¹ I cite the text of the Georgics from Mynors, R.A.B., P. Vergili Maronis Opera (Oxford, 1969)Google Scholar. All translations from Latin and Greek are my own.

² Virgil does not use the term, which appears as the title of a work of uncertain date, authorship and subject-matter at Varro, Rust. 2.5.5. In the same passage, Varro states that the Greeks call bees bugenes, and Greek authors use both that epithet and similar periphrases to indicate that bees are ‘ox-born’. I use the term bugonia out of convenience. For the relevant ancient passages, see Olk, F., ‘Biene’, RE 3 (1897), 431–51, at 434–5Google Scholar, and Mynors, R.A.B., Virgil Georgics (Oxford, 1990), on 4.281–314Google Scholar. Shipley, A.E., ‘The “bugonia” myth’, Journal of Philology 34 (1915–18), 97–105Google Scholar traces belief in the procedure from antiquity up to the seventeenth century.

³ For the purposes of this paper, an aition connects a present reality with a past event, stating or even simply suggesting that the past event constitutes the origin of, or somehow accounts for, the emergence of the present reality. See Codrignani, G., ‘L’‘aition’ nella poesia greca prima di Callimaco’, Convivium 26 (1958), 527–54, at 544Google Scholar; Schechter, S., ‘The aition and Virgil's Georgics’, TAPhA 105 (1975), 347–91, at 351–2Google Scholar; Loehr, J., Ovids Mehrfacherklärungen in der Tradition aitiologischen Dichtens (Stuttgart, 1996), 30Google Scholar. O'Hara, J.J., True Names: Vergil and the Alexandrian Tradition of Etymological Wordplay (Ann Arbor, 2017)²Google Scholar gives an excellent sense of Virgil's interest in aetiology.

⁴ e.g. Mynors (n. 2), 294.

⁵ Habinek, Thus T., ‘Sacrifice, society, and Vergil's ox-born bees’, in Griffith, M. and Mastronarde, D.J. (edd.), Cabinet of the Muses: Essays on Classical and Comparative Literature in Honor of Thomas G. Rosenmeyer (Atlanta, 1990), 209–23Google Scholar. Thomas, R.F., ‘The “sacrifice” at the end of the Georgics’, CPh 86 (1991), 211–18Google Scholar raises serious objections to Habinek's position. Pellegrini, J., ‘Note sur la double description de la bugonia au chant IV des Géorgiques’, Latomus 66 (2007), 336–41Google Scholar tries unpersuasively to answer some of Thomas's objections. Ov. Fast. 1.361–80 traces the origins of ox-sacrifice (not bugonia) back to Aristaeus; Virgil's narrative is quite different.

⁶ For a basic orientation, see Volk, K., The Poetics of Latin Didactic: Lucretius, Vergil, Ovid, Manilius (Oxford, 2002)CrossRefGoogle Scholar; Sider, D., ‘Didactic poetry: the Hellenistic invention of a pre-existing genre’, in Hunter, R., Rengakos, A. and Sistakou, E. (edd.), Hellenistic Studies at a Crossroads: Exploring Texts, Contexts and Metatexts (Berlin, 2014), 13–29Google Scholar.

⁷ For the centrality of the teacher-student relationship, see e.g. Fowler, D., ‘The didactic plot’, in Depew, M. and Obbink, D. (edd.), Matrices of Genre: Authors, Canons, and Society (Cambridge, MA, 2000), 205–19, at 205Google Scholar; Volk (n. 6), 37–9. Sharrock, A., Seduction and Repetition in Ovid's Ars Amatoria 2 (Oxford, 1994), 146–68Google Scholar explores the importance of this relationship in a mythological inset within a didactic poem, not unlike the Aristaeus narrative. In the teacher-student relationship, the poet can also assume the role of student. Thus Sider (n. 6), 16–17 argues that in the case of Parmenides the poet plays the role of student and the goddess functions as his teacher.

⁸ Sider (n. 6).

⁹ Sider (n. 6), 22.

¹⁰ e.g. Aratus’ Dike narrative (Phaen. 96–136) programmatically engages Hesiod's myth of ages (Op. 106–201), and Virgil's myth of ages (G. 1.118–59) programmatically engages both of the earlier texts. Some or all of these passages play significant roles in later didactic poems (e.g. Manilius 1.25–112) and in poems that engage the didactic tradition but lack a second-person addressee (e.g. Ov. Met. 1.89–150). For the centrality of so-called digressions in the didactic tradition, see Sharrock (n. 7), 87–195, especially 89–90; Gale, M.R., ‘Digression, intertextuality, and ideology in didactic poetry: the case of Manilius’, in Green, S.J. and Volk, K. (edd.), Forgotten Stars: Rediscovering Manilius’ Astronomica (Oxford, 2011), 205–21CrossRefGoogle Scholar.

¹¹ Homeric intertexts: Farrell, J., Vergil's Georgics and the Traditions of Ancient Epic (New York, 1991), 104–13, 320–4Google Scholar; Baier, T., ‘Episches Erzählen in Vergils Georgica: Struktur und Funktion des Proteus-Geschichte’, RhM 150 (2007), 314–36Google Scholar emphasizes the significance of Od. 11 as well. Proteus and pastoral: Segal, C., Orpheus: The Myth of the Poet (Baltimore, 1989), 76–7Google Scholar. Orpheus and elegy: Miles, G.B., Virgil's Georgics: A New Interpretation (Berkeley, 1980), 274–6Google Scholar; Conte, G.B., ‘Aristaeus, Orpheus, and the Georgics: once again’, in id., The Poetry of Pathos: Studies in Virgilian Epic (Oxford, 2007), 123–49CrossRefGoogle Scholar; some (e.g. P. Domenicucci, ‘L'elegia di Orfeo nel IV libro delle Georgiche’, GIF 37 [1985], 239–48, Gagliardi, P., Gravis cantantibus umbra: Studi su Virgilio e Cornelio Gallo [Bologna, 2003], 61–94Google Scholar and id., ‘I due volti dell'Orfeo di Virgilio’, Hermes 140 [2012], 284–309) argue that Orpheus represents Gallus. Proteus and/or Orpheus and the georgic poet: Perkell, C., The Poet's Truth: A Study of the Poet in Virgil's Georgics (Berkeley, 1989), 56–9, 80–9Google Scholar; Gale, M.R., ‘Poetry and the backward glance in Virgil's Georgics and Aeneid’, TAPhA 133 (2003), 323–52, at 333–7Google Scholar. Cyrene and the georgic poet: Schiesaro, A., ‘The boundaries of knowledge in Virgil's Georgics’, in Habinek, T. and Schiesaro, A. (edd.), The Roman Cultural Revolution (Cambridge, 1997), 63–89, at 65–8Google Scholar.

¹² Rejection of epic: Kraggerud, E., ‘Die Proteus-Gestalt des 4. Georgica-Buches’, WJA 8 (1982), 35–46Google Scholar. Homeric and Neoteric epic: Morgan, L., Patterns of Redemption in Virgil's Georgics (Cambridge, 1999), 150–211CrossRefGoogle Scholar. Epic and didactic: Baier (n. 11). Didactic and bucolic: Segal (n. 11), 76–7. Erotic elegy and didactic: Segal (n. 11), 20–4; Gale, M.R., Virgil on the Nature of Things: The Georgics, Lucretius and the Didactic Tradition (Cambridge, 2000), 55–6CrossRefGoogle Scholar; Conte (n. 11). Poetic inspiration and composition: Sibona, B., ‘La Bugonia: du fort est sorti le doux, ou des origines infectes de la poésie’, Pallas 60 (2002), 345–61Google Scholar; Baumbach, M., ‘Quae mox ventura trahantur. Eine poetologische Lektüre der Proteusfigur im Vergleich von Homers Odysee, Goethes Faust II und Vergils Georgica’, in Baumbach, M. and Polleichtner, W. (edd.), Innovation aus Tradition: Literaturwissenschaftliche Perspektiven der Vergilforschungen (Trier, 2013), 205–29Google Scholar. Problems with or limitations of didactic itself: Freudenburg, K., ‘Lucretius, Vergil, and the causa morbi’, Vergilius 33 (1987), 59–74Google Scholar; Perkell (n. 11); Schiesaro (n. 11). For recent takes on Virgil's reflections on genre in the Georgics, see Henkel, J., ‘Vergil talks technique: metapoetic arboriculture in Georgics 2’, Vergilius 60 (2014), 33–66Google Scholar; Gowers, E., ‘Under the influence: Maecenas and Bacchus in Georgics 2’, in Hardie, P. (ed.), Augustan Poetry and the Irrational (Oxford, 2016), 134–52CrossRefGoogle Scholar.

¹³ Both Conte (n. 11), 135, 141 and Thibodeau, P., Playing the Farmer: Representations of Rural Life in Vergil's Georgics (Berkeley, 2011), 197Google Scholar stress that Virgil refuses to make Orpheus a teacher, even though that is one of his main traditional identities. Orpheus, then, is simply a subject, rather than a singer or even reader, of didactic, which means that he functions quite differently from Proteus, Cyrene and Aristaeus. I cannot agree with Ross, D.O. Jr., Virgil's Elements: Physics and Poetry in the Georgics (Princeton, 1987), who (at 226–9)CrossRefGoogle Scholar characterizes Virgil's Orpheus as a representative of scientific poetry. Orpheus may play this role elsewhere but not in Proteus’ song; cf. below (n. 16).

¹⁴ For a brief account, see Mackenzie, T., ‘Georgica and Orphica: the Georgics in the context of Orphic poetry and religion’, in Freer, N. and Xinyue, B. (edd.), Reflections and New Perspectives on Virgil's Georgics (London, 2019), 67–77Google Scholar. Morgan (n. 12), however, remains fundamental.

¹⁵ e.g. Mackenzie (n. 14) compares Aristaeus’ binding of Proteus with an episode from the Orphic tradition, in which Zeus gets Kronos drunk on honey and then binds and castrates him, whereupon Kronos reveals how Zeus will come to power.

¹⁶ The oldest fragments of such a text are probably those contained in the Derveni Papyrus. No clear allusions, however, link Proteus’ song itself to the extant fragments of the Orphic theogonies, and it is noteworthy that Proteus does not actually characterize Orpheus as a singer of theogonies or scientific themes more generally, even though ancient readers associated Orpheus closely with this kind of poetry (e.g. Ap. Rhod. Argon. 1.496–511). For possible influence of a lost Orphic katabasis, see section 3.

¹⁷ For Orphic eschatology, see Graf, F. and Johnston, S.I., Ritual Texts for the Afterlife: Orpheus and the Bacchic Gold Tablets (London and New York, 2013)CrossRefGoogle Scholar; Bremmer, J.N., Initiation into the Mysteries of the Ancient World (Berlin, 2014), 55–80Google Scholar.

¹⁸ To recognize in bugonia a promise of (political, social, spiritual, etc.) renewal does not require an Orphic interpretative lens. By contrast, I suspect that those who read the poem as profoundly ambivalent or pessimistic may view my conclusions as confirmation of their own point of view.

¹⁹ The major exception is Habinek (n. 5).

²⁰ Those who invoke Servius include Skutsch, F., Aus Vergils Frühzeit (Leipzig, 1901), 145–7Google Scholar and Jacobson, H., ‘Aristaeus, Orpheus, and the Laudes Galli’, AJPh 105 (1984), 271–301, at 283–4Google Scholar. According to Servius, the fourth Georgic originally contained an extensive celebration of Gallus; after Gallus fell into disfavour and committed suicide, Virgil replaced the laudes Galli with the story of either Aristaeus (on Ecl. 10.1) or Orpheus (on G. 4.1). Whatever one's position on the laudes Galli, we cannot invoke Servius in an attempt to explain away any and all perceived difficulties in the poem; instead, we must confront the text as we have it. For similar assessments of the situation, see Thomas, R.F., Virgil Georgics, 2 vols. (Cambridge, 1988), 1.13–16Google Scholar; Mynors (n. 2), 296; Erren, M., P. Vergilius Maro: Georgica, vol. 2 (Heidelberg, 2003), 906Google Scholar; Baier (n. 11), 317–18; Conte (n. 11), 127; Baumbach (n. 12), 221.

²¹ Gale (n. 12), 52. Critics regularly point out the overlap between Aristaeus’ skills and the topics of the Georgics; e.g. Miles (n. 11), 266–70 and Conte (n. 11), 135–6. At the beginning of the poem, Virgil calls Aristaeus cultor nemorum (1.14).

²² Gale (n. 12), 52. Putnam, M.C.J., Virgil's Poem of the Earth: Studies in the Georgics (Princeton, 1979)Google Scholar repeatedly observes that Aristaeus is someone who learns, and he compares him to the reader of the poem (at 276, 290, 314, 316, 319). Putnam, however, prefers to treat Aristaeus as ‘the georgic artist in the process of learning’ (at 276) and so to contrast him with Orpheus, ‘paradigm of poets’ (at 321).

²³ Schiesaro (n. 11), 67.

²⁴ Miles (n. 11), 277–8 and 288–9 characterizes Proteus’ song as being about the inflexible laws of nature; Gale (n. 11), 335–6 is on the right track when she identifies Proteus as an aetiological uates; cf. P. Hardie, ‘Political education in Virgil's Georgics’, SIFC 97 (2004), 83–111, at 104. Schiesaro (n. 11) 67 is disinclined to connect Proteus with the narrator of the Georgics.

²⁵ Morgan (n. 12), 161–70 with ample discussion of the terminology and defining characteristics. To single out merely the criterion of narrative subject-matter, and without wishing to oversimplify, we can say that Homeric epic tends to celebrate martial valour, while Neoteric epic focusses on love, especially tragic love.

²⁶ See Kraggerud (n. 12), 38. Thomas (n. 20), on 4.387–8 notes that Callimachus may have made Proteus a mantis (SH 254.5).

²⁷ Virgil does not use uates in this sense in the Georgics, where its only other occurrence comes at 3.491. In the Eclogues, however, uates occurs twice in this sense (7.28, 9.34).

²⁸ See Freudenburg (n. 12).

²⁹ Virgil does not spell out the stipulation, stating only that Eurydice follows Orpheus namque hanc dederat Proserpina legem (4.487). Others (e.g. Ov. Met. 10.50–2) are more explicit. Ziegler, K., ‘Orpheus’, RE 18 (1939), 1200–316, at 1269–70Google Scholar stresses that Orpheus was forbidden not from looking back at Eurydice but rather from looking back at all.

³⁰ See, most recently, Gladhill, B., Rethinking Roman Alliance: A Study in Poetics and Society (Cambridge, 2016), 69–96Google Scholar.

³¹ Cf. Miles (n. 11), 277–8, who also notes the Lucretian use of foedera at G. 1.60–1. Note that, when Orpheus actually looks back, he is uictus … animi ‘overcome in his mind’ (uictusque animi respexit, 4.491). The genitive of respect seems to echo similar expressions in Lucretius (e.g. nec me animi fallit at 1.136, 1.922 and 5.97). At G. 3.289–90, Virgil uses animi in exactly the same way (nec sum animi dubius uerbis ea uincere magnum | quam sit et angustis hunc addere rebus honorem).

³² Note that Proteus describes the souls who flock to Orpheus as simulacraque luce carentum (4.472). The phrase comes from DRN 4.35, where Lucretius is rejecting the idea that such nocturnal visions are actually ghosts. Proteus also mentions Tartarus, the Furies and Cerberus (G. 4.481–3), all three of which items figure together in Lucretius’ list of non-existent terrors of the underworld (3.1011–13). That the Georgics constantly critiques and/or rewrites the DRN is well established; see Freudenburg (n. 12) and, more broadly, Gale (n. 12). If Proteus’ song has Orphic undertones, the contrast with Lucretius becomes even more pointed.

³³ In addition to practical information about farming, the Georgics contains considerable theoretical knowledge; see especially Ross (n. 13) and Gale (n. 12). Schiesaro (n. 11), however, downplays the technical nature of Virgil's instruction, and Gowers (n. 12) stresses Virgil's aversion to technical systematization and comprehensiveness.

³⁴ There is some ancient evidence for this kind of division: the Tractatus Coislinianus divides didactic (παιδευτική) into ‘instructional’ (ὑφηγητική) and ‘theoretical’ (θεωρητική). Volk (n. 6), 42 considers ‘instructional’ the Works and Days, the Georgics and the Ars Amatoria, and treats as ‘theoretical’ the poems of Empedocles and Lucretius. For the Tractatus Coislinianus, see Janko, R., Aristotle on Comedy: Towards a Reconstruction of Poetics II (London, 1984)Google Scholar.

³⁵ See Volk (n. 6), 41; cf. Sharrock (n. 7), 146–9.

³⁶ If we read the emergence of bees from the ox-carcass as symbolic of human purification and rebirth, then the ethical thrust of Cyrene's instructions is all the more significant: to produce a new swarm is to atone for the sinful fallen state of humankind/Rome and to achieve a new state of blessedness/social harmony or flourishing. For this kind of Orphic reading, see again Morgan (n. 12) and Mackenzie (n. 14).

³⁷ See again Volk (n. 6), 41.

³⁸ The reminiscence also reinforces the intertextual connections with Iliad Book 1 through which Aristaeus plays the role of Achilles. At the beginning of the Iliad, Achilles proposes consulting a seer or priest (μάντιν … ἢ ἱερῆα, 1.62), and Calchas comes forward. Aristaeus’ interaction with Proteus mirrors Achilles’ interaction with Calchas.

³⁹ Just as Hesiod names Perses in Op. and later didactic poets name their addressees, the Dichterweihe scene names Hesiod (Theog. 22). As Fowler (n. 7) has observed, initiation is a common ‘plot’ that describes the progress of the addressee through the lessons of the text. If, from Virgil's perspective, the Muses ‘initiate’ Hesiod, this will enhance his character as a didactic addressee.

⁴⁰ Note that the author of the Contest of Homer and Hesiod puts the entire formula into Hesiod's mouth (Cert. 8), and a similar phrase may occur at Hes. fr. 204.113 M–W. Poetic knowledge of either present and future things or present and past things occurs elsewhere; see West on Theog. 32. Virgil also expands the Greek in a significant way. Calchas and the Hesiodic Muses know three substantivized participles (τά τ’ ἐόντα τά τ’ ἐσσόμενα πρό τ’ ἐόντα), whereas Proteus knows one substantivized adjective (omnia) in apposition to which stand three indirect questions (quae sint, quae fuerint, quae mox uentura trahantur). The didactic tradition regularly employs strings of indirect questions to summarize the contents of didactic songs; the technique occurs already at Theog. 108–15 and is prominent in Lucretius, and Virgil exploits it (e.g. G. 1.1–5). For discussion, see Brown, R.D., ‘The structural function of the song of Iopas’, HSPh 93 (1990), 315–34Google Scholar.

⁴¹ The classic treatment of this tradition is Buffière, F., Les mythes d'Homère et la pensée grecque (Paris, 1956)Google Scholar.

⁴² For the Aristaeus episode in particular, see Farrell (n. 11), 253–72 and Morgan (n. 12), especially 61–101.

⁴³ Hardie, P., Virgil's Aeneid: Cosmos and Imperium (Oxford, 1986), 83–4Google Scholar.

⁴⁴ Farrell (n. 11), 270–1 offers a reading of Clymene similar to Hardie's (n. 43). Proteus is particularly well suited to embodying theoretical didactic since ancient critics could interpret him allegorically as the primordial state of the cosmos, the Empedoclean elements or even the origins of all life; see especially Heracl. All. 65–7 with Buffière (n. 41), 179–86.

⁴⁵ Supposed parallels between G. 4.464–70 and lines 40–2 of the late antique Orphic Argonautica led Norden (on Aen. 6.120) to hypothesize a common source in the lost katabasis; see, more recently, Nelis, D.P., ‘The reading of Orpheus: the Orphic Argonautica and the epic tradition’, in Paschalis, M. (ed.), Roman and Greek Imperial Epic (Herakleion, 2005), 169–92, at 170–2Google Scholar. The parallels between this same passage of the Orphic Argonautica and Aen. 6.119–20 are much more convincing. For the possible influence of the katabasis on the underworld of the Aeneid, see Bremmer (n. 17), 180–204.

⁴⁶ For the katabasis itself, see Graf and Johnston (n. 17), 174–6; Bremmer (n. 17), 58–65. In Bernabé's edition of the Orphic fragments, testimonia for descent-narratives are frr. 707–16; fr. 717 is a papyrus whose katabasis-narrative may or may not be Orphic.

⁴⁷ In the context of Orphic-Bacchic ritual and belief, the katabasis probably served a practical function, acting as a kind of underworld roadmap upon which the soul of the initiate would rely after death. If Virgil has incorporated parts of that roadmap into Proteus’ narrative, he has done away with the practical function: Proteus does not indicate Orpheus’ route or give advice on how to interact with the rulers of underworld. The author of the Derveni Papyrus gives an impression of just how theoretical some ancient readers thought Orphic poetry was; see e.g. Betegh, G., The Derveni Papyrus: Cosmology, Theology, and Interpretation (Cambridge, 2004)CrossRefGoogle Scholar. I am sceptical of Mackenzie's (n. 14) argument that the entire Aristaeus-narrative has the practical function of showing Octavian how to purify the sins of Rome.

⁴⁸ Thomas (n. 20) on 4.475–94, 4.490 and 4.493–4 is a notable exception: he contends that Virgil first wishes for the success of the Georgics and then, failing that, hopes to be content with the Eclogues. For the possible influence of Hesiod, Parmenides, Empedocles and Aratus in this immediate context, see D. Nelis, ‘Georgics 2.458–542: Virgil, Aratus and Empedocles’, Dictynna 1 (2004) (http://dictynna.revues.org/161).

⁴⁹ See e.g. Schiesaro (n. 11), 85–6; Gale (n. 12), 11, 42–3.

⁵⁰ Given the very anti-Lucretian elements of Proteus’ song, especially its mythology of the underworld, it is perhaps surprising to compare him to the felix who tramples on the terrifying roar of Acheron (4.491–2). But it is not necessary to equate the felix with Proteus. Rather, Proteus represents the commitment to theoretical didactic, and that commitment connects him to the more straightforwardly Lucretian felix.

⁵¹ See Clay, J.S., Hesiod's Cosmos (Cambridge, 2003)CrossRefGoogle Scholar.

⁵² See e.g. Freudenburg (n. 12); Perkell (n. 11), 143–7; Gale (n. 12), 55.

⁵³ So Schiesaro (n. 11), 85–6 regarding the makarismos. He observes that in the Georgics Virgil does not actually engage in rigorous Lucretian-style philosophical analysis. This does not invalidate my argument, however, that Proteus symbolizes or represents the origins of precisely that mode of thinking.

⁵⁴ How an author presents an aition greatly affects the reader's sense of historical continuity or discontinuity. In general, see Graf, F., ‘Römische Kultaitia und die Konstruktion religiöser Vergangenheit’, in Flashar, M., Gehrke, H.-J. and Heinrich, E. (edd.), Retrospektive: Konzepte von Vergangenheit in der griechisch-römischen Antike (Munich, 1996), 125–35Google Scholar. For analyses of aitia that emphasize historical discontinuity, see e.g. Goldhill, S., The Poet's Voice: Essays on Poetics and Greek Literature (Cambridge, 1991), 321–33Google Scholar; Hutchinson, G., Propertius Elegies, Book IV (Cambridge, 2006), 3. Miles (n. 11), 256–7Google Scholar senses a similar sense of historical rupture in the second half of Georgics Book 4. Nelis, D.P., ‘Past, present, and future in Virgil's Georgics’, in Farrell, J. and Nelis, D.P. (edd.), Augustan Poetry and the Roman Republic (Oxford, 2013), 244–62CrossRefGoogle Scholar reads the Georgics as a whole as a sustained meditation on Rome's past, present and future, arguing that the poem forces the reader to contemplate and construct historical narratives. My interpretation complements these approaches.

⁵⁵ Servius on 4.553 makes just this point. More recent proponents of such an interpretation include: Anderson, W.B., ‘Gallus and the fourth Georgic’, CQ 27 (1933), 36–45, at 40CrossRefGoogle Scholar; Wilkinson, L.P., The Georgics of Virgil: A Critical Survey (Cambridge, 1969), 113; cf. 120Google Scholar; Schechter (n. 3), 381–5; Mynors (n. 2), on 4.530; Pellegrini (n. 5), 341; C. Formicola, Da Orfeo a Lavinia (Virgilio: Morte Vita Storia) (Naples, 2008), 21–4. A similar perspective must be adopted by those who see Aristaeus’ slaughter as an aition for animal sacrifice; see above (n. 5). Thibodeau (n. 13), 159 agrees with those who disregard the aetiological nature of the Aristaeus episode as ‘little more than a pretense’.

⁵⁶ When an aition contains instructions, they usually persist into the present. On this point, Ov. Fast. 4.629–76 is instructive. During a time of agricultural difficulty, Numa had a dream in which Faunus told him to slaughter one cow but thereby to offer up two lives. The king's wife realized that he was being asked for the entrails of a pregnant cow, and so the entrails of a pregnant cow were given and the situation improved (4.669–72). In the reader's present, the ritual is much more elaborate (4.635–40), but the only procedural detail the aition contains (exta bouis grauidae dantur, 4.671) persists even now (exta dedere focis, 4.638). In the case of ritual, it is more common for aitia to characterize present practice as formalized imitation or commemoration of past events that were themselves simply spontaneous and unscripted (e.g. Callim. Aet. frr. 22–3c Pfeiffer on Lindian sacrifice, Ov. Fast. 4.133–62 on the Veneralia).

⁵⁷ Ov. Fast. 5.621–32 provides a clear example. Once upon a time, Jupiter ordered a sacrifice to Saturn that consisted of throwing two humans into the river. This ritual was performed regularly until Hercules came by and threw straw effigies into the river instead, and that is what the Romans have done ever since. Although Ovid ultimately rejects this aition, the story itself explains why the original instructions were discarded. Harder, A., ‘The invention of past, present, and future in Callimachus’ Aetia’, Hermes 131 (2003), 290–306, at 299Google Scholar cites four narratives in Callimachus’ Aetia that involve a change in ritual practice. In two of them, the Diegesis explains who initiated the change and when, which may imply that Callimachus included such material in his narrative.

⁵⁸ In general, the prōtos heuretēs topos ignores the details about how any given task was initially performed (e.g. Hom. Hymn 5.12–15, 20.1–7; [Aesch.] PV 436–506; Ar. Ran. 1032–6). Looking at the matter the other way around, in the Aetia Callimachus generally describes the present-day reality quite briefly, ‘with little detail beyond the bare essentials’ (Harder [n. 57], 294). If Virgil had taken this approach, describing Egyptian bugonia vaguely and in only a few lines, the relationship between past and present might seem quite different.

⁵⁹ Contrast Ovid's practice in the Metamorphoses and the Fasti, where various tags (nunc, nunc quoque, etiam nunc, unde, etc.) often signal the end of an aition and reinforce the connection between past and present; see Myers, K.S., Ovid's Causes: Cosmogony and Aetiology in the Metamorphoses (Ann Arbor, 1996), 66–7Google Scholar on these kinds of expression. On occasion, Ovid goes so far as to characterize the present as an imitation of the past or to declare that traces of the past persist in the contemporary world. See e.g. the following passages from the Fasti: forma manet facti (2.379), res latuit, priscique manent imitamina facti (4.211), factum abiit, monimenta manent (4.709). All such commentary is lacking in the Georgics.

⁶⁰ By suppressing Aristaeus’ name in Book 1 and in Book 4 (Arcadii … magistri, 4.283), Virgil signals the identity of the two figures. Moreover, O'Hara (n. 3), 254 detects an etymological wordplay in 1.14 that reveals the identity of the cultor as Aristaeus. Note that, even if we agree with Thomas (n. 20), on 1.14–15 that Varro of Atax may have made Arcadia Aristaeus’ home, that does not diminish the jarring discontinuity. Critics have also puzzled over how bugonia is supposed to have migrated from Arcadia to Egypt.

⁶¹ Aristaeus, indeed, resembles didactic addressees such as Hesiod's Perses and Lucretius’ Memmius, neither of whom would make for good didactic poets; from here, we might connect him with the fortunatus of the makarismos, but he is still a student, not a teacher. Other authors, by contrast, emphasize Aristaeus’ identity as a teacher. At Diod. Sic. 4.81.2, Aristaeus learns certain agricultural activities from the Nymphs and teaches mankind about them. At Nonnus, Dion. 5.215–86, he also learns, discovers and teaches various rural practices. Ap. Rhod. Argon. 2.498–527 is slightly different: the Muses teach Aristaeus about medicine and prophecy, but, despite his beneficent service to mankind, he is not characterized as a teacher.

⁶² Habinek (n. 5), 209–10: ‘In the Egyptian tale the bees are born from liquid that has grown warm (tepefactus, 308); as soon as their wings are formed the bees set to buzzing (stridentia, 310); and their appearance from the putrefied cadaver is likened to a summer cloudburst (ut aestiuis effusus nubibus imber, 312). In the Aristaean bugonia, the slaughtered oxen's innards have turned liquid (liquefacta, 555); from them the bees buzz (stridere, 556) and burst forth (ruptis … costis, 556; cf. erupere at 313); in their swarming they resemble immense clouds (immensaque trahi nubes, 557).’

⁶³ Osorio, P., ‘Vergil's physics of bugonia in Georgics 4’, CPh 115 (2020), 27–46Google Scholar argues that both bugonia narratives draw on theoretical models of spontaneous generation; I refer the reader to this study for more evidence and comprehensive discussion. There are some scholars who acknowledge the scientific character of Virgilian bugonia: e.g. Ross (n. 13), 216–18; Erren (n. 20), 889–90, 902–4, 906, 997–8; Formicola (n. 55), 15–21. Most, however, are reluctant to characterize bugonia as theoretically justified in any way. Thus Perkell (n. 11), 77 n. 66 describes the procedure as ‘more religious than “scientific”’, while Pellegrini (n. 5), 341 attributes the efficacy of Egyptian bugonia to sympathetic magic, and claims that Aristaeus succeeds because the Nymphs answer his prayer.

⁶⁴ The Egyptians also perform the procedure as the weather begins to warm; see Thomas (n. 20), on 4.305–7.

⁶⁵ For the relevant passages, see Campbell, G., Lucretius on Creation and Evolution: A Commentary on De Rerum Natura 5.772–1104 (Oxford, 2003)Google Scholar, Appendix A and Osorio (n. 63). Although we do not know much about the earliest theories, Anaxagoras (DK 59 A1.9) and Empedocles (DK 31 A70) stress the interaction of heat, moisture and earth, and Aristotle repeatedly emphasizes the process of putrefaction (Mete. 4.1 379a–b, Hist. an. 5–6).

⁶⁶ The connection with DRN 3.713–40 has not gone unnoticed; see e.g. Erren (n. 20), 889–90. Lucretius, of course, discusses spontaneous generation elsewhere (2.871–3, 2.897–901, 2.926–30, 5.796–9). Ovid uses similar language when discussing spontaneous generation (Met. 1.416–33) and bugonia (Met. 15.361–7; cf. Fast. 1.377).

⁶⁷ The Egyptian bugonia narrative is purely descriptive: a place is selected (eligitur, 4.296), a calf is sought out (quaeritur, 4.300), etc. By not instructing the reader to perform these tasks through commands, Virgil distances himself from the specifics. Cyrene, by contrast, tells Aristaeus what to do: select four bulls and four heifers (delige, 4.540), set up the altars (constitue, 4.542), and so on, but these are precisely Cyrene's praecepta, not the poet's. Thomas (n. 20), on 4.295–314 concludes that Virgil's description of Egyptian bugonia ‘has the effect of separating it from the real technical advice of the poem’; I am arguing something slightly different.

⁶⁸ For bugonia as authorial fiction, see Ross (n. 13), 215–16; Thomas (n. 20), on 4.281–314.

⁶⁹ Through the Aristaeus narrative, Virgil thematizes his fusion of the two Hesiodic traditions. For the pervasiveness of theoretical knowledge in the poem, see above (n. 33).

⁷⁰ This conclusion can be applied to Orphic readings of the Aristaeus narrative: in this case, Virgil's ‘theoretically informed, efficacious lesson’ is Orphic teaching about death and rebirth, sin and purification, etc.

⁷¹ Cf. the destabilizing role of Fama in the Aeneid and Ossa in epic (e.g. Il. 2.93). In general, see Hardie, P., Rumour and Renown: Representations of Fama in Western Literature (Cambridge, 2012)Google Scholar.

⁷² Thomas (n. 20), on 4.287–94 notes that the location of India is ‘somewhat imprecise’. In light of my overall argument, I read this imprecision as deliberate.

⁷³ Hdt. 2.28–34 provides the oldest evidence of the debate.

⁷⁴ The phrase ‘Alexandrian footnote’ describes certain expressions, such as ut fama (est) (‘as the story goes’) or dicitur (‘it is said’), that appear to be vague references to tradition or common knowledge but that actually signal allusions to specific texts. Latin poets working in the learned Alexandrian tradition employ this device frequently and often quite prominently (e.g. dicuntur in the second line of Catull. 64: Peliaco quondam prognatae uertice pinus | dicuntur liquidas Neptuni nasse per undas). For discussion, see e.g. Hinds, S., Allusion and Intertext: Dynamics of Appropriation in Roman Poetry (Cambridge, 1998), 1–3Google Scholar. No ancient source other than Virgil connects Aristaeus with bugonia. If Virgil invented the link, the footnote ut fama points to nowhere.

⁷⁵ O'Hara, J.J., Inconsistency in Roman Epic: Studies in Catullus, Lucretius, Vergil, Ovid and Lucan (Cambridge, 2007), 83–5CrossRefGoogle Scholar offers a dispassionate overview of the scholarship, which is sometimes heated. For the possibility that such passages, especially the treatment of grafting and the laudes Italiae in Book 2, might have metapoetical significance, see S. Clément-Tarantino, ‘La poétique romaine comme hybridation féconde: les leçons de la greffe (Virgile, Géorgiques, 2, 9–82), Interférences Ars Scribendi 4 (2006) (http://ars-scribendi.ens-lsh.fr/article.php3?id_article=37&var_affichage=vf); Deremetz, A., ‘The question of the marvellous in the Georgics of Virgil’, in Hardie, P. (ed.), Paradox and the Marvellous in Augustan Literature and Culture (Oxford, 2009), 113–25Google Scholar; cf. Henkel (n. 12). My reading of the Aristaeus episode complements this kind of approach to the poem.

⁷⁶ The same story seems to have been told in the Cat. as well: Prometheus’ name occurs in frr. 2 and 4 M–W, and Pandora appears in fr. 5 M–W. For discussion, see Clay (n. 51), 167–8, who also contrasts Hesiod's competing accounts of the origins of human life.

⁷⁷ For instance, van Noorden, H., ‘Aratus’ maiden and the sources of belief’, in Harder, M.A., Regtuit, R.F. and Wakker, G.C. (edd.), Nature and Science in Hellenistic Poetry (Leuven, 2009), 255–75Google Scholar argues that Aratus’ Dike-myth raises questions about the transmission of knowledge, the continuity between past and present, and the possible failure of didactic.

⁷⁸ Wallace-Hadrill, A., Rome's Cultural Revolution (Cambridge, 2008), 213–58Google Scholar.

⁷⁹ Cf. Wallace-Hadrill (n. 78), 236: ‘Antiquarianism, far from enhancing the reassuring sense of continuity provided by traditionalism, subverted it.’ For a similar rejection of antiquarianism as conservative, see Moatti, C. (transl. J. Lloyd), The Birth of Critical Thinking in Republican Rome (Cambridge, 2015), 94CrossRefGoogle Scholar.

⁸⁰ Note esp. Wallace-Hadrill (n. 78), 258.

⁸¹ See Cic. Brut. 62 with Wiseman, T.P., ‘Legendary genealogies in Late-Republican Rome’, G&R 21 (1974), 153–64, at 158–60Google Scholar.