22. INTER INHVMANOS ... GETAS. The same phrase in the same metrical position at EP I v 65-66 'hoc ubi uiuendum est satis est si consequor aruo / inter inhumanos esse poeta Getas' and EP III v 27-28 'quem ... fatum ... inter inhumanos maluit esse Getas'.

23. LAVDES DE CAESARE DIXI. In 1896 J. Gilbert ingeniously proposed the punctuation 'laudes [potential subjunctive]: de Caesare dixi'. But laus de + ablative instead of the more usual objective genitive construction is supported by Tac Ann I 12 'addidit laudem de Augusto'. Nipperdey there explains de by equating laus with oratio and sermo, both of which take de as a normal construction; but it appears from the present passage that laus de may have been a special term for panegyric. Professor E. Fantham notes that Ovid may have been seeking a synonym for laudātĭō.

24. ADIVTA EST NOVITAS NVMINE NOSTRA DEI. Nouitas nostra could mean either 'my novel attempt' (Wheeler, Lewis and Short) or 'my inexperience'; if the latter, adiuta would bear the uncommon but quite valid meaning 'compensated for'; OLD adiuuo 7 cites passages from Cicero (Fam V xiii 5 'ea quibus secundae res ornantur, aduersae adiuuantur'), Livy, and Ulpian.

25-32. Similar catalogues of the imperial family occur at Met XV 834-47, Tr II 161-68, Tr IV ii 7-12, EP II ii 69-74, and EP II viii 29-34; these passages are quoted from below.

25-26. NAM PATRIS AVGVSTI DOCVI MORTALE FVISSE / CORPUS, IN AETHERIAS NVMEN ABISSE DOMOS. Other mentions of the deified Augustus at vi 15-16 'coeperat Augustus detectae ignoscere culpae; / spem nostram terras deseruitque simul' and viii 63-64 'et modo, Caesar, auum, quem uirtus addidit astris, / sacrarunt aliqua carmina parte tuum'. Ovid had predicted Augustus' apotheosis: see Met XV 838-39 'nec nisi cum senior Pylios aequauerit annos, / aetherias sedes cognataque sidera tanget', Tr II 57-58 'optaui peteres caelestia sidera tarde, / parsque fui turbae parua precantis idem', and Tr V ii 51-52, V v 61-62, V viii 29-30 & V xi 25-26.

Augustus' apotheosis was similar to those of Hercules, Aeneas, Romulus, and Julius Caesar: compare the descriptions at Met IX 262-72 'interea quodcumque fuit populabile flammae / Mulciber abstulerat, nec ... quicquam ab imagine ductum / matris habet, tantumque Iouis uestigia seruat ... maiorque uideri / coepit et augusta fieri grauitate uerendus. / quem pater omnipotens inter caua nubila raptum / quadriiugo curru radiantibus intulit astris', Met XIV 603-4 'quicquid in Aenea fuerat mortale, repurgat [sc Numicius] / et respersit aquis; pars optima restitit illi', Met XIV 824-28 'abstulit [sc Mars] Iliaden: corpus mortale per auras / dilapsum tenues ... pulchra subit facies et puluinaribus altis / dignior', and Met XV 844-46 'Venus ... Caesaris eripuit membris neque in aera solui / passa recentem animam caelestibus intulit astris'.

25. PATRIS AVGVSTI. Patris to make it clear that Ovid is not speaking of Tiberius Caesar Augustus.

26. CORPVS ... NVMEN. Precisely the same distinction is found in Velleius' description of Augustus' apotheosis and the start of Tiberius' reign: 'post redditum caelo patrem et corpus eius humanis honoribus, numen diuinis honoratum, primum principalium eius operum fuit ordinatio comitiorum' (II 124 3).

27. PAREM VIRTVTE PATRI. Compare EP II viii 31-32 (to Augustus, about Tiberius) 'perque tibi similem uirtutis imagine natum, / moribus agnosci qui tuus esse potest'.

27-28. FRENA ... IMPERII. The same metaphor at Tr II 41-42 'nec te quisquam moderatius umquam / imperii potuit frena tenere sui', EP II ix 33 'Caesar ut imperii moderetur frena precamur', and EP II v 75 (of Germanicus) 'succedatque suis orbis moderator habenis'.

At Fast I 531-34 Ovid uses the same metaphor, as here, of Tiberius' accession to power: (Carmenta is prophesying Rome's future) 'et penes Augustos patriae tutela manebit: / hanc fas imperii frena tenere domum. / inde nepos natusque dei [Tiberius was the adopted son of Augustus, and therefore the grandson of Julius Caesar], licet ipse recuset, / pondera caelesti mente paterna feret'. In all of these passages Ovid may have had in mind Aen VII 600 (of Latinus) 'saepsit se tectis rerumque reliquit habenas'.

27-28. FRENA ... SAEPE RECVSATI ... IMPERII. At Tr V iv 15-16 Ovid had used frena recusare of a horse: 'fert tamen, ut debet, casus patienter amaros, / more nec indomiti frena recusat equi'. This perhaps influenced his choice of words here.

27. COACTVS excerpta Scaligeri ROGATVS codd. Ovid is referring to the second meeting of the Senate after the death of Augustus (the first meeting had been devoted to funeral arrangements); at this meeting there had been some confusion over Tiberius' intentions. Rogatus is awkward to construe, since Tiberius must already have been asked to accept power: otherwise he could not have refused the offer. The difficulty of rogatus is clearly shown by the description of the scene in Tacitus: 'et ille [sc Tiberius] uarie disserebat de magnitudine imperii sua modestia. solam diui Augusti mentem tantae molis capacem: se in partem curarum ab illo uocatum experiendo didicisse quam arduum, quam subiectum fortunae regendi cuncta onus, proinde in ciuitate tot inlustribus uiris subnixa non ad unum omnia deferrent: plures facilius munia rei publicae sociatis laboribus executuros ... senatu ad infimas obtestationes procumbente, dixit forte Tiberius se ut non toti rei publicae parem, ita quaecumque pars sibi mandaretur eius tutelam suscepturum ... fessus ... clamore omnium, expostulatione singulorum flexit paulatim, non ut fateretur suscipi a se imperium, sed ut negare et rogari desineret' (Ann I 11-13). Scaliger's conjecture is supported by (and is probably based on) the corresponding description at Suetonius Tib 24 'principatum ... diu ... recusauit ... tandem quasi coactus et querens miseram et onerosam iniungi sibi seruitutem, recepit imperium'.

Professor A. Dalzell notes, however, that Suetonius' description is an imperfect parallel, since coactus is there modified by quasi; he suggests to me that rogatus could be accepted, if it is taken closely with recusati—Tiberius finally accepted what he had many times been offered and had many times refused.

29. VESTAM. Ovid similarly equates Livia with Venus and Juno at EP III i 117-18 'quae Veneris formam, mores Iunonis habendo / sola est caelesti digna reperta toro', and implicitly equates her with Juno at Fast I 650 'sola toro magni digna reperta Iouis'. These appear to be instances of metaphor rather than true equations; but PW XIII,1 913-14 cites inscriptions indicating a cult of Livia-as-Juno.

29-30. LIVIA ... AMBIGVVM NATO DIGNIOR ANNE VIRO. Tiberius is mentioned by Ovid in connection with Livia at Fast I 649, a description of the rededication of the temple of Concordia in AD 10: 'hanc tua constituit genetrix et rebus et ara', but does not figure in Ovid's other mentions of Livia (Fast V 157-58, Tr II 161-62, EP II viii 29-30, and EP III i 117-18); these passages would have been written before Tiberius' assumption of power.

For the coupling of both Augustus and Tiberius with Livia, Professor C. P. Jones cites 'ἡ δοίους σκήπτροισι θεοὺς αὐχοῦσα Σεβαστή / Καίσαρας' from an epigram of Ovid's contemporary Honestus.[27]

30. AMBIGVVM. The same use of ambiguum (which may be an Ovidian peculiarity) at Met I 765-66 'ambiguum Clymene precibus Phaethontis an ira / mota magis' and Met XI 235-36 'est specus in medio, natura factus an arte / ambiguum, magis arte tamen'.

30. ANNE. The word is found at Am III xi 49-50 'quicquid eris, mea semper eris; tu selige tantum, / me quoque uelle uelis, anne coactus amem' and Fast VI 27-28 (Juno speaking) 'est aliquid nupsisse Ioui, Iouis esse sororem / fratre magis dubito glorier anne uiro'; the resemblances between this and the present passage are obvious. Bömer ad loc cites instances of anne from Plautus (Amph 173), Terence (Eun 556), Cicero (Fin IV 23, Att XII xiv 2), and Virgil (G I 32 & II 159, Aen VI 864).

31. DVOS IVVENES. Germanicus and Drusus. For other mentions of them, see Tr II 167 'tui, sidus iuuenale, nepotes', Tr IV ii 9 'et qui Caesareo iuuenes sub nomine crescunt', EP II ii 71-72 'praeterit ipse suos animo Germanicus annos, / nec uigor est Drusi nobilitate minor', and EP II viii 33-34.

31. ADIVMENTA. The word is rare in verse (but see Lucretius VI 1022 and Silius XI 605 & XVI 12), and Ovid here seems to be giving a version of the construction in which people are said to be adiumento, as at Cic Att XII xxxi 2 'magno etiam adiumento nobis Hermogenes potest esse in repraesentando ['in making cash payment'—Shackleton Bailey]', Varro LL V 90, and Rhet Her III 29. TLL I 704 1 cites "Caecil. mort. 18" for 'duo minores, qui sint adiumento', which resembles the present passage, but I do not understand the reference: "Caecil." does not appear in the table of authors.

33. NON PATRIA ... SCRIPTA CAMENA. 'Written in a poem that was not in Latin'. This is the only instance in Ovid of this sense of Camena, which seems to have been a Horatian idiom: see Carm II xvi 38 'spiritum Graiae tenuem Camenae', Ep I i 1-3 'Prima dicte mihi, summa dicende Camena ... Maecenas', and AP 275 'tragicae ... Camenae'. Professor R. J. Tarrant cites Martial XII xciv 5 'fila lyrae moui Calabris exculta Camenis', which possibly refers to Horace.

36. MVRMVR. The hum caused by the exchange of approving comments. Compare Met XIII 123-24 'finierat Telamone satus, uulgique secutum / ultima murmur erat'. Livy (XXXII 22 1) has a murmur of mingled praise and dissent following a speech: 'murmur ortum aliorum cum adsensu, aliorum inclementer adsentientes increpantium'. Other murmura are disapproving or anxious, as at Met I 206, VIII 431 & IX 421, and Aen XII 238-39.

The Latin murmur could be quite loud: Martial uses the word of a lion's roar (VIII liii [lv] 1).

40. SEXTA ... BRVMA. The poem must have been written in the winter of 14.

41. NOCVERVNT. Nocere again used of the Ars Amatoria at xiv 20 'telaque adhuc demens quae nocuere sequor?' and Tr IV 1 35.

42. PRIMAQVE TAM MISERAE CAVSA FVERE FVGAE. The second cause was of course Ovid's error (EP III iii 67-72).

43. STVDII COMMVNIA FOEDERA SACRI. Similar references to shared poetic interests at viii 81 'communia sacra tueri', EP II v 60 (to Salanus, a famous orator) 'seruat studii foedera quisque sui', EP II ix 63-64 (to Cotys, king of Thrace, who was a writer of verse) 'haec quoque res aliquid tecum mihi foederis adfert; / eiusdem sacri cultor uterque sumus', EP II x 17 'sunt tamen inter se communia sacra poetis', and EP III iv 67 'sunt mihi uobiscum communia sacra, poetae'.

The foedera would carry the obligation of mutual assistance.

44. PER NON VILE TIBI NOMEN AMICITIAE. 'By the name of friendship which is not cheap in your eyes' (Wheeler). Professor R. J. Tarrant cites similar invocations at Tr I viii 15 'illud amicitiae sanctum et uenerabile nomen', and EP II iii 19-20 'illud amicitiae quondam uenerabile nomen / prostat', III ii 43 & III ii 100.

44-46. AMICITIAE ... INGENIIS. For Ovid's use of quadrisyllable endings for pentameters, see at ii 10 Alcinoo (p 164).

45-46. SIC VINCTO LATIIS GERMANICVS HOSTE CATENIS / MATERIAM VESTRIS ADFERAT INGENIIS. Compare EP II viii 39-40 'sic fera quam primum pauido Germania uultu / ante triumphantes serua feratur equos'. Germanicus celebrated his triumph in 17: see Tac Ann II 41.

Vestris is a true plural referring to Carus and other poets who might be inspired by Germanicus' exploits. For this use of uester to address one member of a collectivity, see Austin on Aen I 140 and Fordyce on Catullus XXIX 20.

45. VINCTO is my restoration for the manuscripts' CAPTO, which I am unable to construe with catenis. Vincto was first corrupted to uicto, which was then displaced by the gloss capto. For the picture compare AA I 215 'ibunt ante duces onerati colla catenis'; for uincto compare Livy VII 27 8 'eos uinctos consul ante currum triumphans egit', and for uincto ... catenis compare Caesar BG I 53 'trinis catenis uinctus'.

47. PVERI. The sons of Germanicus: Nero, Drusus III, and Gaius Caligula.

47. VOTVM COMMVNE DEORVM. Wheeler translates 'the source of universal prayers to the gods'. But it seems difficult to take uotum in this sense, and impossible to construe deorum. André translates 'c'est le voeu de tous les dieux', but it seems strange to have gods forming a uotum. Postgate placed a comma before deorum; but Germanicus and Agrippina were not gods. Heinsius conjectured SVORVM, but this seems rather forced. I suspect that deorum is correct, the sense of the passage being close to that of Fast II 63-64 'templorum positor, templorum sancte repostor, / sit superis opto mutua cura tui'; but what originally stood in place of uotum is not clear.

48. QVOS LAVS FORMANDOS EST TIBI MAGNA DATOS. 'Whose entrustment to you for education is an immense honour'. For the construction Ehwald (KB 68) cites Aen IX 92 (Cybebe asks that Aeneas' ships be rescued from fire) 'prosit nostris in montibus ortas', 'let it profit them that it was in my mountains that they had their origin' (Jackson Knight).

49. MOMENTA. 'Influence'. Compare Caesar BC III 70 2 'ita paruae res magnum in utramque partem momentum habuerunt', Livy I 47 6, Hor Ep I x 15-16 'ubi gratior aura / leniat et rabiem Canis et momenta Leonis', and Manilius II 901 (of the fifth temple) 'hic momenta manent nostrae plerumque salutis'.

49. MOMENTA Vaticanus 1595 (saec xv), sicut coni Scaliger et Gronouius MONIMENTA BCMFHILT. Similarly, most manuscripts have monimenta at Met XI 285-86 (Ceyx to Peleus) 'adicis huic animo ['my kindly nature'] momenta potentia, clarum / nomen auumque Iouem'.

49-50. SALVTI, / QVAE NISI MVTATO NVLLA FVTVRA LOCO EST. A similar qualification of salus at Met IX 530-31 'quam nisi tu dederis non est habitura salutem / hanc tibi mittit amans'; Bömer ad loc cites other word-plays with salus at Her IV 1, XVI 1 & XVIII 1, and at Tr III iii 87-88.

50. MVTATO ... LOCO. See at viii 86 qui minus ... distet (p 284).

XIV. To Tuticanus

In 3-14 Ovid expresses at length his wish to be sent anywhere, even the Syrtes, Charybdis, or the Styx, as long as he can escape Tomis. Such complaints as these have caused the Tomitans to be angry with him (15-22). But he has been misunderstood: he was complaining not of the people but of the land. Hesiod criticized Ascra, Ulysses Ithaca, and Metrodorus Rome, all with impunity, but Ovid's verse has once more caused him trouble (23-44). The Tomitans have been as kind to him as the Paeligni would have been: they have even granted him immunity from taxation, and publicly crowned him (45-56). After this lengthy account of the Tomitans, he moves to an unexpectedly quick summing-up: Tomis is as dear to him as Delos is to Latona (57-60). This conclusion is immediately undercut by the final distich: his only wish is that Tomis were not subject to attack, and that it had a better climate. This type of undercutting is paralleled elsewhere in Ovid's verse: I discuss these passages at 61-62.

At ix 97-104 Ovid had mentioned the Tomitans' sympathy for him; but the present poem is unique for the praise Ovid bestows on them, and furnishes a striking contrast to the horrific picture of Tomis in, for instance, Tr V x. A primary purpose of Ovid's poetry from exile was to secure recall, and so he no doubt intentionally emphasized his hardships; it is clear enough from this poem that at the same time he was in fact reaching an accommodation with his new conditions of life.

3. VTCVMQVE. 'Somehow (in spite of my hardships)'. The word is used by Ovid only in the poetry of exile, and only in this sense: compare Ibis 9-10 'quisquis is est (nam nomen adhuc utcumque tacebo), / cogit inassuetas sumere tela manus' and EP III ix 53 'postmodo collectas [sc litteras] utcumque sine ordine iunxi'. This is a prose sense of utcumque, common in Livy; when the word is used in verse, it generally means 'whenever' (Hor Epod XVII 52, Carm I xvii 10, I xxxv 23, II xvii 11, III iv 29 & IV iv 35) or 'however' (Aen VI 822; the only instance of the word in Virgil).

4. TE Berolinensis Diez. B. Sant. 1, saec xiii Bodleianus Rawlinson G 105^ul ME BCMFHILT. Me seems unlikely to be right, for the phrase 'nil me praeterea quod iuuet inuenies' would not only be awkward in itself, but would also be in apparent contradiction with the following 'ipsa quoque est inuisa salus', where salus refers back to utcumque ualemus.

4. INVENIES. See at ii 10 Alcinoo (p 164).

5. VLTIMA VOTA. 'My utmost wish'. For this sense of ultimus compare Cic Fin III 30 'summum bonum, quod ultimum appello', Livy XXVII 10 11 'aurum ... quod ... ad ultimos casus ['the greatest emergencies'] seruabatur promi placuit', Hor Carm II vii 1-2 'O saepe mecum tempus in ultimum / deducte Bruto militiae duce' (tempus has the same meaning as casus in the passage from Livy), and Petronius 24 'non tenui ego diutius lacrimas ... ad ultimam perductus tristitiam'.

6. SCILICET seems difficult to explain in this context, and the translators ignore its presence. ILICET ('at once') should possibly be read: the corruption of the rarer word to the more common would be easy enough in view of the final s of the preceding istis.

7. MVTER F¹ Bodleianus Canon. lat. 1, saec xiii Barberinus lat. 26, saec xiii. Muter is so much choicer than the better attested mittar that I have followed editors from Ciofanus to Merkel in printing it. Gronovius (Obseruationes III 1) made a strong case for muter, citing Virgil G II 50 (where however the meaning of mutata is disputed), Hor Sat II vii 63-64 'illa tamen se / non habitu mutatue loco peccatue superne', Claudian Rap Pros I 62 'rursus corporeos animae mutantur in artus' (where mittuntur is a variant reading, which Hall prints), and from Ovid Tr V ii 73-74 'hinc ego dum muter, uel me Zanclaea [Politianus: Panchea codd] Charybdis / deuoret aque [Heinsius: atque codd] suis ad Styga mittat aquis', and EP I i 79 'inque locum Scythico uacuum mutabor ab arcu'; compare as well Cic Balb 31 'ne quis inuitus ciuitate mutetur' and Livy V 46 11 'quod nec iniussu populi mutari finibus posset'.

11. SI QVID EA EST. See at i 17 si quid ea est (p 153).

11. BENE. 'Profitably'. Compare Tac Ann III 44 'miseram pacem uel bello bene mutari'. The word in this sense is generally used in describing good commercial investments: see Plautus Cur 679-80 'argentariis male credi qui aiunt, nugas praedicant, / nam et bene et male credi dico', Sen Suas VII v 'si bene illi pecunias crediderunt faeneratores', Cic II Verr V 56 'ut intellegerent Mamertini bene se apud istum tam multa pretia ac munera conlocasse', and Livy II 42 8.

11. COMMVTABITVR. Commutare was a commercial term: it is used of selling at Cic Clu 129 'ad perniciem innocentis fidem suam et religionem pecunia commutarit', Columella XII 26 2 'reliquum mustum ... aere commutato', Dig II xv 8 24 'si uinum pro oleo uel oleum pro uino uel quid aliud commutauit', and CIL I 585 27.

12. SI QVID ET INFERIVS QVAM STYGA MVNDVS HABET. Professor R. J. Tarrant notes another instance of the same idea at Sen Thy 1013-14 'si quid infra Tartara est / auosque nostros'.

13. GRAMINA. 'Weeds'. Compare Met V 485-86 'lolium tribulique fatigant / triticeas messes et inexpugnabile gramen' and Tr V xii 24 'nil nisi cum spinis gramen habebit ager'; TLL VI.2 2165 65 notes as well Columella IV 4 5 'omnesque herbas et praecipue gramina extirpare, quae nisi manu eleguntur ... reuiuiscunt'.

CARMINA, the reading of C, is a frequent corruption of gramina, occurring as a variant at Met II 841 & XIV 44 and Fast VI 749; it gives no obvious sense in this passage. Bentley's FLAMINA is ingenious but unattractive.

14. MARTICOLIS is possibly an Ovidian innovation, being found elsewhere only at Tr V iii 21-22 'adusque niuosum / Strymona uenisti Marticolamque Geten'.

14. NASO. The use of the third person adds to the emotive power of the tricolon 'ager ... hirundo ... Naso'.

15-16. TALIA SVSCENSENT PROPTER MIHI VERBA TOMITAE, / IRAQVE CARMINIBVS PVBLICA MOTA MEIS. For the similar omission of the est of a perfect passive, even in the presence of a parallel finite verb, see Met VII 517-18 'Aeacus ingemuit tristique ita uoce locutus: / "flebile principium melior fortuna secuta est"'.

15. SVSCENSENT. The word is foreign to high poetry. It occurs in Ovid only here and at EP III i 89-90 'nec mihi suscense, totiens si carmine nostro / quod facis ut facias teque imitere rogo'; the only instances from other poetry cited at OLD suscenseo are from Her XVI-XXI and Martial.

SVSCENSENT is the spelling of C; the other manuscripts have SVCCENSENT. I print susc- because that is the spelling given by the ninth-century Hamburg manuscript at EP III i 89 (cited above), where most manuscripts offer succ-. Succ- is, however, quite possibly correct, for although susc- is the spelling of the ancient manuscripts of Plautus and Terence (and of the older manuscripts of the Heroides), succ- is found at Livy XLII 46 8 in the fifth-century Vienna codex.

18. PLECTAR. Similar uses at Tr III v 49 'inscia quod crimen uiderunt lumina, plector' and EP III iii 64 (Ovid to Amor) 'meque loco plecti commodiore uelit'.

18. AB INGENIO is parallel to per carmina in the preceding line; for the idiom, see at x 46 ab amne (p 346).

20. TELAQVE ... QVAE NOCVERE SEQVOR. See at xiii 41 nocuerunt (p 406).

23. SED NIHIL ADMISI. 'But I have committed no crime'—Wheeler. Compare EP III vi 13 'nec scelus admittas si consoleris amicum'. Admittere in this sense belonged to daily speech: TLL I 752 77 cites Plaut Trin 81, Ter HT 956 'quid ego tantum sceleris admisi miser', Lucilius 690 Marx, and Hor Ep I xvi 53.

25. EXCVTIAT. See at viii 17 excutias (p 263).

25. NOSTRI MONIMENTA LABORIS is rather grand, perhaps because Ovid intended the poem to come near the end of the collection. At Tr III iii 78 Ovid's libelli are called his most lasting monimenta, and at EP III v 35 Ovid flatteringly refers to Maximus Cotta's monimenta laboris.

26. LITTERA DE VOBIS EST MEA QVESTA NIHIL. This, of course, is manifestly untrue. See Tr V x entire, and compare for instance Tr V vii 45-46 'siue homines [sc specto], uix sunt homines hoc nomine digni, / quamque lupi saeuae plus feritatis habent'.

28. ET QVOD PVLSETVR MVRVS AB HOSTE QVEROR. Compare EP III i 25 'adde metus et quod murus pulsatur ab hoste'.

30. SOLVM BCFILT LOCVM MH. The interchange is very common (examples at Met I 345 & VII 57); the reverse corruption in some manuscripts at EP II ii 96 'sit tua mutando gratia blanda loco'.

31-40. The argument Ovid here employs ("other have done what I have done, and not suffered for it") is that used at Tr II 361-538 to excuse the Ars Amatoria.

31-40. VITABILIS. A. G. Lee has ingeniously conjectured VITIABILIS (PCPhS 181 [1950-51] 3). It would have the sense uitiosa; Lee compares such words as aerumnabilis, perniciabilis, and lacrimabilis. He argued that Hesiod nowhere said that Ascra was 'always to be avoided' (although this is a natural inference from Op 639-40) and that the variants miserabilis, mirabilis, and mutabilis 'point to the conclusion that the archetype was here difficult to make out'. For uitium used of localities he cited EP III ix 37 'quid nisi de uitio scribam regionis amarae', and for the word uitiabilis (in the sense 'corruptible') Prudentius Apoth 1045 and Ham 215 (there is a variant uitabilis in a ninth-century manuscript of the Hamartigenia).

Lee's argument is a good one, but uitabilis does not seem in itself objectionable enough to be removed from the text. The variant readings he cites are from unnamed manuscripts of Burman, and are not safe evidence for the condition of the archetype. It can be said in Lee's favour that Heinsius and Bentley before him clearly found uitabilis somewhat strange: Heinsius considered the verse suspect, while Bentley conjectured VT ILLAVDABILIS.

31. ASCRA MFILT. I take ASCRE (BCH) to be a hypercorrect formation by the scribes; Ascra is metrically guaranteed at 34 'Ascra suo' and AA I 28 'Ascra tuis'. It is possible that Ascre is correct, although its use would be strange so close to Ascra in 34: Ovid certainly used both nympha and nymphe (Her IX 103; Met III 357).

32. AGRICOLAE ... SENIS. For Hesiod as an old man compare AA II 3-4 'laetus amans donat uiridi mea carmina palma, / praelata Ascraeo Maeonioque seni', Prop II xxxiv 77 'tu canis Ascraei ueteris praecepta poetae', and Ecl VI 69-70 'hos tibi dant calamos, en accipe, Musae, / Ascraeo quos ante seni'.

35. SOLLERTE ... VLIXE. Sollerte could represent either πολυμήχανος (Il II 173) or πολύτροπος (Od I 1). I believe that Ovid was translating πολύτροπος, since Livius Andronicus in translating Od I 1 had used uersutus to represent the adjective: 'Virum mihi, Camena, insece uersutum'. It is clear from Cic Brut 236 'genus ... acuminis ... quod erat in reprehendendis uerbis uersutum et sollers' that the Romans regarded the two adjectives as having much the same force.

At Hor Sat II v 3-5 πολυμήχανος is translated by dolosus: (Tiresias to Ulysses) 'iamne doloso / non satis est Ithacam reuehi patriosque penates / aspicere?'.

36. HOC TAMEN ASPERITAS INDICE DOCTA LOCI EST. At Od IX 27 Ulysses describes Ithaca to Alcinous as 'τρηχεῖ' [=aspera] ἀλλ' ἀγαθὴ κουροτρόφος'.

36. DOCTA (B; C has DOCTVS) seems clearly preferable to DICTA, offered by most of the manuscripts, which cannot be construed with hoc ... indice. The difficulty with docta is that the passive of docere seems in general to have been used of the person taught, not the thing; this is no doubt what induced Riese to print NOTA, found in certain of Heinsius' manuscripts. Still, the construction seems logical enough in view of the double accusative construction of the verb in the active.

38. SCEPSIVS. Metrodorus[28] of Scepsis (a town on the Scamander, about 60 kilometres upstream from Troy) was famous for his hatred of Rome; see Pliny NH XXXIV 34 'signa quoque Tuscanica per terras dispersa quin [Detlefsen: quae codd] in Etruria factitata sint non est dubium. deorum tantum putarem ea fuisse, ni Metrodorus Scepsius, cui cognomen [Professor R. J. Tarrant suggests that 'Μισορωμαῖος' has fallen out of the text around this point] a Romani nominis odio inditum est, propter MM statuarum Volsinios expugnatos obiceret'. According to Plutarch (Lucullus 22) and Strabo (Geog XIII 1 55), he was a close confidant of Mithridates; apparently, when on a mission to Tigranes, he privately advised him not to give Mithridates the requested assistance against Rome. Tigranes reported this to Mithridates; Metrodorus was either executed by Mithridates, or died of natural causes while being sent back to him. Cicero mentions Metrodorus and his phenomenal memory at de Or II 360.

The present passage is more specific than any other surviving reference to Metrodorus' anti-Roman sentiments; Ovid had perhaps read the scripta in question.

As both Cicero and Pliny use the epithet 'Scepsius', Ovid's reference would have been immediately understood: Mētrŏdōrus could not be used in elegiac verse.

38. ACTAQVE ROMA REA EST. Similar verse-endings at RA 387-88 'si mea materiae respondet Musa iocosae, / uicimus, et falsi criminis acta rea est', Fast IV 307-8 'casta quidem, sed non et credita: rumor iniquus / laeserat, et falsi criminis acta rea est', and Tr IV i 26 'cum mecum iuncti criminis acta [sc Musa] rea est'; other instances of reus agi at Her XIV 120, Met XV 36, Tr I i 24, Tr I viii 46, and Her XX 91. See at xv 12 nil opus est legum uiribus, ipse loquor (p 434) for a full discussion of Ovid's use of legal terminology.

39. FALSA ... CONVICIA has a place in the rhetoric of Ovid's argument, balancing uerissima crimina at 29.

40. OBFVIT AVCTORI NEC FERA LINGVA SVO. Obesse is used of Ovid's own situation at Tr I i 55-56 'carmina nunc si non studiumque quod obfuit odi, / sit satis', IV i 25 'scilicet hoc ipso nunc aequa [sc Musa], quod obfuit ante', IV iv 39 'aut timor aut error nobis, prius obfuit error' & V i 65-68. Compare as well Tr II 443-44 'uertit Aristiden Sisenna, nec obfuit illi / historiae turpis inseruisse iocos'.

41. MALVS = malignus.

41. INTERPRES. The word probably combines the senses of 'translator' and 'interpreter'; that is, the person intentionally misconstrued the meaning of certain passages.

As André points out, Ovid's statement here that his Latin poems have caused him difficulty in Tomis indicates that Latin was not as completely unknown in the city as Ovid claims at, for example, Tr III xiv 47-48, V vii 53-54 'unus in hoc nemo est populo qui forte Latine / quamlibet [Heinsius: quaelibet codd] e medio reddere uerba queat' & V xii 53-54 'non liber hic ullus, non qui mihi commodet aurem, / uerbaque significent quid mea norit, adest'; compare as well Tr III xiv 39-40.

42. INQVE NOVVM CRIMEN CARMINA NOSTRA VOCAT. In crimen uocare was a normal idiom: compare Cic Scaur (e) 'custos ille rei publicae proditionis est in crimen uocatus' and Fam V xvii 2 'ego te, P. Sitti, et primis temporibus illis quibus in inuidiam absens et in crimen uocabare defendi'.

42. NOVVM CRIMEN. The uetus crimen was of course the accusation that the Ars Amatoria was immoral. Professor E. Fantham suggests to me that nouum could have the meaning 'unprecedented', as at Cic Lig 1 'Nouum crimen, C. Caesar, et ante hunc diem non auditum propinquus meus ad te Q. Tubero detulit'. Ovid would therefore be saying that the kind of geographical maiestas the Tomitans were accusing him of did not constitute a proper charge.

43. PECTORE CANDIDVS. 'Kind of heart'. This sense of candidus is constantly misunderstood by modern commentators. The basic transferred sense of the word is 'kind' or 'generous towards others'. This can be clearly seen in such passages as Tr III vi 5-8 'isque erat usque adeo populo testatus, ut esset / paene magis quam tu quamque ego notus, amor; / quique est in caris animi [codd: animo fort legendum] tibi candor amicis— / cognita sunt ipsi quem colis ipse uiro', Tr IV x 130-32 'protinus ut moriar non ero, terra, tuus. / siue fauore tuli siue hanc ego carmine famam, / iure tibi grates, candide lector, ago', Tr V iii 53-54 'si uestrum merui candore fauorem, / nullaque iudicio littera laesa meo est', EP II v 5, EP III ii 21-22 'aut meus excusat caros ita candor amicos, / utque habeant de me crimina nulla fauet', and EP III iv 13 'uiribus infirmi uestro candore ualemus'.

For pectore candidus compare from other authors Hor Epod XI 11-12 'candidum / pauperis ingenium', Val Max VIII xiv praef 'candidis ... animis' and Scribonius Largus praef 5 26 'candidissimo animo'.

44. EXTAT ADHVC NEMO SAVCIVS ORE MEO. Ovid makes similar claims at Tr II 563-65 'non ego mordaci destrinxi carmine quemquam ... candidus a salibus suffusis felle refugi' and Ibis 1-8 'Tempus ad hoc, lustris bis iam mihi quinque peractis, / omne fuit Musae carmen inerme meae ... nec quemquam nostri nisi me laesere libelli ... unus ... perennem / candoris titulum non sinit esse mei'. André says of the present passage, 'C'est oublier le poème Contre Ibis', but Housman wrote 'Who was Ibis? Nobody. He was much too good to be true. If one's enemies are of flesh and blood, they do not carry complaisance so far as to chose the dies Alliensis for their birthday and the most ineligible spot in Africa for their birthplace. Such order and harmony exist only in worlds of our own creation, not in the jerry-built edifice of the demiurge ... And when I say that Ibis was nobody, I am repeating Ovid's own words. In the last book that he wrote, several years after the Ibis, he said, ex Pont. IV 14 44, "extat adhuc nemo saucius ore meo"' (1040). Housman is wrong to adduce this line as though it were a statement made under oath (compare the claim made in 26 'littera de uobis est mea questa nihil'). It is nonetheless true that in the extant poems of reproach Ovid does not identify the person he is addressing.

45. ADDE QVOD. See at xi 21 adde quod (p 368).

45. ILLYRICA ... PICE NIGRIOR. For the formula, Otto (pix) cites this passage and Il IV 275-77 'νέφος ... μελάντερον ἠύτε πίσσα' and from Latin poetry AA II 657-58 'nominibus mollire licet mala: fusca uocetur / nigrior Illyrica cui pice sanguis erit', Met XII 402-3 'totus pice nigrior atra, / candida cauda tamen', EP III iii 97 'sed neque mutatur [uar fuscatur] nigra pice lacteus umor', Her XVIII 7 'ipsa uides caelum pice nigrius', and Martial I cxv 4-5 'sed quandam uolo nocte nigriorem, / formica, pice, graculo, cicada'.

45. ILLYRICA ... PICE. A famous mineral pitch was produced near Apollonia; André cites Pliny NH XVI 59 'Theopompus scripsit in Apolloniatarum agro picem fossilem non deteriorem Macedonica inueniri', NH XXXV 178, and Dioscorides I 73.

45. NIGRIOR. The man who was niger had qualities opposite to those of the man who was candidus; that is, he habitually thought and spoke evil of others. This is illustrated by Hor Sat I iv 81-85 'absentem qui rodit amicum, / qui non defendit alio culpante, solutos / qui captat risus hominum famamque dicacis, / fingere qui non uisa potest, commissa tacere / qui nequit—hic niger est, hunc tu, Romane, caueto'. The same sense is seen at Sat I iv 91 & 100, and at Cic Caec 28 'argentarius Sex. Clodius cui cognomen est Phormio, nec minus niger nec minus confidens quam ille Terentianus est Phormio'. A similar sense of ater is seen at Hor Epod VI 15-16 'an si quis atro dente me petiuerit, / inultus ut flebo puer'; Lindsay Watson ad loc (in an unpublished University of Toronto dissertation) cites Hor Ep I xix 30 'nec socerum quaerit quem uersibus oblinat atris' for the same meaning.

A specific connection is often made between blackness and envy: compare Met II 760 (the home of Inuidia is nigro squalentia tabo) and Statius Sil IV viii 16-17 (atra Inuidia).

Catullus XCIII 2 'nec scire utrum sis albus an ater homo' and similar passages at Cic Phil II 41 and Apuleius Apol 16 are examples of an unrelated idiom meaning 'I know absolutely nothing about you'.

46. MORDENDA. For biting as an image of malice, Watson at Hor Epod VI 15 'atro dente' cites Cic Balb 57 'in conuiuiis rodunt, in circulis uellicant; non illo inimico, sed hoc malo dente carpunt', and Val Max IV 7 ext 2 'malignitatis dentes'; Professor R. J. Tarrant cites Hor Sat II i 77 and Martial V xxviii 7 'robiginosis cuncta dentibus rodit'. The image is of course used at times specifically of jealousy; Watson cites Tr IV x 123-24 'nec, qui detrectat praesentia Liuor iniquo / ullum de nostris dente momordit opus' and EP III iv 73-74 'scripta placent a morte fere, quia laedere uiuos / liuor et iniusto carpere dente solet', and Professor Tarrant cites Hor Carm IV iii 16 'et iam dente minus mordeor inuido' and Pindar P II 52-53 'ἐμὲ δὲ χρεὼν / φεύγειν δάκος ἀδινὸν κακαγοριᾶν'.

47. MEA SORS = ego sortem grauem passus.

48. GRAIOS. The more poetic Graius is more than four times as common in Ovid as Graecus, which, apart from Her III 2, is only found in the Fasti (I 330, IV 63 & V 196) and the Tristia (III xii 41, V ii 68 & V vii 11).

49. GENS MEA PAELIGNI REGIOQVE DOMESTICA SVLMO. This line is a type of hendiadys, the first half of the line being redefined by the second. The other cities of the Paeligni were Corfinium and Superaequum.

51-52. INCOLVMI ... SALVOQVE. The two words, equivalent in meaning, were used together as a common Latin phrase; see Caesar BC I 72 3 'mouebatur etiam misericordia ciuium ... quibus saluis atque incolumibus rem obtinere malebat' & II 32 12 'saluum atque incolumem exercitum', Cic Fin IV 19, Diuin in Q Caec 72, Inuen II 169, and Livy XXIII 42 4 'saluo atque incolumi amico', XXIX 27 3 & XLI 28 9.

53. IMMVNIS is also used without a qualifying word or phrase at Plautus Tr 354, Sall Iug 89 4 'eius [sc oppidi] apud Iugurtham immunes', Cic Off III 49 'piratas immunes, socios uectigales habemus', Cic Font 17, Livy XXXIV 57 10 'urbes ... liberas et immunes' & XXXVII 55 7, and CIL XIV 4012 4. For a recent discussion of immunitas, see V. Nutton, "Two Notes on Immunitas: Digest 27,1,6,10 and 11", JRS 1971, 52-63.

54. EXCEPTIS SI QVI MVNERA LEGIS HABENT. The phrase is difficult. Perhaps legal magistrates enjoyed immunity from taxation; if this is what Ovid is saying, munera legis is related to such expressions as consulatus munus (Cic Pis 23) and legationis munus (Phil IX 3). Munus by itself of magistrates' duties is quite common.

Professor E. Fantham suggests to me, however, that munera legis is a reference to civic duties, or liturgies, that Greek cities imposed on certain of their citizens, and Ovid may be saying that citizens performing such liturgies at Tomis procured exemption from regular taxation.

Wheeler translates 'those only excepted who have the boon by law'. This seems difficult; but Professor A. Dalzell notes that the strangeness of the phrasing may be the results of Ovid's striving for a play on munera/immunis.

55. CORONA. Professor C. P. Jones notes that the corona indicates that Ovid was probably invested with a local priesthood.

57-58. DELIA TELLVS, / ERRANTI TVTVM QVAE DEDIT VNA LOCVM. Accounts of this at Met VI 186-91 (Niobe speaking) 'Latonam ... cui maxima quondam / exiguam sedem pariturae terra negauit! / nec caelo nec humo nec aquis dea uestra recepta est: / exul erat mundo, donec miserata uagantem / "hospita tu terris erras, ego" dixit "in undis" / instabilemque locum Delos dedit' and in the passages cited by Williams at Aen III 76 and Tarrant at Sen Ag 384f.

61-62. DI MODO FECISSENT PLACIDAE SPEM POSSET HABERE / PACIS, ET A GELIDO LONGIVS AXE FORET. In this final distich Ovid unexpectedly reverts from his gratitude to the Tomitans to the subject of the first part of the poem, the inhospitality of the region.

This passage provides an example of the technique pointed out in the Amores by Douglass Parker ("The Ovidian Coda", Arion 8 [1969]) whereby Ovid unexpectedly modifies a poem's tone in the concluding distich. In Am I x Ovid rails against his girl because she has asked him for a present: 'nec dare, sed pretium posci dedignor et odi; / quod nego poscenti, desine uelle, dabo!' (63-64). In Am II xiv Ovid scolds his girl for having an abortion: 'di faciles, peccasse semel concedite tuto, / et satis est; poenam culpa secunda ferat!' (43-44). In II xv, Ovid imagines that he becomes the ring he is giving his girl: 'inrita quid uoueo? paruum proficiscere munus; / illa data tecum sentiat esse fide!' (27-28). Am I vii, I xiii, I xiv, and II xiii are other examples of the device.

62. A GELIDO ... AXE. Compare XV 36 'dura iubet gelido Parca sub axe mori' and Her VI 105-6 (Hypsipyle to Jason) 'non probat Alcimede mater tua—consule matrem— / non pater, a gelido cui uenit axe nurus'.

XV. To Sextus Pompeius

The poem effectively combines a number of commonplaces of the works of exile, subordinating them to the central theme of Ovid's indebtedness to Pompeius. The topic of Ovid as Pompeius' property is to a certain extent foreshadowed in EP I vii, throughout which Ovid refers to himself as a client of Messalinus' family: 'ecquis in extremo positus iacet orbe tuorum, / me tamen excepto, qui precor esse tuus?' (5-6); it is found explicitly at i 35-36 'sic ego sum rerum non ultima, Sexte, tuarum / tutelaeque feror munus opusque tuae'. Syme (HO 156) believes that the addressing of the first and penultimate letters to Pompeius constitutes a dedication of the book to Pompeius. However, as Syme recognizes, the abnormal length of the book indicates that it may be a posthumous collection (see page 4 of the introduction); if so, the arrangement of the poems is presumably by Ovid's literary executor.

The poem is remarkable for the cluster of legal terms at 11-12. The passage is evidence for Ovid's expertise and interest in law. For other indications of this in his works, see at 12 (p 434).

1. SI QVIS ... EXTAT. Pompeius is kept in the third person through line 10; Ovid thereby indicates that he is making a public declaration.

1. EXTAT. As Riese pointed out, the choice in 1-2 is between extat ... requirit and extet ... requirat; the problem is that the manuscripts give extat ... requirat, requirit being found only in a few manuscripts of Heinsius, while extet is a conjecture of Guethling. Owen (1894) thought that the ending of extat caused requirit to be corrupted to requirat; on the other hand, the alteration of extet to extat would be all but automatic. There is a similar difficulty at Tr I i 17-18 'si quis ut in populo nostri non immemor illi [=illic], / si quis qui quid agam forte requirat erit', where most manuscripts have requiret. Both passages seem to involve the assimilation of requirere to the mood of the verb immediately following. I print extat ... requirit in consideration of Tr III x 1-2 'Si quis adhuc istic meminit Nasonis adempti, / et superest sine me nomen in urbe meum' (cited by Lenz), Tr III v 23-24 'si tamen interea quid in his ego perditus oris— / quod te credibile est quaerere—quaeris, agam' and Tr V vii 5 'scilicet ut semper quid agam, carissime, quaeris'.

3. CAESARIBVS = Augusto et Tiberio. Augustus is similarly given primary credit for Ovid's survival at v 31-32 'uiuit adhuc uitamque tibi debere fatetur, / quam prius a miti Caesare [=Augusto] munus habet'.

4. A SVPERIS ... PRIMVS. The same idiomatic use of ab 'after' at v 25-26 'tempus ab his uacuum Caesar Germanicus omne / auferet; a magnis hunc colit ille deis' and Fast III 93-94 (of the month of March) 'quintum Laurentes, bis quintum Aequiculus acer, / a tribus hunc primum turba Curensis habet'.

5. TEMPORA ... OMNIA. Compare i 23 'numquam pigra fuit nostris tua gratia rebus'.

5. COMPLECTAR. Complecti in the weak sense 'include, take in' is found in Ovid only here and at Tr I v 55 'non tamen idcirco complecterer omnia uerbis'. The usage is common in prose (OLD complector 8).

6. MERITIS. Compare i 21-22 'et leuis haec meritis referatur gratia tantis; / si minus, inuito te quoque gratus ero'.

7-10. QVAE NVMERO TOT SVNT. Ovid is very fond of using this type of catalogue to indicate great number. Compare AA I 57-59 ('tot habet tua Roma puellas'), AA II 517-19 ('tot sunt in amore dolores'), AA III 149-50 (the many ways women can ornament themselves), Tr V vi 37-40 (the number of Ovid's ills), and EP II vii 25-28 ('nostrorum ... summa laborum').

8. LENTO CORTICE. 'Tough skin'.

8. GRANA. Ovid does not use pomegranates in his similar catalogues elsewhere. Professor R. J. Tarrant points out to me how Ovid elaborates the novel item of comparison in a full distich with several picturesque details (Punica, lento cortice, rubent), then reviews familiar elements rather more quickly in 9-10, with geography the ordering principle.

9. AFRICA QVOT SEGETES. Compare EP II vii 25 'Cinyphiae segetis citius numerabis aristas' (the Cinyps was a river in Libya).

9. SEGETES ... RACEMOS. Compare AA I 57 'Gargara quot segetes, quot habet Methymna racemos'.

9. TMOLIA TERRA = Lydia. The adjective Tmolius (from Tmolus, a mountain in Lydia famous for its wines) occurs only here.

10. QVOT SICYON BACAS. Compare AA II 518 'caerula quot bacas Palladis arbor habet'. For Sicyonian bacae compare Virgil G II 519 'Sicyonia baca' and Ibis 317 'oliuifera ... Sicyone'.

10. QVOT PARIT HYBLA FAVOS. Fauos stands by a type of metonymy for apes; compare AA II 517 'quot apes pascuntur in Hybla', AA III 150 'nec quot apes Hybla nec quot in Alpe ferae', and Tr V vi 38 'florida quam multas Hybla tuetur apes'. For a similar metonymy, see EP II vii 26 'altaque quam multis floreat Hybla thymis'.

11. CONFITEOR; TESTERE LICET. 'I make a public deposition; you, Pompeius, may be a witness'. The deposition is to the effect that Ovid is now Pompeius' property by virtue of the many gifts Pompeius has made to him.

11. TESTERE ... SIGNATE. André cites Dig XXII v 22 'curent magistratus cuiusque loci testari uolentibus et se ipsos et alios testes uel signatores praebere'.

11. SIGNATE, QVIRITES. After addressing Pompeius directly (testere licet), Ovid addresses those witnessing the mancipatio. As Professor A. Dalzell points out, this was achieved ex iure Quiritium; there is a similar direct address to the witnessing Quirites in the formula for establishing a will (Gaius II 104).

Professor Dalzell also notes the abrupt change of audience; typical of Propertius, this is a very unusual procedure in Ovid.

For signare used without an object, compare Suet Cl 9 2 'etiam cognitio falsi testamenti recepta est, in quo et ipse signauerat' & Nero 17 'cautum ut testamentis primae duae cerae testatorum modo nomine inscripto uacuae signaturis ostenderentur'.

Ovid uses testis and signare in a similarly metaphorical sense at EP III ii 23-24 (he forgives those friends who deserted him in his disaster) 'sint hac [M (Heinsius): hi codd] contenti uenia, signentque [uarr sientque; fugiantque] licebit / purgari factum me quoque teste suum'

12. NIL OPVS EST LEGVM VIRIBVS, IPSE LOQVOR. Ehwald (KB 52) aptly cites Quintilian V vii 9 'duo genera sunt testium, aut uoluntariorum aut eorum quibus in [in add editio Aldina] iudiciis publicis lege denuntiari solet ['or those who are summoned sub poena in trials']'.

The reference in this passage to a legal procedure is rather curious, as is the connected reference in 41-42. But it is clear from Ovid's verse that he had a solid practical expertise and interest in law. In his youth he had been one of the tresuiri monetales or capitales (Tr IV x 33-34), and had also served in the centumviral court (Tr II 93-94; EP III v 23-24). He must have been known for his knowledge of law as well as for his fairness in order to be selected as arbitrator in private cases: 'res quoque priuatas statui sine crimine iudex, / deque mea fassa est pars quoque uicta fide' (Tr II 95-96). E. J. Kenney has presented some interesting statistics concerning the frequent occurrence of legal terms in Ovid's poetry ("Ovid and the Law", Yale Classical Studies XXI [1969] 241-63) comparing the number of occurrences of certain legal terms in Ovid and in Lucretius, Catullus, Virgil, Propertius, Tibullus, and the Odes of Horace. Ius and lex are not much more common in Ovid than in the other poets (the proportions being 134:59 and 74:60 respectively for Ovid and the other poets combined); this is not surprising, since these common words could hardly be considered technical terms. Arbiter (7:4) and lis (23:10) are not much more common in Ovid than in the other poets. But it will be seen from the following list how fond Ovid was of legal terminology: legitimus (16:0), iudex (47:12), iudicium (39:7), index (26:1), indicium (36:8), arbitrium (23:6), reus (23:5), uindex (26:5), uindicare (16:6), uindicta (11:0), asserere (3:0), assertor (1:0). Compare as well the play on legal terminology at AA I 83-86 (with Hollis's notes), and the use of such terms as addicere (Met I 617), fallere depositum (Met V 480 & IX 120), usus communis (Met VI 349), transcribere (Met VII 173), primus heres (Met XIII 154), rescindere (Met XIV 784), accensere (Met XV 546), subscribere (Tr I ii 3), sub condicione (Tr I ii 109), and acceptum referre (Tr II 10).

13. OPES ... PATERNAS. Pompeius appears to have been very wealthy. Seneca speaks of the wealth of a Pompeius (presumably the son of Ovid's patron—so Syme Ten Studies 82, HO 162), who was murdered by Gaius Caligula (Tranq 11 10).

13. REM PARVAM MHIT PARVAM REM BCFL. Either reading is possible enough. On balance, I believe paruam rem to be an intentional scribal alteration to avoid the incidence of a spondaic word in the fourth foot of the hexameter; for a discussion of the phenomenon, see at i 11 uellem cum (p 150).

In an older poet, the alliteration of paruam pone paternas would be a strong argument for the reading (see page 15 of Munro's introduction to his commentary on Lucretius), but Ovid did not use the device in his poetry.

15. TRINACRIA = Sĭcĭlĭa, unusable because it begins with three consecutive short vowels; compare Met V 474-76 (of Ceres) 'terras tamen increpat omnes / ingratasque uocat nec frugum munere dignas, / Trinacriam ante alias'.

André avoids the literal meaning of the passage, joining terra with Trinacria as well as with regnataque ... Philippo and taking it to mean 'estate': 'ta terre de Trinacrie et celle où régna Philippe'. But this sense of terra is rare in Latin (Martial IX xx 2, Apuleius Met IX 35), it is difficult to see how regnataque ... Philippo could stand as an epithet in such a case, and it is clear enough that Ovid is imitating Aen III 13-14 'terra ... acri quondam regnata Lycurgo', as he does at Her X 69 'tellus iusto regnata parenti', Met VIII 623 'arua suo quondam regnata parenti', and Met XIII 720-21 'regnataque uati / Buthrotos Phrygio'. In these lines Ovid states that Pompeius owns Sicily, Macedonia, and Campania, and by the hyperbole indicates the size of Pompeius' holdings. Seneca similarly mentions how the Pompeius murdered by Gaius Caligula possessed 'tot flumina ... in suo orientia, in suo cadentia'.

16. QVAM DOMVS AVGVSTO CONTINVATA FORO. Compare v 9-10 'protinus inde domus uobis Pompeia petetur: / non est Augusto iunctior ulla foro'.

18. QVAEQVE RELICTA TIBI, SEXTE, VEL EMPTA TENES. The line seems rather prosaic. For the thought, compare Cic Off II 81 'multa hereditatibus, multa emptionibus, multa dotibus tenebantur sine iniuria'; for this sense of relicta, compare Nepos Att 13 2 'domum habuit ... ab auunculo hereditate relictam', Livy XXII 26 1 'pecunia a patre relicta', and Martial X xlvii 3 'res non parta labore, sed relicta'.

19. TAM TVVS EN EGO SVM. Professor A. Dalzell notes the play on the dual sense of tuus (devoted/belonging to you) which is probably the basis of the entire poem. For tuus 'devoted' compare Tr II 55-56 '[iuro ...] hunc animum fauisse tibi, uir maxime, meque, / qua sola potui, mente fuisse tuum' and the other passages cited at OLD tuus 6.

19. MVNERE. The word is difficult. 'Gift' seems strange in view of the stress placed on Pompeius' ownership of Ovid. Professor E. Fantham suggests to me that the phrase could mean 'by virtue of whose sad service you cannot say you own nothing in the Pontus', while Professor R. J. Tarrant suggests that munere could mean 'responsibility, charge', with cuius (=mei) as an objective genitive.

21. ATQVE VTINAM POSSIS, ET DETVR AMICIVS ARVVM. This elliptical use of posse seems to be colloquial. The only instance cited by OLD possum 2a from verse is Prop IV vii 74 'potuit [uar patuit], nec tibi auara fuit'; there as well the tone is that of lively speech.

21. AMICIVS ARVVM. The same phrase at Met XV 442-43 (Helenus to Aeneas) 'Pergama rapta feres, donec Troiaeque tibique / externum patrio contingat amicius aruum'. The use of the adjective amicus of things rather than person is in the main a poetic usage, but compare Cic Quinct 34 'breuitas postulatur, quae mihimet ipsi amicissima est', ND II 43 'fortunam, quae amica uarietati constantiam respuit', and Att XII xv 'nihil est mihi amicius solitudine'; other instances in the elder Pliny and Columella.

22. REMQVE TVAM PONAS IN MELIORE LOCO. Compare EP I iii 77-78 'liquit Agenorides Sidonia moenia Cadmus / poneret ut muros in meliore loco'.

24. NVMINA PERPETVA QVAE PIETATE COLIS. Tiberius and Germanicus are meant. For Pompeius' devotion to Germanicus, compare v 25-26 'tempus ab his uacuum Caesar Germanicus omne / auferet; a magnis hunc colit ille deis'.

25-26. ERRORIS NAM TV VIX EST DISCERNERE NOSTRI / SIS ARGVMENTVM MAIVS AN AVXILIVM. This distich does not belong in the text: it is in itself unintelligible, and interrupts a natural progression from 24 to 27. I am not certain that the distich is a simple interpolation, since there is nothing in the context to which it is an obvious gloss. Possibly it has been inserted from another letter from exile, in which its meaning would have been clear from context.

Argumentum is difficult. Wheeler translates, 'For 'tis hard to distinguish whether you are more the proof of my mistake or the relief', and notes 'Apparently Pompey could prove (argumentum) that "error" which Ovid regarded as the beginning of his woes'. But this seems a strange thing to say, for Ovid's error was hardly in need of demonstration.

Auxilium is used in its medical sense, erroris being equivalent to morbi or uulneris; compare RA 48 'uulneris auxilium' and the passages collected at OLD remedium 1.

25. DISCERNERE. Gronovius argued (Obseruationes III xiii) that DECERNERE (MI¹) should be read here, since decernere has the required sense 'uel decertare uel iudicare et certum statuere', whereas discernere means 'separare, dirimere, distinguere, diuidere'. On the evidence of the lexica, however, Gronovius' distinction breaks down, since discernere meaning 'decide, determine, make out' is common enough: compare Sallust Cat 25 3 'pecuniae an famae minus parceret haud facile discerneres', Cic Rep 2 6 'ne nota quidem ulla pacatus an hostis sit discerni ac iudicari potest', Varro LL VII 17 'quo discernitur homo mas an femina sit', and Livy XXII 61 10 'quid ueri sit discernere'. I therefore let discernere stand.

29-30. ET PVDET ET METVO SEMPERQVE EADEMQVE PRECARI / NE SVBEANT ANIMO TAEDIA IVSTA TVO. Compare EP III vii entire (an apology to his friends for the monotony of his verse), and especially the opening lines: 'Verba mihi desunt eadem tam saepe roganti, / iamque pudet uanas fine carere preces. / taedia consimili fieri de carmine uobis, / quidque petam cunctos edidicisse reor'.

30. SVBEANT ANIMO. Subire animo occurs also at Tr I v 13. Ovid uses subire with the dative several times in the poetry of exile (Tr I vii 9, II 147, III iii 14 & V vii 58; EP I ix 11, II x 43 & IV iv 47), but not beforehand; earlier he has the accusative (Met XII 472) or the simple verb (Met XV 307). The dative construction is taken up by the author of the later Heroides (XVI 99, XVIII 62).

31. RES IMMODERATA CVPIDO EST. Cupido similarly called immoderata at Apuleius Plat II 21; elsewhere qualified as immodica (Livy VI 35 6) and immensa (Aen VI 823, Tac Ann XII 7).

33. DELABOR. Cicero uses the word for moving from one subject to another (OLD delabor 5b); here the metaphorical sense 'fall' is still active.

34. IPSA LOCVM PER SE LITTERA NOSTRA ROGAT. This line as it stands is clearly corrupt. I do not understand Wheeler's 'my very letters of their own accord seek the opportunity'; André's 'c'est la lettre qui, d'elle-meme, demande le sujet' seems equally difficult, although locus can certainly have the meaning 'subject, topic of discussion' (OLD locus 24b).

The only parallel I have found is Fast II 861 'iure uenis, Gradiue: locum tua tempora poscunt'. If littera is retained in the present passage, this parallel is of little assistance, since locum there means 'a place within a larger work', and Ovid's poetry cannot ask for a locus in that sense. Taking the passage from the Fasti as a parallel, I once thought that Ovid wrote ipsa locum pro se tristia nostra rogant (or petunt); for the noun triste compare Fast VI 463 'scilicet interdum miscentur tristia laetis', Ecl III 80-81 'triste lupus stabulis, maturis frugibus imbres, / arboribus uenti, nobis Amaryllidis irae', and Hor Carm I xvi 25-26 'nunc ego mitibus / mutare quaero tristia'. I now consider this unlikely, since the personal adjective nostra with tristia seems unidiomatic; but I still believe that littera is the key to the corruption.

Professor R. J. Tarrant has tentatively suggested something like inque locum ... redit, but questions whether in locum, even just after eodem, can have the sense in eundem locum. Professor Tarrant also points out to me the possible relevance of locus in the sense locus communis (compare Sen Suas I 9 'dixit ... locum de uarietate fortunae'); Ovid might be saying that his poetry had made rather frequent use of the locus de exilio. In this case, rogat would require emendation.

One of Heinsius' manuscripts read per se ... facit, which is just possibly correct. Heinsius proposed pro se ... facit, which I do not understand.

35. HABITVRA is a good instance of the future participle used to express what is inevitably destined to happen (with Parca balancing in the pentameter); for the sense, see Tarrant on Sen Ag 43 'daturus coniugi iugulum suae'.

37. INOBLITA = memori. Apparently the only instance of the word in classical Latin.

39. CAELO ... SVB VLLO. Bentley oddly conjectured ILLO, the reading of M^ac, which gives the sense 'under the Tomitan sky'. This obviously contradicts the following transit nostra feros si modo Musa Getas.

41. SERVATOREM occurs in Ovid only here and at Met IV 737-38 (of Perseus) 'auxiliumque domus seruatoremque fatentur / Cassiope Cepheusque pater'. In prose it is several times used in a civic context (Cic Pis 34, Planc 102, Livy VI 20 16 & XLV 44 20; CIL IX 4852 in a dedication to Ioui optimo maximo seruatori conseruatori ... ex uoto suscepto). The solemn overtones of seruatorem must be part of what Ovid means for his own land and for the rest of the world to hear and know; the poem thus ends with an implied pronouncement to balance the public statement of the opening.

42. MEQVE TVVM LIBRA NORIT ET AERE MAGIS. This line clearly refers to mancipatio, the receiving of property (including slaves), which is described by Gaius as follows: 'adhibitis non minus quam [Boeth.: quod cod] quinque testibus ciuibus Romanis puberibus, et praeterea alio eiusdem condicionis qui libram aeneam teneat, qui appellatur libripens ['scale-holder'—de Zulueta], is qui mancipio accipit, aes [aes add Boeth.] tenens, ita dicit: "hunc ego hominem ex iure [Boeth.: iūst cod] Quiritium meum esse aio isque mihi emptus esto hoc aere aeneaque libra", deinde aere percutit libram, idque aes dat ei a quo mancipio accipit quasi pretii loco' (I 119).

MAGIS is found as a secondary reading in F and in the thirteenth-century Barberinus lat. 26; the reading of most manuscripts is MINVS, which seems to me impossible. Several explanations of minus have been advanced:

(i) Gronovius took the line to mean 'tuus sum, immo mancipium tuum, nisi quod sola libra et aes mea mancipatione abfuerunt'. This retention of minus, however, involves Ovid in a qualifying retraction just when he seems to be aiming for a ringing conclusion. As well, the instances of minus cited by Gronovius do not in fact illustrate this passage: among them are EP I vii 25-26 'uno / nempe salutaris quam prius ore minus', Met XII 554-55 'bis sex Herculeis ceciderunt me minus uno ['except for me alone'] / uiribus', and Manilius I 778 'Tarquinio ... minus reges', 'the kings, except for Tarquin'.

Gronovius seems to have realized that difficulties remained, and proposed to read NOVIT in 42 and make 41-42 a relative clause dependent on tellus in 38, so that the concluding lines of the poem would mean 'mea tellus, Sulmo, Roma, Italia, me tuum esse audiet. sed audiet idem etiam, quaecumque sub alia quauis caeli parte terra posita est, et te, meum seruatorem, meque, libra et aere tuum, minus nouit'. Once again, minus seems to weaken the poem fatally.