CHAPTER 10
The Eleusinian Mysteries and Vergil’s “Appearance-of-a-Terrifying-Female-Apparition-in-the-Underworld” Motif in Aeneid 6

RAYMOND J. CLARK

More than two and a half centuries ago, in 1745, in the second book of his The Divine Legation of Moses, Bishop William Warburton put forth the hypothesis that Aeneas’ Descent into the Underworld was an allegorical representation of an initiation into the Eleusinian mysteries.1 The bishop considered Aeneas to be a grand legislator (in his capacity as founder of Lavinium) within a tradition of ancient heroes and lawgivers who were initiated in the mysteries;2 he noted that Caesar Augustus, whom he says Aeneas anticipates, was likewise initiated at Eleusis;3 and he concluded that Vergil worked into Aeneas’ journey the doctrine of a “future state of rewards and punishments” that was the foundation and support of ancient politics. This hypothesis evoked a skillful adversary in Gibbon, who, objecting that Aeneas was no legislator, set out to expose Warburton’s many unproved assumptions—“probably repelled not more by the arrogant dogmatism of the untrained scholar,” as Conington put it, “than by the zeal of the ecclesiastic in proving that even pagan times witnessed to the alliance between religion and civil government.”4 Conington, for his part, granted Gibbon that Aeneas was not a mere anticipation of Augustus, despite his many Augustan traits, and he further conceded that Aeneas’ descent was not simply a sustained allegory of the mysteries as though there were an authorized doctrine. But Conington nevertheless considered it quite possible that several of Vergil’s details, if not his general conception, may have been drawn from the mysteries—that is to say, from such ancient literature as alludes to them.

My purpose here is not to review the whole topic of correspondences between Aeneas’ infernal journey and the Eleusinian mysteries, but rather to examine a single incident in book 6 of the Aeneid at verse 290, where Aeneas raises his sword in terror against the phantoms of the Gorgons and other monsters who appear before him in the darkness of Pluto’s house at 282-289. In this examination I shall draw attention to just three of several “motifs,” or themes, cited by Warburton as evidence that Aeneas underwent an initiation. The bishop contends (1) that tradition obliged the hero Aeneas to be initiated, just as (to name one other) Herakles was initiated into the Eleusinian mysteries;5 (2) that Aeneas in the Gorgon scene encountered imaginary false terrors no differently from all initiates in the mysteries, who are subjected to the phantoms of Hekate;6 and (3) that Aeneas was soon found in a “fright” resembling that experienced by other initiates at the mysteries according to the writings of Themistius and Proclus.7 In Warburton’s argument, these are three separate motifs having Eleusinian associations without any other connection between them.

Yet there are other connections between these motifs. They exist in some versions of Herakles’ descent to fetch the Hell-dog Cerberus. To introduce them, I adduce what is clearly a summary made in Bibliotheca 2.5.12 by Apollodorus of Athens of an earlier source telling how Herakles went to Eumolpus at Eleusis in order to be initiated, presumably (we are not told the reason) as the means of ensuring success in his quest for Cerberus.8 But first Herakles had to be adopted by an Athenian (Pelius) in order to qualify for the rite, which he was the first foreigner to undergo. And before Eumolpus could initiate him,9 Herakles had also to be purified by him from his slaughter of the Centaurs.10 After initiation, Herakles descended through a Hades entrance in Taenarum in Laconia. Upon seeing him in the lower world, the souls of the dead all fled, save Meleager and the Gorgon Medusa. Herakles drew his sword against Medusa as if she were alive, but desisted when his underworld companion Hermes told him that she was but an empty phantom. Herakles then found Theseus and Peirithoos near the gates of Hades and rescued Theseus. Continuing his journey, some details of which I omit, Herakles obtained Pluto’s permission to capture Cerberus, on the condition that he not use against the dog the weapons he was carrying. So Herakles throttled Cerberus, whom he found at the gates of Acheron, into submission,11 and ascended with him to the upper world through Troezen. Herakles later returned the Hell-dog to Hades after showing him to Eurystheus.

So goes Apollodorus’ Greek narrative, composed in the second century CE. As this mythographer consistently ignores Roman literature,12 he is unlikely to have modeled his narrative on an earlier scene of Aeneas’ meeting with the Gorgon in the Aeneid, from which, in any case, Apollodorus differs in detail. Eduard Norden, in his commentary on the sixth book of the Aeneid, made a strong case for believing that Apollodorus drew instead for this episode on a lost epic version of Herakles’ descent that he claimed influenced, in addition to the Apollodoran narrative itself, Bacchylides’ fifth Dithyramb, Aristophanes’ Frogs, and the sixth book of the Aeneid,13 to which a passage in the fourth book of Vergil’s Georgics, to be mentioned later, should be added. In his fifth Dithyramb, Bacchylides at 71-84 describes a scene resembling Apollodorus’ in that the descending hero Herakles is warned not to shoot at a mere wraith. But in the highly compressed scene by this Greek lyric poet, neither the Gorgon nor Herakles’ underworld guide is mentioned. Instead, Meleager’s ghost admonishes Herakles against shooting an arrow at itself. Its assurance that there is nothing to fear (ο τοι δος) from a ghost underscores the fright that Herakles in fact experiences as he aims his weapon at the underworld shade. When at Aeneid 6.290-294 the Cumaean Sibyl warns the terrified Aeneas not to use his sword against Gorgons and other bodiless shapes as well, Vergil assigns to Aeneas’ august guide the function performed by both Herakles’ guide Hermes and Meleager’s ghost in the comparable versions so far mentioned. Yet Vergil cannot have derived his knowledge of the Gorgon episode from Bacchylides, even if he read him, since the Greek lyric poet did not mention the Gorgon. Nor was Vergil’s source Apollodorus, who wrote long after him.

Nor indeed could Vergil have exploited Aristophanes’ Frogs for his Gorgon scene, since Aristophanes did not include such a scene, even though, as I believe, one episode in his comedy—I now raise a matter not noticed by Norden—presupposes the existence of the standard Gorgon scene in Aristophanes’ source. I refer to verses 564ff., where the Greek playwright seems to have transformed the motif of Herakles’ frightened encounter with one or more Gorgons into what appears to be a comic parody of the theme. In the comic parody, two formidable female keepers of the kitchen tell Dionysus, after he knocks on Pluto’s door, how Herakles had drawn his sword upon them. I take these keepers of the kitchen to be comic doublets of the Gorgons. The correspondence between the two sets of formidable females, which I observed more than thirty years ago with the later approval of Dover in his commentary on the Frogs ,14 illustrates a further influence of the lost version of Herakles’ descent upon Aristophanes beyond the points of contact noticed by Norden.

In a brilliant article, Hugh Lloyd-Jones adduces a fragment of Greek lyric poetry preserved in P.Oxy. 2622 ascribed to Pindar (which has a commentary upon it partially preserved in PSI 139) together with the Herakles of Euripides at 610-613, where Herakles reports that he saw the ργια of the initiates, as additional works influenced by the lost epic postulated by Norden. Lloyd-Jones infers from Herakles’ pro-Athenian sympathies and connection with Eleusis that the lost epic was composed around 550 BCE by an Athenian or a person belonging to the orbit of Athenian culture.15 The partially preserved poem by Pindar agrees with the Apollodoran narrative in telling how Herakles was initiated by Eumolpus at Eleusis before recovering Cerberus. It also alludes to Herakles’ meeting with Meleager among innumerable ghosts in Hades, as related by both Apollodorus and Bacchylides. Unfortunately, the fragmentary remains of Pindar’s poem do not tell us whether Herakles was frightened by any Gorgon or Gorgons in the underworld.

How, then, might Aeneas’ terror at seeing the Gorgons have been drawn from the Eleusinian mysteries? The question involves consideration of comparative figures. Let us first recall what has just been noticed, that Aristophanes omitted the Gorgon scene from the Frogs, having transmuted it into a comic parody that takes place at the front door of Pluto’s house. Let us also bear in mind that Herakles’ directions to Dionysus based on his own experiences are the playwright’s indirect acknowledgment that a version of Herakles’ descent to the lower world in the living flesh underlies Dionysus’ in this play. But Aristophanes has made changes. Dionysus and his slave Xanthias in the Frogs are terrified not by a Gorgon, as was Herakles in the lost epic used by Aristophanes, but by Empousa, another female monster who appears in the infernal region just where Herakles told the descending pair they would meet snakes and monsters (143-144, 278-279). According to Herakles’ directions, they must pass these before they reach the region where the wicked are punished in mud and dung (145ff., cf. 273ff.), and beyond that region again, says Herakles, are myrtle groves, where deceased Eleusinian initiates are seen and heard singing and dancing (154ff., cf. 312-459); nearby lies Pluto’s house (163, cf. 431-436 and 460). Both the place where the wicked are punished by lying in mud and the myrtle groves of Hades as home to the initiates evoke associations with Eleusinian mysteries.16 Struck by the general correspondence between the Aristophanic and Apollodoran descent versions, Lloyd-Jones has suggested that the first two stages mentioned by Herakles parallel those in Apollodorus’ narrative, where Herakles meets the Gorgon (in the region of monsters) and then sees Theseus and Peirithoos undergoing punishment (in the region of the wicked).17 I shall return to certain specific matters of location presently. More pertinent to our immediate purpose is Lloyd-Jones’ further inference that the underlying common source, the sixth-century Attic epic katabasis of Herakles, which stresses this hero’s Eleusinian connections, influenced also the Empousa scene.

The existence of this Eleusinian source and the collocation of Empousa’s appearance with Eleusinian bliss in the Frogs have in turn led to the hypothesis that Empousa’s appearance in the Frogs alludes to a specific Eleusinian cultic event. In its support, Brown cites Borthwick’s observation that Xanthias compares Empousa to a weasel (γαλν) in language derived from a hieratic formula of the sort associated with mystery religions, to which Dionysus reacts in ritual terms. He also adduces evidence from the partially surviving work On Demagogues by the fourth-century BCE historian Idomeneus of Lampsacus (FGrH 338.F2) and from Lucian’s Cataplus 22.18 In the former, Empousa appears from out of the darkness to initiates (π σκοτεινν τπων νεφανετο τος μυουμνοις); the brief surviving fragment does not identify the initiates as Eleusinian, but this they are likely to be, since the work from which the fragment comes focuses on Athens, and Graf has shown that references to mysteries within Athenian contexts always refer to Eleusis.19 In Lucian’s Cataplus, it is the dread figure of an Erinys that appears from out of the darkness, in the same region as Empousa in the Frogs, that is to say, as soon as the infernal travelers reach the far shore of the underworld lake. Lucian, moreover, gives his satire a specifically Eleusinian context, since a deceased cobbler is made to exclaim, “By Herakles!” to other dead men who have just disembarked with him from Charon’s boat, and he asks the philosopher Cyniscus if the appearance of the Erinys in the darkness resembles Cyniscus’ earlier experience when he was initiated in the Eleusinian mysteries. Cyniscus affirms that it does, torch-bearing female with frightful menacing aspect and all. Since apparitions, φσματα, are also much spoken of in the celebrations of the Greater Mysteries at Eleusis20 —they are at times said to be sent by Hekate, with whom Empousa is sometimes identified21 —Brown thinks that at a relatively early point in the proceedings, initiates were terrified by the appearance of a specter, as were Dionysus and Xanthias, and he suspects that Empousa (perhaps not her official name) was one of the names given to it by individual worshipers.22 Accordingly, he assigns to this terrifying female in the Frogs a cultic origin together with, through the lost Eleusinian Herakles katabasis, a literary origin. He further suggests that, like the Homeric Hymn to Demeter, the lost epic contained aetiological passages alluding to and glossing the δρμενα at Eleusis.23

I would not, however, care to see it taken for granted that the lost Herakles epic itself contained an Empousa scene. It strikes me as far more likely that Aristophanes modeled his Empousa scene upon the Gorgon scene in his source, which, if true, has just provided one more point of influence upon Aristophanes. We could not, of course, have inferred this direction of influence had Norden not conjectured the existence of the epic katabasis, which Lloyd-Jones then dated to the mid-sixth century BCE, since all extant references to one or more Gorgons seen by Herakles in the underworld, and by Aeneas in imitation of Herakles, are post-Aristophanic.

The conclusion so far reached is that Empousa can be added to the Gorgon and to the two female keepers of the kitchen and to the Erinys in Lucian’s Cataplus (as well as perhaps to Hekate, with whom Empousa is sometimes identified) in a list of variants on the appearance-of-a-terrifying-female-apparition-in-the-underworld motif with strong Eleusinian associations. Another source supports this conclusion. Elsewhere I have expressed the view that Aristophanes as well as Vergil would have taken every chance to read, in addition to the lost epic version, whatever they could on the articulate and well-developed tradition about Herakles’ descent, and that both did by exploiting Euripides’ (or, less likely, Critias’) Peirithoos,24 which now survives in only a few fragments. In one fragment, P.Oxy. 3531, Peirithoos refers to a female he can hear but not see.25 If the Peirithoos precedes the production of the Frogs in 405 BCE, as I think it does,26 Cockle, the editor of this fragment, may well be right when he observes, “Perhaps this creature, whatever her precise nature, is reflected in Empousa.”27 Since the chorus of this play, as in the Frogs, is composed of deceased Eleusinian initiates,28 Euripides may have borrowed this female from Eleusinian cult for his Peirithoos. Aristophanes could have taken her from Euripides’ play or from Eleusinian cult, or from both.

When weighing the evidence of Lucian’s Cataplus, Brown cautions that the satirist may also have had Aristophanes’ Frogs in mind, since in addition to the similarities between the two works already pointed out, Cyniscus, like Dionysus, has to row Charon’s boat across the infernal lake.29 More can be adduced to support this supposition and to strengthen Brown’s claim that Empousa has Eleusinian ties. For it looks to me as though Lucian even chose for his satire characters appropriate to the shapes assumed by Aristophanes’ Empousa—among them a dog and a copper leg. The very name of the “cynic” philosopher Cyniscus means “dog,” and recognition of the copper leg would have helped the deceased cobbler clinch the apparition’s identity. Since, moreover, Sophocles gave the attribute “copper-footed” to the avenging Erinys in Elektra 491, Dionysus, as Stanford notes,30 may be jestingly alluding to it when he asks in the Frogs if shape-shifting Empousa has a copper leg. It seems to me, then, a small leap if Lucian, observing the Sophoclean underpinning of the Aristophanic attribute of Empousa, makes the characters in his satire identify Aristophanes’ Empousa as an Erinys. If we combine this identification with Cyniscus’ association of the frightful Erinys with Eleusis, Lucian indirectly gives Empousa, too, an Eleusinian setting. I have already argued that Empousa in the Frogs is Aristophanes’ substitute for the Gorgon encountered by the Eleusinianized Herakles in the sixth-century source. By giving the female apparition the identity of a Gorgon, the author of the lost epic portrayed the terrified Herakles as actually encountering a Hellish female apparition of the sort that even Odysseus feared to meet at the end of the Nekyia.31

Such, then, is the convergence of Eleusinian associations underlying Aeneas’ encounter with the Gorgons in Aeneid 6. I turn now to some issues of infernal topography that indicate deviations by both Aristophanes and Vergil from their common sixth-century source. Observe that in the Frogs, the Gorgons and their comic doublets, the formidable keepers of Pluto’s kitchen, both occupy the same residence, namely Pluto’s palace. The first of these two sets of females, despite being snaky-haired like their further counterparts Hekate, the Erinyes, and Empousa,32 do not, after all, reside in the region of snakes and monsters, where Lloyd-Jones assumed them to be in his comparison (see above) between the two parallel stages in the Aristophanic and Apollodoran descents. Nevertheless, in Aristophanes’ lost source, this is where they belonged. Several matters to be raised in the next few paragraphs make this clear.

We know that the Aristophanic Gorgons have their dwelling in Pluto’s palace because the doorkeeper Aeacus goes inside to search for them at 472-478. Yet their snaky hair makes them natural compatriots with the snakes and monsters that Herakles leads Dionysus to expect to meet immediately after reaching the far side of the bottomless lake. This is also where the sinners and monsters in Polygnotus’ mid-fifth-century wall-painting must have been depicted as described in Pausanias 10.28.1-7—on Acheron’s far side, since Odysseus is said to be already in Hell before these are listed. No mention, incidentally, is made of Pluto’s palace in Polygnotus’ mural. Insofar as Aristophanes’ Gorgons are placed not with the snakes and monsters immediately across the lake, but in the company of some other snake-like creatures of torture in Pluto’s palace much deeper within the underworld, it is as though they have been attracted away from their sixth-century location to the residence of the two kitchen-keepers, who have assumed the Gorgons’ formidable attributes in the comic parody.

Nor is this Aristophanes’ only departure from what might have been expected. Though Herakles tells Dionysus that he will encounter the snakes and monsters first after crossing the lake (143ff.), Dionysus actually meets them second, after he has encountered the wicked (273ff.). Much commentary has been written on the reversal of the two regions as described by Herakles, in contrast to Dionysus’ actual experience of them. But it has not hitherto been observed that we can ascertain the sixth-century sequence of events by comparing the versions of Aristophanes and Apollodorus, who both drew upon the lost Herakles descent. The comparison reveals that Herakles’ directions to Dionysus preserve the original sequence, so that Aristophanes deviates from his source when he makes Dionysus and Xanthias meet the wicked first. Aristophanes no doubt had dramatic reasons for reversing the order. The two infernal travelers barely mention the wicked, save with a glance at the audience, and this region is passed through first and quickly, perhaps to suppress possible conflict with the location of the main dead beyond the door of Pluto’s house (760), from which the deceased Aeschylus and Euripides exit onto the stage at 830ff. Aristophanes also evidently judged it more dramatically effective to put second the region of the snakes and monsters, in which the two panicstricken travelers are made to linger by Empousa’s frightful apparition.

The detail provided by Apollodorus to which I just alluded, which enables us to infer in which order Aristophanes’ source narrated these infernal experiences, raises an issue of its own that needs sorting out. It is widely held that the gates of Hades near which Apollodorus says Herakles found Theseus and Peirithoos are located at the entrance to the underworld. For instance, Brown in his article on Empousa says that according to Apollodorus 2.5.12, “As soon as Heracles enters Hades with his guide, Hermes, all souls flee before them with the exception of Meleager and the Gorgon, Medusa” (my emphasis).33 The same misunderstanding infects also Lavecchia’s summary of Apollodorus’ scene thus: “Subito dopo il suo arrivo nell’ Ade [i.e., at the start of his infernal journey], Eracle incontra Medusa e Meleagro.”34 Similarly, a popular commentary on Apollodorus disseminates the view that the gates of Acheron at 2.5.12 “are the gates of Hades mentioned above, symbolizing the boundary between the lands of the living and the dead.”35 Or, to cite the editor of the papyrus fragment of Euripides’ underworld scene again, Cockle infers from Apollodorus’ mention of Hades’ gates that in Euripides’ Peirithoos, Herakles’ conversation with Hades’ doorkeeper Aeacus must have taken place near the entrance of the underworld also.36 But the Apollodoran gates of Hades are not near the entrance to the underworld. The fact that Apollodorus does not provide for these gates a specific reference point in his brief summary of Herakles’ descent should not be taken to imply that Herakles in his account meets the Gorgon as soon as he enters the underworld, or finds Theseus and Peirithoos near the gates of Hades at the entrance to the underworld also. On the contrary, since Apollodorus reports that Herakles sees Theseus and Peirithoos as he approaches Hades’ gates after thrusting his sword at the Gorgon, the gates can hardly be at the entrance separating the world of the living from the world of the dead, where Cockle and others imagine them to be. They must belong instead to Hades’ palace across the lake, which Apollodorus in his brief summary omits, and where Aristophanes, too, depicts Hades’ palace in the Frogs. Moreover, in the scene depicted by Bacchylides (5.64), Herakles meets innumerable ghosts, including Meleager, with whom Apollodorus couples the Gorgon, beside the infernal waters of Cocytus, in all likelihood in their final resting-place on Cocytus’ far bank. This is where Vergil’s Orpheus, in imitation of Herakles, sees the corresponding ghosts in the fourth book of the Georgics at 471-480—the other passage influenced by the epic Herakles katabasis to which I alluded earlier. In sum, the related texts support the inference drawn from the Apollodoran narrative that Herakles in the common source encounters the terrifying apparition of the Gorgon as soon as he has crossed the infernal water, not as soon as he enters Hades, and that he has to travel deeper into the underworld before finding Theseus and Peirithoos near the gates of Hades.

In verse 290 in the sixth book of the Aeneid, Aeneas is near the beginning of his infernal journey when he experiences terror in the face of the frightening specters of the Gorgons and other shades. The occurrence of Aeneas’ fright at this point might tempt us to postulate a direct connection between this order of events in Aeneas’ underworld journey and the early part of the proceedings in the Eleusinian mysteries, when initiates are said to be frightened, according to Brown and various passages of late antiquity not cited by him (see note 7 above). But another explanation forces itself upon us as soon as we realize how much earlier the Gorgon scene occurs in the Aeneid than in the lost Herakles epic as here reconstructed from related texts: Herakles in the lost epic katabasis encountered one or more Gorgons after crossing the infernal water, whereas Aeneas meets them before his crossing. In another respect, Vergil’s Gorgons at Aeneid 6.273-294 retain their Aristophanic abode — since they still dwell within Pluto’s palace, quite precisely, as I have inferred elsewhere,37 in the stable rooms beside its main entrance. No inconsistency exists between saying that Aeneas’ encounter with the Gorgons is both earlier and in the same place, since Vergil has relocated Pluto’s palace and translated the Gorgons with it, to the antechamber of the Vergilian underworld. Aeneas and his guide, the Sibyl, thus reach the palace shortly after they pass through “the gate of Dis,” which is synonymous with the cave beside Avernus (Aen. 6.127 and 237ff.).38 This is the gate that separates the land of the living from the world of the dead in Vergil’s underworld, in contrast to the Apollodoran gates of Hades and the palace gates guarded by Aeacus in the Frogs. The Vergilian location of Pluto’s palace at the very beginning of the underworld rather than at its far end is not an error on Vergil’s part. In a recent article, I undertook to show how Vergil expanded the underworld by displacing forward exploit after exploit that in his sources occurred later in the underworld, in order to put more space between the beginning of the underworld and the near shore of the infernal bank, and to heighten the horrors Aeneas faces at the very beginning of his ordeal.39 The details are repeated here as a cautionary note in the present task of investigating the relationship between Aeneas’ terror in the Gorgon scene and its comparable cultic event in the Eleusinian mysteries. To show the existence of this relationship, I have traced the many paths connecting Aeneas’ experience to Eleusis. I have also taken pains to point out how Vergil has rearranged the infernal topography he inherits, to judge from reconstructed details in the lost sixth-century Attic epic katabasis of an Eleusinianized Herakles. Because Vergil has rearranged what he has read to suit his poetry, it would be misleading to treat Aeneas’ infernal journey, however deeply it is imbued with Eleusinian associations, as a poetic document from which to reconstruct the order of events in the mysteries. For this reason, Aeneas’ descent as concerns the Gorgon episode cannot be regarded allegorically as “no other than an enigmatical representation of his initiation into the mysteries,” as Bishop Warburton claimed in 1745.40

Notes

1. Warburton 1745: 270ff., esp. 288.

2. Ibid.: 288-291. Among initiated “ancient heroes,” Warburton includes Jason, the Dioscuri, Herakles, and Orpheus as named by Diodorus (4.43.1 and 5.49.6); among “lawgivers,” he lists both the kings of Eleusis named in the Homeric Hymn to Demeter 474-476 and such other figures as Tarquinius Priscus (Macrobius Sat. 3.4.8), Augustus Caesar (Suet. Aug. 93), and the later founders of empire who received instructions concerning their office from the mysteries. With regard to all of the foregoing, observe (1) that Warburton’s list of Eleusinian kings can be supplemented by Polyxenus and Dolichus in the Hymn to Demeter at 154-155 and 477; (2) that “lawgivers” for kings is a late term used, for instance, for Triptolemus by Porphyry (De abst. 4.22); and (3) that heroes and kings merge in Warburton’s political theory because Herakles, for example, is regarded (Xen. Hell. 6.3.6) by the torch-bearer Callias in the Eleusinian mysteries as the founder of the Spartan state. In addition, observe that in Warburton’s sources, Tarquinius and the list of heroes are presented as Samothracian initiates. The Dioscuri and Herakles—and Dionysus, too—are, however, called Eleusinian initiates in other sources (found in notes 9-10 below).

3. Suet. Aug. 93. Here Suetonius explains how Augustus’ Eleusinian initiation (Athenis initiatus) led to his recognition of the need for secrecy in a dispute involving the privileges of the priests of Attic Ceres in a court case at Rome. Cf. also Dio Cass. 51.4.1 and 54.9.7.

4. Conington 1872: 425.

5. Warburton 1745: 291-294; pertinent references for Herakles, with additions, are now assembled in notes 8-10 below.

6. Warburton 1745: 305-306, referring to Schol. in Ap. Rhod. Argon. 3.861. The passage is quoted in note 21 below, which offers a collection of passages on Hekate’s phantoms.

7. Warburton 1745: 309, referring to Themist. Or. 20.235a (2 p. 5 Downey-Norman; p. 287 Dind.): μν ρτι προσιν τος δτοις φρκης τε νεπμπλατοκα λγγου, δημονι ε ξυνεχετο τε κα πορ ξυμπσ, οδ χνους λαβσθαιος τε ν οδ ρχς στινοσον πιδρξασθαι ισω φεροσης, τε δ προφτηςκενος ναπετσας τ προπλαια το ναο. . . .(”Entering now into the mystic dome he is filled with horror and amazement. He is seized with solitude, and a total perplexity: he is unable to move a step forward, and at a loss to find the entrance to that road which is to lead him to the place he aspires to. Till the Prophet [the vates] or Conductor, laying open the vestibule of the temple … ,” trans. Warburton). Similarly Proclus Theol. Plat. 3.18: σπερ ν τας γιωτταις τελετας πρτν μυστικν θεαμτων κπληζις τν μυουμνων, οτω. . . . (“As in the most holy Mysteries, before the scene of the mystic visions, there is a terror infused over the minds of the initiated, so …,” trans. Warburton.) For more on fear and terror in the mysteries, see notes 20-21 below.

8. In iconographical representations of his capture of Cerberus, an Eleusinianized Herakles receives a more favorable reception in the underworld. The earliest such representation appears on a black-figure amphora (fr. Reggio 4001) from Locri c. 540 BCE, which Robertson (1980: 274-300, esp. 275-276) thinks relies on the same lost Eleusinian source as Apollodorus. I refer to this source at notes 13-15 below. For more on the Reggio fragments and on Athenian vases from about 530 BCE that show the Eleusinianized Herakles, see Boardman 1975: 1-12, pls. I-IV; and cf. Sourvinou-Inwood 1974.

9. Herakles’ need for adoption is narrated also by Plutarch, in a passage (Thes. 33.2) that does not name Eumolpus. In a fragment edited by Lloyd-Jones (1967) providing the earliest extant literary reference to Herakles’ Eleusinian initiation, Pindar names Eumolpus in agreement with Apollodorus against Diodorus (4.2526), who says Herakles was initiated at Eleusis by Musaios. The agreement lends support to Lloyd-Jones’ completion of the Pindaric lacuna at v. 8, πρτω[ι ξνων, and to his interpretation of what Eumolpus gave to Herakles “first” in the completed lacuna, “probably the privilege of being initiated in spite of being a foreigner.” Plutarch (Thes. 33.2) and Xenophon (Hell. 6.3.6) remark that Herakles’ adoption paved the way for the later adoption and Eleusinianization of the Dioscuri, also foreigners. Schol. in Aristoph. Plutus 845 and 1013 also remarks upon their common treatment by the Athenians, but then attributes to Herakles’ status as a non-Athenian the institution of the Lesser Mysteries at Agrae. This assertion contradicts the usual view that these mysteries were instituted to deal with Herakles’ need to be purified from bloodshed.

10. See also Plutarch Thes. 30.5, and Diodorus, who at 4.14.3 gives this as the reason why the Lesser Mysteries were founded by Demeter. The act of Herakles’ purification before initiation is shown in many artistic representations listed in, e.g., Richardson 1974: 211-213. It was perhaps then commemorated in Eleusinian ritual, which was regarded for others as a preliminary to initiation. For the rites, see Kerényi 1967: 45-60; cf. also Nelson 2000: 25-43, esp. 31ff. Additional references to Herakles as an Eleusinian initiate are found in the fourth-century Ps.Plato Axiochus 371e (Dionysus is coupled with Herakles here); Schol. in Homer Il. 8.368; Lycophr. 1328; Tzetz. Chil. 2.394; and passages cited throughout this article.

11. Robertson (1980: 275-276) interprets a bearded figure in the earliest extant Eleusinianized scene (mentioned in note 8 above) as the recently freed Theseus holding the club and weaponry that Herakles has undertaken not to use against Cerberus. By contrast, in a non-Eleusinianized underworld scene described by Barron (1972: 44), Theseus carries weapons of his own, no club included.

12. As pointed out by Bowra (1952: 116).

13. Norden 1926: 5 and other pages mentioned in his note 2.

14. Clark 1970: 252 n. 22; Dover 1993: 263.

15. Lloyd-Jones 1967: 206-229 (= Lloyd-Jones 1990: 167-187). For the views of later editors on the text of P.Oxy. 2622 (= Pindar fr. 346 S-M), see Lavecchia 1996: 1-26. Robertson (1980: 274-300) thinks the lost Herakles katabasis formed part of the Hesiodic Aegmius frr. 294-301 M-W attributed to Cecrops of Miletus. On the possibility that the Herakles epic survived to Vergil’s day, see Clark 2000: 192-196, esp. 195 n. 17.

16. Sommerstein (1996: 169), agreeing with West (1983: 23-24), against the doubts of Graf (1974: 103-107), argues that “lying in the mud” was a punishment recognized in Eleusinian “doctrine” (cf. Pl. Phd. 69c and Rep. 363d). He draws attention to the same triad of wrongdoings against gods, parent, and host or guest incurring this same punishment in Frogs 145-153 and other sources having Eleusinian connections. On Aristophanes’ initiates, see note 28 below.

17. Lloyd-Jones 1967: 219 (= 1990: 179).

18. Brown 1991: 41-50. Borthwick’s hypothesis (1968: 200-206) is that contemporary ritual, as well as superstition concerning weasels, underlies the language of the Empousa scene.

19. Graf 1974: 29-30 n. 36.

20. Brown (1991: 42) cites Plato Phaedrus 250b-c as the earliest explicit reference to φσματα, but not all apparitions are frightful. In this Platonic passage they are εδαμονα, as in Plut. Περ Ψυχς fr. 178 (Sandbach), where Plutarch speaks of the initiate’s fear and terror (sc. in the darkness) followed by the vision of blissful φσματα in the light; cf. Aristid. Or. 22.3 (Keil) and Procl. In Pl. Rep. 2.185.4 (Kroll). For fear and terror felt by the initiate before initiation, see also the passages instanced in note 7 above. Similar emotions are aroused by the epiphany of Demeter in the Homeric Hymn to Demeter 190, with parallels noted by Richardson (1974: 208-211, 252-256 and 306ff.); but this event, too, should be distinguished from the experience of fearful φσματα of the Empousa type.

21. Empousa is a frightful demonic shape-shifting apparition that (1) is sent by the goddess Hekate and (2) is sometimes even identified with Hekate. Several sources support (1): Schol. in Aristoph. Frogs 293 explains Empousa as a φντασμαδαιμονιδες π κτης πιπεμπμενον (Dübner p. 283) κα τς μορφς ναλλτον (Dindorf; Dübner prints instead: ο δ [φασιν] τε ξηλλττετο τν μορφν). Schol. in Ap. Rhod. Argon. 3.861 (Wendel) names Empousa among the φσματα . . . τκαλομενα καταα Cf. Bekker, Anec. Graeca 1.249.27-28: “μπουσα φσμα σττν π κτης πεμπομνων. Suidas, s.v. κτην, says that Hekate strikes fear in those who see her snaky-headed φσματα. Cf. also Plut. Mor. 166a. The following sources support (2): Hesychius s.v.”μπουσα· ριστοφνες δ τν κτηνφη μπουσαν. Schol. in Aristoph. Frogs 293 similarly names Aristophanes among those who identify Empousa with Hekate: νιοι δ [φασιν sc. τν μπουσαν] τνατν τ κτ, ς ριστοφνης ν τος Ταγηνιστας. The scholiast then pinpoints where in this partially surviving comedy the identification is made, by one speaker saying χθονα θ’ κτη / σπερας φεων ελιξαμνη and another replying τ καλεςτν μπουσαν; (fr. 515 PCG). Brown (1991: 47-49) thinks Hekate’s presence at Eleusis is attested by the Homeric Hymn to Demeter at 25, 440, and by archaeological evidence, which he discusses at length; he further links passages in (2) with other evidence such as Idomeneus (FGrH 338F2), to show Empousa’s cultic identity.

22. Brown 1991: 42-43; Brown uses (p. 50) the Plutarchan fragment (note 20 above) as the source for his view of Eleusinian proceedings.

23. Ibid.: 49.

24. Vergil’s indebtedness is emphasized in Clark 2000: 192ff., and Aristophanes’ in Clark 2001: 103-116, esp. 108.

25. P.Oxy. 3531, vv. 14-20, ed. Cockle (1983: 29-36 = F4a in Snell and Kannicht 1986: I. 349-351 and Critias IIa in Diggle 1998: 174-175).

26. In Clark 2001: 109-111, I argue for the priority of Peirithoos on several grounds, including the treatment of Aeacus in the two plays.

27. Cockle (1983: 35) raises the further possibility that this female may be an Erinys, comparable to the Furies visible at Aesch. Cho. 1048ff. to no one onstage except Orestes, but he notes that the hypothesis of Peirithoos makes no mention of a Fury.

28. Cockle (1983: 34) suggests “dead Eleusinian priests,” citing Peirithoos F2 (Ath. 11.496A), ed. Snell and Kannicht. Since the initiates in the chorus of the Frogs have led virtuous lives on earth (457-458) and now have their own sunshine (455), they must be in Hades. This view is defended against others by Lloyd-Jones (1967: 219-220), who thinks that Aristophanes’ initiates, though dead, nevertheless suggest the atmosphere of Eleusinian cult.

29. Brown (1991: 46) notes that Lucian draws on the Frogs also at Philopatris 25, Contemplantes 24, Cataplus 14, and Fugitivi 28.

30. Noticed by Stanford (1958: 98 ad 289-295).

31. My argument assumes that the author of the lost epic knew the Nekyia and wished Herakles’ performance to be an improvement upon that of Odysseus. Lloyd-Jones (1967: 227) remarks of Herakles, “Instead of being frightened, he threatens her with his sword.” I infer rather from Meleager’s words ο τοι δος in Bacchylides’ account that Herakles uses his weapon because he is afraid.

32. For the Gorgons’ serpentine hair, see, e.g., Pind. O. 13.63 and Pyth. 10.47. Pausanias tells us that Aeschylus (Cho. 1049-1050) was the first to represent the Erinyes with snakes in their hair. Ar. Ταγηνιστα, fr. 515 PCG, quoted in note 21 above, suggests that Empousa as well as Hekate is snaky-headed. The snaky-headed φσματα in the Suidas passage reported in the same note also include Empousa. Hekate is similarly represented in Sophocles ιζοτμοι, TrGF F535.5-6 ed. Radt: στεφανωσαμνη δρυϊ κα πλεκτας / μν σπεραισι δρακντων.

33. Brown 1991: 49.

34. Lavecchia 1996: 25. In Clark 1970: 250, I corrected a similar error of interpretation regarding Herakles’ journey in Bacchylides 5. But the error persists when Robertson (1980: 295) asserts that Bacchylides’ Herakles sees the ghosts and Meleager “on entering the underworld.”

35. Hard 1977: 211.

36. Cockle (1983: 30) makes this comparison and cites Lucian Dial. Mort. 20 and De Luctu 4 in its support, in order to interpret P.Oxy. 2078. In Clark 2001: 105 and 107, I concluded from an examination of all passages by Lucian concerning Aeacus’ infernal functions that this satirist followed different traditions in different places; for instance, in Dial. Mort. 6.1, Aeacus is the gatekeeper on the far side of the infernal river Pyriphlegethon and Charon’s lake. In the present chapter I have added some new insights on infernal topography with the focus on Apollodorus.

37. Clark 2003: 308-309.

38. Vergil’s use of synonymous expressions to reveal every aspect of this chthonic cave as the transition path from the upper world to the lower is treated more fully in Clark 1992: 167-178.

39. Clark 2001: 114.

40. Warburton 1745: 288 and 294.