¹ A section of this paper was read to the Laurence Seminar on the Fasti held at the Cambridge Classics Faculty in May 1990. I am grateful to Professor M. D. Reeve for the invitation to attend and to the participants for their interest and suggestions. It is a special pleasure to acknowledge the criticisms offered by A. Barchiesi, I. Du Quesnay, E. Fantham, P. Hardie, C. Newlands, S. Oakley and N. Wright. None of these scholars should be assumed to be endorsing this final version in any way. References are to Fasti 2 unless the context obviously suggests otherwise. I cite the Fasti from the Teubner edition of Alton, Wormell and Courtney (Leipzig, 1978), and use the editor's name only in citing the commentaries of Bömer, F. (Heidelberg, 1957–1958)Google Scholar, Frazer, J. G. (London, 1929)Google Scholar and Gierig, G. E. (Leipzig, 1812)Google Scholar. The following books and articles are usually referred to by the author's name: Fantham, E., ‘Sexual Comedy in Ovid's Fasti: Sources and Motivation’, HSCP 87 (1983), 185–216Google Scholar; Fantham, E., ‘Die Schlacht am Cremera in Ovids Fasten 2, 195–242’, RhM 123 (1980), 152–62Google Scholar; Littlewood, R. J., ‘Ovid's Lupercalia (Fasti 2.267–452): a Study in the Artistry of the Fasti’, Latomus 34 (1975), 1060–72Google Scholar; Ogilvie, R. M., A Commentary on Livy, Books 1–5 (Oxford, 1965)Google Scholar; Richard, J.-Cl., ‘Historiographie et Histoire: L'Expédition des Fabii à la Crémere’, Latomus 47 (1988), 526–53Google Scholar = Richard¹; Richard, J.-Cl., ‘Ovide et le Dies Cremerensis’, RPh 62 (1988), 217–25 = Richard²Google Scholar; Syme, R., History in Ovid (Oxford, 1978)Google Scholar = Syme HO; Syme, R., The Augustan Aristocracy (Oxford, 1986)Google Scholar = Syme AA; Wissowa, G., Religion und Kultus der Römer ² (Munich, 1912).Google Scholar

² The indispensable study remains Syme HO, but ch. VIII ‘Paullus Fabius Maximus’ can now be supplemented by ch. XXVIII of AA. Although these chapters range widely over Ovidian references to the Fabii, it is odd to find Syme ignoring entirely Ovid's celebration of the family's most famous heroic action and of their most distinguished religious privileges, both centrally located in Fasti 2.

³ The literary typology fulfilled the emperor's own wishes. There has been a proposal in 27 to have Octavian entitled Romulus ‘quasi et ipsum conditorem urbis’ (Suet. Aug. 7.2, cf. 95 for the conveniently supportive omen), and Dio (53.16.7) attributes the move to the keen wish (έπεθύμει ίσχυρς) of Octavian himself. ‘Augustus' was adopted instead in case the parallels with Romulus’ kingship and fratricide were pressed too far, a scruple which failed to inhibit contemporary poets (see Binder, G., Aeneas und Augustus: Interpretationen zum 8 Buch der Aeneis [Meisenheim, 1971], pp. 150ffGoogle Scholar. and now Cairns, F., Virgil's Augustan Epic [Cambridge, 1989], p. 97CrossRefGoogle Scholar with bibliography in note 41). When Ovid later contrasted the ‘achievements’ of Romulus and Augustus (Fast. 2.133–44), we should remember that he was exposing the ironies implicit in a comparison once endorsed by the emperor himself.

⁴ Suet. Aug. 31.4, Res G. 10.4. Holleman, A. W. J. (‘Ovid and the Lupercalia’, Historia 22 [1973], 260–8Google Scholar) develops an ‘anti-Augustan’ thesis out of interpreting Ovidian burlesque as an attack on Augustus' ritual restoration. Fantham (in her notes 27 and 79) is right to criticize this approach.

⁵ They were established as two colleges for the purpose. For the story cf. Fast. 2.365–80 (a concise version) and Wissowa 559. The ancient evidence is assembled (with much else) by Frazer (ii.328–41) and more briefly by Bömer on Fast. 2.375.

⁶ The short-lived addition of a third college, the Iulii, in the mid-1st century (Suet. Iul. 76.1) may be connected with a crisis in the fortunes of the Fabii at that time which I discuss below. Certainly, the two original colleges survived to be recorded by Ps.-Aurelius OGR 22.1 (‘utrumque nomen etiamnunc in sacris manet’).

⁷ Lack of a persuasive explanation of this connection is a weakness in the recent studies of Lefèvre and Richard². See pp. 220–1 of the latter.

⁸ An instance is the intervening fable of the Raven, Bowl and Snake at 243–6. Inaccurate though its astronomical position may be, it is not entirely irrelevant here, as I shall explain below. Ovid was as little concerned with precision in such matters as Lucan was, and for the same reason: our confidence in the poet's range and skill is enhanced by the inclusion of narratologically apposite myths from the wider cosmic framework.

⁹ See most recently Richard¹, 531 (‘Il nous semble plus important que le recit d'Ovide doive beaucoup à celui de Tite-Live’) with a summary of the Livian parallels familiar since Niebuhr and Mommsen. Richard provides a thorough analysis of the different versions in the historians, but admits that Ovid's version ‘fasse figure de corps etranger dans l'ensemble de nos sources’ (542).

¹⁰ This brief summary gives preference for the moment to Livy's version in the case of disputed details. I shall presently discuss the discrepancies in Ovid. A full list of testimonia is given in Richard² 217 n. 1.

¹¹ So Livy explicitly at 6.1.11, though a summer date has also been assumed for the Book 2 narrative from 2.52.3 and also Tac. Hist. 2.91, Plut. Camillas 19.1–3 and the Fasti Antiates CIL i². 248, 322. On the parallel with the Allia see Richard² 217–18, 223–4.

¹² So Bömer (p. 96 on 195), whose Fabian ‘Familientradition’ would juxtapose the Cremera and the Lupercalia. This view is derived from Mommsen's later thesis (Römische Forschungen ii [Berlin, 1879], 255) and is now discredited if Richard¹ is right (541–2) in holding that one tradition of the Fabian gens was in fact precisely the opposite – that the Cremera and the Allia were synchronous and that the association between the two was transmitted to the family archives from the annalist Q. Fabius Pictor. See further Richard² 220–1.

¹³ Mommsen's earlier thesis, which Frazer (ii.323) follows. For fuller background and references to earlier literature see Lèfevre 153–4, Richard¹ 542–3 and Richard² 219–20. Lefèvre himself argues for an exilic dating of the Cremera narrative (pp. 156–60) and thinks Ovid placed it in February simply to put it next to the Lupercalia. But if there were no basis for such a juxtaposition the point would lose much of its effect. See Richard² 222 for a critical assessment. The solution offered in Richard² 223–5 develops the connection between the Cremera and the Gallic attack of 390, but in a way which is unnecessarily speculative in view of what I believe can be made of the material in Fasti 2.

¹⁴ See Ogilvie 359, echoing Dionysius' judgement μύθοις γρ δή τατά γε κα πλάσμασιν έοικε θεατρικοῖς (AR 9.22.3). Richard is also judiciously sober here, unlike Heurgon, J., The Rise of Rome to 264 B.C. (London, 1973), 181.Google Scholar

¹⁵ The two battles are already linked in Gellius (17.21.12–13). See Ogilvie 359–60 and Richard¹ 533–4.

¹⁶ 3rd ed. (Leipzig, 1889), pt II, p. 28 on 195. See Richard² 219, n. 8.

¹⁷ See pp. 360–1. Ogilvie's insight is noted briefly by Richard¹ 543 and Richard² 220–1 but its significance is missed, so that Richard ends up having to argue for a prior tradition by a less convincing route.

¹⁸ The discrepancies seem of a minor kind, as one would expect in a skilful stitching together of different sources, but Ogilvie's explanation of them seems quite convincing. The decisive evidence occurs in chapters 51–2. In 51.1 T. Menenius, consul for 477, is ‘confestim missus’ against the Etruscans who ambushed the Fabii. This suggests (says Ogilvie) that there was no consular army already prepared in the field, but that a ‘scratch force’ had to be contrived on an ad hoc basis. And the Etruscan advance into Latium after the Cremera ‘precipitates a corn shortage’ (51.2) ‘because the Romans were prevented from harvesting their crops… [which] would have been harvested well before 18 July’. Livy is therefore here following a tradition which located the Cremera in the winter (the consular year of office began on 1 August), presumably on 13 February. At 52.3, however, Menenius already has his stativa in the field at the time of the Cremera, presumably now on the summer date of 18 July, and he remains consul until 31 July. Ogilvie therefore divides the two narrative sources at 51.4, the beginning of the next consular year.

¹⁹ An idea ‘hazarded’ by Ogilvie (p. 361) with good supporting evidence. Cf. Richard¹ 541–2. We are elsewhere (Richard² 218, n. 5) promised a paper on Licinius and the Cremera, which I have not seen.

²⁰ ‘Causation and the Authority of the Poet in Ovid's Fasti’, CQ 38 (1989), 164–85.Google Scholar

²¹ The basic details are in Sofer, E., Livius als Quelle von Ovids Fasten (Vienna, 1905–1906)Google Scholar, and see most recently Richard¹ 531.

²² The familiar parallels are respectively: (reading exercitus with Bentley), In the first of these parallels Ovid and Livy are making the same point, and the objections of Elter, A., Cremera und Porta Carmentalis (Bonn, 1910)Google Scholar and the latest Teubner editors (p. xx) are unfounded.

²³ They are quoted by Frazer, ii.322 n. 2.

²⁴ For quisquis used in quasi-religious admonitions in verse see Ibis 98–9 with the parallels (especially Tibullus 2.1.1 and 2.2.2) given in La Penna's note ad loc, but the usage is sometimes extended in even more solemn addresses to the gods themselves (Pease on Aen. 4.577). Ovidian practice and formulaic religious language can be followed up in Bömer's note on Fast. 6.731.

²⁵ The dilemma posed by the contrasts in this couplet was succinctly expressed by Gierig (p. 73): ‘Parum mihi placet; neque tamen abesse potest, ne manca sit oratio.’ In fact, the apparent oddness of 204 is explained by the line's reference to the fama of the fateful arch of the Carmentine Gate, contradicting its title ‘Porta Scelerata’ (Florus 1.6.2). Festus groups together several examples of the transference of this epithet in Roman place-names, ‘Sceleratus Campus’ and ‘Sceleratus Vicus’ along with the ‘Scelerata Porta’ (495–7, pp. 449–51 Lindsay).

²⁶ For repeated phrases of this kind cf. ‘una dies’ in 235–6 below and the far from exhaustive list given by Bömer on 202–4.

²⁷ A catalogue of such acts of editorial rashness in dealing with lines 196–204, as inconclusive as they were inappropriate, is given in Frazer's critical note on 203 (i.66–7). When we recognize that these lines exploit stylistic incongruity and are not merely a bland versification of Livy, the critical problems disappear.

²⁸ Cf. OLD s.v. lb and Ars am. 1.388 (ventus), DRN 1.17 (fluvios), 3.650 (falces).

²⁹ Cf. OLD s.v. 2a and Tib. 1.3.65, Hor. Carm. 2.18.30. The theme of destruction in the swollen rivers of winter will return in connection with the Tiber (389–90) and will be discussed in a later section.

³⁰ Hymn 2.108–9.

³¹ The tension in the Cremera passage between the ‘epic’ subject and the elegiac medium is discussed briefly by Barchiesi, A., ‘Voci e istanze narrative nelle Metamorfosi di Ovidio’, Materiali e Discussioni per l'Analisi dei Testi Classici 23 (1989), 55–97, p. 62.CrossRefGoogle Scholar

³² The fuller Greek tradition represented to us by Dionysius has a succession of battles spread over an even longer period (Richard¹ 527–30), while Ovid's compression skilfully heightens the tension.

³³ Ovid suppresses reference to the larger force in order to concentrate attention on the central role of the Fabii. That the 306 were not alone is clear from Livy 2.49.5, Diod, 11.53.6, Gellius 17.21.13 and Servius on Aen. 6.845, while the number of their clientes is given as 4000 in Dion. 9.15.3 and 5000 in Festus 496 (p. 450 Lindsay). (For discussion see Richard¹ 545ff.). Ovid cannot be trying to conceal anything. As we shall see, he has his reasons for making this an exclusive venture of the Fabii.

³⁴ Lèfevre supplies the most recent evidence for this attitude, believing that Ovid's version is a barely concealed eulogy of the Fabii designed to impress Paullus Fabius Maximus from exile.

³⁵ This view has some textual support; cf. Livy 2.50.3 in which the trap is a ‘consilium… insidiis ferocem hostem captandi’ with Fast. 2.213–14 and especially 226–7 ‘fraude perit virtus’. But it is only one side of the coin.

³⁶ ‘Quo fessum rapitis, Fabii?’ Like Ovid, Virgil also is about to introduce the character of Fabius Cunctator. I shall develop this parallel later.

³⁷ Livy (2.50.11) does not have to specify Fabius' policy here, but ‘saepe… maximum futurum auxilium’ suggests a higher level of success than their ancestors at the Cremera enjoyed.

³⁸ There is a fuller list from Roman verse in TLL 1.2A 168.45ff.

³⁹ The most important feature of the parallel is the destruction of the flocks. Virgil intensifies the already sharp impact of a simile from the Homeric Doloneia (Il. 10.485–6; cf. also Sarpedon at 12.299–301 and the less dramatically forceful version at Od. 6.130–4).

⁴⁰ Bömer's list of parallels is here undiscriminating; the essential point had already been made (with other pertinent remarks) by Gierig (p. 74 n. 219).

⁴¹ For the significance of the relationship between Mezentius and his son in both life and death see Heinze, R., Virgils epische Technik ⁴ (Stuttgart, 1957), pp. 213–15.Google Scholar

⁴² See McKeown on Am. 1.15.24 (ii.408–9) for references and parallels, to which Ex P. 1.2.4 (referring to Fast. 2.235–6) and Prop. 3.11.70 (the destruction of Antony's forces at Actium) should be added.

⁴³ See Ogilvie 120. The suggestion is ignored in O. Skutsch's discussion of the possible parallels between Ennius and the Livy passage (The Annals of Q. Ennius [Oxford, 1985], pp. 279–80Google Scholar), though it is clear from Ann. fr. 258 Sk (‘multa dies in bello conficit unus’) that Ennius used the motif elsewhere. Gierig (p. 74, note on 235) thought that Livy's account of the destruction of Alba was very likely Ovid's source for the motif in Fast. 2.235–6, though it is equally possible that if Ennius was the source he was used directly. Skutsch's point (p. 279, note 9) ‘[Livy] probably knew [the Annales] by heart from his school days’ would of course apply to Ovid too.

⁴⁴ πργμα ού μόνον πίθανον λλ κα δύνατον (AR 9.22.1). His supporting arguments seem irrefutable (ibid. 2–6), but there is of course no trace of such scepticism in Livy (2.50.11), unless ‘satis convenit’ may suggest a reluctance to endorse a received opinion which the author reports.

⁴⁵ Cf. the ironic eulogy of the astrologers in Fast. 1.295–310: ‘credibile est illos pariter vitiisque locisque altius humanis exseruisse caput.’ The expression ‘credibile est’ is noticeably Ovidian (see McKeown on Am. 1.11.11–12).

⁴⁶ For the parallel see Lefèvre 157, though my reading differs from his.

⁴⁷ Kl. Pauly s.v. ‘Fabius’ I, 33. Syme HO 136, AA 16. 104, 419.

⁴⁸ Kl. Pauly s.v. 39, Syme HO 136, AA 16, 419.

⁴⁹ For this danger see Syme HO 136 ‘The Fabii were saved from extinction by the device of adoption from families in alliance or at social parity’, AA 419 and Kl. Pauly ii.489.22ff. ‘Die Maximi starben in 2 Jh. aus, wurden aber durch Adoptionen aus den Familien der Aemilii und Servilii fortgeführt.’

⁵⁰ HO 135. For AUobrogicus see Kl. Pauly s.v. ‘Fabius’ I, 34.

⁵¹ AA 419–20.

⁵² For this procedure see most recently Powell, J. G. F., Cicero, De Senectute (Cambridge, 1988), p. 151.Google Scholar

⁵³ Kl. Pauly s.v. 32; he had been aedile in 57 and was to die during his consulship (Syme HO 135, AA 420). The genealogical table of the Fabii in Pauly-Wissowa is reproduced by Syme as table XXVII at the end of AA.

⁵⁴ Nep. Att. 18.3 (where ‘pari modo’ extends this description to all the lives listed). One wonders what line Atticus took on the date of the Cremera.

⁵⁵ For Africanus see Syme HO 136 (‘The record is sparse indeed’), and for Fabia AA 417. It is not known whether Africanus married (ibid.).

⁵⁶ HO 144; cf. AA 403, but there is no basis for the speculation that Fabius was already a widower when he married Marcia. Hor. Carm. 4.1 tells against the idea.

⁵⁷ Cf. Fast. 6.801–10 and Ex P. 1.2.137–40. Sources on Marcia's family background are collected by Bömer on 6.801 (the story is neatly summarized by Syme AA 403–4), but the Cypriot inscription recorded in Syme HO 137 should be added to Bömer's list.

⁵⁸ Cf. Syme AA 413.

⁵⁹ Syme HO 154.

⁶⁰ Ben. 4.30.2, quoted by Syme ibid, with other facts which I repeat in this paragraph; cf. AA 417. Adding to the irony is the apparent extinction of the Marcii with the generation of Paullus' wife (HO 145).

⁶¹ The evidence for the date is fully analysed by Peeters, F., Les Fastes d'Ovide: histoire du texte (Brussels, 1939), pp. 19–31Google Scholar. Syme agrees, HO 146.

⁶² Servius (ad loc.) paraphrases Anchises' question differently: ‘cur me, o Fabii, fessum ad vestram trahitis narrationem?’ Neither Norden nor Austin follows up this approach, in which Servius interprets the narratio as the long list of the 306 Fabii killed at the Cremera. The reference seems too particularized for the context, but it is tempting to think that Ovid may have seen the same possibility in Anchises' question when he echoed it in Fasti 2, and that it may have helped to cement the association in his mind between the Cremera and the Cunctator.

⁶³ He had already used Virgil's introductory unus to begin his previous pentameter (240).

⁶⁴ Evidently so from its occurrences in Cicero and Livy; see Otto Sprichwörter, s.v. cunctari (p. 101) and Skutsch ad loc. (pp. 529–31).

⁶⁵ For the wider influence of Ennius on Anchises' speech see Hardie, P. R., Virgil's Aeneid: Cosmos and Imperium (Oxford, 1986), pp. 77–8Google Scholar with further examples collected at the foot of p. 77.

⁶⁶ It has often been noted that a motif from Ennius (Ep. 18V) which Virgil takes up in the proem to the third Georgic (see R. Thomas on 3.9) makes its mark along with other influences on the epilogue to the Metamorphoses (see Bomer on 15.871ff.).

⁶⁷ To the identical list of mythographical sources in Frazer (ii.326 n. 7) and Bömer (on 243) add the version in the scholia on Aratus Phaen. 449 (p. 282 Martin).

⁶⁸ Ovid omits the Crab and the Lion, which Vitruvius (9.51) links to the three used in the fable. The correct risings (apparent and real) of these constellations are recorded by Frazer (ii. 327), but Ovid was as little concerned with precision in such matters. The intended effect is to make the poet's narrative strategy appear to receive support from an independent calendrical/celestial sequence, and astronomical accuracy is subordinated to literary advantage. It so happens that the next astronomical passage, the movement of the sun from Aquarius to Pisces (453–74), is accurate enough in its dating (Frazer, iii.390).

⁶⁹ Note, for example, the chiasmus in 250 and the contrasting word order in 264, and the double alliteration in 262.

⁷⁰ Met. 2.531–632, Call. Hec. fr. 74 Hollis, 15–20. On that first occasion, of course, the raven had tried telling Apollo the truth, when it detected Coronis committing adultery. For a wide familiarity with this second punishment as characteristic of ravens see the passages collected in Aristotle fr. 343 Rose.

⁷¹ For an Ovidian treatment of the motif from Hor. Carm. 3.30.1 see Met. 15.871ff. and Tr. 5.14.1ff.

⁷² Art. cit. (n. 20), 166–7.

⁷³ For the way in which these traditions were conflated see F. Spaltenstein's note on Sil. Pun. 6.627, though his view that the conflation goes back no further than Silius should be treated with scepticism, despite the lack of literary evidence.

⁷⁴ See Bömer on 4.10 for some Ovidian parallels, and A. S. Hollis on Ars 1.39–40 for their Greek ancestry and further references. Their literary significance in the light of Callimachus' Aelia fr. 1 is discussed by Wimmel, W., Kallimachos in Rom (Wiesbaden, 1960), pp. 105–11.Google Scholar

⁷⁵ See most recently J. N. Bremmer (n. 88), p. 33. I have not seen Duval, Y. M., ‘La victoire de Rémus à la course des Lupercales chez Ovide’, Caesarodunum 71 (1972), 201–19.Google Scholar

⁷⁶ Cf. Fast. 3.707, 4.69, 6.765. A few non-Ovidian instances are listed by Littlewood (1975), 1061 n. 6, with some interesting observations on the tone of 269–76.

⁷⁷ Bömer on 273; see Fedeli on Prop. 3.15.11–12 (p. 478).

⁷⁸ See Bömer on Fast. 1.471, Eden on Aen. 8, p. III and Wissowa 208–10.

⁷⁹ Cf. the many parallels with Tib. 1.3.35–48, whose sources are discussed by K. F. Smith loc. (pp. 244–51) and by Ball, R. J., Tibullus the Elegist (Göttingen, 1983), pp. 54–6Google Scholar. In Tibullus and most similar instances description of the Saturnian Age introduces a σχετλιασμός against contemporary society (cf. Tib. 1.3.49 ‘nunc love sub domino…’), but uniquely at Fast. 2.301–2 ‘nunc quoque…’ introduces the continuation of the rites of the Saturnian Age in the present 289).

⁸⁰ This contrast in the connotation of opes recurs early in the third book. At 3.50 opes refers to the glittering incentive which induced Amulius to kill Numitor, while at 56 the more restricted reference is suggested by the oxymoron in the second half of the line ‘nee taceam vestras, Faustule pauper, opes’.

⁸¹ The problem of reconciling a σχετλιασμός against the corrupting influence of opes with asserted continuity in primitive ritualistic observance at Rome arises in a slightly different form in Evander's meeting with Aeneas and the περιήγησις in Aen. 8, but the parallels with Evander's Saturnian Age (8.314–32) are sufficiently strong to confirm the association. For the σχετλιασμός cf. 8.327, 348, 364; for continuity see the Lupercal (342–4), Argiletum (345–6) and the presence of Jupiter (353), not to mention the wider setting of the festival of Hercules (see Eden on 102 for its contemporary relevance).

⁸² Cf. Servius auct. on Aen. 8.343 (ii.250 T–H), Wissowa 210 with references collected in note 3, and the discussion in Ogilvie 51–2.

⁸³ A similar effect is achieved with the woods at 2.165f. and 6.41 Iff.; cf. also Met. 3.175ff. and the standard example at Am. 3.1.1. At Fast. 2.449 the religious significance of the lucus is emphasized by deriving Juno's title Lucina from it, a unique departure from standard opinion (Varro, Cicero, Plutarch; see Frazer ad loc. for references) which associated Lucina with lux (cf. also [Tib.] 3.4.13), an opinion which Ovid follows without question elsewhere in this poem (3.255, 6.39–40 and even 2.450).

⁸⁴ If Ovid derived the story of the repulse of the rustlers (369–72) from a source which gave the etymology of Lupercal as lupus/arcere as Servius did (on Aen. 8.343), he suppressed the etymology to be able to accommodate without apparent contradiction the story of the wolf and the twins which he placed immediately after that of the rustlers. For the etymology see Wissowa 559 (with note 1), Frazer ii.337 and Fowler, W. Warde, The Religious Experience of the Roman People (London, 1911), pp. 478–9Google Scholar, and for its possible connection with this episode see Bömer on 2.359.

⁸⁵ Ovid's emphasis can be judged from the space allotted to the different stages of the story. The version in this book devotes 36 of its 40 lines to the threat to the life of the twins and their rescue by the wolf, while the version in the third book (11–58) devotes only 6 of its 48 lines to the same episode.

⁸⁶ Romulean etymologies are naturally in evidence here; cf. 412 with Bömer's note and Ogilvie on Livy 1.4.5 for an explanation of the fiction.

⁸⁷ The exclamatory quid facis? is regularly used by Ovid as a reprimand. So Triptolemus' mother rebukes Ceres as she pulls her son from the fire in which the goddess has put him (4.556), Ariadne warns Bacchus against taking another lover (3.496), and Ovid himself tries to warn Juno against punishing the innocent Callisto (2.178). For further examples see Met. 3.641, 13.225, Heroid. 5.115, 8.7, Am. 2.5.29, 3.2.70, 71, Ars am. 3.735, Ex P. 4.3.29.

⁸⁸ Recent landmarks in discussing the wider issues raised by the story of the twins are Binder, G., Die Aussetzung des Konigskindes: Kyros und Romulus (Meisenheim, 1964)Google Scholar, Cornell, T. J., ‘Aeneas and the Twins; the Development of the Roman Foundation Legend’, PCPS 21 (1975), 1–32Google Scholar, and Bremmer, J. N., ‘Romulus, Remus and the Foundation of Rome’Google Scholar, in Bremmer, J. N. and Horsfall, N. M., Roman Myth and Mythography, BICS Suppl. 52 (London, 1987), pp. 25–48Google Scholar. The last two items record an extensive bibliography.

⁸⁹ Mirum (413) and quis credat? (414) have the effect of distancing the reader from the subjectmatter, just as the weeping of the servants (388) and the screaming of the infants (405) suggest a warm human intimacy with it.

⁹⁰ For the background and a survey of the versions see Frazer ii.393–6. No version is like Ovid's in every respect. Ps.-Eratosthenes Catast. 38 (p. 43 Olivieri) offers a similar story about Derceto, while Hyginus Astron. 2.30 (72–3 Le Bœuffle) and Manilius 4.579–81 (800–1 are too brief to establish a context) both involve Venus in a temporary metamorphosis. As the evidence is so slight, it is impossible to gauge the extent of Ovid's independence.

⁹¹ ‘The intentions of deities are swiftly accomplished’, McKeown on Am. 1.6.13 ‘nee mora, venit amor’. Bömer's list (on 471) of other instances of nec mora in the Fasti is undiscriminating. The one interesting parallel case of its phrase being followed by a sudden break at the third foot caesura is at 4.843 ‘nee mora, transiluit’, where the abruptness of the expression suggests the sudden recklessness with which Remus (lacking the security conferred by divinity) leaps to his destruction.

⁹² The difference between the earlier rescues and that of Venus and Cupid by the fish is all the clearer when we consider the parallel between the latter and Arion's rescue by the dolphin on 3 February (79–118). At first, Ovid may seem to be following Herodotus (1.24) in allowing the gods no interest in Arion's rescue until the story is complete: ‘di pia facta vident’ (117) introduces Jupiter at the last moment, to arrange the dolphin's translation to the heavens. Despite this, it would be misguided to see Arion as effecting his escape merely through his own mortal skill, for uniquely among narrators of this story Ovid presents the bard as virtually transforming himself into a second Apollo. This transformation is already foreshadowed at 91–2 and is completed in 105–10, and it is as the garlanded Phoebus (cf. 106) that Arion descends to the water to be rescued by the dolphin, which is sacred to the god. Like Venus', Arion's rescue has all the characteristics of a magic escape, with a suitable vehicle divinely conjured out of the water and subsequently rewarded with a future among the stars.

⁹³ See the full discussion of the background in Fantham 196–201. Littlewood makes the interesting point that Ovid's description of the grotto fulfils exactly Vitruvius' specifications for the satyric stage-set (1064, referring to De Arch. 5.6.9).

⁹⁴ Other ancient testimonia for this ancestry are collected by Bömer ad loc. See also Syme HO 147 n. 3.

⁹⁵ The locus amoenus is here the setting for a festival of Bacchus (329–30), whose cultic presence throughout the passage has been analysed by Turcan, R., ‘A propos d'Ovide Fasti II 313–330: conditions préliminaires d'une initiation dionysiaque’, REL 37 (1959), 195–203.Google Scholar

⁹⁶ See Frazer on 2.329 (ii.364).

⁹⁷ Commentators compare ‘improbe amor, quid non mortalia pectora cogis?’ (Aen. 4.412).

⁹⁸ Cf. Aen. 9.236 and p. 156 above for an earlier allusion in the Cremera section to the Euryalus episode and to a simile in Aeneid 2.

⁹⁹ For brief references to this and other Virgilian parallels see Littlewood 1066–7 and Fantham 194–5.

¹⁰⁰ Cf. pp. 157 above.

¹⁰¹ For the double entendre see Adams, J. N. on ‘weapons’, The Latin Sexual Vocabulary (London, 1982), pp. 19–20Google Scholar. It is tempting to see similarly hilarious possibilities in applying ‘visc…angue’ from Ovid's version of the Aeneid 2 simile to Faunus' discovery of Hercules in bed, but Adams (30–1) can find ‘no certain example’ of the phallic significance of the snake in Latin.