27. A TVRBA RERVM. 'De ces multiples affaires' (André). Heinsius conjectured CVRA, citing ix 71 (addressed to Graecinus as consul) 'cum tamen a rerum cura propiore uacabit'. The conjecture is elegant enough, but the manuscript reading seems sufficiently supported by Her II 75-76 (Phyllis to Demophoon) 'de tanta rerum turba factisque parentis / sedit in ingenio Cressa relicta tuo' and EP III i 144 'per rerum turbam tu quoque oportet eas'; compare as well Columella XI 2 25.

28. MANSVETAS ... MANVS. The same phrase in the same position at Prop III xvi 9-10 'peccaram semel, et totum sum pulsus in annum: / in me mansuetas non habet illa manus'. Mansuetus is foreign to poetic vocabulary, not being found in Virgil or Horace, and only three times in Propertius (I ix 12, I xvii 28, III xvi 10): in Ovid it occurs elsewhere only at Tr III vi 23 'numinis ut laesi fiat mansuetior ira' and Ibis 26.

28. PORRIGET ILLE MANVS. Manus = manum; for the latter, compare Her XVIII 15-16 'protinus haec scribens "felix i littera" dixi, / "iam tibi formosam porriget illa manum"'. Alternatively, the phrase could be taken to indicate Pompeius' gesture of welcoming to a suppliant: at Met III 458 Narcissus, saying how he wished to embrace his reflection, says 'cumque ego porrexi tibi bracchia, porrigis ultro'.

31-32. VIVIT ADHVC VITAMQVE TIBI DEBERE FATETVR, / QVAM PRIVS A MITI CAESARE MVNVS HABET. See on i 2 debitor ... uitae, and compare Tr V ix 11-14 'Caesaris est primum munus, quod ducimus auras; / gratia post magnos est tibi habenda deos. / ille dedit uitam; tu quam dedit ille tueris, / et facis accepto munere posse frui': the similarity of phrasing makes it all but certain that the poem was addressed to Pompeius.

33. MEMORI ... ORE. The phrase belongs to high poetic diction: compare Met VI 508 'absentes pro se memori rogat ore salutent', Met X 204 (Apollo to the dead Hyacinthus) 'semper eris mecum memorique haerebis in ore', and AA III 700 'auditos memori detulit ore sonos'.

35. SANGVINE BISTONIVM QVOD NON TEPEFECERIT ENSEM. Another instance of high poetic diction: compare Her I 19 'sanguine Tlepolemus Lyciam tepefecerat hastam', Aen IX 333-34 'atro tepefacta cruore / terra', Aen IX 418-19 'hasta ... traiecto ... haesit tepefacta cerebro', and Hor Sat II iii 136.

37-38. ADDITA PRAETEREA VITAE QVOQVE MVLTA TVENDAE / MVNERA. The dative expresses purpose. For the sense of tueri 'sustain', compare Tr V ix 13 'uitam ... quam dedit ille tueris', Cic Deiot 22 'atque antea quidem maiores copias alere poterat; nunc exiguas uix tueri potest', Livy V 4 5, XXIII 38 12 & XXXIX 9 5, and Pliny NH XXXIII 134 'M. Crassus negabat locupletem esse nisi qui reditu annuo legionem tueri posset'.

38. NE PROPRIAS ATTENVARET OPES. This may be a reference to the financial burden of living in exile, but more probably refers to the actual financial loss Ovid suffered in exile: 'ditata est spoliis perfida turba meis' (EP II vii 62). It is clear from Tr I vi 7-8 that Ovid had feared such losses from the beginning of his exile.

Attenuare is a very strong verb: compare Met VIII 843-45 (of Erysichthon) 'iamque fame patrias altique uoragine uentris / attenuarat ['had exhausted'—Miller] opes, sed inattenuata manebat / tum quoque dira fames'.

39. PRO QVIBVS VT MERITIS REFERATVR GRATIA. Similar language to Pompeius at i 21 'et leuis haec meritis referatur gratia tantis'.

40. MANCIPII ... TVI (CB²) 'belonging to your property' seems a much more elegant construction than the other manuscripts' MANCIPIVM ... TVVM 'your slave', and was conjectured by Heinsius; in support of mancipium ... tuum Burman cited viii 65-66 'si quid adhuc igitur uiui, Germanice, nostro / restat in ingenio, seruiet omne tibi'.

41-44. Ovid uses the common device of listing adynata; the second version of the device at Tr I viii 1-10, where Ovid says that now his friend has betrayed him he expects to see the adynata occur. Comprehensive listings of adynata in ancient literature given by Smith on Tib I iv 65-66, Shackleton Bailey on Prop I xv 29, Nisbet and Hubbard on Hor Carm I ii 9, xxix 10 & xxxiii 7, and by Gow on Theocritus I 132-36.

42. VELIVOLAS occurs once more at xvi 21 'ueliuolique maris uates', and nowhere else in Ovid's poetry. It is found at Lucretius V 1442 and Aen I 224 'mare ueliuolum', and was from old Latin poetry: Macrobius (Sat VI v 10) cites instances from Livius Andronicus (Morel 58) and Ennius (Ann 380 Vahlen³; Andromache 74 Ribbeck³).

43. SVPINO. 'Backwards'; almost the reverse of praeceps. The same sense at Med Fac 40 'nec redit in fontes unda supina suos'.

45. DIXERITIS. See on 6 transieritis.

45. SVA DONA. Compare Her XII 203 (Medea to Jason) 'dos mea tu sospes' and Sen Med 142 'muneri parcat meo [=uitae suae]' & 228-30.

46. SIC FVERIT VESTRAE CAVSA PERACTA VIAE. 'So you will have carried out the reason for your journey'. The same sense of causa at Met VI 449-50 'coeperat aduentus causam, mandata referre / coniugis' and of peragere (always with mandata as object) at Met VII 502, XI 629 & XIV 460, Fast III 687, and Tr I i 35-36 'ut peragas mandata, liber, culpabere forsan / ingeniique minor laude ferere mei'.

Professor E. Fantham points out to me that Ovid may here be playing on a second sense of causam peragere, 'end a speech [in court]', for which see Met XV 36-37 'spretarumque agitur legum reus ... peracta est / causa prior ['the case for the prosecution'—Miller], crimenque patet' and Her XXI 152.

VI. To Brutus

1. QVAM LEGIS. See at ii 1 quod legis (p 162).

3-4. SED TV QVOD NOLLES, VOLVIT MISERABILE FATVM; / EI MIHI, PLVS ILLVD QVAM TVA VOTA VALET. For the play on nolle/uelle and the thought of 4, compare Met IX 757-58 'quodque ego, uult genitor, uult ipsa socerque futurus, / at non uult natura, potentior omnibus istis'.

5. QVINQVENNIS. Ovid often mentions the time he has spent in exile: see Tr IV vi 19-20 (AD 10) 'ut patria careo, bis frugibus area trita est, / dissiluit nudo pressa bis uua pede', Tr IV vii 1-2 'Bis me sol adiit gelidae post frigora brumae, / bisque suum tacto Pisce peregit iter', Tr V x 1-2 (AD 11-12) 'Vt sumus in Ponto, ter frigore constitit Hister, / facta est Euxini dura ter unda maris', EP I ii 25-26 (AD 12-13) 'hic me pugnantem cum frigore cumque sagittis / cumque meo fato quarta fatigat hiemps', EP I viii 27-28 'ut careo uobis, Stygias detrusus in oras, / quattuor autumnos Pleias orta facit', EP IV x 1 (AD 14) 'Haec mihi Cimmerio bis tertia ducitur aestas', and EP IV xiii 39-40 'sed me iam, Care, niuali / sexta relegatum bruma sub axe uidet'.

Ovid's first full year of exile was AD 9; since Augustus died on 19 August 14, this poem can be securely dated to the final few months of that year.

5. OLYMPIAS in Latin can mean a period of four or of five years; Ovid may have used quinquennis to remove the ambiguity. Olympias elsewhere in classical poetry apparently only at Manilius III 596, where it also denotes a five-year period.

5-6. OLYMPIAS ACTA / IAM Housman OLYMPIAS ACTA EST. / IAM edd. The subject of transit must be Olympias, since otherwise the pentameter is without a subject. Wheeler offers 'the time is now passing to a second lustrum', which does not account for the genitive lustri ... alterius (a second tempus, in the accusative, would have to be understood), while André gives 'et déjà j'entre dans un second lustre', which does not explain the person of transit. The editors' reading could be retained, and Olympias understood as the subject of the pentameter; but it seems simpler to follow Housman in omitting est (with L and T) and joining the two lines in a single sentence.

Transit is in strict terms illogical, since an Olympiad once completed (acta) cannot pass into a second period of time, but the idiom seems natural enough in view of Ovid's use of transire with seasons at Met XV 206 'transit in aestatem post uer robustior annus'; compare as well Fast V 185 (to Flora) 'incipis Aprili, transis in tempora Maii'.

7. PERSTAT ENIM FORTVNA TENAX. In Ovid's case, Fortune does not show her typical inconstancy.

8. OPPONIT NOSTRIS INSIDIOSA PEDEM. Otto pes 7 cites this passage and Petronius 57 10 'et habebam in domo qui mihi pedem opponerent hac illac'.

9-10. CERTVS ERAS ... LOQVI. 'You had made up your mind to speak'. The same idiom at Her IV 151-52, Her VII 9 'certus es, Aenea, cum foedere soluere naues ...?', Met IX 43, X 394 & XI 440; the impersonal construction at Met V 533, IX 53 'certum est mihi uera fateri' & X 38-39.

9. FABIAE LAVS, MAXIME, GENTIS. Similar phrasing at EP III iii 2 'o sidus Fabiae, Maxime, gentis, ades'. This passage seems to be the earliest instance of laus 'object of praise; reason for praise' used of a person: TLL VII.2 1064 73 ff. cites from classical Latin only Eleg Maec 17-18 'Pallade cum docta Phoebus donauerat artes; / tu decus et laudes huius et huius eras', Valerius Flaccus II 243-44 'decus et patriae laus una ruentis, / Hypsipyle', Silius XIII 824, and Martial I xlix 2-3 'nostraeque laus Hispaniae ... Liciniane'. LVX (F²), printed by Burman, is acceptable enough (compare Cic Cat IV 11 'hanc urbem, lucem orbis terrarum'), but is clearly a guess based on F¹'s DVX.

For a full discussion of the career of Paullus Fabius Maximus, consul ordinarius in 11 BC, see Syme HO 135-55. He is the recipient of EP I ii, a request to plead for Ovid with Augustus, and EP III iii, an account of Ovid's vision of Amor which ends with a plea for Fabius' assistance. He is prominently mentioned at Hor Carm IV i 9-12 as a suitable prey for Venus, and it appears from Juvenal VII 94-95 that he was a famous patron of literature: Ovid mentions his scripta at EP I ii 135. We learn from the same poem that Ovid's wife was a member of Fabius' family: 'ille ego de uestra cui data nupta domo est' (136).

10. SVPPLICE VOCE LOQVI. Similar phrasing at Met VI 33 'supplice uoce roga: ueniam dabit illa roganti'. The adjectival use of supplex is not confined to verse; OLD supplex 2 cites instances from Caesar and Suetonius.

11. OCCIDIS ANTE PRECES. 'You died before making your request'. Since Fabius is named in an inscription (CIL VI 2023a, line 17; cited by Froesch 209) as having participated in the election of Drusus to the Arval Brotherhood on 15 May AD 14, he must have died very shortly before Augustus.

11-12. CAVSAMQVE EGO, MAXIME, MORTIS ... ME REOR ESSE TVAE. The death of Fabius, so soon before that of Augustus, seems to have raised popular suspicions. Tacitus (Ann I 5 1-2) mentions a rumour that Fabius had secretly accompanied Augustus to Planasia to visit Agrippa Postumus and that his wife had warned Livia of this; Augustus heard of this, and at Fabius' funeral she was heard blaming herself for his death. If Fabius' death occurred under strange circumstances, Ovid's accusation against himself of having been its cause may have special point.

For a full discussion of the circumstances of Fabius' death, see Syme HO 149-51.

12. NEC FVERAM TANTI. 'But I was not worth this much'. Fueram has the sense of the imperfect, as at AA I 103-4 'tunc neque marmoreo pendebant uela theatro, / nec fuerant liquido pulpita rubra croco'; other instances at Her V 69, AA II 137, AA III 429 & 618, and Tr III xi 25. A full discussion at Platnauer 112-14: he cites thirteen instances from Propertius, who seems to have been fondest of the idiom, and only one certain instance from Tibullus, II v 79 'haec fuerant olim'.

FVERO (BC) gives the sense 'but I will be discovered not to have been worth this much'; the tense seems difficult to fit to the context.

FVERIM (British Library Burney 220, saec xii-xiii) 'but I hope I was not worth so much' is quite possibly correct, and would account for the corruption to fuero.

12. NEC ... TANTI. Similar phrasing at Met X 613 (Atalanta ponders Hippomenes' willingness to risk death to gain her hand) 'non sum me iudice tanti'.

13. MANDARE. 'Consign'; a legal term for charging others with carrying out business on one's behalf, which carried certain obligations with it. See Gaius III 155-62, Just Inst III 26, and the discussion at Buckland 514-21.

15. DETECTAE ... CVLPAE scripsi DECEPTAE ... CVLPAE codd. Me decipit error is a phrase used by Ovid to mean 'I am making a mistake'; see EP III ix 9-12 'auctor opus laudat ... iudicium tamen hic non decipit error ['I do not make this error of judgment'], / nec quicquid genui protinus illud amo'. Ovid uses the expression very often for the "mistake" which led to his exile: see Tr I iii 37-38 (Ovid to his friends on the night of his exile) 'caelestique uiro quis me deceperit error / dicite pro culpa ne scelus esse putet', Tr IV i 23 'scit quoque [sc Musa] cum perii quis me deceperit error', and EP II ii 61 'quasi me nullus deceperit error'. He uses decipere once when speaking of the other cause of his exile: 'o puer [sc Amor], exilii decepto causa magistro' (EP III iii 23). Wheeler took deceptae to refer to Ovid: 'Augustus had begun to pardon the fault I committed in error'. This kind of extreme hypallage, with the true modified noun not expressed, does not however seem to be Ovid's practice, although found in the Silver poets: Statius Theb IX 425 'deceptaque fulmina' means 'the thunderbolts thrown by Jupiter at the request of Semele, who had been deceived by Juno'. Professor J. N. Grant suggests DECEPTI to me; but the genitive of the first person is rare in Ovid, and the perfect participle without expressed noun seems difficult. Owen saw the difficulty with deceptae, and in his second edition referred to Livy XXII 4 4 'id tantum hostium quod ex aduerso erat conspexit; ab tergo ac super caput deceptae insidiae'. But deceptae (which has been variously emended) there means occultae, as explained by Housman (521-22), who cited Prop II xxiv 35-36 'Phrygio fallax Maeandria campo / errat et ipsa suas decipit unda uias' and Sen HF 155 for the same sense; and occultae is clearly not the meaning here required, since Ovid's misdemeanour was all too visible.

Being unable to explain deceptae, I have conjectured detectae. Ovid seems to have committed his error in two stages. First he committed the original misdemeanour; then he kept silent about it when it might have been better for him to speak. Compare Tr III vi 11-13 'cuique ego narrabam secreti quicquid habebam, / excepto quod me perdidit, unus eras. / id quoque si scisses, saluo fruerere sodali'. Later this misdemeanour was discovered: for the arrival of the news of this discovery when Ovid was visiting Elba with Cotta Maximus, see EP II iii 83-90. It is to this discovery that detectae refers: 'Augustus had begun to forgive the misdemeanour that had been revealed'. For this use of detegere compare Met II 544-47 'ales / sensit adulterium Phoebeius [coruus, the raven], utque latentem / detegeret culpam, non exorabilis index, / ad dominum tendebat iter' and Livy XXII 28 8 'necubi ... motus alicuius ... aut fulgor armorum fraudem ... detegeret'.

Professor R. J. Tarrant points out to me the parallel problem at Met IX 711 'indecepta pia mendacia fraude latebant', where context requires indecepta to have the meaning 'undetected'. Indecepta might be taken to support deceptae in the present passage, but I am more inclined to read indetecta for indecepta: of the various conjectures made, Zingerle's inde incepta is most commonly accepted.

At Her IX 101-2 'tolle procul, decepte, faces, Hymenaee, maritas / et fuge turbato tecta nefanda pede!', detecte should similarly be read. Detecte better explains why Hymenaeus should flee; also, Hymenaeus has not been deceived, for it appears from 61-62 'spes bona det uires; fratris [Palmer: fratri codd] nam nupta futura es; / illius de quo mater, et uxor eris' that Macareus had fully intended to marry Canace.

16. SPEM NOSTRAM TERRAS DESERVITQVE SIMVL. The -que should of course be taken with terras.

This is a typical instance of Ovid's love of syllepsis, of giving a single verb two objects (or more), each of which uses a different meaning of the verb. Compare, from many instances, ix 90 'nec cum fortuna mens quoque uersa mea est', Her VII 9 'certus es, Aenea, cum foedere soluere naues', Met II 601-2 'et pariter uultusque deo plectrumque colorque / excidit', Met VIII 177, Fast III 225, Fast III 857 'hic [the messenger of Ino] ... corruptus cum semine', Fast V 652 'montibus his ponunt spemque laremque suum', and EP II vii 84 'meque simul serua iudiciumque tuum'.

16. DESERVITQVE. Ovid does not use deserere with things as object until his poetry of exile: compare Tr I ix 65 'nec amici desere causam'. Instances in the later Heroides at XV 155 'Sappho desertos cantat amores' and XVI 260 'orantis medias deseruere preces'; in both cases the objects are virtually equivalent to persons.

17. TAMEN. 'In spite of my dejection'.

17-18. DE CAELITE ... RECENTI ... CARMEN. The poem does not survive. At xiii 25-32 Ovid describes a similar poem on the apotheosis of Augustus, written in Getic.

17. RECENTI. 'New, freshly created'. Used in similar contexts at Met IV 434-35 'umbraeque recentes ... simulacraque functa sepulcris', VIII 488 'fraterni manes animaeque recentes', X 48-49 'Eurydicenque uocant: umbras erat illa recentes / inter', and especially XV 844-46 'Venus ... Caesaris eripuit membris nec in aera solui / passa recentem animam caelestibus intulit astris'.

18. VESTRA = 'of you [plural] at Rome'.

18. CARMEN IN ORA DEDI. 'I sent a poem for you to recite from and speak of'. Dare meaning 'send' is usually restricted to use with litteras (OLD do 10; compare Cic Att II i 12 & IX viiB 1, Livy XXVII 16 13).

For in ora, compare Catullus XL 5 'an ut peruenias in ora uulgi [sc hoc facis]?', Hor Ep I iii 9 '... Titius, Romana breui uenturus in ora', Prop III ix 32 (to Maecenas) 'et uenies tu quoque in ora uirum', Tr V vii 29-30 'non tamen ingratum est quodcumque obliuia nostri / impedit et profugi nomen in ora refert', and Livy II 36 3. The only instance I have found of the expression being used of a thing rather than a person other than this passage is also from Ovid: 'illud opus ... nunc incorrectum populi peruenit in ora, / in populi quicquam si tamen ore mei est' (Tr III xiv 21-24). Neither passage would have seemed strange to the Romans, given the close identification between poet and work: compare Ennius' famous 'uolito uiuo' per ora uirum' and Met XV 878 'ore legar populi'.

19. QVAE PIETAS. 'This demonstration of loyalty'.

20. SACRAE ... DOMVS. Augustus' house called 'magni ... Iouis ... domum' at Tr III i 38; compare as well EP III i 135 'domus Augusti, Capitoli more colenda'.

20. MITIOR IRA. Compare EP III iii 83 'pone metus igitur: mitescet Caesaris ira'.

21. LIQVIDO POSSVM IVRARE. 'I can swear unambiguously'. The only other instance of this sense in verse apparently III iii 49-50 'scis tamen et liquido iuratus dicere possis / non me legitimos sollicitasse toros'. From prose compare Cic II Verr IV 124 'confirmare hoc liquido, iudices, possum, ualuas magnificentiores ... nullas umquam ullo in templo fuisse', II Verr III 136, Fam XI 27 7 'alia sunt quae liquido negare soleam', and Sen Ben VII 9 5.

22. NON DVBIA ... NOTA. The phrase logically belongs with the preceding line: on the firm evidence of Brutus' past behaviour (described in 23-42), Ovid can confidently state that Brutus prays for his restoration. Non dubia by litotes for certa (for which see Her XX 207 'te ... nimium miror, nota certa furoris'); nota 'tangible sign, evidence' similarly used at Met I 761 (generis). FIDE (LTM^2ulF^2ul) is an obvious gloss for nota.

23. VERVM ... AMOREM. 'Sincere love' (Wheeler); compare Met V 61 'ueri non dissimulator amoris' and Tr IV iv 71 'et comes exemplum ueri Phoceus amoris'.

25. TVAS ... LACRIMAS NOSTRASQVE. The tears of Ovid's friends at his departure described at Tr III iv 39-40, EP I ix 17-18, and EP II xi 9-10 (to Rufus) 'grande uoco lacrimas meritum quibus ora rigabas, / cum mea concreto sicca dolore forent'.

26. PASSVROS POENAM CREDERET ESSE DVOS. Compare Tr V iv 37-38 (Ovid's letter speaking) 'quamuis attonitus, sensit tamen omnia, nec te / se minus aduersis indoluisse suis'.

27. LENEM TE MISERIS GENVIT NATURA. Compare Cic Tusc II 11 'te natura excelsum quendam uidelicet et altum et humana despicientem genuit' and Ennius Ann 112 Vahlen³ (of Romulus) 'qualem te patriae custodem di genuerunt'.

29. MARTE FORENSI. Similar metaphor for the lawcourts at Fast IV 188 'et fora Marte suo litigiosa uacent', Tr III xii 17-18 'ludis / cedunt uerbosi garrula bella fori' and Tr IV x 17-18 'frater ... fortia uerbosi natus ad arma fori'. According to Ovid real wounds were suffered in the forum at Tomis: 'adde quod iniustum rigido ius dicitur ense, / dantur et in medio uulnera saepe foro' (Tr V x 43-44).

30. POSSE TVO PERAGI VIX PVTET ORE REOS. Similar language at Tr I i 23-24 'protinus admonitus repetet mea crimina lector, / et peragar populi publicus ore reus'. Peragere refers to the prosecution of a defendant carried to its end, but does not imply success for the prosecutor: see Pliny Ep III ix 30 and Ulpian Dig XLVIII v 2 1 'non alias ad mulierem possit peruenire, nisi reum peregerit [sc adulterii]; peregisse autem non alias quis uidetur, nisi et condemnauerit'.

31. QVAMVIS PVGNARE VIDENTVR BMFH. Given the dependent pugnare, it seems hardly possible to read the VIDETVR given by the other manuscripts. The same problem arises at Met VIII 463-64 'pugnant materque sororque, / et diuersa trahunt unum duo nomina pectus', where the manuscripts divide between pugnant and pugnat; for an unambiguous parallel, see Her XIX 173 'nunc, male res iunctae, calor et reuerentia pugnant'.

Heinsius further suggested deleting est from the preceding scilicet eiusdem est 'cum tribus libris', but the change in number does not seem unduly harsh.

32. SVPPLICIBVS FACILEM. See on iv 30 faciles in tua uota, and compare Am II iii 5-6 (to his girl's eunuch) 'mollis in obsequium facilisque rogantibus esses, / si tuus in quauis praetepuisset amor' and Her XVI 197-98 'da modo te facilem, nec dedignare maritum ... Phrygem'.

Ovid is here indirectly referring to his own situation: compare EP III iii 107-8 'at tua supplicibus domus est adsueta iuuandis, / in quorum numero me precor esse uelis'.

33. LEGIS VINDICTA. 'The exacting of punishment on behalf of the law'. The law has been broken, and therefore demands retribution; Brutus acts on its behalf. For the sense of the genitive compare Val Max I 1 ext 3: (Dionysius of Syracuse committed many acts of sacrilege, but punishment was visited on him after his death in the form of his son's ignominious career) 'lento enim gradu ad uindictam sui diuina procedit ira tarditatemque supplicii grauitate pensat'.

33. LEGIS ... SEVERAE. Seuerae here serves as a standard epithet and has no such special force as at EP III iii 57-58 'uetiti ... lege seuera / credor adulterii composuisse notas'.

34. VERBA VELVT TAETRVM SINGVLA VIRVS HABENT. The same image at EP III iii 105-6 'ergo alii noceant miseris optentque timeri, / tinctaque mordaci spicula felle gerant'.

34. TAETRVM R. J. Tarrant TINCTV Ehwald TINCTVM codd. Tinctum is impossible: if the word were used, it would have to go with uerba. Compare Ibis 53-54 'liber iambus / tincta Lycambeo sanguine tela dabit', Ibis 491 '[tamque cadas domitus ...] quam qui dona tulit Nesseo tincta ueneno', EP III i 26 'tinctaque mortifera tabe sagitta madet', and EP III iii 106 'tinctaque mordaci spicula felle gerant'. Ehwald's tinctu is linguistically and palaeographically somewhat better than Merkel's tinguat: for similar corruptions compare Fast III 612 'flet tamen admonitu motus, Elissa, tui', where many manuscripts read admonitus, and Tr I iv 9 'pinea texta sonant pulsu [Rothmaler: pulsi codd], stridore rudentes'. Even so, 'Each of your words carries poison, as though it had been dipped in it' seems awkward. For Professor Tarrant's taetrum compare Lucretius I 936 'absinthia taetra', Dirae 23 'taetra uenena', and Hal 131 'nigrum ... uirus'.

34. VIRVS HABENT. Compare Tr IV i 84 'aut telo uirus habente perit' & III x 64 'nam uolucri ferro tinctile uirus inest'.

35-36. HOSTIBVS EVENIAT QUAM SIS VIOLENTVS IN ARMIS / SENTIRE. Hostibus eueniat is a common phrase in Ovid: compare Am II x 16-17 'hostibus eueniat uita seuera meis! / hostibus eueniat uiduo dormire cubili', Am III xi 16, AA III 247, Fast III 493-94 'at, puto, praeposita est fuscae mihi Candida paelex! / eueniat nostris hostibus ille dolor [recc quidam: color codd plerique]!', and Her XVI 219-20 (Paris to Helen) 'hostibus eueniant conuiuia talia nostris, / experior posito qualia saepe mero!'.

37. QVAE TIBI TAM TENVI CVRA LIMANTVR. 'Which are sharpened by you with such painstaking care'. For this meaning of limare compare Pliny NH VIII 71 'cornu ad saxa limato' and Cic Brut 236 '[M. Piso ...] habuit a natura genus quoddam acuminis, quod etiam arte limauerat'.

37-38. VT OMNES / ISTIVS INGENVI PECTORIS ESSE NEGENT. 'So that all would deny that they are the product of your kindly spirit'; for this sense of ingenuus compare Catullus LXVIII 37-38 'quod cum ita sit, nolim statuas nos mente maligna / id facere aut animo non satis ingenuo'. Ingenui pectoris is my correction for the manuscripts' INGENIVM CORPORIS, which could only mean 'so that all would deny that the talent of your body exists'; Ovid can hardly be identifying the tela of 36 with Brutus' ingenium. Wheeler translates 'On these [the missiles of your tongue] you use the file with such extreme care that none would recognize in them your real nature', and André 'que personne ne croirait qu'un tel esprit habite ton corps'; neither translation fits the Latin. Shackleton Bailey's INGENIVM NOMINIS still leaves unsolved the problem of ingenium.

The corruption of ingenui to ingenium (or rather, ingeniū) is simple enough; and the interchange of pectus and corpus is a common error.

42. NOTITIAM ... INFITIATA. Infitiari used similarly at EP I vii 27 'nec tuus est genitor nos infitiatus amicos'.

43. IMMEMOR ... IMMEMOR. Professor R. J. Tarrant points out the similar epanalepsis at Hor Ep I xi 9 'oblitusque meorum, obliuiscendus et illis'.

44. SOLLICITI BCM^2ul SOLLICITE M¹FHILT. The adjective with adverbial meaning would be especially liable to corruption. The same construction at Am II iv 25 'dulce canit flectitque facillima uocem'.

44. LEVASTIS Barberinus lat. 26, saec xiii LEVATIS BCMFHILT. If 44 were taken in isolation, leuatis, which most editors print, would be acceptable enough; compare Tr IV i 49 ' iure deas igitur ueneror mala nostra leuantes' and EP III vi 13-14 'nec scelus admittas si consoleris amicum, / mollibus et uerbis aspera fata leues'. But it is clear from 42 'est infitiata' and 49 'doluistis' that Ovid is speaking of the time of his banishment, and so leuastis must be read. Compare Tr I v 75 'me deus oppressit, nullo mala nostra leuante', EP II vii 61-62 'recta fides comitum poterat mala nostra leuare: / ditata est spoliis perfida turba meis', and EP III ii 25-26 'pars estis pauci melior, qui rebus in artis / ferre mihi nullam turpe putastis [uar putatis] opem'.

45-50. Compare the listing of adynata at the end of v (41-44), which again illustrates Ovid's eternal gratitude (to Sextus Pompeius). Here the personal detail (hic nimium nobis conterminus Hister) makes the adynaton reflect Ovid's own circumstances.

46. DE MARE. The same form of the ablative at Tr V ii 20 'pleno de mare'. Compare Ovid's frequent use of the metrically convenient ablative in -e of third-declension adjectives.

47-48. Thyestes' feast cited as a proverbial example at Met XV 62 (Pythagoras is urging a vegetarian diet) 'neue Thyesteis cumulemus uiscera mensis', Tr II 391-92 'si non Aeropen [Politianus: Meropen uel Europen codd] frater sceleratus amasset, / auersos Solis non legeremus equos', Lucan I 534-44, and Martial III xlv 1-2 'Fugerit an Phoebus mensas cenamque Thyestae / ignoro: fugimus nos, Ligurine, tuam'.

47. VTQVE ... SI = et, quasi. All of the instances of the idiom cited by Lewis & Short ut II A 2e and OLD ut 8d are from prose, except for Ter Eun 117 and Lucilius 330 Marx. In none of these passages is ut separated from si: the hyperbaton elevates the phrase and makes more natural its use in verse.

49. QVI ME DOLVISTIS ADEMPTVM. 'Who mourned my exile' is the meaning imposed by context, but the phrase would usually mean 'who mourned my death': compare EP I ix 41 'iure igitur lacrimas Celso libamus adempto', and the similar use of raptus for the exiled Ovid at xi 5 and xvi 1. For Ovid's considering his exile as his death, see xvi 1-4, Tr III iii 53 'cum patriam amisi, tunc me periisse putato', and EP I ix 56 'et nos extinctis adnumerare potest'.

VII. To Vestalis

Vestalis, a younger son of Cottius, monarch of a small kingdom in the Alps (see at 29 [p 253]), was primipilaris of the legion of the area (perhaps the V Macedonica). He had just been named administrator of the region around Tomis (see at 1); as an important local official, he was a natural choice as recipient of one of Ovid's letters.

The poem starts with a description of the harsh climate of Tomis, to which Vestalis along with Ovid can now testify, and of the savagery of the inhabitants (1-12). This serves as a bridge to a compliment to Vestalis on being named primipilaris (13-18), and to the main body of the poem, a long and rather conventional description of how Vestalis led the final attack in the recovery of Aegissos (19-52). In the concluding distich Ovid declares that he has rendered immortal the deeds of Vestalis.

1. ORAS (CI) seems more suited to the nature of Vestalis' command than VNDAS, the reading of the other manuscripts. After Euxinas, corruption from oras to undas would be very easy, the inverse less so. Ovid does not elsewhere use Euxinae orae, the usual substantives with Euxinus being aquae, mare, fretum, and, closest in meaning, litus, for which see iii 51 'litus ad Euxinum ... ibis', Tr V ii 63-64 'iussus ad Euxini deformia litora ueni / aequoris', and Tr V iv 1.

2. POSITIS ... SVB AXE in effect acts as a single adjective meaning 'northern'; axe plays a subordinate role and so does not require an epithet. The phrasing may be based on Accius 566-67 Ribbeck² '[ora ...] sub axe posita ad Stellas septem, unde horrifer / Aquilonis stridor gelidas molitur niues'. Lycaonio ... sub axe at Tr III ii 2.

3. ASPICIS EN PRAESENS. Compare ix 81-86, where Ovid invites Graecinus to ask his brother Flaccus, recently stationed in the Pontus, about conditions of life in the area.

3. IACEAMVS. 'Lie suffering': similarly used at EP I iii 49 'orbis in extremi iaceo desertus harenis', I vii 5, II ix 4 & III i 85 'ut minus infesta iaceam regione labora'.

4. FALSA ... QVERI. Perhaps a common phrase: Professor R. J. Tarrant cites Sallust Iug 1 'Falso queritur de natura sua genus humanum'.

5-6. ACCEDET ... FIDES. 'People will believe'. Compare Cic Diu I 5 'Cratippusque ... isdem rebus fidem tribuit, reliqua diuinationis genera reiecit' and Tac Germ 3 4 'ex ingenio quisque demat uel addat fidem' 'each can believe or disbelieve this according to his disposition'.

5-6. NON IRRITA ... FIDES = rata fides, a phrase meaning 'trustworthiness', rata having no special force. Compare Met III 341 'prima fide [genitive] ... ratae temptamina', Tr I v 49-50 'multa credibili tulimus ratamque, / quamuis acciderint, non habitura fidem', and Tr III x 35-36 'cum sint praemia falsi / nulla, ratam debet testis habere fidem'. Note the hyperbaton in all these passages.

6. ALPINIS IVVENIS REGIBVS ORTE. See at 29 progenies alti fortissima Donni (p 253). For the language, compare Hor Carm I i 1 'Maecenas atauis edite regibus'.

7. IPSE VIDES CERTE GLACIE CONCRESCERE PONTVM. At ix 85-86 Ovid tells Graecinus to ask his brother Flaccus 'mentiar, an coeat duratus frigore Pontus, / et teneat glacies iugera multa freti'.

Similar language at Tr III x 37-38 'uidimus ingentem glacie consistere pontum, / lubricaque [codd: lubrica cum fort scribendum] immotas testa premebat aquas'.

8. IPSE VIDES RIGIDO STANTIA VINA GELV. The same picture more explicitly given at Tr III x 23-24 'nudaque consistunt, formam seruantia testae, / uina, nec hausta meri, sed data frusta bibunt'.

9-10. IPSE VIDES ONERATA FEROX VT DVCAT IAZYX / PER MEDIAS HISTRI PLAVSTRA BVBVLCVS AQVAS. Similar descriptions at Tr III x 33-34 'perque nouos pontes, subterlabentibus undis, / ducunt Sarmatici barbara plaustra boues' and Tr III xii 29-30 'nec mare concrescit glacie, nec ut ante per Histrum / stridula Sauromates plaustra bubulcus agit'.

9. IAZYX. The Iazyges Sarmatae are mentioned by Pliny (NH IV 80) and by Strabo (VII 3 17), who describes them as one of several tribes living between the Borysthenes (Dnepr) and the Danube. They are also listed by Pompey, under the name of 'Iazyges Metanastae', the Wandering Iazyges (Geog III 7); the 'Iazyges' he describes as living along the shore of the Maeotis (III 5 19). Tacitus mentions the nation at Ann XII 29 4 (Vannius, king of the Suebi, is under attack) 'ipsi manus propria pedites, eques e Sarmaticis Iazygibus erat' and at Hist III 5 (the principes Sarmatarum Iazygum are enlisted to ensure the defence of Moesia in the absence of the regular troops; their offer to raise infantry as well as supplying their usual force of cavalry is rejected because of the fear of future treachery).

The name of the tribe was difficult metrically, so here Ovid calls them Iazyges, while at Tr III xii 30 (cited in the previous note) he calls them Sauromatae. At EP I ii 77 he solves the difficulty through hendiadys: 'quid Sauromatae faciant, quid Iazyges acres'.

11. ASPICIS. Ovid here uses verbs of seeing in an interesting way. At 7 and 9 he has uides; then aspicis suggests continuity but at the same time movement toward a new subject, and with a military detail introduced so as to introduce Vestalis' experience of war; then in 13-14 the emphasis is changed by the contrary-to-fact past optative utinam ... spectata fuisset.

11. ASPICIS ET MITTI SVB ADVNCO TOXICA FERRO. 'You behold how poison is hurled on the barbed steel' (Wheeler). The telum of 12 should be taken to be a spear, since mittere never seems to be used of arrows. At Ibis 135 the hasta is mentioned as the special weapon of the Iazyges.

11. ADVNCO. The spear had hooks. Compare Met VI 252-53 'quod [sc ferrum] simul eductum est, pars et pulmonis in hamis / eruta cumque anima cruor est effusus in auras', where Bömer cites among other passages Curtius IX 5 23 'corpore ... nudato animaduertunt hamos inesse telo nec aliter id sine pernicie corporis extrahi posse quam ut secando uulnus augerent' and Prop II xii 9 'et merito hamatis manus est armata sagittis'.

13-14. ATQVE VTINAM PARS HAEC TANTUM SPECTATA FVISSET, / NON ETIAM PROPRIO COGNITA MARTE TIBI. A similar opposition at Met III 247-48 (of Actaeon) 'uelletque uidere, / non etiam sentire canum fera facta suorum'.

15. TENDITVR Owen TENDITIS codd. The number of tenditis is inappropriate to the context. Owen's tenditur, independently conjectured two years later by Ehwald (KB 84), seems a somewhat more elegant solution to the problem than Merkel's TENDISTI. It puts the weight of the line on ad primum ... pilum rather than on Vestalis himself; the pentameter, with its emphasis on the honor, suggests that this is right.

15. PRIMVM PILVM. Compare Am III viii 27-28 'proque bono uersu primum deducite pilum! / nox [A. Y. Campbell: hoc uel hic codd] tibi, si belles [Madvig: uelles codd], possit, Homere, dari'. The primipilaris was the commander of the first century of the first cohort of the Roman legion, and hence first in rank among the legion's centurions.

17. PLENIS is the reading of all but two of the manuscripts collated. For this sense of plenus ('abundant'), compare Am I viii 56 'plena uenit canis de grege praeda lupis', Nux 91-92 'illa [the tree that is not near a road] suo quaecumque tulit dare dona colono / et plenos fructus adnumerare potest', Hor Sat I i 57, and Cic Sex Rosc 6 'alienam pecuniam tam plenam atque praeclaram'. Ehwald read PLENVS (F^acI), joining ingens with uirtus in the following line, arguing that the honour would not seem a great one to a member of a royal family. But Ovid devoted four lines to describing Vestalis' new rank: he must have believed that Vestalis would consider it a very great honour indeed. As well, if ingens is connected with titulus, uirtus ... maior gains point.

17. PLENIS ... FRVCTIBVS. For the wealth of the primipilaris, see Am III viii 9-10 'ecce recens diues parto per uulnera censu / praefertur nobis sanguine pastus eques'. In that poem the newly-rich primipilaris, Ovid's rival in love, is given a character very different from that of Vestalis.

17. INGENS is used at ix 65 of another office, the consulship.

18. IPSA TAMEN VIRTVS ORDINE MAIOR ERIT. A similar sentiment at EP II ix 11-14 (to king Cotys) 'regia, crede mihi, res est succurrere lapsis ... fortunam decet hoc istam ['this befits your position'], quae maxima cum sit, / esse potest animo uix tamen aequa tuo'.

19. NON NEGAT HOC HISTER. For the device of calling to witness the scenes of military exploits compare Catullus LXIV 357 'testis erit magnis uirtutibus unda Scamandri' and the passages there cited by Fordyce. For non negat Professor A. Dalzell cites Catullus IV 6-7 'negat ... negare'.

20. PVNICEAM GETICO SANGVINE FECIT AQVAM. Similar language at ix 79-80 (of Flaccus) 'hic raptam Troesmin celeri uirtute recepit, / infecitque fero sanguine Danuuium'.

21. AEGISSOS. The city, the modern Tulcea, is situated about 110 kilometres directly north of Tomis (Constanţa) on the southernmost branch of the Danube, 60 kilometres from the mouth of the river. At EP I viii 11-20 Ovid describes the recapture of the city from the Getes; evidently the city had been lost once again.

Aegissos is the spelling certified by three of the five sources cited by Mommsen (CIL III page 1009), namely Hierocles Synecdemus 637 14, Notitia dignitatum 99, and Procopius Aed IV 7 20. The Itinerarium Antoninianum (226 2) offers Aegiso (ablative); Ehwald (KB 41), citing Mommsen, took this as sufficient justification for retaining the single s of the Ex Ponto manuscripts, although the now lost Strasbourg manuscript had egissus at I viii 13 (and an indication of an alternative ending in -os). The Ravenna Cosmography (4 5), Mommsen's final source, reads Aegypsum.

27. TE SVBEVNTE RECEPTA. 'Recaptured on your attack'. Intransitive subire in this sense belongs to military vocabulary: compare Caesar BG VII 85 'alii tela coniciunt, alii testudine facta subeunt' and Curtius IV 2 23. For instances from military prose of subire with a direct object see Caesar BG II 27 'subire iniquissimum locum', Hirtius BG VIII 15, Bell Alex 76 2 'subierant iniquum locum', and Bell Hisp 24 2.

22. INGENIO ... LOCI. 'The nature (i.e. difficulty) of its terrain'. The same standard phrase at Tac Ann VI 41 'locorumque ingenio', Hist I 51 'diu infructuosam et asperam militiam tolerauerant ingenio loci caelique ['climate']', and from Ovid Tr V x 17-18 'tumulus defenditur ipse / moenibus exiguis ingenioque loci' and EP II i 52 '[oppida ...] nec satis ingenio tuta fuisse loci'.

22. NIL OPIS. The expression is rather prosaic: compare Cic Fam IV i 1 'aliquid opis rei publicae tulissemus'.

23. DVBIVM BMFHIT DVBIVM EST CL. The same variant in many manuscripts at EP III i 17-18 (Ovid is addressing Tomis) 'nec tibi sunt fontes laticis nisi paene marini, / qui potus dubium sistat alatne sitim'.

24. NVBIBVS AEQVA. 'As high as the clouds'. For this use of aequus compare Aen IX 674 'abietibus iuuenes patriis in [Heyne: et codd; cf Il XII 132 'ἕστασαν ὡς ὅτε τε δρύες οὔρεσιν ὑψικάρηνοι'] montibus aequos', Statius Ach I 173 'aequus uertice matri', Sen Ep 94 61 'aequum arcibus aggerem ... et muros in miram altitudinem eductos', and Aen IV 89 'aequataque machina caelo'.

25. SITHONIO = Thracio.

25. INTERCEPERAT. Intercipere 'capture' common in Livy (IX 43 3, XXI 1 5, XXVI 51 12, XXXVI 31 10); compare Ammianus XX 7 17 & XX 10 3 'locis ... recuperatis quae olim barbari intercepta retinebant ut propria'.

26. EREPTAS VICTOR HABEBAT OPES. Similar phrasing at Fast III 49-51 'hoc ubi cognouit contemptor Amulius aequi / (nam raptas fratri uictor habebat opes), / amne iubet mergi geminos'.

27. FLVMINEA ... VNDA. Flumineus does not occur elsewhere in the Tristia or Ex Ponto; fluminea ... aqua at Fast II 46 & 596.

27. VITELLIVS. This Vitellius is presumably one of the four sons of Publius Vitellius, grandfather of the emperor. Suetonius wrote of the sons, Aulus, Quintus, Publius, and Lucius, that they were 'quattuor filios amplissimae dignitatis cognomines ac tantum praenominibus distinctos' (Vit 2 2). Heinsius suggested Aulus (cos AD 32) was the one here meant, 'nisi ad L. Vitellium patrem [sc principis] referre mauis'. 'On the general and reasonable assumption', wrote Syme (HO 90), 'this is P. Vitellius'. But Suetonius calls P. Vitellius 'Germanici comes', and he is heard of in 15 assisting Germanicus in a campaign (Tac Ann I 70 1): it is perhaps more likely that Publius would have been with Germanicus at the time of the capture of Aegissos, and that another of the brothers is meant. Certainty is in any case not attainable.

29. PROGENIES ALTI FORTISSIMA DONNI. For the phrasing, compare EP II ix 1-2 'Regia progenies, cui nobilitatis origo / nomen in Eumolpi peruenit usque ['goes back to'], Coty'.

The Donnus here referred to is Vestalis' grandfather (CIL V 7817), or possibly a more distant ancestor. Vestalis' father, Cottius, became a client of Augustus; at XV 10 7 Ammianus mentions the worship still accorded Cottius 'quod iusto moderamine rexerat suos, et ascitus in societatem rei Romanae quietem genti praestitit sempiternam'. At Nero 18 Suetonius mentions as one of the few additions to the empire under Nero the 'regnum ... Alpium defuncto Cottio'. This Cottius would probably have been Vestalis' older brother; André is therefore right to infer that Vestalis 'n'était pas l'héritier du trône, ce qu'Ovide n'aurait pas manqué de signaler'.

30. IMPETVS. Impetus + infinitive usually indicates a mad impulse: the only other exception in Ovid is Met V 287-88 (one of the Muses speaking) 'impetus ire fuit; claudit sua tecta Pyreneus / uimque parat, quam nos sumptis effugimus alis'.

31. CONSPICVVS LONGE FVLGENTIBVS ARMIS. Modelled on Aen XI 769 'insignis longe Phrygiis fulgebat in armis'.

32. FORTIA NE POSSINT FACTA LATERE CAVES. Vestalis would in any case have fought bravely; so that his deeds would not pass unnoticed, he led the attack.

33. INGENTIQVE GRADV. When Ovid elsewhere use ingens gradus (passus) he gives the phrase a humorous tone: see Am III i 11 'uenit et ingenti uiolenta Tragoedia passu', AA III 303-4 'illa uelut coniunx Vmbri rubicunda mariti / ambulat ingentes uarica fertque gradus', and Met XIII 776-77 (of Polyphemus) 'gradiens ingenti litora passu / degrauat'. The straightforwardness of this passage is of a piece with the rest of the poem.

For an example of the normal epic use of this detail, see Aen X 572 'longe gradientem'.

33. FERRVM LOCVMQVE reflects 23 'dubium positu melius defensa manune'.

34. SAXAQVE ... GRANDINE PLVRA. The same phrase in the same metrical position at Ibis 467-68 'aut te deuoueat certis Abdera diebus, / saxaque deuotum grandine plura petant'.

35. MISSA SVPER IACVLORVM TVRBA. 'The crowding missiles hurled from above' (Wheeler).

38. FERE. Heinsius' FERO would involve the repetition of fero in 44; and fero uulnere would be rather feeble when applied to a shield.

Professor R. J. Tarrant points out to me that Ovid's description of Vestalis' exploit may have served as a distant model for Lucan's account of how a centurion named Scaeua rallied Caesar's forces and led an attack against Pompey's encampment (VI 140-262). Scaeua was made primipilaris in reward for his bravery (Caesar BC III 53 5).

40. SED MINOR EST ACRI LAVDIS AMORE DOLOR. Similar language of a similar exploit at Met XI 525-28 'ut miles, numero praestantior omni, / cum saepe adsiluit defensae moenibus urbis, / spe potitur tandem laudisque accensus amore / inter mille uiros murum tamen occupat unus'. Ovid's description of Vestalis' exploit is little more than a string of conventional phrases.

40. ACRI. 'Sharp'. Compare ii 36 'immensum gloria calcar habet'.

41-42. TALIS APVD TROIAM DANAIS PRO NAVIBVS AIAX / DICITVR HECTOREAS SVSTINVISSE FACES. Compare Met XIII 7-8 (Ajax speaking of Ulysses) 'at non Hectoreis dubitauit cedere flammis, / quas ego sustinui, quas hac a classe fugaui' and Met XIII 384-85 (the death of Ajax) 'Hectora qui solus, qui ferrum ignesque Iouemque / sustinuit totiens, unam non sustinet iram'. All three passages are drawn from Il XV 674-746, the description of how Ajax repulsed Hector's attempt to set the Greek ships afire, and in particular from 730-31 'ἔνθ' ἄρ' ὅ γ' ἕστήκει δεδοκημένος, ἔγχεϊ δ' αἰεὶ / Τρῶας ἄμυνε νεῶν, ὅς τις φέροι ἀκάματον πῦρ'.

41. PRO NAVIBVS. 'In front of the ships'; a reminiscence of Il XV 746 (the final line of the book) 'δώδεκα δὲ προπάροιθε νεῶν αὐτοσχεδὸν οὖτα'.

43. DEXTERA DEXTRAE. Ovid used syncope in dextera where metrically convenient. Elsewhere when he employs the two forms he is usually describing the joining of hands in pledge or friendship. See Her II 31 'commissaque dextera dextrae', Her XII 90 'dextrae dextera iuncta meae', and Met VI 447-48 'dextera dextrae / iungitur'. For a different use, see Met III 640-41 'dextera [uar dextra] Naxos erat: dextra mihi lintea danti / "quid facis, o demens? quis te furor," inquit "Acoete?"'.

45-46. DICERE DIFFICILE EST QVID MARS TVVS EGERIT ILLIC, / QVOTQVE NECI DEDERIS QVOSQVE QVIBVSQVE MODIS. As Professor E. Fantham points out to me, this praeteritio takes the place of a full aristeia detailing Vestalis' exploits.

46. QVOSQVE QVIBVSQVE MODIS. Compare quotque quibusque modis in an erotic context at Am II viii 28, and Tr III xii 33-34 'sedulus occurram nautae, dictaque salute, / quid ueniat quaeram quisue quibusue locis'.

47. ENSE TVO FACTOS CALCABAS VICTOR ACERVOS. Compare Met V 88 (of Perseus) 'extructos morientum calcat aceruos'.

50. MVLTAQVE FERT MILES VVLNERA, MVLTA FACIT. A similar conjunction of verbs at Fast II 233-34 'non moriuntur inulti, / uulneraque alterna dantque feruntque manu'.

52. IBAT. IBIT (BP) is printed by all modern editors except André, and is possibly correct: compare Am II iv 31-32 'ut taceam de me, qui causa tangor ab omni, / illic Hippolytum pone, Priapus erit' for the future tense used of a mythological character, and EP II xi 21-22 'acer et ad palmae per se cursurus honores, / si tamen horteris, fortius ibit [uar ibat] equus' for the corruption of future to imperfect.

53. TEMPVS IN OMNE. Similar promises of immortality at Tr I vi 36 (to his wife) 'carminibus uiues tempus in omne meis', EP II vi 33-34 (to Graecinus) 'crede mihi, nostrum si non mortale futurum est / carmen, in ore frequens posteritatis eris', and EP III i 93 (to his wife) 'nota tua est probitas testataque tempus in omne'.

Vestalis is known to us only through this poem.

VIII. To Suillius

This poem, nominally addressed to Suillius, husband of Ovid's stepdaughter, is in fact directed to Germanicus, of whose staff Suillius was a member (see at 23 [pp 264-65]).

Ovid begins the poem by expressing his pleasure at receiving, at last, a letter from Suillius, saying he hopes that Suillius does not feel ashamed of being related to him by marriage (1-20). He then asks him to address Germanicus on his behalf (21-26). In 27-30 he says how grateful he will be if Germanicus assists him; at 31 he begins to address Germanicus directly in a tripartite defence of poetry. The first part (31-42) builds on 34 'Naso suis opibus, carmine, gratus erit': Ovid is now poor, but can still offer Germanicus his poetry. The second section (43-66) builds on 43-44 'nec tamen officio uatum per carmina facto / principibus res est aptior ulla uiris', and explains how verse brings immortality to great men and their deeds. The third section (67-78) offers culminating evidence for the value of poetry: Germanicus is himself a poet. Ovid moves from this to a final plea that Germanicus help his fellow-poet: once removed from Tomis, he will praise him in verse (79-88). In the final distich of the poem, he asks Suillius to assist his prayer.

The structure of the poem is similar to that of Tr V ii. In that poem Ovid addresses his wife for the first thirty-eight lines, telling her of his misery and asking her to approach Augustus on his behalf. In the six lines that follow, he asks himself what he will do if she fails him; he answers that he will make his own direct approach to Augustus. The final thirty-four lines are his prayer to Augustus, in which he describes the hardships he endures at Tomis and begs for a mitigation of his punishment. It is remarkable that in both poems direct addresses to members of the imperial family should be disguised in this way: it seems probable that Tr II, Ovid's long defence of his conduct, had been received by Augustus with hostility, and that he was thenceforth more circumspect.

1-2. SERA QVIDEM ... GRATA TAMEN. Tamen goes with grata, balancing quidem. For instances of the separate serus tamen idiom ('it is late in happening, but it does in fact happen') see Nisbet and Hubbard at Hor Carm I xv 19.

1. SERA QVIDEM. It seems that in spite of his being a close relative of Ovid, Suillius, like Sextus Pompeius (see the introduction to i), had been reluctant to be openly associated with him.

1. STVDIIS EXCVLTE. 'Refined'. Studiis adds little to the force of exculte: the same idiom at Quintilian XII ii 1 'mores ante omnia oratori studiis erunt excolendi' and Cic Tusc I 4 'ergo in Graecia musici floruerunt, discebantque id omnes, nec qui nesciebat satis excultus doctrina putabatur'.

1. SVILLI. P. Suillius Rufus (PW IV A,l 719-22; PIR¹ S 700) is otherwise chiefly known to us from three passages of Tacitus: Suillius is presented as 'strong, savage, and unbridled' (Syme Tacitus 332). At Ann IV 31, Tacitus describes how, in 24, Tiberius insisted that Suillius, convicted of accepting a bribe, be relegated to an island rather than merely be exiled from Italy; what seemed cruelty at the time later seemed wisdom in view of his later behaviour as a favourite of Claudius. At Ann XI 1-7 Tacitus describes how Suillius' excesses resulted in a proposal in the Senate to revive the lex Cincia of 204 BC, by which advocates had been forbidden remuneration: the proposal was modified by Claudius at the instance of Suillius and others affected so as to establish a maximum fee of ten thousand sesterces. At Ann XIII 42-43 (AD 58) Tacitus tells how Suillius, 'imperitante Claudio terribilis ac uenalis', was charged with extortion as proconsul of Asia and with laying malicious charges under Claudius. Banished to the Balearic islands, he led a luxurious existence, remaining unrepentant.

3-4. PIA SI POSSIT SVPEROS LENIRE ROGANDO / GRATIA. Compare 21 'si quid agi sperabis posse precando'.

5-6. ANIMI SVM FACTVS AMICI / DEBITOR. 'Your friendly purpose has placed me in your debt' (Wheeler). The genitive similarly used for the cause of indebtedness at i 2 'debitor est uitae qui tibi, Sexte, suae' and Tr I v 10 'perpetuusque animae debitor huius ero'.

6. MERITVM VELLE IVVARE VOCO. 'I call the desire to help a favour already given'. Otto uelle 2 cites EP III iv 79 'ut desint uires, tamen est laudanda uoluntas', Prop II x 5-6 'quod si deficient uires, audacia certe / laus erit: in magnis et uoluisse sat est', Pan Mess 3-7, Laus Pisonis 214; the same proverb at Sen Ben V 2 2 'uoluntas ipsa rectum petens laudanda est'.

7. IMPETVS ISTE TVVS LONGVM MODO DVRET IN AEVVM. Similar phrasing at EP II vi 35-36 (Graecinus has been rendering Ovid assistance) 'fac modo permaneas lasso, Graecine, fidelis, / duret et in longas impetus iste moras'.

9. IVS ALIQVOD. 'A certain claim on each other'. The same phrase for a similar situation at EP I vii 60 (to Messalinus, elder brother of Cotta Maximus) 'ius aliquod tecum fratris amicus habet'.

9. ADFINIA. The adfinis was a relative by marriage, commonly, as here, a son-in-law; a relative by common descent was a cognatus.

9. ADFINIA VINCVLA. Vinculum used of family relationships at Met IX 550 (Byblis wishes to marry her brother) 'expetit ... uinclo tecum propiore ligari' and Cic Planc 27 'cum illo maximis uinclis et propinquitatis et adfinitatis coniunctus'.

10. INLABEFACTA. The word elsewhere in Latin only at xii 29-30 'haec ... concordia ... uenit ad albentes inlabefacta comas'. Ovid is fond of using negative participles of this type.

11-12. NAM TIBI QVAE CONIVNX, EADEM MIHI FILIA PAENE EST, / ET QVAE TE GENERVM, ME VOCAT ILLA VIRVM. The same type of circumlocution at Her III 45-48 (Briseis to Achilles) "diruta Marte tuo Lyrnesia moenia uidi; ... uidi ... tres cecidisse quibus [Bentley: tribus codd] quae mihi, mater erat'.

11. EADEM MIHI FILIA PAENE EST. This is presumably Perilla, the recipient of Tr III vii, whom Ovid there speaks of in terms appropriate to a stepfather.

13-14. EI MIHI, SI LECTIS VVLTVM TV VERSIBVS ISTIS / DVCIS, ET ADFINEM TE PVDET ESSE MEVM. A similar lament at EP II ii 5-6 'ei mihi, si lecto uultus tibi nomine non est / qui fuit, et dubitas cetera perlegere!'; both passages are followed by defences of Ovid's character.

For uultum ... ducis see at i 5 trahis uultus (p 149).

15. NIHIL BCMFHLT NIL I. Copyists were more prone to alter nil to nihil than the inverse; but in 1919 Housman demonstrated that nihil was Ovid's invariable form for the latter half of the first foot by pointing out that in all of the twenty-odd passages where the manuscripts offer nihil or nil at that position the following word invariably begins with a vowel (Collected Papers 1000-1003). There would be no reason for such an avoidance of consonants if Ovid had allowed nil in this position; he must therefore have used nihil alone.

16. FORTVNAM, QVAE MIHI CAECA FVIT. The image of Fortune being blind to a single individual seems very strange. Professor R. J. Tarrant suggests that caeca could mean 'unforeseeing', and by fortunam Ovid could be referring to his own previous circumstances; alternatively, caeca might be a corruption induced by the familiar image of the blind goddess, replacing an original SAEVA (Riese) or LAEVA, for which compare Silius III 93-94 'si promissum uertat Fortuna fauorem, / laeuaque sit coeptis'.

17-18. SEV GENVS EXCVTIAS, EQVITES AB ORIGINE PRIMA / VSQVE PER INNVMEROS INVENIEMVR AVOS. A similar claim at Tr IV x 7-8 'usque a proauis uetus ordinis heres, / non modo fortunae munere factus eques'. The status of eques was not hereditary except in the case of a senator's son. The Paeligni did not receive the citizenship until after the Social War; to be born to equestrian status, and to assume that he could have had a senatorial career (Tr IV x 35), Ovid must have belonged to one of the dominant families of the region.

17. EXCVTIAS. 'Examine'. Ovid plays on the primary meaning of the word, 'shake out', at Am I viii 45-46 'has quoque quae frontis rugas in uertice portant [Burman: quas ... portas codd] / excute; de rugis crimina multa cadent'. The transferred meaning had lost any sense of metaphor by Ovid's time, however; see especially Tr II 224 'excutiasque oculis otia nostra ['the product of my leisure hours'—Wheeler] tuis'.

19-20. SIVE VELIS QVI SINT MORES INQVIRERE NOSTRI, / ERROREM MISERO DETRAHE, LABE CARENT. A similar claim of no fault beyond his error at EP II ii 15-16 'est mea culpa grauis, sed quae me perdere solum / ausa sit, et nullum maius adorta nefas'.

20. ERROREM ... DETRAHE. At Met II 38-39 the same phrase with a different meaning: (Phaethon to his father) 'pignora da, genitor, per quae tu uera propago / credar, et hunc animis errorem ['doubt'] detrahe nostris*.

20. LABE CARENT. The same sense of labes at Tr I ix 43 'uitae labe carentis' and Prop IV xi 41-42 'neque ulla labe mea nostros erubuisse focos'; compare as well the phrase sine labe at Tr II 110 (domus), Tr IV viii 33 (decem lustris ... peractis), EP I ii 143 (praeteriti anni), EP II vii 49 (uita prior), Her XVII 14 (tenor uitae), and Her XVII 69 (fama).

22. QVOS COLIS ... DEOS. A similar definition of the imperial family at EP II ii 123 'quos colis ad superos haec fer mandata sacerdos'.

23. DI TIBI SVNT CAESAR IVVENIS. BCFM^2ul read SINT; but the indicative seems to be required by the preceding 'quos colis ... deos' and the following 'tua numina placa' and 'hac certe nulla est notior ara tibi'.

23. CAESAR IVVENIS. Germanicus; he would have acquired the cognomen Caesar on his adoption by Tiberius in AD 4. Iuuenis probably refers to Germanicus' title of princeps iuuentutis, which EP II v 41-42 indicates he must have held: 'te iuuenum princeps, cui dat Germania nomen, / participem studii Caesar habere solet'. Germanicus' holding of the title is not elsewhere attested.

At Ann IV 31 5, Tacitus identifies Suillius as 'quaestorem quondam Germanici'; at Ann XIII 42 4, he represents Suillius as saying of himself and Seneca 'se quaestorem Germanici, illum domus eius adulterum fuisse'. His service under Germanicus was clearly a principal fact of his life.

25-26. ANTISTITIS ... PRECES. Here antistes is virtually equivalent to cultor, as at Tr III xiv 1 'Cultor et antistes doctorum sancte uirorum'; compare as well Met XIII 632-33 'Anius, quo ... antistite Phoebus / rite colebatur'.

27-28. QVAMLIBET EXIGVA SI NOS EA IVVERIT AVRA, / OBRVTA DE MEDIIS CVMBA RESVRGET AQVIS. Ovid here mixes two nautical metaphors: if a ship is overwhelmed by high seas, a favouring breeze will not be of great assistance.

28. OBRVTA DE MEDIIS CVMBA RESVRGET AQVIS. Similar wording at [Sen] Oct 345-48 '[cumba ...] obruta ... ruit in pelagus rursumque salo / pressa resurgit'.

29. TVNC EGO TVRA FERAM RAPIDIS SOLLEMNIA FLAMMIS. Perhaps a verbal reminiscence of Aen IX 625-26 'Iuppiter omnipotens, audacibus adnue coeptis. / ipse tibi ad tua templa feram sollemnia dona'.

29. TVRA ... SOLLEMNIA. The phrase does not occur elsewhere in Ovid; but compare the passage from Aen IX quoted above, as well as the conjunction of words at Tr III xiii 16 'micaque sollemni turis in igne sonet'.

29. RAPIDIS is here used as a standard epithet; its full force ('destructive') at Met II 122-23 'tum pater ora sui sacro medicamine nati / contigit et rapidae fecit patientia flammae', Met XII 274-75 'correpti rapida, ueluti seges arida, flamma / arserunt crines', and EP III iii 60 (to Amor) 'sic numquam rapido lampades igne uacent'.

31-32. NEC TIBI DE PARIO STATVAM, GERMANICE, TEMPLVM / MARMORE. Professor R. J. Tarrant points out to me the reference to Virgil G III 13-16 'et uiridi in campo templum de marmore ponam ... in medio mihi Caesar erit templumque tenebit'; Parii lapides are mentioned at III 34. Here Ovid makes the temple literal, and conducts his recusatio in the terms used by love-poets.

32. CARPSIT OPES ... MEAS. 'Has destroyed my wealth'. This is not strictly true, since Ovid at v 38 says that Pompeius give him gifts (Ovid's letter speaking) 'ne proprias attenuaret opes'.

The same use of carpere at ix 121-22 'fortuna est impar animo, talique libenter / exiguas carpo munere pauper opes' and Am I viii 91 'et soror et mater, nutrix quoque carpat amantem'.

34. NASO SVIS OPIBVS, CARMINE, GRATVS ERIT. Compare Am II xvii 27 'sunt mihi pro magno felicia carmina censu' and Am I iii entire.