1. For ancient definitions of ekphrasis see e.g. Aelius Theon Progymnasmata 118.6–120, Webb (1999) and Eisner in the Introduction to this volume.

2. The passage from Diodorus Siculus which I discuss is in fact a fragment found in Photius.

3. That is, specifically as literary description of a visual work of art; see Eisner in the Introduction to this volume.

4. On this and Diodorus’ universal project generally see Clarke (1999). On Diodorus as a historian see also Marincola (1997) and Sacks (1990). On Ovid’s Metamorphoses as drawing on the genre of universal history for structure and content see Wheeler (2002). On the similarity in structure between the two works, see Graf (2002), 119.

5. See especially Hardie (2002) on Ovid’s ‘poetics of illusion’ with further bibliography. See Sharrock (1991) and Eisner and Sharrock (1991) on artistry, realism and the poet’s self-reflection in Ovid’s Pygmalion story. See Schiesaro (2002) on Ovid’s awareness of the role of language in constructing the world around us and his problematisation of ‘the very notion of knowing’.

6. E.g. at 20.43f. he expresses awareness that the capacity of the text to reproduce the truth of history is limited because language must relate sequentially events that actually occurred simultaneously.

7. For instance, in an extended programmatic passage at the end of the first part of Book 1 (32–41) Diodorus uses a description of the extraordinary river Nile to reflect on his own historical project and particularly to explore the line between belief and incredulity. Like his own work, the Nile has aspects which at first seem implausible, contrary to what a reader believes to be possible (παράδοξον), that stretch the boundaries of belief (). Following in a long historiograph-ical tradition, Diodorus encourages his readers to believe in these even when their instinct is scepticism. He often works through troubling accounts in order to explain why they deserve credence, e.g. a violent consequence of the Egyptian reverence for cats that he, in one of only two instances of ‘autopsy’ in his work, claims to have witnessed with his own eyes (1.83f.); here, his own visual experience has caused him to believe what at first hearing sounds unbelievable, and as a result he is able to reassure his reader that even stranger things are true (for autopsy as a strategy used by ancient historians to confirm the truth of something that seems incredible see Marincola [1997], 82f.). See also 3.30 for similar reasoning designed to convert those who are sceptical about the truth of historical accounts. For the term (‘incredible’, ‘stretching belief’) in Herodotus, Thucydides and Xenophon and a discussion of the themes of credulity and disbelief in these authors see Packman (1991), who argues that incredulity is presented by these historians as a temporary state and an inappropriate response to (being told about) an amazing phenomenon; this is the case in Diodorus’ work too.

8. Diodorus 3.31: . The translation here is from the Loeb edition of C.H. Oldfather. Elsewhere, where no attribution is made, the translation is mine. All translations of Ovid are my own. The term peripeteia is used in a technical sense by Aristotle to describe unexpected change in circumstances as a plot device (alongside that of recognition) characteristic of tragedy and also found in other literary genres (Poet. 1452, and also Rhel. 1371bl0 for peripeteia as a cause of wonder); for discussion see Cave (1988), 28–35. Cave summarises it as ‘designed to surprise and amaze the audience by an outcome which defies expectations yet appears in retrospect to be a logical outcome’ (31). Peripeteia is also a key term in Diodorus’ work, describing an unexpected turn of events of just the kind that provides a challenge to human understanding; it is found at 3.29.7, 3.31.2, 3.57.8, 3.58.2, 4.9.7, 4.34.2, 4.42.5, 43.2.6,4.55.1,4.82.3,5.11.1,5.62.1,8.10.4, 13.33.2, 13.35.5, 14.112.5, 17.27.7, 17.47.1, 17.47.5, 17.86.4, 18.59.5, 31.18.3, 33.11.1, 34/35.2.41 and 37.16.1, as well as in the passage discussed below—see n.19.

9. E.g. Phlegon of Tralles Book of Marvels 6–9; Pliny NH 7.4.36 (both authors employ the strategy of autopsy to confirm the truth of such transformation); Livy 24.10. See also Gleason (1995), 39f., and Hansen (1996), 112–26. Note that male to female sex change, on the other hand, is not represented in antiquity as a real phenomenon and appears only in myth (such as the myth of Tiresias.)

10. In nouafert animus mutatas dicere formaslcorpora (‘My spirit leads me to tell of forms changed into new bodies’, Ovid Met. 1.1f.).

11. In this respect it compares to the historical material of the final book. There are other instances of sex change in the Metamorphoses—Tiresias’ double transformation (3.324–31) and the female Caenis’ tranformation into the male Caeneus (12.189–209), as well as the fusion of Hermaphrodite and Salmacis (4.274–388); these are, however, less close to the historical accounts.

12. The episode of sex change in Book 32 that I discuss in this paper may be read as unpacking and resolving a tantalising and inconclusive reference to hermaphrodites and the belief that they are ill-omened monsters that is found trailed much earlier at 4.6.5.

13. Compare Heliodorus’ description of a giraffe, discussed by Morgan (1994) as Heliodorus’ teasing invitation to the reader to take ah active role in the interpretation of the text.

14. Indeed fragmentary description is often deliberately employed as a literary strategy in ancient ekphrastic descriptions of bodies: see e.g. Ovid Amores 1.5 or Lucian Imagines esp. 6, with recognition at 10.

15. Ruth Webb, paper delivered at the Passmore Edwards Conference on Ekphrasis, 11th-13th September 2002.

16. On the theme of recognition and anagnorisis in ancient literature, paticularly as a focus for metatextual reflection on issues of fiction and deception, see Cave (1988), esp. 10–54. The locus classicus, much drawn on by Aristotle in his Poetics, is Homer’s account of the homecoming of Odysseus in the Odyssey.

17. For an illustrative account of the social unease that can arise today from being unable to tell the sex of a person, see Prosser 1998, esp. 1–4, and passim on issues in contemporary trans-sexuality.

18. See e.g. Epictetus’ Discourses 1.16.9–14 on the minor works of nature such as the visible signs enabling viewers to distinguish male and female. ‘Does not the nature of each one among us cry aloud immediately from afar: “I am a man, approach me on this basis, speak to me on this basis, ask for nothing else, behold the signs!”’ (llf.; I have slightly adapted W.A. Oldfather’s Loeb translation here). Epictetus follows by saying that it important not to confuse the sexes; for horror at the idea of such confusion see also Aesch. Tim. 131. For discussion of ancient visual signifiers of gender see Gleason (1995). The ancient statue type of the sleeping hermaphrodite reflects the visuality of sex recognition in ancient Greece and Rome and is also designed to play games with the viewer (Beard and Henderson [2001], 132–39) and ‘engage the viewer in a process of discovery’ (Ajootian [1997], 231), in which one initially encounters what seems from behind to be a reclining woman and then subsequently happens upon the male genitals. The element of surprise and incongruity was significant and may have been apotropaic (Ajootian 1997). See also Ovid Fasti 2.303–58 where there is a nasty shock for Faunus when he is deceived by the superficial signs of sex.

19. In this episode of Herais’ sex change the term peripeteia appears at 32.10.2, 32.10.4, 32.10.5, 32.11.1, and twice at 32.12.1. See also n.8 above.

20. It is significant that the context for this digression is the life of Alexander the Great, a historical figure who was also the subject of fiction such as the Alexander Romance and a point of interpenetration between myth and history in the ancient literary tradition.

21. (32.10.2).

22. 32.10.2.

23. At 32.10.3 Herais’ disease is described with an only slightly different formulation as (‘a paradoxical and utterly incredible ailment’). Her transformation is also described as a paradoxon at 32.10.4, 5, and 7. A search in the TLG s.v. paradoxon and related terms yields 345 uses in Diodorus’ work; many of these are employed at moments where the author is concerned to address the boundaries between amazement and disbelief.

24. Our first information about the protagonist, then, is that she is dimorphous or ‘two-formed’ with respect to her parentage and cultural background, an Arabian-Macedonian; perhaps there is a teasing suggestion by Diodorus that we might find the solution to his enigma here.

25. This action of lifting up the skirts to display the genitals recalls that of Baubo in the Homeric Hymn to Demeter, who thus distracts Demeter from her grief (on this see Olender 1990), and of the doctor Agnodike in Hyginus Fabulae 274 who displays her genitals in a courtroom in order to prove that she is a woman. Ajootian (1997) describes such passages as the literary equivalent of the anasyromenos figure (lifting up skirt to display genitals) in art which is associated with the goddess Isis (instrumental in Iphis’ transformation in Ovid’s account) and the gender indeterminate figures of Hermaphroditus and Attis, as well as Baubo.

26. The term does not occur often in the Metamorphoses: the constellations seen by Phaethon are miracula (2.195); in Acoetes’ account of Bacchus turning the sailors into dolphins one sailor asks another in amazement ‘in quae miracula uerteris?’ (‘into what amazing things are you being turned?’, 3.673); in some manuscripts (see the Teubner edition ad loc.) Thescelus describes Perseus’ use of Medusa’s head as miracula in scathing terms, only to fall victim to it immediately (5.181); Bacchus witnesses Medea taking away Aeson’s old age as a miracle (7.294); Ceyx’s story of the transformation of Daedalion is miracula (11.346). One may note that in all these cases the amazement is focalised through a viewer (and in the last case through listeners) within the text, suggesting that miracula are dependant on subjective responses to phenomena. In the case of Iphis the viewer is (the population of) Crete itself.

27. In this respect there are clear parallels between this tale and the story of Pygmalion and his statue in the following book; here too an apparently insurmountable barrier to romance (the fact that the beloved is no more than a statue) is overcome through the appeal to and intervention of a goddess, and the outcome is marriage and celebration.

28. For an analysis of the Iphis narrative that focuses on the significance of names in the story see Wheeler (1997).

29. At 32.11 Diodorus tells in more condensed form the story of a similar transformation that took place thirty years later in Epidauros to a girl called Callo (later Callon). The description of the physicians’ work in this case is even more detailed: ‘first of all, cutting into the glans he made a passage into the urethra, and inserting a silver catheter drew off the liquid residues…’ (32.11.3, tr. Walton {Loeb}). There are some similarities with Galen’s later medical accounts of of male and female genitals (de usu partium 14.2).

30. Tr. Walton (Loeb).

31. Ibid.

32. Ibid.

33. One of Ovid’s ‘invitations to view’ as discussed by Hinds (2002), 136.

34. Cultus is also, of course, a term that evokes the sophistication of Ovid’s own writing; see e.g. Hardie (2002), 1.

35. For a discussion of this omission see Wheeler (1997) who argues that Ovid alludes discreetly to Iphis’ male sexual potency with the terms uires and uigor.

36. See n.18 above.

37. The imagery resonates with ancient rites of passage; on this see Wheeler (1997).

38. Par aetas, par forma fuit, primasque magistris/accepere artes, elementa aetatis, ab isdem (‘equal in age, equal in beauty, they had their first education during the formative years from the same teachers’).

39. For the relationship between male genitals and masculine identity see e.g. Catullus 63; also Henderson (1975), 5 and 77–82 on the word phusis (nature) to mean male genitals.

40. The language in this passage again echoes Aristotle on plot and the effects of literature; see n.8 above.

41. See Webb (1997).

42. For pseudography see Polybius 16.14.78.

43. Does she really visit Telethusa when she is about to give birth? Aut stetit aut uisa est (‘she either stood there—or appeared to’, 688).

44. The text here again resonates with the story of Pygmalion which Ovid will tell in the following book of the Metamorphoses; there too Ovid explores metatextual issues through the figure of the craftsman within the narrative. See Leach (1974); Sharrock (1991); Eisner & Sharrock (1991); Hardie (2002), 173–93.

45. At Iliad 18.592, Daedalus’ mortal craftmanship is compared with Hephaestus’ depiction on the work of art that is the subject of this earliest extant ekphrasis, the shield of Achilles.

46. Ovid compares Daedalus’ extraordinary labyrinth to the river Maeander, taking us back to Diodorus’ description of the Nile as universal history at the start of his own work.