Isis in the Greco-Roman World: Cultural Memory and Imagination

Juliette Harrisson

‘... priscaque doctrina pollentes Aegyptii... appellant vero nomine reginam Isidem.’

Th and the Egyptians, excelling in ancient learning... call me by my true name, Queen Isis.'

Apuleius, Metamorphoses, 11.4.

The Greco-Roman Isis cult was a deliberately Egyptianizing cult. Actively engaging with Egypt and with cultural memories of ancient Egypt was one of the cult's aims, as the above quote from Apuleius' novel demonstrates; his protagonist, Lucius-as-ass, encounters a goddess who is known by many names, but it is the ancient Egyptians who know her ‘true' name. Only the Egyptians, according to Lucius/Apuleius, maintain the ancient memory of this name, and the true form of the goddess is passed down through the Isis cult that Lucius encounters at Cenchrea. However, as this paper will demonstrate, actual resemblances between Isis worship in ancient (pre-Ptolemaic) Egypt and Isis worship in the Greco-Roman world are few. What Greco-Roman devotees of Isis were actually doing for the most part was not remembering ancient Egyptian rites, values or even myth, but creating an imaginary Isis out of a blend of Greco-Roman Egyptian elements and aspects of Greek ‘mystery' religion. In this way, they created a new Isis within their shared cultural imagi­nation - that is, within the notions of divinity, spirituality and the extra-natural understood, if not believed, by the majority of the inhabitants of the Roman Empire - that was consciously (falsely) believed to be a shared memory.¹

The only area of Isis-worship in which elements appear to have been genuinely, accurately remembered from ancient Egypt right through to the Roman period is in her mythology (though we must add to this the caveat that it is difficult to determine how much her myth may have changed over such a long period of time, because our best evidence for the myth comes from the Roman period).

There is, however, one major aspect of Isis' myth and divine character that was much newer and was especially prominent during the Hellenistic and Roman periods. Although Osiris is not forgotten and still frequently appears as Isis' consort, she also gains another consort - Serapis. Serapis appears to be a combination of Osiris and the Apis-bull and seems to have come into being under either Alexander the Great or the early Ptolemies.⁴ The Romans referred to him as Sarapis, but I have chosen to use the earlier spelling because it may reflect the god's origin; ser means to foretell or prophesy in Egyptian, so the name Ser-Apis may indicate that Serapis was originally a ‘foretelling Apis’⁵ It might be argued that the inclusion of Serapis, a ‘younger' god, in Roman period accounts of the myth of Isis indicates a significant change from pre-Ptolemaic Egypt. However, Serapis as an individual in fact makes little impact on Plutarch and Diodorus' versions of the myth. Plutarch devotes most of the space given to Serapis to trying to work out who or what this particular god is, associating him with, among others, Osiris, Pluto, Dionysus, a son of Heracles and the name of the coffin of the Apis-bull (Plutarch, De Iside et Osiride, 28-9).

The exact origins of the cult of Serapis have been widely disputed. Most agree that Serapis was originally a combination of Osiris and the Apis-bull from Memphis.⁶ However, there is much disagreement concerning whether the transference of the god to Alexandria, a city with which he is strongly connected, and his emergence as Serapis occurred during the reigns of the early Ptolemies or earlier, possibly under Alexander the Great. Stambaugh, for example, has suggested that Alexander ‘came across' a shrine of Osiris and commissioned a new structure for Osiris, at the same time linking him with Osiris-Apis from Memphis and thereby introducing the Hellenized Serapis.⁷ Heyob agrees that this is the most likely origin of the cult and adds that both the name and the character of the god bear this out.⁸ Stiehl also argues, chiefly on the basis of Roman-period evidence, that the cult pre-dated Ptolemy.⁹ She argues, for example, that an attempted incubation at the temple of Serapis carried out as Alexander the Great was dying and recorded by both Plutarch and Arrian provides evidence for the pre-Ptolemaic existence of the cult, on the assumption that Plutarch and Arrian are correct in suggesting that the temple was already in existence by that point (Plutarch, Alexander, 76; Arrian, Anabasis, 7, 25, 1-26, 3).¹⁰

Fraser, on the other hand, argues that Serapis came into being under the early Ptolemies, in Alexandria.¹¹ He acknowledges that Serapis was ‘adapted from an aspect of an Egyptian deity, Osiris', but suggests that the new god ‘was freshly conceived in terms of both Egyptian and Greek theological beliefs’.¹² Like Stiehl, he also bases his assessment on Roman-period sources, but chooses to focus on the two sources that actively describe an origin for the cult: Plutarch's On Isis and Osiris and Tacitus' Histories.

Plutarch’s version of the story records a dream apparently dreamt by Ptolemy I Soter, which demonstrates certain links between Pluto, Serapis and Osiris. In Ptolemy’s dream, a colossal statue of Pluto from Sinope orders him to move it to Alexandria, which Ptolemy does, ‘not without the help of divine providence’ (Plutarch, On Isis and Osiris, 28).¹⁷ Plutarch goes on to discuss the debate concerning which god the statue represented that ensued when it was displayed in Egypt, demonstrating that Pluto is to be identified with Serapis, and further identifying Serapis with Osiris. This does not seem to fit his suggestion in his Life of Alexander that a temple to Serapis was already present in Alexandria. It is important to remember here that Plutarch is not unwilling to change historical details for dramatic effect in his biographies; for example, he moved Julius Caesar’s well-known dream of having sex with his mother from his quaestorship in Spain to the night before he crossed the Rubicon, apparently for dramatic effect.¹⁸

Tacitus records the same story in his Histories, but there it is not the statue that appears to Ptolemy, but decore eximio et maiore quam humana specie iuvenum, ‘a youth with exceptional splendour and greater than human appearance’, who is described as one who oblatum per quietem, ‘appeared in the course of his rest’, which presumably refers to his sleep, though the specific word somnus is not used (Tacitus, Historiae, 4.83).

We are left, then, with a handful of sources, all post-dating the event by a number of centuries, some of which contradict each other (including two written by the same author). Certainty is impossible, but there is one clue that might point to a solution. As is clear from the sources so far cited, dreams and incubation (which I have previously defined as a practice in which a person performs a ritual act and then sleeps in a sacred place, with the deliberate intention of receiving a divine dream, often related to sickness and medicine) play a central role in the mythology and worship of Serapis.¹⁹ However, as I have argued elsewhere, incubation was an essentially Greek practice for which there is limited evidence in pre-Ptolemaic Egypt, though it became extremely popular in Egypt in the Ptolemaic and Roman periods.²⁰ It seems likely, therefore, that this story of carrying out an incubation ritual in an attempt to save Alexander has probably become attached to him at a later date, when the practice had become more common. This would seem to imply that Plutarch’s account in On Isis and Osiris is the more accurate. Further, this seems to be a logical assumption, since Plutarch would be much more likely to invent or move incidents in his biographies, works intended to entertain and improve morals by example, than in a philosophical work in which the aim is to get at the truth concerning these Egyptian gods.

Although Serapis is therefore probably a Ptolemaic ‘invention’, it is important to note that, as described above, he was often remembered not as Ptolemaic, but as Alexandrian. This is significant because among the many legends surrounding Alexander were several tying him ever more closely to Egypt, including some which made him part Egyptian. The Alexander Romance, a novel written in Greek between the first and third centuries ad, claims that Alexander was the son of the last native king of Egypt, Nectanebos (Pseudo-Callisthenes, The Alexander Romance, 1.1).²¹ As rightful pharaoh, he is also claimed as the son of the Egyptian god Ammon (Pseudo-Callisthenes, The Alexander Romance, 1.30). According to the Romance, in the process of founding Alexandria, he was led by dreams and oracles to found the Serapeum at Alexandria; it is implied that he is rediscovering an ancient god who has been neglected (Pseudo-Callisthenes, The Alexander Romance, 1.31-3). Serapis is remembered as Alexandrian and Alexander is remembered as Egyptian; thus Serapis becomes, in the cultural memory, a properly ancient Egyptian god rediscovered by Alexander, a native Egyptian pharaoh.

I will briefly outline some basic aspects of how Isis was depicted and worshipped in ancient Egypt and in the Hellenistic world, to provide a point of comparison with the Greco-Roman Isis of the period of the Roman Empire.²² In ancient Egypt, Isis was chiefly associated with marriage and motherhood, and one of the most common depictions of her shows her suckling Horus or Harpocrates, in an image not dissimilar to the later Christian Virgin and Child.²³ Isis was also known as ‘great in magic'; for example, in the Metternich Stela, in which she is described as ‘the possessor of magic'.²⁴ Her association with magic and spells probably stems from the part of her myth describing the conception of her son Horus. Although there are some variations in the details - for example, Faulkner has argued that the version recorded in Spell 148 of the Middle Kingdom Coffin Texts describes Horus being conceived through a miraculous flash of lightning, rather than post mortem through necrophilia as is usually the case - the conception of a child from a dead father in all versions requires some use of magic in order to be successful.²⁵ She was sometimes depicted as a kite or a woman with long wings, hovering over the dead body of Osiris in order to conceive Horus.²⁶ Perhaps because of this association with magic, or with life-giving powers, she could also be connected with medicine, as she is in Papyrus Ebers, which is a medical text.

As the mother of Horus, a god associated with the divine nature of the pharaoh, Isis was also associated with royalty; her name may have origi­nally meant ‘seat'.²⁷ Her other common headdress was a solar disc between cow horns; she was the Isis-cow, the mother of the Apis-bull, and from the New Kingdom (c. 1550-1070 bc) onwards she was closely associated with the goddess Hathor.²⁸ Hathor was a cow-goddess, sometimes depicted as a woman with cow's ears, sometimes as a cow and sometimes as a woman with horns and a solar disc between them. Although she could be a representation of divine malevolence against humanity, she was more often associated with sexuality and music and her cult was closely associated with the sistrum, a rattling instrument usually played by women which often included an image of her head on the handle. Hathor was sometimes worshipped under the name ‘lady of Byblos'.²⁹

Fraser has suggested that the cult of Isis had already been dispersed by Egyptian merchants before the Hellenistic period, but the Hellenistic cult of Isis appears to have come into being at Alexandria.³⁰ The earliest evidence for the Hellenized Isis cult outside of Egypt is the aretalogy found inscribed at Maroneia, near the Black Sea. This inscription, from the late second or early first century bc, seems to be strongly influenced by the Eleusinian mysteries, as it concludes with praise of Athens and of Eleusis specifically.³¹ As Gasparro has observed, within the text, only the Egyptian names of the goddess and

her husband, Serapis, and a brief reference to Egypt as her favourite place of residence give any clue to her non-Greek origin.³² In the early third century bc, the Hellenized cult of Isis-Serapis appeared in Delos. An inscription on Delos describes how a priest called Apollonius built the Serapeum on the island. He describes his family as being of the Egyptian priestly class, though, as Austin has noted, their names are Greek, the inscription is in Greek and the cult of Serapis they established was that of the Hellenized Serapis.³³ From Delos, the cult eventually spread to Italy, brought to Campania by Italian merchants.³⁴

Both members of the Greco-Roman cult of Isis and Greco-Roman writers in general often show a keen desire to engage with ancient Egypt, which was venerated for its great antiquity and the wisdom thought to be found there. Diodorus Siculus opened his universalizing work by stating that ‘since Egypt is the country where mythology places the origin of the gods, where the earliest observations of the stars are said to have been made... we shall begin our history with the events connected with Egypt' (Diodorus Siculus, Library, 1.9). On the subject of Isis specifically, he claims, ‘as for Isis, when translated the word means “ancient” [παλαιάν], the name having been given her because her birth was everlasting and ancient' (Diodorus Siculus, Library, 1.11).

Despite a certain amount of lip-service to the wisdom and antiquity of Egypt, however, Greek writers always find themselves coming back to the conclusion that whatever they are talking about is really Greek in origin. Plutarch, in particular, has a tendency to temper any authority given to Egypt with the suggestion that the gods and their myths are really Greek anyway, claiming that ‘Isis is a Greek word, and so also is Typhon, her enemy, who is conceited, as his name implies, because of his ignorance and self-deception' (Plutarch, On Isis and Osiris, 2/351f). His many references to the gods' Greek identities seem to go beyond the usual syncretism and almost imply that they are Greek. Twice he refers to words brought to Egypt from Greece in ancient times and then transferred back to Greece again from Egypt (Plutarch, On Isis and Osiris, 29/362e, 61/375e-376a). Intriguingly, he refers to Serapis as ‘foreign' (ξενικός) but Osiris as Greek (Plutarch, On Isis and Osiris, 61/376a). Overall, the implication is that even Egypt's traditional place as the oldest of cultures is denied, as everything Egyptian is revealed, ultimately, to be Greek. Alvar has suggested that one of the aims of this treatise is to justify the relationship between the mysteries of Isis and the Eleusinian mysteries, and this may have been a factor in Plutarch’s work in this regard, but generally speaking, this seems to be part of a much wider attempt on the part of Greek writers such as Plutarch and Diodorus Siculus to claim the heritage of most of the Mediterranean for the Greeks.³⁵ The cult of Isis, as represented by Plutarch, is a Greek cult of ultimately Greek origin which embraces Greek values. The Egyptian element of the cult remains superficial. Even the language and the names of the gods, in Plutarch’s account, are claimed for Greece.

Isis was still associated with Hathor in the Hellenistic period in Egypt, but by the Roman period, in the form of the cult that spread through Italy, the memory of Hathor as a separate but connected goddess seems to have started to fade (though in Egypt itself she remained important, with a temple built at Dendera during the Roman period, for example).³⁶ Diodorus Siculus remembers that Isis is associated somehow with a cow, and suggests that she is given horns on her head by the Egyptians because the cow is ‘held sacred to her’ in Egypt; but he also suggests that they do this partly because of ‘the appearance which she has to the eye when the moon is crescent­shaped’ (Diodorus Siculus, Library, 1.11). He does not appear to connect the horns with Hathor at all. Herodotus had stated that Isis was represented as a woman with cow’s horns exactly like the Greek images of Io, presumably having misunderstood or misrepresented Egyptian Isis/Hathor images, and this may be the ultimate source of the connection sometimes drawn by later sources between Isis and Io (Herodotus, Histories, 2.41).³⁷ Isis was frequently identified with Io during the Roman period. A fresco from Pompeii, for example, shows Io with horns on her head, being greeted by Isis.³⁸ According to Apollodorus, when Io had recovered her son and settled in Egypt, she set up an image of Demeter, whom the Egyptians call Isis, and Io herself is also called Isis in Egypt (Apollodorus, The Library, 2.1.3). Both Apollodorus and Plutarch associate Isis with the town of Byblos; Apollodorus claims Io recovered her son from the queen of Byblos (Apollodorus, The Library, 2.1.3) while Plutarch records that Isis found Osiris’ sarcophagus in Byblos (Plutarch, On Isis and Osiris, 15, 357a). However, neither make any reference to Hathor, the Egyptian ‘lady of Byblos’. Hathor’s name and her separate identity are forgotten, swallowed up into Isis and Io, and cow imagery is almost all that is left of her.

The Roman period engagement with Egyptian culture in general was enthusiastic, but only skin-deep. Swetnam-Burland has argued that it was the image of a thing that made it Egyptian to Roman eyes and suggests that it did not matter particularly whether an object was made from Egyptian materials or used an Egyptian design - it only mattered that the image was of something Egyptian.³⁹ Even in the Temple of Isis at Pompeii, Egyptian-themed images appear, but in a distinctly Roman style. The frescoes are chiefly in the Fourth Style of Roman interior decoration, a style characterized by large images of views across the landscape, framed by architectural details.⁴⁰ These landscapes were usually idealized images of vaguely Mediterranean feel, featuring hilly regions and mountains with temples dotted around them. In the case of the frescoes from the Temple of Isis, these images also include scattered Egyptian elements: an ibis bird wanders across the bottom of an image from the ecclesiasterion, for example, or Isis appears with Io in the scene described above.⁴¹ In the sacrarium, the paintings are more heavily Egyptianized and less professionally painted that in the outer rooms; Moorman has suggested that this indicates they may have been the work of a non-professional painter who was an initiate.⁴² Whether this is the case or not, the same basic features are present as in the outer rooms. Those who used the temple clearly wanted to engage with Egyptian culture on a conscious level, but the form of the decoration remained squarely Roman.

The form and practice of the cult itself was also only superficially Egyptian, though adherents did their best to keep the group memory of ancient Egypt alive. As a mystery cult, the Hellenized cult of Isis bore some similarities to the Eleusinian Mysteries of Demeter, one of the oldest and most important mystery cults. Isis, as a goddess, had been associated with Demeter at least as far back as Herodotus, pre-dating the Hellenistic development of the Greco- Roman Isis cult, so it is natural that her cult should be particularly similar to that of Demeter (Herodotus, Histories, 2.59). Of the five mystery cults Burkert identified as the most prominent and important in his monograph on the subject, those of Demeter and Isis show the greatest similarity, sanctuaries of Isis being modelled after those of Demeter and Persephone.⁴³ We have already seen how important the Eleusinian mysteries were to the writer of the inscription at Maroneia.

In addition to the similarities with the Eleusinian Mysteries, other Greek elements also found their way into the cult. There were certain elements of Egyptian religion that were not considered appropriate in Greek religion. For example, aspects of the Egyptian cult of Isis involving animals did not fit Greek practice and were omitted, while other typically Greek practices found their way in to take their place.⁴⁴ For example, we know from Pausanias, from Apuleius and from the epigraphic evidence that Isis was particularly associated with prophecy and healing through dreams. Pausanias describes a temple of Isis which one is only allowed to enter if invited in a dream (he does not mention if there was any kind of system for checking this) (Pausanias, Description of Greece, 10.32.13). Apuleius makes many references to divine dreams of Isis in the final chapter of his Metamorphoses, including instruc­tions (11.3-6) and invitations (11.22, 11.26, 11.27, 11.30) received through dreams, which may be a reference to this sort of practice.⁴⁵ This is a particu­larly Greek and Near Eastern notion of how mankind might interact with the gods.⁴⁶ Although there is evidence for an interest in prophetic dreams in ancient Egypt, this interest greatly increased following the conquest of Egypt by Alexander, and, in particular, the practice of incubation, developed in Greece from Near Eastern predecessors, became very popular in Egypt from this period onwards and became a major feature of Isiac sites. The connection of Isis with dreams and incubation, then, is just one of many aspects of the cult that was entirely Greek in origin.

Those involved in the organization of the cult did want to remember, pass on and maintain certain aspects of Egyptian religious ritual. The cult was based chiefly in temples with attached clergy, like Egyptian religious practice, but unlike some of the other mystery cults.⁴⁷ The Egyptian tradition of keeping the cult image in the innermost part of the temple, accessible only to a few, was, of course, particularly appropriate for a mystery cult and was maintained, and some of the Egyptian ritual was retained as well.⁴⁸ However, as we have seen, the form and function of the cult was much more Greek mystery religion than a reflection of Egyptian religious practice.

I have defined the term cultural imagination as referring to ideas shared, but not necessarily commonly ‘believed in’, among a very large group of people and which may endure for a long time.⁴⁹ The interaction between the memory of ancient Egypt and the ancient Egypt of the Greco-Roman cultural imagination may best be demonstrated with a short case study of the ‘Isis book' of Apuleius' Metamorphoses. Apuleius wrote the Metamorphoses, structured around the story of a man who turns into an ass but also incorpo­rating a lengthy narrative of the story of Cupid and Psyche and a description of a festival of Isis, in the second half of the second century ad. It is the only Latin novel from the ancient world to have survived in full. In the Isis cult as described by Apuleius, we can see both some genuinely ‘remembered' Egyptian elements and many ‘newer' elements. Overall, the will among the participants to Egyptianize is strong, but actual memories of ancient Egypt are weak.

The question of how seriously we should take Book 11 of Apuleius' Metamorphoses, the final Book in the novel and the one in which he describes his protagonist's ‘conversion' and initiation into the cult of Isis, has been the centre of a great deal of debate for many years. Some argue that Book 11 is a sincere, autobiographical story of religious conversion.⁵⁰ Others argue that the scene should be read purely as a comic parody of cult devotees like Lucius, and S. Harrison has even suggested a possible real-life target for this sort of humour: Aelius Aristides.⁵¹ Harrison also points to the many different genres and influences that come together in the novel.⁵² Taking the third option, one of the most influential theories in the last couple of decades has been that of Winkler, who argues that the story is not necessarily autobiographical and that the sequence can be read both ways, as the narrator deliberately leaves enough ambiguity that either one of two possible interpretations is equally possible.⁵³

However, this question is largely irrelevant for our current purposes. There is no need to assume that the Metamorphoses, or any part of it, is autobio­graphical. Aside from the obvious fantastical content, there is no reason to assume that Apuleius did not invent his hero from the basic outline provided in older versions of the story of Lucius, or the Ass (which tells a shorter, more broadly comic version of the story, and pseudo-Lucian's version of which has also survived in its entirety).⁵⁴ However, it is known that authors of fiction frequently draw on their own experiences and we also know from his Apologia that Apuleius was initiated into several mystery cults (Apuleius, Apologia, 55), so it is reasonable to assume that Apuleius had some personal knowledge of the Isis cult, possibly specifically at Corinth and Rome, and used this in writing his novel. His description of the practical features of the cult, therefore, we assume to be broadly accurate.

The climax of the Metamorphoses takes place at Cenchrea and at Corinth, and this does not appear to have been a meaningless choice on Apuleius' part. Leaving aside for the moment one statement referring to the ‘man from Madauros' (Apuleius, Metamorphoses, 11.27), Lucius is identified at the beginning of the novel as a Corinthian and Apuleius' version of the story takes his hero there instead of on a journey through Thessaly to Thessaloniki, as is the case in pseudo-Lucian's Lucius, or the Ass. Lucius' story is one of coming home (before his journey to Rome interrupts his homecoming; see below). Why Corinth and Cenchrea? It is possible that Apuleius himself had visited the place and witnessed the festival he describes there, and that this was the only reason he chose this setting and this festival for the climax of his novel. However, this place may also have seemed especially suitable for Apuleius' story, because of the different elements of Greco-Roman Isis worship that could be found together there.

According to Pausanias, there were two precincts of Isis on the Acrocorinth, one of Isis Aegyptia and one of Isis Pelagia, that is, Egyptian Isis and Isis of the sea (Pausanias, Description of Greece, 2.4.6).⁵⁵ There were also some other temples and shrines to Isis in the area, one of which was at Cenchrea (Pausanias, Description of Greece, 2.2.3). The Isis temple at Cenchrea, dating from the second century ad, has only tentatively been identified, but seems to have included the usual wall paintings of Egyptian scenes heavily featuring ibises and the Nile (these are a later renovation of presumed similar earlier friezes), a building which may have been constructed, Susan Handler has suggested, in an Egyptianizing style, and two remaining figures, of Homer and Plato.⁵⁶ We may tentatively imply from this that Isis was worshipped under two main titles, or aspects, in Corinth and perhaps in Cenchrea - Egyptian Isis, Isis Aegyptia, and Isis of the Sea, Isis Pelagia.

It is a festival of Isis Pelagia that Apuleius describes, culminating on the sea shore at Cenchrea. The presence of Isis Pelagia in this coastal area will surprise no one, and is borne out by the numismatic evidence; the strongest associ­ation of Isis in coins from Corinth and, particularly, from Cenchrea, is that of Isis Pelagia, and the motif of Isis with a sail clearly identifies her with the sea and marine activity.⁵⁷ This image of a woman with a large, curved sail, may ultimately be derived from ancient Egyptian images of Isis with long wings but this is a tentative assertion and, even if this is the case, all associations of the wings and their accompanying myth are lost, totally subsumed by the importance of her association with the sea.

The other aspect of Isis at Corinth is more interesting: Isis Aegyptia. The name itself suggests that this is a tradition that is more interested than most in actively trying to remember the Egyptian origins of the goddess. Coins depicting Isis carrying a sistrum and situla appear in a number of Peloponnesian cities, including Corinth, from the reign of Hadrian through to that of Severus Alexander. Bricault and Veymiers have suggested that these may have been connected to places where a particularly Egyptian Isis was worshipped, under the name Isis Aegyptia, because the more Hellenized image of the goddess would present her carrying a cornucopia and a sceptre.⁵⁸ The association of the situla with Isis goes back at least to the Ptolemaic period and may, therefore, represent a remembered Egyptian element of her cult. The association of Hathor and, through her, Isis with the sistrum certainly goes back to pre-Ptolemaic Egypt, as we have seen. As these coins demonstrate, Isis was frequently associated with the sistrum in the Roman period, showing that, just like the connection made between Isis and Io springing from the depictions of Isis with cow horns, an element of her ancient connection with Hathor has been represented, but without any memory of the second goddess.

Apuleius' description of the Isis cult at Corinth (Apuleius, Metamorphoses, 11.7-18) emphasizes three main aspects of the cult: the association with the sea (it is a festival of Isis Pelagia that Lucius attends, 11.7-18); the connection with the Underworld that the mystery cult provides (emphasized by the reference to Lucius coming to the gates of Proserpina, 11.23); and the Egyptian origin of the cult (particularly in the description of the paraphernalia carried in the procession, 11.7-9). Isis Pelagia and her association with the sea appears to be an entirely Greco-Roman development, but the association of Isis (and especially of Osiris) with death, rebirth and the Underworld is as ancient as the gods themselves, and the nature of the cult of Isis Aegyptia certainly demonstrates a desire to reach back into the Egyptian past on the part of the worshippers.

Some of the genuine links to Isis' Egyptian past described above can also be seen in the Metamorphoses. For example, when Lucius, the protagonist, sees a vision of Isis, Apuleius describes her as holding a sistrum (with a boat-shaped gold dish decorated with an asp in the other hand) (Apuleius, Metamorphoses, 11.4). Clearly, Corinth was not the only place where this association was made, but Corinth does seem to have been one of the places where the Egyptian elements of the Isis cult were particularly prominent and where some limited aspects of her Egyptian nature were perhaps remembered, or some effort was made to remember them, as well as copying obviously Egyptian motifs from images. Much of Apuleius' description of the Egyptian elements of the cult here, however, displays the usual superficial acknowl­edgement of its Egyptian origins. When he describes the statue of a cow that is carried in the procession, for example, he says that it represents the goddess as fruitful Mother of everyone, and makes no mention of Hathor (Apuleius, Metamorphoses, 11.11). Like Diodorus Siculus, he associates Isis with the moon rather than the sun (the moon is shining brightly when Lucius falls asleep and the lunar goddess features in his prayer; Apuleius, Metamorphoses, 11.1-2). Any ‘memories' of ancient Egypt remain fragmentary, with essential elements from their Egyptian origins missing.

We can see, in Apuleius's description of elements of the Isis cult in Corinth and Cenchrea, some deliberate attempts on the part of those involved in the cult to engage with the goddess's Egyptian past. We have already seen that one of the aspects in which Isis was worshipped at Corinth was Isis Aegyptia, Egyptian Isis, and this very name indicates the desire on the part of initiates to reach back, remember and revive the Egyptian past. Apuleius' description of the rites of Isis Pelagia also shows some deliberate attempts to evoke the goddess's Egyptian past. The procession includes a man dressed as the god Anubis, one person carrying a situla with a snake-handle, and others carrying palm leaves and other Egyptian objects (Apuleius, Metamorphoses, 11.11). However, the importance of the marine aspect of Isis to this port community is always uppermost in the minds of the people Apuleius describes, as this festival exists in order to ask for a good sailing season (Apuleius, Metamorphoses, 11.16).⁵⁹ Egyptian elements are, ultimately, secondary to the much more immediate purpose of the festival, upon which, if one sincerely believed in the goddess and her ability to help those at sea, lives could depend.

In Corinth and Cenchrea, Apuleius' story is dominated by Isis and by her mysteries, and the results for the hero Lucius are uniformly positive. The goddess's cult is celebrated as a mystery cult of Egyptian origin, the goddess herself as a saviour queen. Lucius has his possessions returned to him and is cautioned to wait and be patient until he is really ready to become an initiate of her cult (Apuleius, Metamorphoses, 11.20-1). He is eventually told, by Isis in a dream, that he may undergo initiation and enthusiastically does so. During the initiation ceremony, which is not described in detail for obvious reasons, Lucius claims he goes to the gates of the Underworld and back, and his family rejoice at his unexpected return after a year's disappearance (while he was trapped in the body of an ass; Apuleius, Metamorphoses, 11.22-5). Emotionally, the Isis cult in Corinth and Cenchrea represents belonging in this story. Lucius is home, among his own people, restored to human form and initiated into a supportive cult.

However, Lucius has one more metamorphosis to undergo before the end of the novel. Instructed by further dreams from Isis, he goes to Rome, where things change. Once he leaves Corinth, Isis is no longer dominant - she remains an important presence, but her consort Osiris becomes the dominant partner. Osiris, as we have seen, was Isis' ancient Egyptian consort, preferred here to Serapis, who was an important part of her cult in Corinth but who is mentioned only briefly by Apuleius.⁶⁰ Osiris takes over Lucius' life entirely, to such an extent that, as Lucius' narrative voice tells us, even the first­person narrator-protagonist worries briefly that he is being cheated (Apuleius, Metamorphoses, 11.29). Where the cult in Corinth provided reassurance and a sense of belonging, the cult of Isis and Osiris in Rome, emotionally, repre­sents difference. Lucius' narrative voice emphasizes that he is in a foreign country, speaking a less familiar language and without the support of friends and family (Apuleius, Metamorphoses, 11.28). Rather, he is at the mercy of the cult and, at one point, of a priest with the significant name ‘Asinius', donkey (Apuleius, Metamorphoses, 11.27).

It is here that Lucius undergoes his final metamorphosis, into the person he will finally become. This last metamorphosis is the most intriguing and controversial of all, and the ultimate source of the argument that the novel is autobiographical, because Lucius seems, at this point, to become more like Apuleius himself. He makes a living pleading in the law courts, just as Apuleius saved his own life and earned notoriety in a legal case (Apuleius, Metamorphoses, 11.28; Apuleius, Apologia). Lucius pleads in Latin, one of Apuleius' native languages (as opposed to the Greek language of Corinth and Cenchrea, in which the only other surviving version of the story is written).

Furthermore, it is here that we come across the ever difficult line that states he is from Madauros, Apuleius' home town, rather than from Corinth, his hero's home, which is especially perplexing considering Lucius has only just come from Corinth to Rome. However, there may be another explanation for this strange turn of events.

Lucius' final transformation creates a new cultural identity for himself, and part of this new identity is that he is becoming more Egyptian. Apuleius pulls together elements of his own cultural identity - Latin, Madauros, the courts - but Lucius' transformation is not into Apuleius, but into an Egyptian. The preface to the novel, which appears to refer to the document itself, describes the story as written on Egyptian papyrus with a Nilotic reed and claims this as the reason for the use of animal metamorphosis within the story, which the preface identifies as Egyptian (despite the apparently Greek origin of the story as told by pseudo-Lucian) (Apuleius, Metamorphoses, 1.1). Egypt is then left to one side for much of the novel until the appearance of the Egyptian gods at the end. Here, the reader is brought full circle, as the narrator of the story becomes Egyptian as well and begins a new life, remembering an old and foreign culture within the heart of Rome. Whereas, in Corinth and Cenchrea, the goddess appeared chiefly to help Lucius, in Rome he must become totally subservient to the powerful and ancient god Osiris. This is symbolized by his shaving of his head, and in the final lines of the novel Lucius proudly describes how he displays his baldness for all to see, marking him out as an Egyptian priest, as different from everyone around him (Apuleius, Metamorphoses, 11.30). The attempt to evoke cultural memories of ancient Egypt feeds into a cultural identity that embraces Egyptianess for Lucius' final transformation into an Egyptian devotee.

This final transformation of Lucius transforms, not his physical body, but his cultural identity. Apuleius was a person who embraced at least three cultural identities. He was a Greek philosopher: he spoke, read and wrote in Greek; he was a student of Greek philosophy and he had studied in Greece. He was a Roman: his father was a duumvir, he wrote as much in Latin as in Greek (perhaps more so, since all his surviving works are in Latin) and the Metamorphoses itself is specifically presented as a Latin - that is, Roman - reworking of a Greek/Egyptian tale. He did not usually emphasize his identity as a North African but, when his home town was dismissed as insignificant by his opponents at his trial on charges of witchcraft, he defended it and made no attempt to deny the association (Apuleius, Apologia, 24). Cultural identity, for Apuleius, is a complex issue and any one aspect of it can be played up against the others when the need arises.

It is this interest in playing with cultural identity that comes through in the final transformation of the Metamorphoses. By drawing on cultural memories of ancient Egypt and placing them in contemporary Rome in the person of a Greek hero, Apuleius draws on all three cultures that go to make up his own identity - Greek, Roman and African. Although Egypt was often considered rather separate from other parts of North Africa, it remains closer in terms of climate and lifestyle. The high scholarly reputation of Greek-speaking Egypt and the presence of the library at Alexandria would also hold an extra appeal for the scholar Apuleius. The memory of ancient Egypt provides a way to form a new identity for his hero, that is neither Greek nor Roman, and Apuleius' hero carves out a new cultural identity for himself from the memories of ancient Africa.

This conclusion has implications for the wider study of cultural memory and identity. Cultural memories, as Assmann has emphasized, feed and reinforce cultural identity. One of the most important aspects of cultural memory, as Assmann defined it, is that it exists in order to cement a sense of cultural identity, and the cultural memory of a group establishes and reinforces that group's sense of identity.⁶¹ In Corinth, attempts to remember Egypt are connected with the sea, because it is a port. In Rome, Lucius' trans­formation is more associated with power, money and profit, and it is more masculine, because that is how Apuleius represents the dominant, military centre of Rome. This goes beyond just Apuleius. Although actual ‘memories' of Egypt among initiates into the cult of Isis in the Roman world are weak and much of the appropriation of Egyptian culture is superficial and image-based, even these weak memories still feed into a sense of cultural identity, as the cult of Isis bases itself around the sea in a port town and in the Campus Martius, centre for culturally vital military activity, in Rome itself.⁶² The memories of another culture altogether reinforce the pre-existing cultural identity of each individual city.

In the case of the cult of Isis, few if any of these memories relate to actual Egyptian practice. It might, then, be disputed how connected they really are to cultural memory.⁶³ However, the important factor here is that, in the cultural imagination, these were memories. It was the ‘belief’ of initiates of Isis that they were continuing an ancient Egyptian tradition in their choice of patron goddess and the way they worshipped her. Cultural memory here has been superseded by cultural imagination in the form of ‘memory’. This may perhaps be best compared with the transmission of myth and legend. We have moved into a different area from Assmann’s historical events commemo­rated to cement cultural identity, into the area of myth which can be used to question and probe cultural problems, questions and assumptions. Like cultural memory, group identity is central to the process, but unlike cultural memory, the substance of what is remembered may be open to greater change and more fluctuations over time.

Notes

1 A note on dates: various references are made in this paper to ‘the Roman period’ or ‘Greco-Roman writers’ - this period is roughly defined as the period from the Roman conquest of Greece in the mid-second century bc through to the end

of the third century ad (after which the term ‘Late Antiquity’ is usually used). The period between the death of Alexander the Great in 323 bc and the Roman conquest of Greece is usually referred to as the Hellenistic period in Greece and Rome. In Egypt, this period extends from the death of Alexander to the death of the last Greek pharaoh, Cleopatra VII, in 30 bc and is usually referred to as the Ptolemaic period. The phrase ‘ancient Egypt’ is sometimes used in this paper to indicate pre-Ptolemaic Egypt.

2 See Tobin 1991: 194 and Bommas 2012: 419-35. On the Pyramid Texts in general, see Shaw and Nicholson 1995: 235-6. On PT 576 = Pyr. §1500a-c, see J. Assmann 2008: 110 (with parallels).

3 Fraser 1972: 246.

4 For a catalogue of monuments to Serapis in Egypt in the Hellenistic and Roman periods, see Kater-Sibbes 1973: 1-64 and Schmidt 2010: 127-46.

5 See Bommas 2005: 24-5.

6 See Stambaugh 1972: 5; Tran Tam Tinh 1982: 101; Heyob 1975: 3; Stiehl 1963: 29-30; Fraser 1972: 250.

7 Stambaugh 1972: 13. Stambaugh argues for an Alexandrian origin for the cult partly on the basis that the legends which place the introduction of Serapis to Alexandria during the reign of Ptolemy I ascribe only the introduction of his statue to this period (Stambaugh 1972: 8). However, given the tendency among ancient religions in general and Egyptian religion in particular to consider the statue to be the living embodiment of the god, this seems to be unnecessarily splitting hairs. He further argues that ‘the ancient sources never say that Ptolemy created Sarapis', but this does not tell us anything about the historical origin of the god - of course the ancient authors would not imply that any god had been deliberately created by a human being, unless they were Epicureans arguing for an early form of atheism (Stambaugh 1972: 12).

Heyob 1975: 3-5.

Bommas agrees with Stiehl on this point; see Bommas 2005: 23-5.

Stiehl 1963: 22-3.

At one point, Fraser appears to agree that the cult originated in Memphis (something Tacitus mentions briefly) and suggests that the sacred bull of Memphis became associated with Apis (associated in life with Ptah, Memphis' city-god) and then with Osiris (Fraser 1972: 250). However, a little further on he suggests the cult may have been created in Alexandria and that ‘the Memphian cult of Sarapis' ‘was probably inaugurated very soon after the Alexandrian one' (Fraser 1972: 253). Fraser suggests that the ‘most plausible explanation' for the cult is that Ptolemy Soter created it to give the Greek population of Alexandria a patron deity, incorporating Egyptian and Greek elements (Fraser 1972: 252).

Fraser 1972: 246.

Fraser, 1972: 246-7.

Fraser 1972: 149.

Alvar 2005: 53.

See Bommas 2005: 24.

Plutarch tells a remarkably similar story in his Lucullus: while besieging Sinope, Lucullus dreams that a figure tells him to step forward because Autolycus has come to meet him (Plutarch, Lucullus, 23). The next day, he took the city and found a statue of Autolycus, the founder, lying on the beach, and was reminded of Sulla's advice to him in his Memoirs, that nothing is so certain as what is signified in dreams (Plutarch, Sulla, 6; Lucullus 23). The purpose in this case seems to be to demonstrate the importance and efficacy of divine dreams. For the more usual placement of this dream in Spain, see Suetonius, Divus Julius, 7; Cassius Dio, Histories, 41.24.2. Brenk has referred to Plutarch's use of this dream as ‘blatant manipulation of a dream for biographical purposes'(Brenk 1975: 346). See also Pelling 1997: 200-1.

See Tran Tam Tinh 1982: 111; Harrisson 2010: 252-3. For the definition of incubation, see Harrisson 2010: 253.

Harrisson 2012, forthcoming. See also Szpakowska 2003: 3-4 on the general lack of interest in conceptualizing dreams in Egypt until well after the New Kingdom period.

Dowden has suggested that this story may have originated in a lost section of the Egyptian tale The Dream of Nectanebos (Dowden 1989: 655).

Isis was well known in the Greek classical period, but the Greco-Roman mystery cult focused around her really came into being during the Hellenistic period. On Isis' presence in classical Athens, see Vasunia 2001: 27-9.

See, for example, a statuette of Isis suckling Horus from the Myers Collection at Eton College, ECM 1717 (Bommas 2010a: 37).

Ritner 1994: 34.

Faulkner 1968: 40. On the Coffin Texts in general, see Shaw and Nicolson 1995: 69.

See Shaw and Nicholson 1995: 142.

See Shaw and Nicholson 1995: 142; Bommas 2011a: 26.

Shaw and Nicholson 1995: 142.

See Shaw and Nicholson 1995: 119.

Fraser 1972: 260; Heyob 1975: 9; Bommas 2005: 16. See further, Bommas 2010b: 25-47.

Henrichs 1984: 157.

Gasparro 2007: 42-3.

Austin 1981: 226-7.

See Takacs 1995: 5; Bommas 2005: 77-8.

Alvar 2005: 41.

See the Hellenistic statue assimilating her to Hathor in Witt 1971: 73. On the Roman temple at Dendera, see Bommas 2005: 103.

Vasunia has gone one step further, and suggested that Egyptian Hathor imagery may, in fact, be the inspiration for fifth-century Greek depictions of Io (Vasunia 2001: 37).

On Isis imagery in Pompeii, see Tran Tam Tinh 1964.

Swetnam-Burland 2007: 118-19. Against this view, see Bommas in this volume. On the Fourth style of Roman interior decoration, see Ling 1991: 71ff.

Eric Moorman has discussed the wall paintings at the Temple of Isis at Pompeii in detail in his recent article, and includes several illustrations (Moorman 2007). Moorman 2007: 152.

See Burkert 1987: 41, 47-8.

44 See Heyob 1975: 2.

45 On the epigraphic evidence, see Harrisson 2010: 252.

46 See Harrisson 2012: forthcoming; see also Szpakowska 2003: 3-4.

47 See Burkert 1987: 38.

48 See Heyob 1975: 2.

49 Harrisson 2010: 314.

50 See, for example, Festugiere 1954: 72, 76-7, 84; Kee 1983: 133.

51 Harrison 2000-1. See also Keulen 2003: 107.

52 Harrison 2005: 218.

53 Winkler 1985: 226-7.

54 See Sullivan 1989.

55 See further Bommas 2011b: 81-2.

56 Handler 1971: 62.

57 See Bricault and Veymiers 2007: 404.

58 Bricault and Veymiers 2007: 405. There are, of course, problems with over-reliance on numismatic evidence as coins can be issued from anywhere and taken all over the Roman world, and I thank Martin Bommas for pointing this out. However, the point bears mentioning.

59 Isis Pelagia was also worshipped in inland places, such as Benevento and the Nemi-Lake.

60 This is particularly interesting given that Isis temples in Rome were often referred to as Serapeia. I am grateful to Martin Bommas for this observation.

61 Assmann 1995: 129. See also Bommas 2011b, in which this issue is explored in detail.

62 The Campus Martius was starting to fill with buildings by Apuleius' time, but its earlier significance as a site of military triumphs is remembered through the name.

63 See further Bommas in this volume.

Bibliography

Adlington, W. (ed. and trans.), 1915. Apuleius: The Golden Ass. London/New York.

Alvar, J., 2005. Romanising Oriental Gods: Myth, Salvation and Ethics in the Cults of Cybele, Isis and Mithras (trans. and ed. R. Gordon). Leiden/Boston.

Assmann, J., 1995. ‘Collective Memory and Cultural Identity' (trans. J. Czaplicka), New German Critique, No. 65, 125-33.

—2000. Religion und kulturelles Gedächtnis. Munich.

—2006. Religion and Cultural Memory: Ten Studies (trans. R. Livingstone). Stanford. —2008. Osirisliturgien in Papyri der Spätzeit. Altägyptische Totenliturgien 3,

Heidelberg.

Austin, M. M., 1981. The Hellenistic World from Alexander the Great to the Roman Conquest. Cambridge.

Babbitt, F. C., 1936. Plutarch: Moralia Vol. 5. Cambridge, MA.

Bommas, M., 2005. Heiligtum und Mysterium: Griechenland und seine ägyptischen Gottheiten. Mainz.

—2010a. ‘Travels to the Beyond in Ancient Egypt', in E. Georganteli and M. Bommas (eds), Sacred and Profane: Treasures of Ancient Egypt from the Myers Collection, Eton College and the University of Birmingham. London, 37-62.

—2010b. ‘Isis in Alexandria', Biblische Notizen Vol. 147, 25-47.

—2011a. ‘Isis: Von der Konigsmacherin zur Universalgottin', Epoc Vol. 5, 24-9.

—2011b. ‘Pausanias' Egypt', in M. Bommas (ed.), Cultural Memory and Identity in Ancient Societies. London, 79-108.

—2012. ‘Isis, Osiris and Serapis', in C. Riggs (ed.), Oxford Handbook for Roman Egypt. Oxford, 419-35.

Brenk, F. E., 1975. ‘The Dreams of Plutarch's Lives, Latomus, Vol. 34, 336-49.

Bricault, L. and R. Veymiers, 2007. ‘Isis in Corinth: The Numismatic Evidence. City, Image and Religion', in L. Bricault, M. J. Versluys and P. G. P. Meyboom (eds), Nile into Tiber: Egypt in the Roman World. Leiden/Boston, 392-413.

Burkert, W., 1987. Ancient Mystery Cults. Cambridge, MA/London.

Butler, H. E. (trans.), 1909. The Apologia and Florida of Apuleius of Madaura. LaVergne, TN.

Dowden, K., 1989. ‘Pseudo-Callisthenes: The Alexander Romance', in B. P.

Reardon (ed.), Collected Ancient Greek Novels. Berkeley/Los Angeles/London, 650-735.

Faulkner, R. O., 1968. ‘The Pregnancy of Isis', The Journal of Egyptian Archaeology 54, 40-4.

Festugiere, A-J., 1954. Personal Religion Among the Greeks. Berkeley/Los Angeles. Fraser, P. M., 1972. Ptolemaic Alexandria Volume 1: Text. Oxford.

Gasparro, G. S., 2007. ‘The Hellenistic Face of Isis: Cosmic and Saviour Goddess', in L. Bricault, M. J. Versluys and P. G. P. Meyboom (eds), Nile into Tiber: Egypt in the Roman World. Leiden/Boston, 40-72.

Graves, R. (trans.), 1990. Lucius Apuleius: The Golden Ass. London.

Handler, S., 1971. ‘Architecture on the Roman Coins of Alexandria', American Journal of Archaeology 75, No. 1, 57-74.

Harrison, S. J., 2000. Apuleius: A Latin Sophist. Oxford.

—2000-1. ‘Apuleius, Aelius Aristides and Religious Autobiography', Ancient Narrative 1, 245-59.

—2005. ‘The Novel', in S. J. Harrison (ed.), A Companion to Latin Literature. Oxford, 213-22.

Harrisson, J. G., 2010. Cultural Memory and Imagination: Dreams and Dreaming in the Roman Empire. Unpublished PhD thesis, University of Birmingham.

—2012. ‘The development of the practice of incubation in the ancient world', in Proceedings of the conference on Medicine in the Ancient Mediterranean World held at the University of Cyprus, Nicosia, September 2008. Forthcoming.

Henrichs, A., 1984. ‘The Sophists and Hellenistic Religion: Prodicus as the Spiritual Father of the Isis Aretalogies', Harvard Studies in Classical Philology 88, 139-58.

Heyob, S. K., 1975. The Cult of Isis Among Women in the Graeco-Roman World. Leiden.

Kater-Sibbes, G. J. F., 1973. Preliminary Catalogue of Sarapis Monuments. Leiden. Kee, H. C., 1983. Miracle in the Early Christian World: A Study in Sociohistorical Method. Yale/New Haven/London.

Keulen, W. H., 2003. ‘Comic Invention and Superstitious Frenzy in Apuleius' Metamorphoses: The Figure of Socrates as an Icon of Satirical Self-Exposure', American Journal of Philology 124, 107-35.

Ling, R., 1991. Roman Painting. Cambridge.

Moormann, E., 2007. ‘The Temple of Isis at Pompeii', in L. Bricault, M. J. Versluys and P. G. P. Meyboom (eds), Nile into Tiber: Egypt in the Roman World. Leiden/ Boston, 137-54.

Pelling, C., 1997. ‘Tragical Dreamer: Some Dreams in the Roman Historians', Greece and Rome, 2nd Ser. 44, No. 2, 197-213.

Perry, B. E., 1966. ‘The Egyptian Legend of Nectanebus', Transactions and Proceedings of the American Philological Association 97, 327-33.

Ritner, R. K., 1994. The Mechanics of Ancient Egyptian Magical Practice. Chicago.

Schmidt, S., 2010. ‘Der Sturz des Serapis. Zur Bedeutung paganer Gotterbilder im spätantiken Alexandria', Biblische Notizen Vol. 147, 127-46.

Shaw, I. and P. Nicholson, 1995. British Museum Dictionary of Ancient Egypt. London.

Stambaugh, J., 1972. Sarapis Under the Early Ptolemies. Leiden.

Stiehl, R., 1963. ‘The Origin of the Cult of Sarapis', History of Religions 3, No. 1, 21-33.

Sullivan, J. P., 1989, ‘Pseudo-Lucian: The Ass', in B. P. Reardon (ed.), Collected Ancient Greek Novels. Berkeley/Los Angeles/London, 589-618.

Swetnam-Burland, M., 2007. ‘Eastern Objects, Roman Contexts: A Taste for

Aegyptiaca in Italy', in L. Bricault, M. J. Versluys and P. G. P. Meyboom (eds), Nile into Tiber: Egypt in the Roman World. Leiden/Boston, 113-36.

Szpakowska, K., 2003. Behind Closed Eyes: Dreams and Nightmares in Ancient Egypt. Swansea.

Takacs, S., 1995. Isis and Sarapis in the Roman World. Leiden/New York/Koln.

Tobin, V. A., 1991. ‘Isis and Demeter: Symbols of Divine Motherhood', Journal of the American Research Center in Egypt 28, 187-200.

Tran Tam Tinh, 1964. Essai sur le culte d’Isis a Pompei. Paris.

—1982. ‘Sarapis and Isis', in B. F. Meyer and E. P. Sanders (eds), Jewish and Christian Self-Definition: Vol. 3: Self-Definition in the Graeco-Roman World. London, 101-17.

Vasunia, P., 2001. The Gift of the Nile: Hellenizing Egypt from Aeschylus to Alexander. Berkeley/Los Angeles.

Winkler, J. J., 1985. Auctor and Actor: A Narratological Reading of Apuleiuss Golden Ass. Berekeley/Los Angeles/London.

Witt, R. E., 1971. Isis in the Ancient World (originally published as Isis in the Graeco­Roman World). Baltimore/London.

10