Horace's politics of literary revision: Gurd, S., Work in Progress: Literary Revision as Social Performance in Ancient Rome (Oxford, 2012), 77–103Google Scholar.

²⁵ The phrase is from Tarrant (n. 23), 71. Cf. Freudenburg, K., The Walking Muse: Horace on the Theory of Satire (Princeton, 1993), 44Google Scholar: Horace's new poetry in the Satires ‘must depart from cliché-Callimacheanism (i.e. by constructing Callimacheanism as cliché) in some real sense’.

²⁶ Cf. especially Henderson (n. 9), with Henderson, J., Writing down Rome: Satire, Comedy, and Other Offences in Latin Poetry (Oxford, 1999), 202Google Scholar: ‘this text, besides, composes a poetic (of “anti-aesthetic” plebeian vigour adapting Callimachean angularity and litotes, not without paradox)’.

²⁷ As Gowers comments (n. 1 [2012] on Sat. 1.9.22–3), the pest has apparently been paying close attention to Horace—knowing that Horace has strong opinions on the quantity and speed of poetic composition, on the celebrity Hermogenes Tigellius, on showy dancing and singing—but has failed to spot that all these opinions are negative. At 1.4.9–21 and 1.10.9–10 Horace emphatically denounces outpourings of lengthy poetry (and note garrulus, 1.4.12 ~ 1.9.33, with garriret, 1.9.13). Hermogenes is a favourite punching-bag, Horace's despised ‘artistic antitype’ (Gowers [n. 1 (2012)], on Sat. 1.2.3; cf. on Sat. 1.2.1–3; 1.3.3–19, 129–30; 1.4.72; 1.10.17–19, 80, 90). The all-singing, all-dancing pest unknowingly convicts himself of suspicious effeminacy and tastelessness (cf. Gowers [n. 1 (2012)] on Sat. 1.9.24–5) and further aligns himself with the grotesque Hermogenes (cf. 1.2.1–3).

²⁸ Lucilius’ sloppy sermones are the target of Horace's most stringently Callimachean criticisms (cf. 1.4.6–13, 1.10.50–71).

²⁹ Three consecutive lines (57–9), for example, all feature a so-called ‘Catullan molossus’ (on which, see Ross, D.O., Virgil's Aeneid: A Reader's Guide [Oxford, 2007], 151–2)CrossRefGoogle Scholar, in all three cases a decisive future verb (corrumpam … desistam … deducam). Cf. Freudenburg (n. 25), 209–10: ‘the “bore” is, after all, a Neoteric’.

³⁰ On Horace's distaste for the sterility and banality of recitation, cf. Gowers (n. 1 [2012]) on Sat. 1.10.17–19, 19. Cucchiarelli, A., La satira e il poeta: Orazio tra Epodi e Sermones (Pisa, 2001), 78 n. 76Google Scholar suggests that the simius and the pest may even be one and the same.

³¹ Cf. the doctores inepti of Satires Book 2 (Freudenburg [n. 25], 47): Damasippus (2.3), Catius (2.4; note docte Cati, 2.4.88), Davus (2.7). Freudenburg (n. 25), 17 sees satire itself as a parody of learning.

³² mille … uersus: cf. Lucilius (Hor. Sat. 1.4.9–10: in hora saepe ducentos … uersus dictabat) and the pest (1.9.23–4). deduci (2.1.4) perhaps adds an element of Callimachean fineness (cf. Ecl. 6.5 deductum), pace F. Muecke, Horace Satires II, with an Introduction, Translation and Commentary (Warminster, 1993), ad loc.

³³ Callimachus famously praised the Phaenomena as ‘a token of Aratus’ vigil’ (Ἀρήτου σύμβολον ἀγρυπνίης, Epigr. 27.4 Pf.) in a poem which seems to respond to Aratus’ ΛΕΠΤΗ acrostic from Phaen. 783–7 (note λεπταὶ ῥήσιες, 3–4). Horace here aligns himself with Aratus in his wakefulness, and so points to the approaching acrostic. (I owe this point to Thomas J. Nelson.)

³⁴ We know from Cicero that Trebatius was a keen swimmer (Fam. 7.10), and that he enjoyed a drink or two (Fam. 7.22). Trebatius’ recommendation of imperial panegyric (Sat. 2.1.10–12) is perhaps a response to the praise he apparently received from Julius Caesar for his own legal skill (Fam. 7.10: te Caesari nostro ualde iure consultum uideri).

³⁵ J.A. Simon, Akrosticha bei den augustischen [sic] Dichtern: Exoterische Studien: Zweiter Teil, mit einem Anhang: akrostichische und telestichische Texte aus der Zeit von Plautus bis auf Crestien Von Troies und Wolfram Von Eschenbach (Cologne and Leipzig, 1899), 164–5 finds elaborate nonsensical telestichs in Sat. 2.1 (reverse telestich MEO O INSIM DA, 2–11; telestich SIMOIM SVMPTOS M MAEMM, 14–31; reverse telestich SESST MI MIME, 76–86), but misses OTIA.

³⁶ Though contrast 1.4.138–9: ubi quid datur oti, | illudo chartis …

³⁷ TLL s.v. I.1.a.β; cf. Muecke (n. 32), ad loc.

³⁸ This time his command precedes the acrostic (compare confice, 1.9.29).

³⁹ The stuttering satirist is familiar from the previous book of Satires: singultim pauca locutus, 1.6.56 (cf. Balbinum 1.3.40, balbutit 1.3.48). Aaron Kachuck points out to me that in Latin capital and uncial scripts ‘L’ and ‘I’ (and ‘T’) are very similar, and are often confused for one another; ALTO therefore very nearly encodes OTIA in reverse, when read backwards from the end of line 8. This perhaps explains why it is sleep that is ‘deep’ here, not the Tiber (compare Verg. G. 4.560–1: altum … Euphraten, discussed below): Tiberim … altum would not give the same slantwise mirror-image of the vertical acrostic.

⁴⁰ The telestich was first reported by Schmidt, W., Vergil-Probleme (Göppingen, 1983), 317–18Google Scholar.

⁴¹ Cf. Katz (n. 7 [2016]), with Grishin (n. 7) on Ecl. 9.34–8.

⁴² Cf. Carter, M.A.S., ‘Vergilium vestigare: Aeneid 12.587–8’, CQ 52 (2002), 615–17, at 616 n. 4CrossRefGoogle Scholar. Clauss, J.J., ‘Once upon a time on Cos: a banquet with Pan on the side in Theocritus Idyll 7’, HSPh 101 (2003), 289–302, at 298 nGoogle Scholar. 36 suggests that the Georgics telestich OTIA confirms his FONS acrostic at Ecl. 1.5–8 (Clauss [n. 7]), since otia in Ecl. 1.6 occurs within the bounds of the acrostic. Regarding the telestich of the sphragis, I would add that, while the Caesar reference does identify Octavian as the god who creates otium (otia fecit, Ecl. 1.6), it is still the poet who creates OTIA in the form of the telestich here—a typical tug-of-war of creative authority between poet and emperor (cf. perhaps certissimus auctor, G. 1.432?).

⁴³ Gowers (n. 1 [2012]), 22: the relationship between Satires Book 1 and the Eclogues is ‘sometimes neatly arithmetic, sometimes pointedly oppositional’.

⁴⁴ Robinson, M., ‘Augustan responses to the Aeneid’, in Clarke, M.J., Currie, B. and Lyne, R.O.A.M. (edd.), Epic Interactions: Perspectives on Homer, Virgil and the Epic Tradition: Presented to Jasper Griffin by Former Pupils (Oxford, 2006), 185–216, at 190CrossRefGoogle Scholar observes that audacia frames the Georgics in reference to Virgil's poetic endeavours (G. 1.40 audacibus adnue coeptis, 4.565 audaxque iuuenta), a feature Horace perhaps recognizes when he repeats audax … audax at Carm. 1.3.25–7. Robinson (n. 5 [2019b]), 298–300 argues for acrostic AVSVM at Aen. 7.178–82; Virgil here develops not only the use of audere by Horace and himself as a marker of Virgilian poetics but also the use of aude/audax in acrostic interchange that I identify in the OTIA acrostic/telestich.

⁴⁵ Maro produces six Latin anagrams, all of relevance to Virgil's works: amor, Roma, ramo, armo, oram, mora. The most famous pair is, of course, amor/Roma, which appears as a suppressed pun in Virgil's hic amor, haec patria est (Aen. 4.347); cf. Hanses, M., ‘Love's letters: an amor-Roma telestich at Ovid, Ars Amatoria 3.507–10’, in Mitsis, P. and Ziogas, I. (edd.), Wordplay and Powerplay in Latin Poetry (Berlin, 2016), 199–211Google Scholar and Reed, J., ‘Mora in the Aeneid’, in Mitsis, P. and Ziogas, I. (edd.), Wordplay and Powerplay in Latin Poetry (Berlin, 2016), 88–105Google Scholar on mora/amor/Roma throughout the Aeneid. There are no explicit examples of Maro/amor in antiquity—for that we must wait for Herbert's virtuosic poem, Lucus 25, in the 1620s—but it surely underpins Virgil's own anagrammatic improvisations on these four letters. Hanses ([this note], 208) suggests that Ovid's amor/Roma palindrome-cum-telestich at Ars am. 3.507–10 suppresses a nod to Virgil; Malamud, M., ‘Gnawing at the end of the rope: poets on the field in two Vergilian catalogues’, Ramus 27 (1998), 95–126, at 116CrossRefGoogle Scholar finds an (overly?) elaborate bilingual pun-anagram of Maro in Ocnus (Aen. 10.198); Virgil's MARS acrostic (n. 7 above) misspells his cognomen by a single letter: coincidence or (attempted) signature?

⁴⁶ Horatian scholarship more frequently compares the opening of Sat. 2.1 to Virgil's recusatio of Augustan panegyric in G. 3.1–48: e.g. Fraenkel, E., Horace (Oxford, 1957), 149 n. 2Google Scholar; Muecke (n. 32), on Sat. 2.1.11; Freudenburg, K., Satires of Rome: Threatening Poses from Lucilius to Juvenal (Cambridge, 2001), 77–8CrossRefGoogle Scholar.

⁴⁷ The latest terminus post quem is Sat. 2.6.55–6, referring to a minor crisis of veterans’ land allocations in 30 b.c.e.: cf. Muecke (n. 32), ad loc.

⁴⁸ With few explicit termini post or ante quem, and no reliable external evidence (VSD 25 and 27 are dubious; cf. Horsfall, N., ‘Virgil: his life and times’, in id. [ed.], A Companion to the Study of Virgil [Leiden, 1995], 1–25)Google Scholar, this seems the most sensible guess.

⁴⁹ Compare Propertius’ prospectus of the Aeneid around 24 b.c.e. at 2.34.61–6: he reworks Virgil's two proemic utterances (Aen. 1.1–4, 7.44–5) and picks up suscitat arma (2.34.63, of Virgil writing the Aeneid) from Virgil's use of it throughout his epic (of Jupiter: 2.618, 11.727–8; of leaders: 9.463, 10.263, 12.108, 12.498–9). Extraordinarily detailed imitation of another poet's unfinished work was not unusual: cf. Starr, R.J., ‘The circulation of literary texts in the Roman world’, CQ 37 (1987), 213–23Google Scholar; Gurd (n. 24), 13.

⁵⁰ The reverse dynamic is not entirely impossible. Recall Virgil's appropriation (from low to high genre, satire to epic) of Horace's description of a hostel kitchen fire (Sat. 1.5.73–4) in his description of Troy aflame (Aen. 2.310–12); cf. Austin, R.G., P. Vergilii Maronis Aeneidos liber secundus (Oxford, 1964)Google Scholar, on Aen. 2.312, 360.

⁵¹ On molle atque facetum, cf. Zetzel (n. 23), 46; Gowers (n. 1 [2012]), ad loc.

⁵² On the interplay between the Eclogues and Satires Book 1, cf. (e.g.) Rooy, C.A. Van, ‘“Imitatio” of Virgil, Eclogues in Horace, Satires, Book 1’, AClass 16 (1973), 69–88Google Scholar; Putnam, M.C.J., ‘Pastoral satire’ (review of Freudenburg [n. 25]), Arion 3 (1995), 303–16Google Scholar; Henderson, J., ‘Virgil, Eclogue 9: Valleydiction’, PVS 23 (1998), 149–76, at 169–71Google Scholar; Welch (n. 15); Gowers (n. 1 [2012]), 22.

⁵³ The epithets are somewhat ambiguous—do they apply to Horace or Virgil or both? Gowers (n. 1 [2012]), ad loc. notes the overtones of poetic reluctance. crudis also suggests not only youth (cf. iuuenta, G. 4.565) and rusticitas but outright immaturity: like a querulous child, Virgil must get his sleep on the journey.

⁵⁴ Volk, K., The Poetics of Latin Didactic: Lucretius, Vergil, Ovid, Manilius (Oxford, 2002), 148–65CrossRefGoogle Scholar observes that the beginning of Georgics 3 hints at, but does not firmly promise, a martial epic; pace Propertius (2.34.60–1) and Servius (Aen. praef.), Virgil's eventual epic does not narrate direct and extended praise of Augustus. The view that Virgil here defers his epic project rests on a post-Virgilian perspective; within the Georgics the ambition remains suppressed and the recusatio in earnest.

⁵⁵ Hilberg (n. 8), 295 spots another OTIA acrostic at Ov. Met. 15.478–81, beginning with the last line of Pythagoras’ speech and stretching into the narration of Numa's return to Rome:

Numa's first actions are to restore peace (pacis … artes, 484)—that is, otium? Ovid's distribution of the acrostic with one line separate from the next three perhaps mimics Virgil's and Horace's OTIA patterns too. Ovid's only explicit mention of otium in Met. 15 alludes to the Georgics: in otia natam | Parthenopen (15.711–12; cf. Golden Age otia at 1.100). If this acrostic is deliberate, Ovid's peace-giving Numa is aligned with Virgil's and Horace's Octavian, a fitting compliment as the Metamorphoses nears its imperial endpoint. And might the recollection of the Metamorphoses at Tr. 1.7.25–6 nod back to Ovid's OTIA, as much as to his audience's leisurely reading (nunc precor ut uiuant et non ignaua legentem | otia delectent admoneantque mei)? OTIA, midway through Met. 15, lays the groundwork for Ovid's final ironic acrostic INCIP- at Met. 15.871–5: see Barchiesi, A., ‘Endgames: Ovid's Metamorphoses 15 and Fasti 6’, in Roberts, D.H., Dunn, F.M. and Fowler, D. (edd.), Classical Closure: Reading the End in Greek and Latin Literature (Princeton, 1997), 181–208, at 195Google Scholar.