Among the splendid miniatures found in the so-called ‘Vergilius Romanus’ (codex Vat. lat. 3867, from the early sixth century) is a remarkable representation of the tempest which ushers in the narrative of the Aeneid. On the upper register of fol. 77r, to the right and left of Juno, two winds are depicted in personified form, bust-length and in profile, blowing their trumpets.Footnote ¹ Such an illustration freely builds on the text of the Aeneid, which does not explicitly state what the winds look like. As Philip Hardie writes, ‘Virgil is deliberately unspecific about the precise physical forms of his winds, but they are at least partly anthropomorphic; Neptune can use rational speech to them.’Footnote ²

Scholars have focussed on two aspects in particular. First, at lines 52–6 (especially 55–6 illi indignantes magno cum murmure montis | circum claustra fremunt) Virgil alludes to a passage from the De rerum natura in which Lucretius compares the winds trapped within clouds to caged beasts: magno indignantur murmure clausi | nubibus, in caueisque ferarum more minantur; | nunc hinc nunc illinc fremitum per nubila mittunt | quaerentesque uiam circum uersantur (Lucr. 6.197–200).Footnote ³ Second, this scene evokes the Hesiodic description of the Titans imprisoned in Theog. 717–35. Through this mythological characterization, the winds prove to be gigantomachic forces that threaten the stability of the cosmos: were Aeolus not to keep them under control, they might upset the universe (Aen. 1.58–9 maria ac terras caelumque profundum | quippe ferant rapidi secum uerrantque per auras).Footnote ⁴

This article aims to highlight a further, overlooked characterization: the winds as horses. Some scholars have rightly noted the presence of equine elements in this episode, interpreting them as metaphors.Footnote ⁵ Instead, I will argue that we are probably dealing with more than just metaphors: within Aeolus’ cave we could (and perhaps should) envision real horses—albeit very special ones. This characterization fits within a long-standing tradition in which the winds are represented as horses, which it is worthwhile to re-examine in its entirety (§2).Footnote ⁶ Let us begin by examining the passage in greater detail.

1. HORSE TRACKS (AEN. 1.50–86)

In the section spanning from Juno’s arrival in Aeolia to the unleashing of the storm (Aen. 1.50–86), various textual elements hint at an equine depiction of the winds—or, at the very least, allow us to envision them also as horses.Footnote ⁷ Let us begin with the opening lines (Aen. 1.52–63):

At line 54 (imperio premit ac uinclis et carcere frenat) Aeolus undoubtedly acts as a jailer: he ‘holds them down with his authority, curbing them with chains and prison-bars’, as Austin translates.Footnote ⁹ However, frenat suggests an animal image, since frenum primarily means ‘horse’s bridle or harness’ (OLD s.v. 1) and frenare ‘to equip or control (horses, etc.) with a bridle’ (OLD s.v. 1). To keep the winds under control, Aeolus puts reins—more or less metaphorically—on them. Thus, as Michael Putnam convincingly suggests,Footnote ¹⁰ we can see Aeolus as a charioteer calming and reassuring his team of horses.

To a careful reader, frenare acts as a trigger: in the same sentence, both uinclis and carcere can be related to the same imagery. Besides indicating bonds or chains used to restrain prisoners, uinc(u)la can designate a halter (OLD s.v. 2a, ‘applied to ropes, etc., for tethering or securing animals’), particularly a horse’s halter. This draws attention to the epic image of the horse breaking free by snapping its halter, as seen in Homer (Il. 6.507 = 15.264 δϵσμὸν ἀπορρήξας), Ennius (Ann. 535–6 Skutsch sicut equos … | uincla suis magnis animis abrumpit ) and Virgil (Aen. 11.493–4 qualis ubi abruptis fugit praesepia uinclis | tandem liber equus). Similarly, carcere, which also occurs at line 141 (‘illa se iactet in aula | Aeolus et clauso uentorum carcere regnet’ says Neptune), primarily meaning ‘prison’, can also signify the ‘cage’ or ‘starting gate’ from which horses began their race in the chariot races at the circus,Footnote ¹¹ as in Ennius (Ann. 80–1 Skutsch omnes auidi spectant ad carceris oras | quam mox emittat pictos e faucibus currus; 463–4 Skutsch quom a carcere fusi | currus cum sonitu magno permittere certant) and Virgil (G. 3.104 = Aen. 5.145 ruuntque effusi carcere currus). Varro writes that in Naevius’ time the carceres of the Circus Maximus were called oppidum Footnote ¹² and were thus equated with a fortified citadel, either because the architectural complex was equipped with towers, as Varro states,Footnote ¹³ or because it was fenced off by a wooden palisade.Footnote ¹⁴ In any case, an analogy can be established with the ‘fortress’ on which Aeolus sits and reigns (Aen. 1.56 celsa sedet Aeolus arce).

At line 55, indignantes may well apply to horses reluctant to accept bridles or any form of control, like Pegasus in Ovid (Fast. 3.455 iamque indignanti noua frena receperat ore ) or a steed in Silius Italicus (16.352 ibat campo indignatus habenas).Footnote ¹⁵ At line 56, circum claustra fremunt also points in the same direction. While claustra primarily means ‘cage’ or ‘prison’ (OLD s.v. 2), it can also designate the ‘enclosure’ of horses in the circus, as shown by Manilius (5.76 cum laxato fugerunt [sc. equi] cardine claustra ) and Sidonius Apollinaris (Carm. 23.331 illi [sc. equi] ad claustra fremunt repagulisque | incumbunt); and fremere is commonly associated with horses (cf. Verg. G. 1.12–13 frementem … equum; Aen. 7.638–9 frementis [trementis FMγRabenru] … equos and 12.82 equos … frementis).Footnote ¹⁶ Both carcer and claustra, in short, allow the coexistence of different planes of imagery. Incidentally, Sidonius’ phrase ad claustra fremunt seems to allude directly to the Virgilian line and at the same time interpret it, making explicit the imagery only hinted at in Aeneid Book 1.

From line 63, we know that Jupiter has instructed Aeolus to ‘pull or loosen the reins at his command’ (et premere et laxas … dare iussus habenas). This wording confirms the association, since habenae are primarily the ‘reins’ used to control horses and chariots (OLD s.v. 1). A few lines later, Juno commands Aeolus: ‘incute uim uentis submersasque obrue puppes’ (1.69). In the phrase incute uim uentis we may recognize the image of ‘whipping’ the winds, as if urging on a horse.Footnote ¹⁷

Finally, at lines 81–6 Aeolus releases the winds:

On the phrase uelut agmine facto (82) Austin remarks: ‘the winds come pouring from the “gate” like an army in column of the march, moving with precision.’Footnote ¹⁸ The noun agmen is rooted in the military domain (OLD s.v. 5: ‘an army on the march, a column’), thus suggesting violence and hostility. Yet another image seems equally, if not more, fitting: a team of horses forming at the starting gate and, upon receiving the signal, dashing off in a race.Footnote ¹⁹ agmen may indeed be ‘applied to groups or lines of animals, often w[ith] strong reminiscence of mil[itary] sense’, and can designate ‘a team (of horses)’ (OLD s.v. 9 and 9b), as happens in Gellius (NA 19.8.4 equorum quattuor iunctorum agmen).Footnote ²⁰ The four cardinal winds could be seen as corresponding to the four horses of a quadriga: according to Servius, three of them are mentioned at lines 85–6 (Eurus … Notus … | Africus), while the northern wind appears at line 102 (stridens Aquilone procella).Footnote ²¹ The use of ruere (83 ruunt) to describe the swift and impetuous departure of chariots in the circus is already found in Verg. G. 3.104 = Aen. 5.145 ruuntque effusi carcere currus.Footnote ²² Even the polyvalent term porta, in the phrase qua data porta (83), may refer to the gates in the animal pens at the circus.Footnote ²³

But why does Aeolus behave like a charioteer? First of all, there are genealogical and onomastic reasons.Footnote ²⁴ In the Odyssey, Aeolus is referred to as Ἱπποτάδης (10.2 Αἴολος Ἱπποτάδης; 10.36 παρ’ Αἰόλοο μϵγαλήτορος Ἱπποτάδαο), that is, ‘son of Ἱππότης’,Footnote ²⁵ ἱππότης meaning ‘driver’ or ‘rider of horses’ (LSJ s.v.). Additionally, his very name, Αἴολος, is nothing but a proparoxytone form of the adjective αἰόλος, meaning ‘nimble’, ‘swift’, as well as ‘changeable’, qualities that suit the king of the winds, which are both swift and proverbially changeable. This adjective is also associated with the speed of horses in epic tradition: Achilles’ steed Xanthos, son of the wind Zephyrus (see below), is said to be πόδας αἰόλος ἵππος in Homer (Il. 19.404; αἰόλος ἵππος is then found in Quint. Smyrn. 1.338 and 4.563); the adjective also appears in the Homeric compound αἰολόπωλος ‘with swift horses’ (Il. 3.185 and Hymn. Hom. Ven. 137; then Theoc. Id. 22.34). So, Aeolus is the ‘son of a horse-driver’ and behaves as such in the Aeneid.Footnote ²⁶

Equine imagery influenced later poets who dealt with the Virgilian tempest. Ovid’s flood in Metamorphoses Book 1, for example, applies similar images to rivers (1.280 ‘fluminibus uestris totas immittite habenas’; 281 fontibus ora relaxant; 282 defrenato uoluuntur in aequora cursu ; 285 exspatiata ruunt per apertos flumina campos).Footnote ²⁷ Closer to his model is Valerius Flaccus, when rewriting the Virgilian episode in the first book of his Argonautica (1.608–14):

Here too the winds wait eagerly to be set free (608 fremere ∼ Aen. 1.56 fremunt). However, above all, as Andrew Zissos has well noted, at lines 610–11 ‘[t]he language makes available an attractive metaphoric equation between the escaping winds and racehorses pouring out of a starting cage’: in addition to equi (611), the ambivalence of carcere (610) and the expression se fundunt (610), used for the chariot race in Verg. G. 1.512 (cum carceribus sese effudere quadrigae; cf. Enn. Ann. 463–4 Skutsch a carcere fusi | currus), contribute to the same effect.Footnote ²⁸ Indeed, it is not entirely clear whether we are dealing with winds as horsesFootnote ²⁹ or whether rather each wind is to be imagined as a charioteer/ἱππότης leading his horses, as the plural equi (611) would suggest.Footnote ³⁰ In the preceding scene, Boreas, flying from Thrace to Aeolia, is unmistakably depicted as a winged deity (1.577 omne dei rapidis nemus ingemit alis). Here, Thraces equi would seem to designate not so much the god himself as his brute force, chained and now freed, that is, the horses hitched to his chariot.

In any case, it seems clear that Valerius Flaccus makes explicit the equestrian and circus imagery that was only hinted at in Aeneid Book 1; the choice of naming Aeolus Hippotades (610) points to the awareness of such an operation. Fifteen centuries later, Teofilo Folengo’s macaronic genius will transfigure the Virgilian Aeolus into a very down-to-earth horse-groomer: castigatque suos uentos de more pedanti, | moreque cozzoni magris dat fraina cauallis ‘he chastizes his winds like a schoolmaster and like a stable-boy gives fodder to his scrawny horses’; cui data cura est | a Ioue striggiandi uentos stallasque sgurandi ‘to whom Jupiter assigned the task of grooming the winds and cleaning their stables’ (Baldus 13.36–7 and 59–60). Retracing the long tradition that, starting from Homer, associates the winds with horses will allow us to determine whether we are dealing solely with a metaphor or, instead, with something more ‘concrete’. Literary accounts will be considered,Footnote ³¹ concluding with a brief comparison with iconographic evidence.

2. WINDS AND HORSES, WINDS AS HORSES/HORSEMEN/CHARIOTEERS

In the Iliad, the world of winds and the world of horses communicate with each other through a permeable boundary. Xanthus and Balius, Achilles’ two steeds, are the offspring of Ποδάργη (‘swift-footed’)Footnote ³² —one of the three Harpies, creatures associated with the winds and identified with the θύϵλλαι (‘hurricanes’, ‘squalls’) in the Odyssey Footnote ³³ —and the wind Zephyrus (Il. 16.148–51, 19.400 and 19.415–16). From this extraordinary unionFootnote ³⁴ taking place at the edges of the world, in a meadow along the currents of the Ocean, two incredibly swift horses were born: they could fly in the company of the winds (Il. 16.148 τὼ ἅμα πνοιῇσι πϵτέσθην) and run like the breath of their father (νῶϊ δὲ καί κϵν ἅμα πνοιῇ Ζϵφύροιο θέοιμϵν, Xanthus says to Achilles in Il. 19.415). Similarly, Boreas, having fallen in love with Erichthonius’ mares, assumed the form of a dark-maned stallion (ἵππῳ δ’ ϵἰσάμϵνος … κυανοχαίτῃ) to mate with them; he thus fathered twelve fillies which galloped over the ears of corn and ran on the crest of the waves (Il. 20.221–9).Footnote ³⁵

Similar miraculous unions proliferate in later epics, giving birth to several extraordinary horses: in Antimachus’ Thebaid, Deimos and Phobos, Ares’ horses, are sons of Thyelle, the ‘Hurricane’ (fr. 34 Matthews = 37 Wyss);Footnote ³⁶ in Silius Italicus, Zephyrus is father of Pelorus, Durius’ racehorse (16.363–5 and 13.426–7); in Quintus Smyrnaeus, the four horses of Ares—Aethon, Phlogius, Conabus and Phobus—are the offspring of Boreas and Erinys (8.241–5), while the mythical Arion, sired by Zephyrus and Harpyia, can compete with his father’s swift hurricanes (4.569–70);Footnote ³⁷ in Nonnus of Panopolis, Boreas and a Sithonian Harpy sired the horse Xanthos and the mare Podarge, gifted to Erechtheus by the northern wind when he abducted Oreithyia (Dion. 37.155–61),Footnote ³⁸ while a young Balius, Scelmis’ horse, is the son of Zephyrus and can run on the sea without getting wet (37.335–6).Footnote ³⁹ The belief that mares could be impregnated by the winds is also widespread outside heroic epics: scientific treatises from Aristotle (Hist. an. 6.18, 572a) until the modern era set such a thaumasion, suitably rationalized, in various geographical contexts, while Virgil’s Georgics (re)shape it poetically (see below).Footnote ⁴⁰

A famous simile in the Iliad associates other outstanding steeds, those of Rhesus, with the winds, thus emphasizing their extraordinary speed: they were ‘whiter than snow and in their running resembled the winds’ (Il. 10.437 λϵυκότϵροι χιόνος, θϵίϵιν δ’ ἀνέμοισιν ὁμοῖοι). This line can be considered the archetype of the widespread Greek and Latin tradition which associates winds and horses with rapidity.Footnote ⁴¹ It may be interesting to note that all Homeric passages referred to are related to Thrace: Boreas and Zephyrus blow from there and return there in Il. 9.4–5 and 23.229–30Footnote ⁴² —Boreas blows ‘through horse-breeding Thrace’ in Hes. Op. 507 (διὰ Θρῄκης ἱπποτρόφου)Footnote ⁴³ —while Rhesus is the lord of the Thracians. However, based on these Homeric images, scholars have deduced that in archaic Greece winds were imagined (also) in the form of horses.Footnote ⁴⁴ Given the widespread presence of such imagery in other cultures, a more remote—or a broader polygenetic—origin can actually be suspected.Footnote ⁴⁵

Winds used to ride in the sky of Attic tragedy, as shown by Hugh Lloyd-Jones.Footnote ⁴⁶ In Sophocles’ Oedipus Tyrannus, the ἀϵλλάδϵς ἵπποι mentioned at lines 467–8 (ἀϵλλάδων | ἵππων) should be interpreted as ‘the horses of the storm’ of Boreas, while in the Antigone (985 Βορϵὰς ἅμιππος ὀρθόποδος ὑπὲρ πάγου) Cleopatra, the daughter of Boreas (Βορϵάς), is depicted as ‘riding the air with’ (ἅμιππος) the other Boreads. Likewise, the βόρϵιον [ἵπ]πον mentioned by Ion of Chios in his satyr play Omphale (fr. 17a Snell) would indicate a winged ‘wind-horse of Boreas’.Footnote ⁴⁷ Lastly, Eur. Phoen. 211–13 is highly relevant: Ζϵφύρου πνοαῖς | ἱππϵύσαντος ἐν οὐρανῷ | κάλλιστον κϵλάδημα, translated by Donald Mastronarde as ‘while Zephyrus rode with his breezes, producing a lovely music in the sky’. The verb ἱππϵύϵιν—like the Latin equitare (see below)—can refer either to a horseman, who ‘rides’, or to a horse, which ‘gallops’. Modern scholars lean towards the second possibility:Footnote ⁴⁸ Zephyrus would be represented as a horse.

This brings us to the fundamental, yet long-disputed, testimony of Callim. fr. 110.52–4 Pfeiffer: καὶ πρόκατϵ γνωτὸς Μέμνονος Αἰθίοπος | ἵϵτο κυκλώσας βαλιὰ πτϵρὰ θῆλυς ἀήτης, | ἵ̣ππ̣ο[ς] ἰοζώνου Λοκρικὸς Ἀρσινόης ‘and straightaway the brother of the Aethiopian Memnon | came rushing on, circling his swift wings, a gentle breeze, | the Locrian horse of Arsinoe with her purple gridle’.Footnote ⁴⁹ These lines describe the movement of Zephyrus, ‘the brother of the Ethiopian Memnon’, depicted as a winged (βαλιὰ πτϵρά) horse (ἵππος). Such reading is confirmed by Catullus’ translation: cum se Memnonis Aethiopis | unigena impellens nutantibus aera pennis | obtulit Arsinoes Locricos ales equos (66.52–4). Both passages have been much debated.Footnote ⁵⁰ Despite the suspicions of scholars such as Housman and Pfeiffer, firmly convinced that winds could be represented only as horsemen,Footnote ⁵¹ modern interpreters agree that Zephyrus appears here as a horse.Footnote ⁵²

It should be noted that the passages mentioned so far associate equine traits with Zephyrus and Boreas (or with their offspring, as in Sophocles and Ion), that is, the two winds that take on horse form in the Iliad.Footnote ⁵³ In Roman literature, this depiction extends to Eurus as well, as we can appreciate from Virgil (Aen. 2.417–18, discussed below) and Horace (Carm. 4.43–4 ceu … Eurus | per Siculas equitauit undas). These lines appear to reference the aforementioned passage from Euripides’ Phoenissae, which is also set in Sicily (210–11 ὑπὲρ ἀκαρπίστων πϵδίων | Σικϵλίας). Just like the Greek ἱππϵύϵιν, the Horatian equitare might refer both to a rider and to a horse.Footnote ⁵⁴ Modern commentators tend to lean towards the second possibility.Footnote ⁵⁵

Further evidence comes from imperial and late antique poetry. The common association between winds and horses takes on an interesting guise in Ov. Fast. 4.391–2 (Circus erit pompa celeber numeroque deorum, | primaque uentosis palma petetur equis ), where the adjective uentosis, besides indicating speed, may no doubt recall the Homeric tradition of horses sired by winds.Footnote ⁵⁶ Also worth mentioning are Quintus’ Posthomerica: at 4.509–21, the chariots, drawn by horses as swift as Harpies, raise dust at their departure as when Notus and Zephyrus raise smoke or mist on the mountains; at 4.548–55, racing horses are compared to the swift gusts of the Boreas and the tempestuous Notus.

More substantial evidence can be found in Greek literature from Egypt. In the so-called ‘Strasbourg Cosmogony’, a fragmentary πάτριον poem preserved in P.Strasb. 481, probably composed in the fourth century in Hermopolis Magna,Footnote ⁵⁷ the blowing of winds is said to be ‘unbridled’ (recto, line 36 ἀχάλινον), thus implying an equine image.Footnote ⁵⁸ In Nonnus’ Dionysiaca, a compound of the Euripidean ἱππϵύϵιν (Phoen. 212, discussed above) is used many times in relation to the winds: at 2.646 the whirlwinds (2.644 στροφάλιγγϵς ἀέλλης) born from Typhon’s limbs unsettle the calm sea with their riding (ἀκλύστοιο καθιππϵύουσι γαλήνης); at 39.377–83, the four winds, engaged in naval battle against Dionysus’ enemies, are depicted as ‘riding their storms’ (καὶ πίσυρϵς κατὰ πόντον ἐφιππϵύοντϵς ἀέλλαις | κύματα πυργώσαντϵς ἐθωρήχθησαν ἀῆται …); at 41.44–6 Zephyrus is depicted as ‘riding’ (λιγυηχέι ταρσῷ | Ἑσπϵρίων Ζϵφύροιο καθιππϵύοντος ἐναύλων).Footnote ⁵⁹ Other equine images referring to (storm) winds can be found in Dion. 28.253 πλωτῆρα συνιππϵύοντα θυέλλαις and 47.358 … μιν ἀσταθέϵσσι συνιππϵύοντα θυέλλαις,Footnote ⁶⁰ both passages clearly recalling Oppian, Halieutica 5.344 αἰϵὶ δυσκϵλάδοισι συνιππϵύοντϵς ἀέλλαις, where fishermen are involved in the impetuous movement of the storm.Footnote ⁶¹ In turn, Nonnus’ poetry, especially Dion. 39.377–83, inspired John of Gaza, who in his Tabula Mundi portrays the four winds personified (Ἀῆται), with winged feet and heads, and their respective four impetuous horses representing the storms (ἄϵλλαι); John also compares the group of the four winds to a charioteer trying to restrain the momentum of his racing quadriga (lines 250–96 Lauritzen/Gigli [= 225–71 Friedländer], especially 257–71 for the comparison).Footnote ⁶² It is worth noting that Nonnus and John are the only remaining literary accounts in which the winds are unequivocally depicted as horsemen, not as horses or charioteers; in both cases, however, the winds–horsemen are one with their mounts, the storm–horses.

We thus come back to Virgil. In my opinion, the equine traces identified in Aeneid Book 1 are to be contextualized within this long-standing tradition, which goes back to Homer. In this regard, other passages from Virgil’s œuvre should be recalled. Take the following passage from the Georgics, where the impregnation of mares by Zephyrus is described (3.271–9):

Virgil gives poetic form to the Aristotelian thaumasion of the mares impregnated by the winds (see above), while at the same time recalling the Iliadic episodes of Boreas and Zephyrus.Footnote ⁶³ Such Homeric reminiscence is prepared, a few dozen lines earlier, by an elaborate simile associating a galloping young stallion with the breath of Aquilo, that is, Boreas (G. 3.193–201):

As much as the comparison between horses and winds is commonplace in Virgil’s time,Footnote ⁶⁵ it should be appreciated how lines 195, 201 and, within the simile, 198–200 give new life to the Homeric image of Zephyrus’ twelve fillies: αἳ δ’ ὅτϵ μὲν σκιρτῷϵν ἐπὶ ζϵίδωρον ἄρουραν, | ἄκρον ἐπ’ ἀνθϵρίκων καρπὸν θέον οὐδὲ κατέκλων· | ἀλλ’ ὅτϵ δὴ σκιρτῷϵν ἐπ’ ϵὐρέα νῶτα θαλάσσης, | ἄκρον ἐπὶ ῥηγμῖνος ἁλὸς πολιοῖο θέϵσκον (Il. 21.226–9). The steed ‘flies’ (194 uolans, 201 uolat), one wonders whether more by mere metaphor or rather by an effect of permeability between comparatum and comparandum, as is characteristic of Virgilian similes.Footnote ⁶⁶ The same image is reworked, with greater breadth and adherence to the model, for Camilla’s debut in the Aeneid (7.803–11):

Lines 808–11 precisely refashion the image of Il. 21.226–9 (four hexameters in both cases), although now applied to Camilla’s running on footFootnote ⁶⁷ —but here she appears in all likelihood on horseback, since she is leading her cavalry squadron (7.804).Footnote ⁶⁸ A final passage from the Aeneid must be cited: a simile likening the Greeks’ onslaught during the final night of Troy to the unleashing of a storm (Aen. 2.416–19):

After Zephyrus and Notus, the eastern wind is mentioned as laetus Eois | Eurus equis (2.417–18). His horses are said to be ‘of Eos/Aurora’ because Eurus blows from the east.Footnote ⁶⁹ In addition, according to Hesiod (Theog. 378), Eos was believed to be the mother of the winds,Footnote ⁷⁰ so the adjective may also have a genealogical meaning. Clearly, here we are not dealing with a wind-horse but rather with a wind driving several horses (equis), that is—we should deduce—a charioteer.Footnote ⁷¹ Not surprisingly, the Homeric phrase ἵπποισιν … ἀγαλλόμϵνος, of which laetus … equis has been seen as a calque,Footnote ⁷² is used of a warrior, Asius, son of Irtacus, ‘proud of the horses and the chariot’ (Il. 12.114 ἵπποισιν καὶ ὄχϵσφιν ἀγαλλόμϵνος).Footnote ⁷³ Furthermore, this expression is taken up, albeit with a significant variation, by Valerius Flaccus in the above-mentioned phrase laeti | Thraces equi (1.610–11), where the winds are seemingly represented as four charioteers.

Aen. 2.417–18 would be the first literary evidence depicting a wind as a charioteer. This representation is also attested by archaeological sources. In the LIMC entry ‘Venti’, Erika Simon cites, as the oldest testimony for the ‘Winde als kosmische Pferdeführer’, a fresco from Nero’s Domus Aurea; the other considered examples also date from the Imperial period (second to fourth centuries).Footnote ⁷⁴ Interestingly, in these depictions the winds drive the horses of the Sun (sometimes also those of the Moon); in particular, in two of them it seems to be Eurus performing such function.Footnote ⁷⁵ This could further motivate the image laetus Eois | Eurus equis in Aen. 2.417–18: Virgil might be alluding to the role of Eurus as the Sun’s charioteer, whose horses ‘rise’ in the east (Eois) and traverse the sky until sunset.Footnote ⁷⁶

The earliest instance of the motif of ‘winds as horses’ cited by Erika Simon is the eastern frieze of the Altar of Zeus in Pergamon: the winged horses yoked to Zeus’s chariot have been recognized since the nineteenth century as the four cardinal winds; the remains of the inscriptions support this interpretation.Footnote ⁷⁷ Late ancient Greek epics portray the four winds similarly yoked: in Quintus, Εὖρος καὶ Βορέης, Ζέφυρος δ’ ἐπὶ τοῖσι Νότος τϵ pull Zeus’s chariot in an Olympian theomachy stopped by his intervention (12.192);Footnote ⁷⁸ according to Nonnus, in the fight between Zeus and Typhoeus ἵπποι δὲ Κρονίωνος ὁμόζυγϵς ἦσαν ἀῆται (Dion. 2.423).Footnote ⁷⁹

To sum up: not all the evidence examined has the same weight; however, if the proposed reconstruction is correct, we may draw the following tentative conclusions. The representation of winds as horses seems to be the oldest ‘equine’ representation: Zephyrus and Boreas appear as horses in Homer, Callimachus and Catullus,Footnote ⁸⁰ while the four winds are harnessed together in the Pergamon gigantomachy. These should be considered as firm points. Charioteers and horsemen are likely to be a later development, combining both an anthropomorphic and a theriomorphic aspect. The first charioteer wind to appear is seemingly Euros in Aeneid Book 2—not surprisingly in Roman literature, where circus imagery is pervasiveFootnote ⁸¹ —while all four winds drive chariots in Valerius Flaccus; contemporary iconographic evidence complements these literary accounts. Concerning the winds as horsemen: Attic tragedy mentions Boreas’ horses, while two fourth-century ceramics depict him as a horseman abducting Oreithyia.Footnote ⁸² No unequivocal literary evidence for the four winds as horsemen can be found before Nonnus and John of Gaza, nor does similar iconographic evidence appear to be known, which nevertheless must have been widespread, as the lost artwork (painting or mosaic) described in John’s poem suggests. In any case, Horace and Euripides cannot support Housman’s and Pfeiffer’s categorical assumption that winds could only be represented as horsemen.Footnote ⁸³

At this point, it should be clear that Virgil had both a literary and an iconographic tradition behind him that represented the winds as horses. Some scholars, as we have seen, have taken the equine elements in the tempest episode as metaphorical. Of course, the boundary between real and metaphorical is problematic and often insidious. Yet in this case such pre-existing tradition may suggest that Virgil—without explicitness, as is typical of his art—is not comparing the winds released by Aeolus to horses, but is describing them as actual horses. It is time now to draw some interpretative conclusions.

3. HORSES IN THE TEMPEST: IMPLICATIONS

First of all, evoking the image of galloping horses and the beginning of a chariot race brings Virgil’s readers back to a fundamental passage in the Georgics, which occurs at the end of the first book (1.511–14):

Damien Nelis has highlighted the political and metapoetic significance of this scene: the chariot symbolizes the res publica, while the charioteer who can no longer control it (Mars or Caesar-Octavian, as Servius already thought?) represents the Civil War; at the same time, Virgil suggests the image of the ‘chariot of song’, of Pindaric origin, as a metaphor for the epic ‘flight’ in the didactic poem.Footnote ⁸⁴ The parallels with Aeneid Book 1, where the storm arouses themes of civil war,Footnote ⁸⁵ are clear. Furthermore, since a tradition dating back at least to the Pergamon Altar depicts the four winds harnessed to Jupiter’s chariot in the gigantomachy, we can read the tempest episode in a contrastive way: here the winds, unleashed by Aeolus on Juno’s orders, (re-)endanger the ordered cosmos ruled by Jupiter.

Envisioning the winds as horses also allows us, I believe, to understand better the overall dynamic opposing Aeolus to Neptune at the beginning of the poem. Since Homer, the sea-god is in fact also the lord and tamer of horses, as well as their creator, as Virgil reminds us in the prologue to the Georgics: tuque o, cui prima frementem | fudit equum magno tellus percussa tridenti, | Neptune (1.12–14). Now, this gesture performed by Neptune—striking the ground with his trident to make the frenzied horse emerge from it—seems precisely imitated by Aeolus when he strikes the mountain rock to release the winds: cauum conuersa cuspide montem | impulit in latus (1.81–2). It is not clear what exactly the term cuspis designates, which most interpreters consider to be a generic hasta.Footnote ⁸⁶ However, I believe that Vinzenz Buchheit and Michael Putnam hit the mark by acknowledging a reference to Neptune’s trident: Virgil seems to be playing on the polyvalence of this term, making the words of the sea-god even more poignant: ‘non illi imperium pelagi saeuumque tridentem, | sed mihi sorte datum’ (1.138–9).Footnote ⁸⁷ In short, the text suggests the image of Aeolus wielding Neptune’s trident, with it repeating—better, usurping—one of his most famous gestures: bringing the horse(s) out of the rock.

Neptune calms the Virgilian storm standing on his horse-drawn chariot, as befits the equestrian god of the sea (cf. 1.156 flectit equos curruque uolans dat lora secundo). A similar scene will occur later in the poem, where the bridling of Neptune’s wild steeds and the quelling of the storm are closely linked (Aen. 5.817–21):

One final point. Late imperial sources—the most recent of which is Tertullian’s De spectaculis—establish an explicit analogy between the Roman circus and the cosmos, both as a whole and in its individual constituents (the arena would correspond to the earth, the euripus to the sea, the four colours identifying the factions with the four elements or the four seasons, etc.); several scholars convincingly date this association much earlier, at least to the beginning of the Imperial era.Footnote ⁸⁸ Was Virgil familiar with this symbolism? If so, the chariot race imagery in the cosmic storm of Aeneid Book 1 would acquire even more significance. When staging a cosmic upheaval, Virgil would be conjuring up the atmosphere of the Circus Maximus, a miniature world in which the social order as well as—symbolically—the cosmic order were temporarily subverted, where chaos and confusion used to reign, where the Roman ciuitas was reduced to angry mobs excited by partisan furor, fighting for the triumph of one faction or colour over the othersFootnote ⁸⁹ —four colours just like the four cardinal winds, one of which, the White, as we learn from Tertullian (De spect. 9.5) and Isidore (Etym. 18.41.2), was consecrated to the Zephyrs.Footnote ⁹⁰

sunt uenti corpora caeca, wrote Lucretius (1.295). Yet Virgil’s mythological epic required them to be visible and have bodies. What animal better to represent them than horses, chthonian creatures and emblems of speed and instinctual violence since Plato,Footnote ⁹¹ which long-standing literary tradition had represented as winds? Like Balius and Xanthus in the Iliad, Virgilian winds are responsive to Neptune’s rational speech: a further clue, perhaps, of their Homeric pedigree.