¹ For a brief history of the myth, see C. Sourvinou-Inwood, OCD ⁴, s.v. ‘Iphigenia’ (at 743).

² See Solmsen, F., ‘The sacrifice of Agamemnon's daughter in Hesiod's Ehoeae’, AJPh 102 (1981), 353–8Google Scholar.

³ Euripides’ Iphigone (El. 1023) is a slight variation.

⁴ On the names of Agamemnon's daughters and their mutability during the Archaic period, see Burgess, J.S., The Tradition of the Trojan War in Homer and the Epic Cycle (Baltimore, 2001), 150–1Google Scholar.

⁵ Schol. D Il. 9.145, Λαοδίκη. μία τῶν Ἀγαμέμνονος θυγατρῶν, ἣν οἱ τραγικοὶ Ἠλέκτραν εἶπον· ὡς καὶ τὴν Ἰφιάνασσαν, Ἰφιγένειάν φησιν ὁ Εὐριπίδης; cf. Ael. VH 4.26.

⁶ Four, if one counts Euripides’ Iphigone (see n. 3 above).

⁷ ἢ Ὁμήρῳ ἀκολουθεῖ εἰρηκότι <τὰς> τρεῖς θυγατέρας τοῦ Ἀγαμέμνονος (Il. 9.145) ἤ, ὡς τὰ Κύπρια, τέσσαράς φησιν, Ἰφιγένειαν καὶ Ἰφιάνασσαν.

⁸ The A scholiast on Il. 9.145 says that he was unaware of it: ὅτι οὐκ οἶδε τὴν παρὰ τοῖς νεωτέροις σφαγὴν Ἰφιγενείας.

⁹ See Griffin, J., ‘The epic cycle and the uniqueness of Homer’, JHS 97 (1977), 39–53CrossRefGoogle Scholar, at 44, noting the reticence of the Homeric epics about human sacrifice and intrafamilial killing. Kullmann detects an allusion to the sacrifice of Iphigenia at Il. 1.106: Kullmann, W., ‘Die Töchter Agamemnons in der Ilias’, Gymnasium 72 (1965), 200–3Google Scholar; cf. Taplin, O., Homeric Soundings (Oxford, 1992), 86Google Scholar.

¹⁰ I am unconvinced by Davidson's suggestion that in line 157 the verb ζώει goes only with Χρυσόθεμις, and that Ἰφιάνασσα refers to the sacrificed daughter; see Davidson, J.F., ‘The daughters of Agamemnon (Soph. El. 153–63)’, RhM 133 (1990), 407–9Google Scholar. Sophocles apparently followed the Cypria in recognizing four daughters; see Finglass, P.J., Sophocles: Electra (Cambridge, 2007), 151Google Scholar.

¹¹ Naevius: fr. 19 Ribbeck = fr. 16 TrRF. Ennius: frr. 177–203 Ribbeck = frr. 213–45 Vahlen = frr. 185–207 Jocelyn = frr. 82–8 TrRF.

¹² See also the probable allusion to Ennius’ play at Rhet. Her. 3.34.

¹³ Lloyd-Jones, H., ‘Artemis and Iphigeneia’, JHS 103 (1983), 87–102CrossRefGoogle Scholar, at 95.

¹⁴ See Taylor, B., ‘Rationalism and the theatre in Lucretius’, CQ 66 (2016), 140–54CrossRefGoogle Scholar, at 150–4.

¹⁵ If it was based on the stories summarized in Hyg. Fab. 120–1.

¹⁶ See Perutelli, A., ‘Ifigenia in Lucrezio’, SCO 46 (1996), 193–207Google Scholar; Taylor (n. 14), 147.

¹⁷ Taylor (n. 14), 147–8.

¹⁸ Taylor (n. 14), 149. The Iphigenia of Ennius is another possible, even likely, source, given the prominence of Ennius elsewhere in the prologue, but verbal parallels are lacking; see Harrison, S.J., ‘Ennius and the prologue to Lucretius DRN 1 (1.1–148)’, Leeds International Classical Studies 1.4 (2002), 1–13Google Scholar, at 4–6.

¹⁹ See Harrison (n. 18), 6–8, following the structural analysis of Jacoby, F., ‘Das Prooemium des Lucretius’, Hermes 56 (1921), 1–65Google Scholar.

²⁰ Cf. Merrill, W.A., T. Lucreti Cari De Rerum Natura Libri Sex (New York, 1907), 282Google Scholar: ‘L. alone of the Latin poets seems to use the form Iphianassa; it is difficult to see why, as metrical reasons did not force him.’

²¹ Brown, P.M., Lucretius: De Rerum Natura 1 (London, 1984), 60Google Scholar: ‘the name is treated as a variant for Iphigeneia, perhaps on account of their similar meaning (“powerful queen”, “born powerful”): as the two are metrically equivalent, L.’s preference is presumably on grounds of euphony.’ Note especially the smooth transition from the ‘i’ of the second syllable to the ‘a’ of the third syllable in Iphianassai as compared to the interpolation of the hard consonant ‘g’ in Iphigeniai; the double ‘s’ preceding the genitive ending also has a softer effect than the nasal ‘n’.

²² As noted by the anonymous referee.

²³ See Leonard, W.E. and Smith, S.B., T. Lucreti Cari De Rerum Natura Libri Sex (Madison, 1942), 167, 210Google Scholar: ‘The choice of the Homeric form of the name indicates Lucretius’ wish to elevate the style and suggest connection with the epic tradition’; Ernout, A. and Robin, L., Lucrèce, De Rerum Natura: Commentaire (Paris, 1962 ²), 1.34Google Scholar. Palmer regards the forms Iphianassai, Danaum and Hymenaeo as ‘glossae’, a category of rare or strange words that Aristotle considered suitable to the epic genre: see Palmer, L.R., The Latin Language (London, 1954), 99, 107Google Scholar.

²⁴ See Sourvinou-Inwood (n. 1). Interpolations have obscured Euripides’ original ending to IA, but it is likely that Lucretius was familiar with a version that included the substitution of a deer for Iphigenia; see Taylor (n. 14), 148–9. For Euripides’ use of the different traditions about Iphigenia's sacrifice, see Hulton, A.O., ‘Euripides and the Iphigenia legend’, Mnemosyne 15 (1962), 364–8CrossRefGoogle Scholar; for his essentially cultic conception of Iphigenia, see Parker, L.P.E., Euripides: Iphigenia in Tauris (Oxford, 2016), xixGoogle Scholar.

²⁵ I am indebted to the anonymous referee for this suggestion and quotation.

²⁶ See Bernays, J., Commentarius in Lucreti librum I, in H. Usener (ed.), Gesammelte Abhandlungen von Jacob Bernays, 2 vols. (Berlin, 1885), 2.1–67Google Scholar, at 8: ‘eruditionem affectando in errorem lapsus esse videtur Lucretius’ (arguing that Lucretius followed ancient scholars, like the D scholiast on Il. 9.145, who, in the view of Bernays, mistakenly identified Iphianassa with Iphigenia); likewise, Duff, J.D., T. Lucreti Cari De Rerum Natura Liber Primus (Cambridge, 1923), 52Google Scholar; Bailey, C., Titi Lucreti Cari De Rerum Natura Libri Sex, 3 vols. (Oxford, 1947), 2.614Google Scholar (‘The confusion was natural’); cf. Brown (n. 21), 60: ‘[Lucretius] has either overlooked or confused Homer, Iliad 9.145 and Sophocles, Electra 158, where Iphianassa is a sister who survived Iphigeneia.’

²⁷ Perutelli (n. 16), passim.

²⁸ Works in which Iphigenia dies at Aulis without being rescued (or at least without the rescue being mentioned) include, in addition to Aeschylus’ Agamemnon, Pindar's Eleventh Pythian Ode, the Electra of Sophocles (where she is unnamed) and the Electra of Euripides.

²⁹ DRN 1.94 is a reminiscence of Eur. IA 1220 (cf. Eur. IT 209); Taylor (n. 14), 148 suggests that 1.100 recalls IA 1575. For Lucretius’ insistence on the impiety of the sacrifice (1.83, impia facta), cf. IA 1105 (ἀνόσια), 1318 (σφαγαῖσιν ἀνοσίοισιν ἀνοσίου πατρός).

³⁰ Taylor (n. 14), 149.

³¹ Their roles (like those of other ductores Danaum such as Calchas and Menelaus) were so well known that naming was unnecessary. But Lucretius’ lack of specificity may also be interpreted as a way of diminishing the male actors and keeping the focus where it belongs, on the suffering of Iphianassa. (I thank the referee for this suggestion.)

³² Aeschylus makes no mention of the marital ploy in Agamemnon, but perhaps hints at it through the matrimonial metaphor of the ‘first rites’ of the ships (Ag. 227) and the saffron colour of Iphigenia's robes (Ag. 239); see Kim On Chong-Gossard, J.H., Gender and Communication in Euripides’ Plays (Leiden, 2008)CrossRefGoogle Scholar, 229 n. 31.

³³ For Lucretius’ simultaneous appropriation and rejection of heroic epic, see Gale, M., Myth and Poetry in Lucretius (Cambridge, 1994), 99–128Google Scholar. There is a large bibliography on Lucretius’ relation to the three authors mentioned above. As starting points, in addition to Gale (this note), see Aicher, P.J., ‘Lucretian revisions of Homer’, CJ 87 (1992), 139–58Google Scholar; Merrill, W.A., ‘Parallelisms and coincidences in Lucretius and Ennius’, University of California Publications in Classical Philology 3 (1918), 249–64Google Scholar; Harrison (n. 18), passim; and for Empedocles, Sedley, D.N., Lucretius and the Transformation of Greek Wisdom (Cambridge, 1998), 1–34CrossRefGoogle Scholar.

³⁴ A conventional trope that Lucretius leaves unexamined here, where he is initiating the reader into the Epicurean world-view. The coming revelation that nothing, including our world, is eternal except atoms and void lends irony, retrospectively, to the above praise of Ennius and to Lucretius’ later reference to poets’ ‘eternal monuments of fame’ (5.329, aeternis famae monimentis; see pp. 7–8 below).

³⁵ Aicher (n. 33), 139. Roy suggests that the opening lines of Book 2 constitute a metapoetic statement, to the effect that, while it is pleasant to read the Odyssey and the Iliad, there is nothing comparable to the joy of Epicurean enlightenment; see SRoy, ., ‘Homeric concerns: a metapoetic reading of Lucretius, De Rerum Natura 2.1–19’, CQ 63 (2013), 780–4CrossRefGoogle Scholar.

³⁶ The pairing of the Trojan and Theban Wars recalls Hes. Op. 161–5, where it comes within Hesiod's description of the heroic race—the fourth in the succession of the five ages. Lucretius’ implicit rejection of the Golden, Silver and Bronze Ages corrects Hesiod's view of history and reinforces his own insistence upon the recent origin of the world.

³⁷ The echo of line 325, semperque aeterna fuere, with its fallacious hypothesis that the world has existed forever, contaminates the use of the adjective in 329.

³⁸ I owe this formulation to the referee.

³⁹ For good discussion of this passage, see Gale (n. 33), 96, 109–10.

⁴⁰ For the latter point, see Sedley (n. 33), 51, though he seems to miss the ironic tone.

⁴¹ In Homer, Helen is called the daughter of Zeus, not of Tyndareus (Il. 3.199, 3.418, 3.426; Od. 4.184, 4.219, 4.227, 23.218); the ‘daughter of Tyndareus’ is Clytemnestra (Od. 24.199). I have wondered if this is another instance of pointed nomenclature, different in kind and less drastic than the use of Iphianassa but similar in its effect of restoring historical accuracy to Homer's account of people and events. But the fact that Tyndareus raised Helen made it natural to think of him as her human parent, and post-Homeric Greek authors frequently describe her as the daughter of Tyndareus without contradicting Zeus's biological fatherhood. Euripides, for example, often employs the same patronymic Tyndaris (Andr. 898; Hec. 269; Hel. 472, 614, 1179, 1546; IA 61, 1335, 1417; Tro. 34), but adheres to the myth of Zeus's parentage even after raising doubts. See Allan, W., Euripides: Helen (Cambridge, 2008), 148Google Scholar on Hel. 17–21; Blondell, R., Helen of Troy (Oxford, 2013), 185CrossRefGoogle Scholar (on Tro. 766–73); for another assertion of Zeus's parentage, see Gorg. Hel. 3. In Rome, Accius refers to her as the offspring of Tyndareus (fr. 609 Ribbeck, Tyndareo gnata), and Virgil followed Lucretius in his use of the name Tyndaris (Aen. 2.569, 2.601). I conclude that Tyndaris was a traditional appellation that Lucretius borrowed from Euripides or from the diction of Latin epic or tragedy—from Ennius even—mainly in order to heighten the style and with little or no corrective intent.

⁴² On Homer's usage, see Kirk, G.S., The Iliad: A Commentary. Volume 1: Books 1–4 (Cambridge, 1985), 266–7CrossRefGoogle Scholar. For the name Paris in Latin prior to and contemporary with Lucretius, see Enn. fr. 38 Ribbeck = fr. 53 Vahlen = fr. 64 Jocelyn = fr. 16 TrRF, quapropter Parim pastores nunc Alexandrum uocant; Accius fr. 561 Ribbeck (the text of Pari dyspari may be wrong, but the adjective inpar seems to pun on Paris); tr. inc. fr. 80 Ribbeck ( = Pacuvius fr. 215 Warmington); Cic. Att. 1.18.3, Nat. D. 3.91. The Ennian fragment, from his Alexander, which was probably based on Euripides’ play of the same name, shows that the name of Alexander was also current; cf. Plaut. Bacch. 947; Cic. Fat. 34.

⁴³ Though the direct inspiration for this and the image of parturition was probably Enn. frr. 60–1 Ribbeck = frr. 76–7 Vahlen = frr. 72–3 Jocelyn = fr. 22 TrRF, nam maximo saltu superauit grauidus armatis equus | qui suo partu ardua perdat Pergama.