The Hellenistic Epyllion and Its Descendants

Adrian Hollis, Keble College, Oxford

In a volume dedicated to the Greek literature of late antiquity and the early Byzantine period it may seem strange to start this contribution from Alexandria in the third century BC. Yet many have sensed a certain community of spirit between Alexandria and Constantinople, and my purpose here is to trace one of the most interesting continuities. The Hellenistic ‘miniature epic’ or ‘epyllion’ [1] is defined by the Oxford Classical Dictionary (third edition, p. 550) as a narrative poem of up to c. 600 hexameters, usually about an episode from the life of a mythological hero or heroine. Some have questioned the very existence of this genre—either in Greek, or in Latin, or in both languages. [2] Everyone knows that the ancients did not use the term in the way that is familiar from modern scholarship, but (in my opinion) it remains useful and does describe a genuine type of poem. We shall here follow the history of the epyllion from the third century BC, as it goes underground for long periods and finally re-emerges c. AD 500 in the Constantinople of the emperor Anastasius. It is a history of strange transformations and combinations with a wide range of other literary genres. We shall alight in most of the intervening centuries, including the third century AD, which seems relatively barren in poetic terms. I will deal with Latin as well as Greek; [3] at different times the one language, and then the other, seizes the initiative.

If we had to choose one poem to represent the epyllion, it would surely be Callimachus’ Hecale. [4] The subject matter is heroic (Theseus’ victory over the monstrous Bull of Marathon) but focuses more on the old woman Hecale, who entertained the hero in her cottage near Marathon. Callimachus’ learning is shown by his adoption of a little-known foundation myth of a small and obscure Attic deme, and his recreation of everyday life in the Attic countryside, with the aid of Old Comedy, commentaries thereon and specialist monographs. [5] Enough survives to show that the Hecale combined learning with strong emotion, fantasy and humour. Its influence was enormous and can be discerned in later epyllia, Latin (e.g. Catullus 64 and the pseudo-Virgilian Ciris) [6] as much as Greek. Above all, Callimachus popularized the hospitality theme (entertainment of a god or hero by a poor old person or couple in the countryside). [7] Composed in Egyptian Alexandria, the poem was known as far away as Alexandria of Arachosia (modern Kandahar in Afghanistan); [8] it was still being copied on papyrus about AD 600 [9] and may have survived intact until AD 1205. [10]

We can only estimate the length of the Hecale, but it seems likely to have been of at least 1000 lines—perhaps appreciably longer. [11] An interesting comparison is now available: the hexameter Hermes [12] by Callimachus’ pupil Eratosthenes contained between 1540 and 1670 lines, [13] more than the average for a book of Apollonius’ Argonautica. Should we put these two poems in the same basket as (to take an extreme example) the mere 75 lines of Theocritus 13 (Hylas), which is indeed regularly classified as an epyllion? [14] Of the other comparable poems by Theocritus, most charming is 24 (Heracliscus, 186 lines but not complete). The picture of Alcmena putting her babies to sleep in her husband’s shield [15] combines the heroic and the intimate (24.1–7):

Ἡρακλέα δεκάμηνον ἐόντα ποχ’ ἁ Μιδεᾶτιϲ
Ἀλκμήνα καὶ νυκτὶ νεώτερον Ἰφικλῆα,
ἀμφοτέρουϲ λούϲαϲα καὶ ἐμπλήϲαϲα γάλακτοϲ,
χαλκείαν κατέθηκεν ἐϲ ἀϲπίδα τὰν Πτερελάου
Ἀμφιτρύων καλὸν ὅπλον ἀπεϲκύλευϲε πεϲόντοϲ.
ἁπτομένα δὲ γυνὰ κεφαλᾶϲ μυθήϲατο παίδων ·
“εὕδετ’, ἐμὰ βρέφεα, γλυκερὸν καὶ ἐγέρϲιμον ὕπνον.”
One night, when Heracles was ten months old, Alcmena, the lady of
Midea, bathed him and his brother Iphicles who was younger by
one night, gave them their fill of milk and laid them to rest in
the bronze shield, that fair piece of armour of which Amphitryon
had spoiled Pterelaus when he fell. And, stroking the boys’ heads,
she said, ‘Sleep sweetly, my babies, and wake again’.

The most interesting to me of these poems is Theocritus 25, ‘Heracles the Lion-slayer’ (281 lines, but, in my unfashionable opinion, seriously defective). [16] Many scholars have denied the poem to Theocritus, whether because of the lack of papyrus fragments and grammatical citations which might confirm the authorship, or the differences from Theocritus’ normal style, or a low estimate of the poem’s quality. [17] That the style should be more Homeric than Theocritus’ norm is no great surprise—the same could be said of Callimachus’ Hecale. As to the quality, my own subjective judgment is closer to that of A.S.F. Gow: [18] ‘Unlike the other poems dubiously ascribed to Theocritus it would, if its authenticity were secure, add appreciably to his stature’. There are many links with Callimachus, more with the Molorcus episode in Aetia book [19] than with the Hecale. An (unfulfilled) hint of the hospitality theme is adumbrated when the unnamed old countryman leads Heracles to his steading, boisterously greeted by the dogs (25.60–77).

Thus it seems that substantial epyllia, as long as a book of epic poetry, may have been a feature of the third century BC. At the same time, however, there grew up a tradition of much slighter hexameter poems, represented by Theocritus 13, 22, 24, 26. [20] This was continued in the second century by works such as the Europa of Moschus, [21] which owes less to Callimachus [22] than to Apollonius Rhodius. The way in which the girl casts off troubling thoughts from the previous night and joins her companions in more cheerful mood (Europa 20 ff.) strongly recalls Argonautica 3.828 ff. In the first century BC we would like to know more about the hexameter poems of Parthenius of Nicaea, [23] but the main interest shifts from Greek to Latin. We have two complete specimens: Catullus 64 and the pseudo-Virgilian Ciris. [24] Also we hear of several epyllia by poets associated with Catullus: the Smyrna of Helvius Cinna, Io of Licinius Calvus, Glaucus of Cornificius. All these myths recur in Ovid’s Metamorphoses; at least the Smyrna and the Io have left a considerable afterglow. [25] The same century saw a new development: combination of the manner and matter of an epyllion with quite different poetic genres. Let us start with didactic poetry. The best-known example is in Virgil’s fourth Georgic, an episode of 244 lines (315–558), linking the myth of Aristaeus with that of Orpheus and Eurydice by means of a connexion which Virgil himself may have invented: while trying to escape from Aristaeus’ pursuit Eurydice failed to see a huge snake (458 ‘immanem…hydrum’). [26] It is worth noting that Aristaeus, benefactor of mankind and inventor of much rural technology, often appears in learned Greek poetry (Callimachus, Apollonius, Euphorion, pseudo-Oppian and Nonnus). Brief mythological digressions are found in the didactic poems of Hesiod, Aratus and Nicander; Virgil seems to have expanded this element in the manner of a Hellenistic epyllion.

Perhaps, however, there was an earlier Latin instance of an epyllion-like section in a didactic poem. A scholiast on Bellum Civile 9.701 [27] informs us that Lucan’s account of African snakes is indebted to the Theriaca of Aemilius Macer. [28] Lucan, in a passage of 81 lines (9.619–699) describes the origin of these snakes from blood which dripped from the severed head of the Gorgon Medusa, carried aloft by Perseus as he flew over Libya. Several features of this passage suggest an epyllion, starting with a rather academic debate about the reliability of the poet’s information (621–623):

non cura laborque
noster scire valet, nisi quod volgata per orbem
abula pro vera decepit saecula causa.

His statement that the mythical explanation is false by no means stops the poet from relating it in full. [29] ‘Volgata’ strikes a note of intellectual disdain (cf. Virgil, Georgics 3.4 ‘omnia iam vulgata’), while ‘decepit’ recalls the common charge against poets of deliberately misleading the public. Even the versification resembles that with which we are familiar from Catullus 64 and the Ciris: hexameters forming a sense-unit, with a molossic word (three long syllables) following the masculine caesura (e.g. 667–668 ‘Persea Phoebeos converti iussit ad ortus/Gorgonos averso sulcantem regna volatu’) and the careful arrangement of two nouns at the end of the line, each with its epithet at the beginning (677 ‘lata colubriferi rumpens confinia colli’). [30]

As observed by Richard Hunter, [31] the inclusion of epyllion-like material in a heroic context cannot unreasonably be said to start during Callimachus’ lifetime with certain episodes of Apollonius’ Argonautica. Hunter cites Cyzicus and Cleite, [32] which begins (1.936) with a geographical ecphrasis and ends (like Callimachus’ Hecale) in the explanation of an annual religious custom (1.1075–1077). The nicest example (though relatively brief at 50 lines) of an epyllion within a martial epic [33] is Silius Italicus’ aetiological account of the origin of the Falernian vineyards (Punica 7.162–211), a theoxeny of Dionysus. The obvious Latin model for this would be Ovid’s tale of Baucis and Philemon in Metamorphoses 8.626 ff., which itself goes back to Callimachus’ Hecale. [34] But an even closer fit is the Erigone of Callimachus’ pupil Eratosthenes (see n. 2 above), which described Dionysus’ visit to the old man Icarius and his daughter Erigone. Although written in elegiacs, the Erigone seems in other respects to conform to the traditions of the epyllion. An interesting possible Greek counterpart to Silius’ theoxeny of Bacchus enclosed within a martial epic may perhaps be recognized in the second or third-century Bassarica by Dionysius. [35] These epyllion-like intrusions into poems of a different genre are mostly of about 80 lines (appreciably shorter than a self-standing epyllion); one could add 81 lines of Manilius on Andromeda (5.538–618) and 78 lines of Valerius Flaccus on Io (4.344–421).

A rare glimpse of Greek hexameter poetry in the early second century AD under the patronage of the cultured Hadrian is afforded by the poet Pancrates. His theme was contemporary, not mythological, but cleverly tailored to please an emperor who delighted in poets such as Antimachus of Colophon and Parthenius of Nicaea—no doubt also Callimachus and Euphorion. His poem on the lion-hunt of Hadrian and Antinous would perhaps invite comparison with Theocritus 25 (though in talent Pancrates was far inferior); it included a flower sprung from the blood of the lion, perhaps also a catasterism. We have papyrus fragments suggesting a poem grandiose in style but of fairly modest length; [36] at the end of the papyrus the lion seems about to expire. One cannot, however, be entirely sure that the unnamed author is the Pancrates from whom Athenaeus (15.677d–f) quotes four quite pretty lines:

οὔλην ἕρπυλλον, λευκὸν κρίνον ἠδ’ ὑάκινθον
πορφυρέην γλαύκου τε χελιδονίοιο πέτηλα
καὶ ῥόδον εἰαρινοῖϲιν ἀνοιγόμενον ζεφύροιϲιν ·
οὔπω γὰρ φύεν ἄνθοϲ ἐπώνυμον Ἀντινόοιο.

Of course it was not so hard to write prettily in a catalogue of attractive flowers, but these lines show that something of the Hellenistic masters lingered in Hadrian’s Alexandria; the fourth line particularly recalls Euphorion. [37] In fact the whole episode of Hadrian and Antinous in Alexandria has a Callimachean air. [38] According to Athenaeus (15.677e) the Emperor was so pleased that he enrolled Pancrates in the Alexandrian Museum.

The name of the poet whom we call pseudo-Oppian is unknown. [39] Despite his low reputation, I would suggest that (after Virgil) he provides the two most interesting examples of the epyllion-style intruding into a didactic poem. Virgil was concerned with bugonia, the alleged production of bees from the carcase of a calf (Georgics 4.281–558), [Oppian] (Cyn. 2.100–158) with the origin of Syrian bulls, fancifully said to descend from those which Heracles captured from Geryon in the far west and gave to his friend Archippus, king of Pella/Apamea (114)—though the city of Apamea-on-the-Orontes was only founded by Seleucus Nicator in approximately 300 BC!

The myth is introduced in the best Alexandrian manner (109):

κεῖνοι, τοὺϲ φάτιϲ ἔϲκε Διὸϲ γόνον Ἡρακλῆα… [40]

The flooding by Orontes of the plain of Pella is ascribed to the river-god’s hopeless passion for a water-nymph (Meliboea); rescue came when Heracles cut through the surrounding hills and directed Orontes towards the sea (134–136). Topographical references abound, none of them known otherwise: the mountains (Emblonus from the West, Diocleion from the East); a ‘plain of Heracles’ (149) and a shrine of Memnon (152–153). The poet speaks with pride of his own city (127 ἐμὴν πόλιν). In concluding he makes his literary position clearer by promising to write a future poem on the glories of his homeland (156–7):

ἀλλὰ τὰ μὲν [41] κατὰ κόϲμον ἀείϲομεν εὐρέα κάλλη
πάτρηϲ ἡμετέρηϲ ἐρατῆι Πιμπληΐδι μολπῆι

In familiar fashion the poet, while appearing to promise something for the future, in fact emphasizes what he is doing at this moment. Poems on the history of cities and states, especially their foundation (κτίϲιϲ) start from the third century BC; they gave ample opportunity for mythical narrative (e.g. Apollonius Rhodius fr. 12 Powell, from the Foundation of Lesbos, on the love of Peisidice for her country’s enemy Achilles), even if—as here—the mythology had to be made up or introduced from elsewhere because the foundation was too recent. [42] A final thought on Apamea: much of what [Oppian] writes is in harmony with the learned poets of the third century BC, and perhaps it is worth mentioning that Euphorion of Chalcis, after moving to become royal librarian to Antiochus III of Syria, is said [43] to have died and been buried in Apamea-on-the-Orontes. Could he have helped to create some of the local mythology?

We remarked earlier that Aristaeus plays a considerable part in the Greek poetry of the Hellenistic and Roman Imperial periods (as well as in Virgil’s Fourth Georgic). He is very much to the fore in my second passage of pseudo-Oppian, more substantial at 90 lines (Cyn. 4.230–319) and even more clearly breathing the atmosphere of a Hellenistic epyllion. The starting point is that leopards can be caught more easily when intoxicated, since they were once wine-loving devotees of Bacchus. [44] The scene is set first in Thebes, where Pentheus already rages and tyrannizes. Dionysus is still an infant and must quickly be removed from the city by his aunts, who here—contrast Euripides, Bacchae 26—play quite a positive role. From time to time we are reminded of Euripides. [45] For example, when the Bacchants implore Dionysus to punish Pentheus (302–303):

ἅπτε ϲέλαϲ φλογερὸν πατρώϊον, ἂν δ’ ἐλέλιξον
γαῖαν, ἀταρτηροῦ δ’ ὄπαϲον τίϲιν ὦκα τυράννου

one cannot fail to remember Bacchae 594–595 ἅπτε κεραύνιον αἴθοπα λαμπάδα·/ ϲύμφλεγε ϲύμφλεγε δώματα Πενθέοϲ. As the narrative progresses we find many motifs at home in a learned epyllion: an etymology of Mount Mēros (241, so named because Dionysus was born from his father’s thigh), concealment of the child in a basket or chest (244, compare Callimachus, Hecale fr. 70.14 and fr. 166 H.), growing of a vine where it was previously unknown (253–4, cf. Eratosthenes, Erigone and Silius’ Falernus), performance of new religious rites and their communication to the locals (246, [46] 248–250), the kindly old person (in 258 a fisherman, compare Hecale, Molorcus, Icarius, Baucis and Philemon, Falernus, Brongus) who accepts and welcomes the strangers though quite unaware of their importance. Aristaeus (265 ff.) is introduced as the First Inventor (πρῶτοϲ εὑρετήϲ) of countless pastoral and agricultural techniques, as in Virgil, [47] Georgics 4.315 ff. ‘Quis deus hanc, Musae, quis nobis extudit artem?/unde nova ingressus hominum experientia cepit?/ pastor Aristaeus…etc.’

The transformation of the women into leopards and Pentheus himself into a bull is described in a manner which seems to us very ‘Ovidian’ [48] —we watch it happening item by item (309–313):

Πενθέα μὲν δὴ ταῦρον ἐδείξατο φοίνιον ὄμμα,
αὐχένα τ’ ἠιώρηϲε, κέραϲ τ’ ἀνέτειλε μετώπου ·
ταῖϲι δὲ γλαυκιόωϲαν ἐθήκατο θηρὸϲ ὀπωπήν,
καὶ γένυαϲ θώρηξε, κατέγραψεν δ’ ἐπὶ νώτου
ῥινὸν ὅπωϲ νεβροῖϲι, καὶ ἄγρια θήκατο φῦλα.
Pentheus he made a bull of deadly eye and arched his neck
and made horns spring from his forehead. But to the women
he gave the grey eyes of a wild beast and armed their jaws
and on their back put a spotted hide like that of fawns, and
made them a savage race.

The most strikingly Callimachean element in the whole episode is the brief literary/mythical polemic at the end (316–319); to this we shall return. Meanwhile one or two small details: note that seven of the ninety hexameters in this episode have a spondaic fifth foot—a proportion not comparable with that in some third-century BC poets, [49] but still significant. Among pseudo-Oppian’s vocabulary, one may single out the epithets ‘Aonian’ = Boeotian (250, 276) and ‘Inoan’ (274), both perhaps invented (and certainly made popular) by Callimachus. [50]

We now know for certain [51] something which several scholars had earlier argued on metrical grounds, that Triphiodorus is a predecessor of Nonnus, not a follower. [52] He achieved the Sack of Ilion in almost 700 lines. The poem has quite a good reputation. [53] I would like to touch on just one point. We have repeatedly seen the strong influence which Callimachus’ Hecale exercised upon the writers of later epyllia, and that is still discernible in Triphiodorus, whose debt to Hellenistic poetry is relatively small. [54] When, after the fall of the city, Menelaus saves the family of his erstwhile host Antenor (657–658):

μειλιχίηϲ προτέρηϲ μεμνήμενοϲ ἠδὲ τραπέζηϲ
κείνηϲ, [55] ἧι μιν ἔδεκτο γυνὴ πρηεῖα Θεανώ,

he must have in mind Hecale fr. 80.1–2 (probably uttered by the neighbours at Hecale’s funeral):

ἴθι, πρηεῖα γυναικῶν
τὴν ὁδόν, ἣν ἀνίαι θυμαλγέεϲ οὐ περόωϲι
Go, gentle among women, on the road along which heartbreaking
pains do not penetrate. [56]

Much the most remarkable figure among Greek poets of the late Imperial/early Byzantine period is Nonnus. A native of Panopolis in Egypt, he composed his poetry in Alexandria. [57] He was probably raised as a Christian, [58] and made a hexameter paraphrase, [59] longer than the original, of St John’s Gospel, showing deeper and more detailed knowledge of Christian theology than would be required of someone who had merely accepted a commission. The forty-eight books of Nonnus’ Dionysiaca represent not only the longest Greek epic surviving from antiquity, but also the most astonishingly varied—the poet’s initial invocation of Proteus (Dion. 1.13 ff.) is appropriate stylistically as well as geographically. [60]

Many episodes in the Dionysiaca reflect both the manner and the matter of a Hellenistic epyllion, [61] and the vast scale of Nonnus’ undertaking allows some of them to be treated at considerable length. Examples: Dion. 5.212–551 (Aristaeus and Actaeon); 38.1–434 (Phaethon); 47.1–264 (Erigone); 17.37–86 (rustic hospitality of Brongus). The last-mentioned imitates Callimachus’ Hecale (and perhaps Eratosthenes’ Erigone) and is explicitly compared to Callimachus’ Molorcus [62] (Dion. 17.52–54):

οἷα Κλεωναίοιο φατίζεται ἀμφὶ Μολόρκου
κεῖνα, τά περ ϲπεύδοντι λεοντοφόνουϲ ἐϲ ἀγῶναϲ
ὥπλιϲεν Ἡρακλῆι
Such as they say Molorcus in Cleonae provided for Heracles
as he hastened on his way to fight the lion. [63]

Let us pause for a moment on Dionysiaca 15.169–16.405 (altogether 659 lines, a respectable number for an old-style epyllion). We find here a foundation myth (a category mentioned earlier) of Nicaea in Bithynia, [64] presented as a pastoral epyllion [65] with all the proper mannerisms, including—astonishingly for an epic poem— a refrain (15.399, 403, 409, 413):

Βούτηϲ καλὸϲ ὄλωλε, καλὴ δέ μιν ἔκτανε κούρη
the fair herdsman has perished; a beautiful girl has killed him

which ‘seemed’ (398 ἔοικε) to come from the mouth of a cow! This takes ad absurdum the pastoral convention that Nature joins humanity in lamenting the deceased. In fact the whole episode is written in a light style, almost amounting to a parody of the traditional bucolic lament.

One feature of the learned epyllion-style which persisted up to the time of Nonnus is literary and mythological polemic. Callimachus’ Hecale preserves just a hint of it: οἵ νυ καὶ Ἀπόλλωνα παναρκέοϲ Ἠελίοιο/χῶρι διατμήγουϲι καὶ εὔποδα Δηωίνην/Ἀρτέμιδοϲ (fr. 103 H.). The speaker seems to express disapproval of those who ‘part asunder both Apollo from the all-shining sun and Deo’s fleet-footed daughter [Persephone] from Artemis’. Puzzling, since (at least in the second case) it was much more common to make the distinction than to deny it. In Euphorion fr. 40 Powell, [66]

Πορφυρέη ὑάκινθε, ϲὲ μὲν μία φῆμιϲ ἀοιδῶν
Ροιτείειϲ ἀμάθοιϲι δεδουπότοϲ Αἰακίδαο
εἴαροϲ ἀντέλλειν γεγραμμένα κωκύουϲαν
Purple hyacinth, one story of poets is that, on the Rhoetean [Trojan]
sands, after the fall of the descendant [Ajax] of Aeacus, you
sprang up from his blood with a lament in your inscription

we may suspect that, having mentioned the better-known story of the hyacinth’s origin, Euphorion will go on to reject it in favour of the other, more obscure, myth about the Spartan boy whom Apollo accidentally killed with a discus.

The pseudo-Virgilian Ciris provides two nice examples: on the question whether or not Scylla daughter of Nisus should be identified with the sea-monster (54 ff.):

complures illam magni, Messalla, poetae [67]
(nam verum fateamur: amat Polyhymnia verum)
longe alia perhibent mutatam membra figura
Scyllaeum monstro saxum infestasse voraci…
…(62) sed neque Maioniae patiuntur credere chartae…

In the sequel we encounter the academic perversity deliberately cultivated in such a context: having decided that the daughter of Nisus is not to be identified with the sea-monster, nonetheless the poet cannot resist adding 23 lines (66–68) on various reasons for the sea-monster’s transformation. [68] The second passage of the Ciris ranges learning against sentiment: Carme’s daughter Britomartis did indeed escape Minos’ amorous pursuit, and may even have achieved divine status—but as far as her mother is concerned she is simply lost (303–306):

unde alii fugisse ferunt et nomen Aphaeae
virginis assignant; alii, quo notior esses,
Dictynnam dixere tuo de nomine lunam. [69]
sint haec vera velim; mihi certe, nata, peristi.

Rival poets, who hold different views, may not merely be mistaken; often they are accused of deliberately misleading the public. Thus at the end of his narrative about Pentheus and the Bacchants [Oppian] waxes indignant (Cyn. 4.316–319):

τοίαδ’ ἀείδοιμεν, τοῖα φρεϲὶ πιϲτεύοιμεν ·
ὅϲϲα Κιθαιρῶνοϲ δὲ κατὰ πτύχαϲ ἔργα γυναικῶν
ἢ μυϲαρὰϲ κείναϲ, τὰϲ ἀλλοτρίαϲ Διονύϲου,
μητέραϲ οὐχ ὁϲίωϲ ψευδηγορέουϲιν ἀοιδωί
Such things let us sing, such things let us believe in our hearts!
But as for the deeds of the women in the glens of Cithaeron, or
the tales told of those wicked mothers, alien to Dionysus,
these are the impious falsehoods of poets.

The cause of [Oppian]’s ire is clearly that, in the version which he condemns, the rending of Pentheus takes place with all participants in human form. That would be impious; on the other hand O.’s version (for which we have no parallel), according to which the Bacchants have been transformed into leopards and Pentheus into a bull, is apparently acceptable.

Unusually, we may be able to identify a target of O.’s polemic. Having described the bloodstained return of Pentheus’ mother and aunts from Cithaeron, Theocritus expresses total indifference—nor would he be any more concerned if the victim were a child of eight or nine years (26.27–30): [70]

οὐκ ἀλέγω · μηδ’ ἄλλοϲ ἀπεχθομένω Διονύϲωι
φροντίζοι, μηδ’ εἰ χαλεπώτερα τῶνδε μογήϲαι,
εἴη δ’ ἐννεατὴϲ ἢ καὶ δεκάτω ἐπιβαίνοι ·
αὐτὸϲ δ’ εὐαγέοιμι καὶ εὐαγέεϲϲιν ἅδοιμι.
I care not. And let not another care for an enemy of Dionysus—
not though he suffer a fate more grievous than this and be in his
ninth year or entering his tenth. But may I myself be pure, and
pleasing in the eyes of the pure.

Nonnus too enters literary-polemical mode, after describing the mass catasterism which concludes his episode of Erigone (Dion. 47.256 ff.):

καὶ τὰ μὲν ἔπλαϲε μῦθοϲ Ἀχαιικὸϲ ἠθάδα πειθώ
ψεύδεϊ ϲυγκεράϲαϲ · τὸ δ’ ἐτήτυμον…
Such is the fiction of the Achaean story, mingling as
usual persuasion with falsehood; but the truth is…

Three categories are recognized here: fiction, plausibility and truth. Thus Callimachus, after rejecting the standard account of the sons of Cronos drawing lots for their spheres of rule, expresses the hope that he could lie more plausibly (hymn. 1.65 ψευδοίμην ἀίοντοϲ ἅ κεν πεπίθοιεν ἀκουήν). In fact, parts of Nonnus’ ‘true’ version are far from clear; things might be different if we possessed Eratosthenes’ Erigone. [71]

Even one or two generations after Nonnus the early Byzantines were still in touch with their Hellenistic inheritance: during the reign of the emperor Anastasius (491–518) Marianus of Eleutheropolis [72] paraphrased Callimachus’ Hecale, Hymns, Aetia and Epigrams in 6,810 iambics; in Egypt Callimachus’ poems were still being copied and annotated. [73] The same period saw the composition of what can reasonably be described as the last two Greek epyllia: the Rape of Helen by Colluthus and Hero and Leander by Musaeus, of 394 and 343 lines respectively. Just as, by comparing Catullus 64 and the pseudo-Virgilian Ciris, we may be able to appreciate the characteristic style and versification of a Latin epyllion in the first century BC, so Colluthus and Musaeus exhibit a common style of Greek hexameter writing—that of Nonnus, who had evolved a metre with even tighter rules than those of Callimachus. [74]

Of the two, I confess to warmer feelings for the Rape of Helen. [75] It is certainly the more learned poem, in the Alexandrian tradition. For example, from where did Colluthus learn that the timber used to build Paris’ ships came from the peak of Mount Ida which was called Phalacra? As far as we can tell, only from Lycophron (Alexandra 24) or possibly Callimachus (fr. 34 Pfeiffer). In 174–5 Colluthus combines two Callimachean reminiscences:

φαϲί ϲε, μῆτερ Ἄρηοϲ, ὑπ’ ὠδίνεϲϲιν ἀέξειν
ἠυκόμων Χαρίτων ἱερὸν χορόν
They say that you, mother of Ares, brought forth with birth-pangs
the sacred band of the fair-tressed Graces.

The phrase in the vocative case, ‘mother of Ares’, [76] appears identically placed in the hexameter of Callimachus, fr. 634, [77] and Hera’s parenting of the Graces was just one of several opinions put forward by the poet himself. As another possible sign of Hellenistic influence, note Colluthus 240–248 on the Spartan boy Hyacinthus, perhaps in imitation of Euphorion. [78] Colluthus also restores one small but significant feature of the epyllion (both Greek and Latin) which is totally absent from Nonnus and Musaeus: the spondaic fifth foot. [79] Clearly there were readers in Colluthus’ time capable of appreciating such subtleties.

In Musaeus, on the other hand, I have found no more than one substantial and wholly convincing imitation of Callimachus. [80] When Musaeus describes Hero and Leander as ἀμφοτέρων πολίων περικαλλέεϲ ἀϲτέρεϲ ἄμφω (22, ‘both outstandingly beautiful stars of their two cities’), he clearly has in mind Callimachus’ Acontius and Cydippe, καλοὶ νηϲάων ἀϲτέρεϲ ἀμφότεροι (fr. 67.8). It would not be surprising if Musaeus turned out to have made more extensive use of the two Callimachean love stories, Acontius and Cydippe (frs. 67–75 Pf.) and Phrygius and Pieria (frs. 80–83). Musaeus is generally thought also to have used an unknown Hellenistic model with which Ovid too was familiar. [81]

We have traced the Greek epyllion, in various forms and combinations, up to approximately AD 500. Could the same be done for Latin? A century earlier Latin poetry enjoyed a significant revival in the hands of such as Claudian, Prudentius, Ausonius, and Rutilius Namatianus. None of these, however, wrote anything that could remotely be described as an epyllion; ‘The Rape of Proserpina’ was a possible subject, but Claudian’s treatment of it is unmistakably epic. The best that I can do [82] for the time of Anastasius is to summon up an inferior figure [83] who lived in Carthage rather than Rome, Blossius Aemilius Dracontius. Some of his subjects are traditional to the epyllion (e.g. Hylas, The Rape of Helen) and he wrote on a scale which matched some Greek epyllia of the third century BC (974 lines on Orestes, 655 on Helen). Dracontius occasionally shows himself capable of the ingenious surprises which we might hope for in a learned poet, for example in the unique setting of the love affair: Menelaus has indeed gone to Crete (441), but Paris and Helen meet not in Sparta but in Cyprus. Helen has gone there, in her husband’s absence, to celebrate a festival of Venus (435–441) and Paris has been driven by adverse gales (425–429). It would, however, be unjust to even the lesser poets whom we have been considering to mention Dracontius in the same breath.

Works Cited

Allen, W. ‘The Epyllion,’ Transactions of the American Philological Association 71 (1940): 1–26.

Bernard, P., G.-J. Pinault and G. Rougemont. ‘Deux nouvelles inscriptions grecques de l’Asie centrale,’ Journal des Savants (2004): 227–332.

Chuvin, P. Mythologie et géographie dionysiaques: Recherches sur l’oeuvre de Nonnus de Panopolis (Clermont-Ferrand: 1991).

Courtney, E. The Fragmentary Latin Poets (Oxford: 1993).

El-Abbadi, Mostafa. ‘The Alexandrian Library in History’, in A. Hirst and M. Silk (eds), Alexandria, Real and Imagined (Ashgate: 2004), 167–183.

Forbes Irving, P.M.C. Metamorphosis in Greek Myths (Oxford: 1990).

Gerlaud, B. Triphiodore, la Prise d’Ilion, ed. Budé (Paris: 1982).

Gow, A.S.F. Theocritus (Cambridge: 1950).

Gutzwiller, K.J. Studies in the Hellenistic Epyllion (Königstein: 1981).

Heitsch, E. Die griechischen Dichterfragmente der romischen Kaiserzeit (2nd ed., Göttingen: 1963).

Hollis, A.S. Metamorphoses. Book VIII (Oxford: 1970).

———. Callimachus: Hecale (Oxford: 1990).

———. ‘[Oppian], Cyn. 2.100–158 and the Mythical Past of Apamea-on-the-Orontes’, Zeitschrift für Papyrologie und Epigraphik 102 (1994): 153–166.

———. ‘Myth in the Service of Kings and Emperors’, in J.A. López Férez (ed.), Mitos en la literatura griega helenística e imperial (Madrid: 2003), 1–14.

———. ‘Hecale’s Babies’, Zeitschrift für Papyrologie und Epigraphik 148 (2004): 115– 116.

———. Fragments of Roman Poetry c. 80 BC–AD 20 (Oxford: forthcoming 2007).

Hunter, R.L. ‘Epic in a Minor Key’, in M. Fantuzzi and R. Hunter (eds), Tradition and Innovation in Hellenistic Poetry (Cambridge: 2004), 191–245.

Hutchinson, G., R.G.M. Nisbet, and P.J. Parsons. ‘Alcestis in Barcelona’, ZPE 52 (1983): 31–36.

Kost, K. Musaios, Hero und Leander (Bonn: 1971).

Lightfoot, J.L. Parthenius of Nicaea (Oxford: 1999).

Livrea, E. Dionysii Bassaricon et Gigantiadis Fragmenta (Rome: 1973).

Lloyd-Jones, H. and P.J. Parsons. Supplementum Hellenisticum (Berlin: 1983).

Lyne, R.O.A.M. Ciris: A Poem Attributed to Virgil (Cambridge: 1978).

Page, D.L. Greek Literary Papyri, Loeb Classical Library (London: 1942).

Pfeiffer, R. Callimachus (Oxford: vol. 1 [Fragments] 1949; vol. 2 [Hymns and Epigrams] 1953).

Powell, J.U. Collectanea Alexandrina (Oxford: 1925).

Rosokoki, A. Die Erigone des Eratosthenes (Heidelberg: 1995).

Spanoudakis, Konstantinos. Philitas of Cos (Leiden: 2002).

West, M.L. Greek Metre (Oxford: 1982).


[ back ] 1. Gutzwiller, 1981; Hunter, 2004: 191–245.

[ back ] 2. Allen, 1940: 1–26, was very negative. The discovery of new evidence has thrown a different light on some of the issues which were of concern to Allen. Although the ancients did not use ‘epyllion’ in the same sense as do modern scholars, the term is still useful provided that certain pitfalls are avoided: e.g. the Greek word epos should often be translated ‘hexameter poem(s)’ or ‘hexameter verse’ rather than ‘epic’ (which makes us think of a work in many books). I would resist the application of ‘epyllion’ to poems written in metres other than hexametric, even though Eratosthenes’ elegiac Erigone (Rosokoki, 1995) and Callimachus’ own account of the old man Molorcus who entertained Heracles (‘Victoria Berenices’, see Lloyd-Jones, 1983: 254–268 C have much in common with Callimachus’ Hecale.

[ back ] 3. I have become increasingly doubtful whether later Greek poets were significantly influenced by earlier Latin poets; perhaps (as some older scholars thought) similarities between Nonnus and Catullus (or Ovid) should rather be explained by common use of Hellenistic models.

[ back ] 4. Hutchinson, forthcoming stresses the importance of the Hecale; he regards its ‘most obvious and natural category’ as epic. There may have been one or two substantial Hellenistic hexameter poems which predate the Hecale, e.g. the Hermes of Philetas (frs. 5–9, Powell, 1925); frs. 1–5, Spanoudakis, 2002).

[ back ] 5. Hollis, 1990: 5–10.

[ back ] 6. Hollis, 1990: 32.

[ back ] 7. Hollis, 1990: Appendix III.

[ back ] 8. See Bernard, 2004: 227–332. The items almost certainly borrowed from Callimachus are κοκύαι = ‘ancestors’ (Hecale fr. 137 H. = 340 Pfeiffer) and τυννόϲ = ‘small’ (see my ‘Hecale’s Babies’, in Hollis, 2004: 115–116).

[ back ] 9. I have in mind P. Oxy. 2258 (pap. 37 Pfeiffer = 2 Hollis, 2004), written in the sixth or seventh century and furnished with rich scholia.

[ back ] 10. The last owner was probably Michael Choniates (see my Hecale, pp. 38–40, with later thoughts based on Michael in ZPE 115 (1997): 55–56 and ZPE 130 (2000): 16).

[ back ] 11. Hollis, 1990 Appendix II. Contrast Oxford Classical Dictionary3 on the typical length of an epyllion, ‘up to c. 600 lines’.

[ back ] 12. Difficult to classify generically: we seem to find elements of an epyllion, a hymn and a scientific/philosophical poem. The only substantial fragment (16 Powell, Coll. Alex., augmented in Suppl. Hell. 397 A) is closely imitated by Virgil, Georgics 1.233 ff.

[ back ] 13. Greater exactitude is not possible because of the damaged state of P. Oxy. 3000 (Suppl. Hell. 397). One may suspect that some of the mythological hexameter poems by Euphorion of Chalcis were on a similarly generous scale, but proof is lacking.

[ back ] 14. On these smaller poems, see Richard Hunter (n. 1 above).

[ back ] 15. This caught the fancy of Ovid (Fasti 3.227–228). We now know that Callimachus’ Hecale, far from remaining unmarried, had been the mother of two boys (for a possible new fragment, see ZPE 148 (n. 8 above)).

[ back ] 16. Can one really accept ‘And to him the old ploughman who guarded the cattle made reply’ as a satisfactory first line? For an argument that one can, see Hunter 2004:. 211 n. 91. Contrast the very conventional opening of Callimachus’ Hecale (fr. 1 H.), ‘Once there lived an Actaean woman in the hill-country of Erectheus’.

[ back ] 17. Thus (generally) P.J. Parsons in ZPE 25 (1977): 1–50.

[ back ] 18. Gow, 1950: 440.

[ back ] 19. Illustrated by Parsons (n. 17 above), p. 44.

[ back ] 20. I omit Theocritus 25 because of uncertainty about its original length.

[ back ] 21. The Oxford Classical Text of the Bucolici Graeci includes other (inferior) poems such as [Moschus], Megara and [Bion], Epithalamium of Achilles and Deidamia, of unknown authorship but similar style and perhaps the same period.

[ back ] 22. Though it has a few very rare words and some recondite mythical genealogy—also what may be a typical feature of the epyllion (compare Catullus 64.50 ff.), use of a work of art (Europa’s basket) to introduce a subsidiary myth (Io).

[ back ] 23. Lightfoot, 1999.

[ back ] 24. Whatever its date, purpose and authorship, the latter (edited by R.O.A.M. Lyne [Cambridge: 1978]) plausibly recreates what we might imagine as the prevailing style of c. 45 BC, when Alexandrian influence on Latin poetry may have been at its height. Another work in the Appendix Vergiliana, Culex, though inferior, has some pretensions to learning, e.g. describing the trees in the wood by reference to their mythological prehistory (Culex 123 ff., cf. Catullus 64.290–291).

[ back ] 25. There are several passages in Augustan poetry which treat Io in a similarly humorous/ ironical manner; they all may go back to Calvus’ epyllion, as I shall argue in my forthcoming Fragments of Roman Poetry c. 60 B.C.–A.D. 20 (Oxford).

[ back ] 26. Perhaps suggested by the πελώριοϲ…ὕδροϲ of Euphorion fr. 58.3 Powell.

[ back ] 27. Fr. 6 of Aemilius Macer in Courtney, 1993: 295.

[ back ] 28. A young Ovid (Tristia 4.10.43–44) used to hear an elderly Macer recite this poem.

[ back ] 29. One might compare [Virgil], Ciris 54 ff.: although Scylla daughter of Nisus is not to be identified with the sea-monster, the poet continues with multiple opinions as to why the latter was transformed.

[ back ] 30. Also worthy of note are the hemiepes ‘Ampitryoniades’ occupying the first half of 644 and the refined anaphora in 662–663 ‘et subitus praepes Cyllenida sustulit harpen,/harpen alterius monstri iam caede rubentem’.

[ back ] 31. n. 1 (above), pp. 192–193.

[ back ] 32. Other epyllion-like sections of the Argonautica are those involving Hypsipyle (1.609– 909), Hylas (1.1207–1357), Amycus and the Bebrycians (2.1–163), Jason and Medea in Phaeacia (4.982–1222).

[ back ] 33. Virgil had to some extent prepared the way when he sent Aeneas to be entertained by Evander (Aeneid 8.90 ff.).

[ back ] 34. Ovid too may have glanced at Eratosthenes’ Erigone, since the wine-miracle by which Jupiter and Mercury announce their divinity (Met. 8.679 ff.) is even more appropriate to Dionysus (thus Silius 7.187 ff.).

[ back ] 35. Appendix II in my edition of Ovid, Metamorphoses 8 (Oxford: 1970). It is unclear whether this particular papyrus fragment belongs to Bassarica or Gigantias; for a complete edition, see Livrea, 1973. Livrea does not comment on the context of this fragment (his 81 verso), which he gives to the Gigantias—in that case the domestic scene would be all the more striking.

[ back ] 36. Text in Heitsch, 1961:51– 54, and Page, 1942: 516–519.

[ back ] 37. Compare fr. 40 Powell (Hyacinthus) for a flower springing from blood, and fr. 84.3 for the ‘not yet…’ motif.

[ back ] 38. Sharing motifs with the Victoria Berenices (see Parsons, n. 17 above). In both cases the lion may have been catasterized. For fuller discussion see my ‘Myth in the Service of Kings and Emperors’, in Férez, 2003: 1–14.

[ back ] 39. He is definitely not pretending to be the Oppian of Cilicia who wrote Halieutica; his similarity of subject matter (Cynegetica) has produced the false ascription. Pseudo-Oppian came from Apamea in Syria and wrote his poem on hunting under Caracalla, probably after AD 212. Almost all the external testimonia (whether or not to be believed) refer to the author of Hal. rather than that of Cyn. The poem on fishing is generally judged to be far superior; Cyn. does indeed have more technical defects (e.g. a distressing insensitivity to Hermann’s Bridge), but also—in my opinion—more positive merits than is commonly allowed. See further Holis, 1994: 153–66.

[ back ] 40. Note the spondaic fifth foot, a mannerism of the learned Hellenistic poets and their Roman admirers, and the stress upon inherited tradition (φάτιϲ).

[ back ] 41. ἀλλὰ τὰ μέν or καὶ τὰ μέν are well-established ways of breaking off a narrative in learned poetry (Pfeiffer on Callimachus fr. 12.6 collects examples).

[ back ] 42. E.g. Apollonius in his Foundation of Alexandria included the story of Perseus shedding drops of Gorgon’s blood (from which sprang snakes) as he flew over Libya.

[ back ] 43. By the Suda (Ε 3801 Adler); others said Antioch. We know practically nothing about Euphorion’s years in Syria, except that he gave currency to the tale that Seleucus Nicator’s mother, before her son’s birth, foresaw that he would become lord of Asia (fr. 174 Powell, from Tertullian).

[ back ] 44. The transformations of the women into leopards and Pentheus himself into a bull are not mentioned by Forbes, 1990.

[ back ] 45. It is worth recalling that, of the three great tragedians, Euripides exerted much the most influence upon Hellenistic poetry.

[ back ] 46. Line 246 περὶ παῖδα τὸ μυϲτικὸν ὠρχήϲαντο may faintly echo Eratosthenes, Erigone fr. 22 Powell τόθι πρῶτα περὶ τράγον ὠρχήϲαντο (the invention of tragedy).

[ back ] 47. Often, too, in Nonnus. Aristaeus is invoked (but not named) in Georgics 1.14–15.

[ back ] 48. Later to be found in Nonnus. It is regrettable that the many Hellenistic transformation poems (such as the Heteroeumena of Nicander) have perished almost completely.

[ back ] 49. For statistics see my Hecale, Introduction, p. 18.

[ back ] 50. Respectively fr. 2a.30 (Pfeiffer vol. II p. 103) and Supplementum Hellenisticum 275.

[ back ] 51. Through the dating of P.Oxy. vol. 41, no. 2946 (third to fourth century).

[ back ] 52. Gerlaud, 1982:9, dates T. between the middle of the third century and the beginning of the fourth.

[ back ] 53. No doubt I am wrong to find it less interesting and attractive than Colluthus’ Rape of Helen.

[ back ] 54. One cannot miss the allusion to Apollonius Rhodius in Triph. 503–505.

[ back ] 55. I prefer the emphatic κείνηϲ (compare the passages of Nonnus and Michael Choniates quoted as part of Hecale fr. 82 H.) to the κοινῆϲ printed by Gerlaud.

[ back ] 56. No doubt by contrast with the sufferings of Hecale’s life—the death of her husband and two sons, and the loss of her property.

[ back ] 57. Anth. Pal. 9.198. This raises the question of where and how Nonnus gained his obvious familiarity with poets such as Callimachus, Apollonius, Euphorion, Nicander and Parthenius, since it seems now to be generally agreed among scholars that the two most important libraries in Alexandria had perished long before the Arab conquest; see El-Abbadi, 2004: 174.

[ back ] 58. The Dionysiaca may seem very pagan in spirit, but Nonnus’ paganism could be merely a reflection of his literary culture—a common uncertainty in that period (Nonnus seems to have flourished c. 450–470).

[ back ] 59. There can hardly be any doubt about the identity of authorship; one indication is that the same bits of learned Hellenistic poetry have left their mark on both the Paraphrase and the Dionysiaca (see my Callimachus, Hecale, p. 35).

[ back ] 60. The latter because Proteus’ mythical island of Pharos was conventionally identified with the island just off the coast of Alexandria, on which stood the famous lighthouse.

[ back ] 61. In this, as in some other respects, the closest surviving parallel would be Ovid’s Metamorphoses, though I hesitate to claim that Nonnus made use of Ovid’s poem (see n. 3 above).

[ back ] 62. From the ‘Victoria Berenices’ (Suppl. Hell. 254–268C) which we now know to have opened Aetia book 3 (including what was previously Call. fr. 177 Pf., on the old man setting his mousetraps).

[ back ] 63. The immediately following reference to olives (line 55) clearly comes from Call., Hecale fr. 36 H.

[ back ] 64. Incidentally the patria of the poet Parthenius (we have no evidence that P. handled this myth). Nicaea in fact was named in the early Hellenistic period after the first wife of Lysimachus, daughter of Antipater. For such invention of myth where the historical reality has been forgotten, see Chuvin, 1991: 149 n. 4.

[ back ] 65. The Io of Calvus (n. 25 above) must have had a strongly pastoral flavour in places; one could also view Ovid’s Pan and Syrinx (Met. 1.689–712) as a pastoral digression in an Io-epyllion.

[ back ] 66. Perhaps the opening words of Euphorion’s poem entitled Hyacinthus?

[ back ] 67. It is unclear whether these ‘several great poets’ are Hellenistic or (as Oliver Lyne favoured in his commentary) Roman.

[ back ] 68. One might compare Callimachus, fr. 75: after rebuking himself for inability to control his tongue (lines 4–9) the learned poet tells Xenomedes’ story of Acontius and Cydippe—then he adds a speedy summary of all the other things to be found in Xenomedes’ chronicle (fr. 75.53–77)!

[ back ] 69. Rather similar is the anonymous fr. 10 in Courtney, 1993: 458 (some have wondered about the Dictynna of Valerius Cato).

[ back ] 70. A strange and disturbing passage; see Gow, 1950 ff. for discussion.

[ back ] 71. One could have fun debating, in the spirit of an Alexandrian ζήτημα (quaestio) whether, in Nonnus, Erigone’s dog was catasterized as Sirius or as Canis Minor.

[ back ] 72. Call. testimonium 24 Pf.

[ back ] 73. See n. 9 above (on P.Oxy. 2258).

[ back ] 74. See Vian 1976; West, 1982: 177 ff.

[ back ] 75. That is not the majority view. In fact I derive pleasure from several poems discussed in this paper which are often disparaged—Theocritus (or [Theocritus]) 25, the pseudo-Virgilian Ciris, [Oppian], Cynegetica, Colluthus, Rape of Helen.

[ back ] 76. Hera was not universally considered to have been Ares’ mother (another candidate was Enyo).

[ back ] 77. In view of the context in Colluthus one might wonder whether Call. fr. 634 came from the debate about the parentage of the Graces in Aetia, book I. Call. himself proposed, as his first suggestion, that they were children of Zeus and Hera, but the Muse Clio declared that their parents were Dionysus and the Naxian nymph Coronis (see the Florentine Scholia, Pfeiffer vol. I, p. 13, lines 29–35). Clio’s view, perhaps direct from Callimachus’ Aetia, is supported by Libanius, Epist. 217.6 F.

[ back ] 78. Fr. 40 Powell, quoted earlier.

[ back ] 79. There are 19 examples in Colluthus’ 394 lines.

[ back ] 80. K. Kost in his 1971 edition of Musaeus, p. 602 s.v. Kallimachos, has a longer list of parallels, most of them fairly general.

[ back ] 81. It is uncertain whether Supplementum Hellenisticum 901 A and 951 could contain fragments of such a poem.

[ back ] 82. Worth a footnote is the ‘Barcelona Alcestis’ (122 hexameters on the myth of Admetus and Alcestis), of uncertain date, perhaps around 400. ‘The poem combines ethopoea and mythological narrative; the type is familiar from the Hylas and Orestis Tragoedia of Dracontius’ (Hutchinson,1983: 31–36 at 31). One may allow this to be a better poem than it seemed to its first editor, but it surely remains very mediocre.

[ back ] 83. He has occasional moments, but is technically far from competent; his prosody is particularly appalling—mainly but not solely with respect to Greek proper names (e.g. ‘Pylades’ is regularly scanned – – ˘).