1. Text of the Tristia is that of Hall, John B. (Stuttgart and Leipzig 1995Google Scholar). Commentary is that of Luck, Georg, Tristia, vol. 2 (Heidelberg 1977Google Scholar), cited in this article as Luck.

2. By ‘book’ I of course mean papyrus roll. When parchment came to be in regular use for books in the Roman world is a contentious area of discussion. Although the use of parchment as a writing material was certainly known in Republican Rome, there seems however no strong evidence that before late in the first century CE the use of parchment was normal practice. See Roberts, Colin H. and Skeat, T. C., The Birth of the Codex (London and Oxford 1983Google Scholar), chapter 4.

3. Hinds, Stephen, ‘Booking the Return Trip: Ovid and Tristia 1’, PCPS 31 (1985), 13–32Google Scholar, n. 11. See also Evans, Harry B., Ovid’s Publico Carmina (Lincoln and London 1983), 52Google Scholar.

4. See especially lines 87–98 where Ovid expresses the hope that the emperor may favourably receive the poetry book in Rome.

5. Edwards, Catharine, Writing Rome (Cambridge 1996), 125Google Scholar.

6. Nagle, Betty Rose, The Poetics of Exile (Brussels 1980), 35Google Scholar, notes that the acknowledgment of Horace may be a gesture designed to propitiate Augustus. Tr. 3.1, however, does no such thing, as we shall see.

7. Bramble, John, Persius and the Programmatic Satire (Cambridge 1974), 59–62CrossRefGoogle Scholar, provides an excellent excursus on the literary-critical vocabulary metaphorically applied in Ep. 1.20 to the boy slave/book. As he points out, Horace ‘manages to get the best of both worlds, by pretending previous devotion to the most esoteric standards, while simultaneously prophesying, with expected false modesty, a popular future for his compositions’ (60).

8. In Transcending Exile: Conrad, Nabokov, I.B. Singer (Miami 1985Google Scholar), chapter 1, Asher Z. Milbauer argues that for Conrad, for instance, writing is clearly seen as a means of survival, a way of balancing the realities of the present with the memories of the past.

9. Thus in the prefatory poem to the Epistulae ex Ponto written some years later, Ovid tells his addressee Brutus that his poetry is safer in private hands; his poems do not yet dare enter a public library (Ex Pont. 1.1.3–10). Nagle (n.6 above, 86) points out that in Ex Pont. 1.1 there are overt allusions to the conclusion of Tr. 3.1.

10. Although the description of the book in Tr. 3.1.13–16 is less detailed than in Tr. 1.1.3–14, the book nonetheless shares some of the same physical ‘defects’ as those mentioned in the very first poem—the lack of cedar oil (Tr. 1.1.7, Tr. 3.1.13); the lack of smoothing pumice (Tr. 1.1.11f., Tr. 3.1.13); the smudging with tears, liturae (Tr. 1.1.13f., Tr. 3.1.16).

11. Williams, Gareth, ‘Representations of the Book-Roll in Latin Poetry: Ovid, Tr. 1, 1, 3–14 and Related Texts’, Mnemosyne 45 (1992), 178–89CrossRefGoogle Scholar, at 181.

12. On Ovid’s punning on pes in Tristia 1.1 see Hinds (n.3 above), 18–20.

13. Williams, Gareth, Banished Voices (Cambridge 1994), 77Google Scholar.

14. Shuckburgh, E.S. (ed.), Ovid Tristia III (London 1928), 37Google Scholar, on line 13; Kenyon, Frederic C, Books and Readers in Ancient Greece and Rome (Oxford 1951), 48Google Scholar.

15. Horace, Ep. 1.20.2 describes his published book as a boy prostitute pumice mundus. See the note of Mayer, Roland, Horace Epistles Book 1 (Cambridge 1994), 269Google Scholar.

16. Williams (n.ll above), 186.

17. Hinds (n.3 above), 15f.

18. On Ovid’s deployment of the topos of stylistic deterioration in the exile poetry see Williams (n.13 above). Williams argues that stylistic deterioration is a carefully deployed pose on the part of Ovid.

19. Shuckburgh (n.14 above), 37, on line 13. Pliny (NH 13.27) tells the story of the finding of the Pythagorean books in Numa’s coffin 535 years after the king’s supposed burial. The books had not decayed because they had been treated with cedar oil.

20. Cf. Williams (n.11 above), 186, who suggests that Ovid’s dispensing with cedar oil implies his resignation to the defects of his poetry book. But the book itself is not resigned. It is on a mission to find a refuge where it can survive, since the matter of its care cannot be properly looked after by its author.

21. Instead of timide, accepted by the edition of Georg Luck (1967), Hall offers his conjecture miseri. The descriptive epithet thus passes from the book to the author. Timide however directs the reader’s attention to the nature of the book, the subject of the poem’s first eighteen lines. Moreover, it is an apt adverb for a book new to the city, hesitant in speech, and uncertain of its reception. On Hall’s edition see the judicious review of Knox, Peter in BMCR 8.4 (1997), 329–36Google Scholar, esp. 332–36 on Hall’s own conjectures.

22. The Atrium Libertatis: Fasti 4.623f.; Domus Augusta and the Palatine Complex: Fasti 4.943–54; the Temple of Vesta: Fasti 6.255–82.; the Temple of Jupiter Stator: Fasti 6.793f. Bishop, John H., ‘Palatine Apollo’, CQ 6 (1956), 187–92CrossRefGoogle Scholar, traces the route which the personified book follows through Rome. See also-the helpful Map 2 in Ovid, , Sorrows of an Exile, trans. Melville, A.D. and ed. Kenney, E.J. (Oxford 1992Google Scholar).

23. See Newlands, Carole, Playing with Time: Ovid and the ‘Fasti’ (Ithaca, NY, 1995), 57Google Scholar.

24. This procedure begins in Book 1, when the narrator interviews Janus at length (Fasti 1.89–288).

25. Harries, Byron, ‘Causation and the Authority of the Poet in Ovid’s Fasti’, CQ 39 (1989), 164–85CrossRefGoogle Scholar, at 182.

26. A good example of this second procedure also occurs in Book 1, when Ovid gives six explanations for the meaning of the Agonalia (F. 1.317–36). See Miller’s, John discussion of this passage in ‘The Fasti and Hellenistic Didactic: Ovid’s Variant Aetiologies’, Arethusa 25 (1992), 11–31Google Scholar, esp. 20f.

27. Fowler, W. Warde, Aeneas at the Site of Rome (Oxford 1917Google Scholar).

28. Livy (1.12) tells us that the temple of Jupiter Stator was vowed by Romulus in return for Jupiter’s aid in arresting the flight of the Roman armies from the Sabines. On the temple of Jupiter Stator see Richardson, L., A New Topographical Dictionary of Ancient Rome (Baltimore 1992), 225Google Scholar.

29. Cicero, Cat. 1.11,33.

30. At the end of his Res Gestae, Augustus proudly describes the decoration of the front of his house, the doorposts wreathed with laurel and the civic corona above the door (34). Wiseman, T.P., ‘Conspicui postes tectaque digna deo: The Public Image of Aristocratic and Imperial Houses in the Late Republic and Early Empire’, in L’Urbs: Espace Urbain et Histoire (Rome 1987), 393–413Google Scholar, esp. 398f., points out that the decoration of the conspicuous entrance to the Domus Augusta included the honours that the emperor was most proud of, for he mentions them at the end of the Res Gestae. See also Richardson (n.28 above), 117f.

31. In the preface to the Fasti, written in exile, Ovid describes his work as fearful in its relationship to imperial authority (F. 1.4, 19f.). Thus through their narrators these late works of the Augustan principate express a new, cautious relationship between the poet and the state based upon the recognition of the emperor’s power of censorship.

32. See for instance Ovid’s treatment of imperial Concordia in F. 1.645–54, and my discussion (n.23 above), 44–47.

33. Bishop (n.22 above), 110; Fowler (n.27 above), 75.

34. Kenney (n.22 above), 140.

35. Edwards (n.5 above), 120.

36. On the modesty of the emperor’s home see Suet. Aug. 72–73. On the fire and the rebuilt domus see Wiseman (n.30 above), 398–406.

37. As Kenney remarks (n.22 above, 140): ‘Whereas, however, Aeneas was vouchsafed a hopeful vision of future greatness and splendour, for Ovid all this is something lost—unless Augustus relents.’ Nothing in this poem suggests that Augustus will change his mind.

38. Thus in Tr. 1.1 the poet cautiously speculates that his book might be received by the emperor on the Palatine, if the emperor’s mood seems conciliatory (93–104).

39. On the Greek and Roman libraries attached to the temple of Palatine Apollo, see Suet. Aug. 29. On the temple in general see Richardson (n.28 above), 117f.; Carter, John, Suetonius Divus Augustus (Bristol 1982), 130Google Scholar.

40. Cf. Tr. 1.1.105–12, where Ovid imagines that his poetry book will visit his ‘brothers’ in his bookcase at home. The three that teach the art of love will lie hidden, however.

41. The fullest description of the temple and the portico occurs in Propertius 2.31.

42. For a full acount of the myth and its variants see Roscher, , Lexicon (Leipzig 1886–8Google Scholar), i.949–52. Detienne, Marcel, ‘Les Danaïdes entre elles ou la violence fondatrice du mariage’, Arethusa 21 (1988), 159–75Google Scholar, offers an interesting discussion of this myth from an anthropological point of view. He notes that the Greeks linked the Danaids with those other murderers of their husbands, the Lemnians, but adds that the Danaids were more terrible, for they murdered their cousins on their wedding night.

43. See Horace C. 3.11.21–32; Ovid Met. 4.462; Tibullus 1.3.79f.

44. LIMC III. 2, 249–54.

45. O’Gorman, Ellen, ‘Love and the Family: Augustus and Ovidian Elegy’, Arethusa 30 (1997), 103–23CrossRefGoogle Scholar, at 109.

46. O’Gorman (n.45 above), 117.

47. Davisson, M.H.T., ‘Parents and Children in Ovid’s Poems from Exile’, CW 78 (1984), 111–14Google Scholar. She argues that the metaphorical description of the author as parent should be distinguished from his description as slave master in that the representation of Ovid as ‘father’ and his poetry as ‘sons’ and ‘brothers’ emphasises the continuity between his works. See also Hinds (n.3 above), 17–20.

48. Luck, 169.

49. Hinds (n.3 above), 20.

50. Thus although he begins Tristia 2 by deriding his books as infelix cura (1), he concludes the opening couplet by also blaming his own talent: ingenio perii qui miser ipse meo (2). Cf. too Tr. 3.1.13–16: the book looks shabby in part because it is loyal to its author and does not want to embarrass him, in part because the author has injured it.

51. See here the commentary of Owen, S.G., P. Ovidi Nasonis Tristium Liber Secundus (Oxford 1924), 168fGoogle Scholar.

52. The title of pater patriae was granted Augustus in 2 BCE, the year of the publication of the Ars Amatoria. See Suet. Aug. 58. We are told in Res Gestae 35 that the title was inscribed in the porch of Augustus’ house as well as in the Curia Julia and the Forum Augustum.

53. Cf. Am. 1.6 where the lover, as in Tr. 3.1, uses the plea of physical debility—thinness (5f.), tears (17f.)—as a playful part of his argument to the doorkeeper..

54. Suet. Aug. 29 mentions both these buildings as important examples of Augustus’ building programme, for they were erected by people allowed to attach their own name to them.

55. The Porticus enclosed temples of Jupiter Stator and Juno Regina. See Carter on the Porticus Octaviae (n.39 above, 131). Carter cautions that the Porticus Octaviae, mentioned by Suet. Aug. 29, is not to be confused with the Porticus Octavia or Octavi, mentioned by Augustus in Res Gestae 19. The Porticus Octaviae ‘was a replacement or reworking, some time after 27 B.C., of a portico of Metellus which existed from 146 B.C. in the same area.’ See also Clark, J.W., The Care of Books (Cambridge 1909), I3fGoogle Scholar., and Richardson (n.28 above), 317f.

56. On the art collections in the Porticus Octaviae and the Atrium Libertatis see Isager, Jacob, Pliny on Art and Society (London and New York 1993), 160–62Google Scholar and 163–67 respectively. On the layout of the Atrium Libertatis in general see Clark (n.55 above), 12; Richardson (n.28 above), 41.

57. Barchiesi, Alessandro, Il poeta e il principe (Rome 1994), 76–79Google Scholar. On the date of the foundation of the Atrium Libertatis see Sir James Frazer’s commentary on the Fasti (London 1927), iii.314f.

58. On Mutina, and the rumour that Octavian there fled the battlefield, see Suet. Aug. 10.4 and Frazer (n.57 above), iii.315f.

59. Barchiesi (n.57 above), 78.

60. Pliny NH 36.33–34. See Isager’s discussion of the eclectic character of Pollio’s collection (n.56 above, 166f.).

61. Clark (n.55 above), 12.

62. Barchiesi (n.57 above), 78. As Barchiesi says, the joke works well in Latin—as it does in Italian.

63. Doblhofer, Ernst, Exit und Emigration: Zum Erlebnis der Heimatferne in der Römischen Literatur (Darmstadt 1987), 211Google Scholar.

64. Cf. also Met. 1.163–261, the Lycaon episode. Jupiter is first introduced in the Metamorphoses as the punisher of mankind—to excess.

65. See the discussion of fas and nefas in Bömer’s, Franz edition of the Fasti (Heidelberg 1957Google Scholar), i.36.

66. Feeney, Denis, ‘Si licet etfas est: Ovid’s Fasti and the Problem of Free Speech under the Principate’, in Powell, Anton (ed.), Roman Poetry and Propaganda in the Age of Augustus (Bristol 1992), 1–25Google Scholar.

67. Luck, 227.

68. Cf. Tr. 3.1.15. Hinds (n.3 above, 28) emphasises the importance of the Heroides to the reading of the Tristia: ‘What are the Tristia but…a rewriting in exile of the Heroides’!’

69. Nabokov, Vladimir, Poems and Problems (New York 1970Google Scholar).