¹ See Austin, R. G., P. Vergili Maronis Aeneidos Liber Sextus (Oxford, 1977), pp. 272–3Google Scholar; Edgeworth, R. J., ‘The Purple Flower Image in the Aeneid’, ZKP 127 (1983), 143–7Google Scholar; Brenk, F. E., ‘Auorum Spes et Purpurei Flores: The Eulogy for Marcellus in Aeneid VI’, AJP 107 (1986), 218–28; esp. 224 n. 13.Google Scholar

² See, in particular, Lattimore, R., Themes in Greek and Latin Epitaphs (Urbana, 1942), pp. 135–41, esp. 138–41Google Scholar. He felt insecure about the origin of the Rosalia but supposed its probable origin as Cisalpine Gaul, with it spreading later to Greece, (p. 140)Google Scholar. He cites (p. 130 and n. 293; p. 131 and n. 297) two Greek epitaphs in which roses appear with other flowers, but they are in the Roman world: Kaibel, G., Epigrammata Graeca e Lapidibus Conlecta (Berlin, 1878), 548, 3–4Google Scholar (= Peek, W., Griechische Grabgedichte [Berlin, 1960], 341Google Scholar = Peek, W., Griechische Vers-Inschriften i [Berlin, 1955], 1409Google Scholar) (Nîmes, 2nd cent. a.d.?)-roses with ia (violets) and narkissos, and EG 547a 1–6Google Scholar ( = GG 463Google Scholar) (Carales in Sardinia, 1st cent. a.d., found with another Greek epigram and 8 Latin ones)–with ia, krina (lilies), narkissos, and hyakinthos. The only other appearance of the rhodon in Kaibel's index is 570 = CIG 6201 (Rome, no earlier than 2nd cent. a.d.), where a young girl is compared to a budding rose cut down in the spring. Seaford, R., ‘The Tragic Wedding’, JHS 107 (1987), 106–30CrossRefGoogle Scholar, cites (112) an epitaph from Leontopolis, 1st cent. a.d., for a young girl buried in her wedding attire, ‘like a rose in a garden’ (GV 1238Google Scholar; cf. 1162.8, 1801), noting Alexiou, M., The Ritual Lament in Greek Tradition (Cambridge, 1974), p. 195Google Scholar. Vérilhac, A.-M., ПАΙΔΕΣ АΩΡΟΙ. Poésie funéraire, i (Athens, 1978), pp. 63–4 n. 39Google Scholar, gives an inscription from Antioch of Pisidia, 2nd–3rd cent. a.d. (GG 383Google Scholar = GV 575Google Scholar), where an eight-year-old boy is compared to a rose, ‘the beautiful flower of the Brotes, ’; and she mentions (ii [Athens, 1982], pp. 343–4)Google ScholarGG 341Google Scholar, GV i. 1238, 1245, 1335, and 1401Google Scholar. Most of these verses relate the rose to untimely death.

³ See Vérilhac, , ii. 344Google Scholar, citing the inscription at Antioch for an eight-year-old boy (GG 383Google Scholar = GV i. 575)Google Scholar, that for a twenty-two-year-old youth in Syria (GV i. 1335–2nd or 3rd cent. a.d.Google Scholar) and (i.64–5 n. 40) that for a four-year-old boy at Rome (= GV i. 401– 2nd or 3rd cent. a.d.Google Scholar). She notes, ii.344, that the rose is usually reserved for young girls or women, with no exceptions before the Imperial period, and that the evocation of a precise flower–only in the Imperiai period–is always a rose, with only two exceptions, one for a youth named Hyakinthos (ii.345).

⁴ Gow, A. S. F. and Page, D. L., The Greek Anthology. Hellenistic Epigrams i (Cambridge, 1965), p. 89Google Scholar (Dioskorides, 25.1623–1628Google Scholar = AP 7.485Google Scholar); commentary, ii.258–9, with reference to Nikandros, fr. 74.70.

⁵ Austin, (p. 204, on line 6.641, and pp. 272–3)Google Scholar, has some good observations on this passage. He prefers purple or bright red for purpureas at 6.884; see also Edgeworth, , 143–7Google Scholar. Egan, R. B., ‘Λειρι⋯εις κτλ in Homer and Elsewhere’, Glotta 63 (1985), 14–24Google Scholar, argues that since λε⋯ριον, originally associated with moisture, later received the meaning ‘flower, lily, etc.’, it is hazardous to designate it ‘white lily’ (23–4). Thus, some lilia could be regarded as ‘purpurei flores’.

⁶ See, for example, Lattimore, , p. 136 and n. 334; p. 137 n. 346.Google Scholar

⁷ Lattimore, , pp. 137–41Google Scholar; Toynbee, J. and Perkins, J. Ward, The Shrine of St. Peter and the Vatican Excavations (London, 1956), p. 76Google Scholar; Toynbee, J. M. C., Death and Burial in the Roman World (Ithaca, 1971), pp. 62–3Google Scholar. Thomas, P. L., ‘Red and White: A Roman Color Symbol’, Rh. Mus. 122 (1979), 310–16Google Scholar, cites (312 n. 3) Propertius 1.17.21 (‘molliter et tenera poneret ossa rosa’) and Tibullus 1.3.61–2, where roses growing in the underworld are associated with youth and death.

⁸ For scepticism about the etymology, see Gippur, H., ‘Purpur. Weg und Leistung eines umstrittenen Farbworts’, Glotta 42 (1964), 39–69, esp. 41–4Google Scholar (including, 68, a swatch coloured by the original dye!); Rowe, C., ‘Conceptions of Colour and Colour Symbolism in the Ancient World’, Eranos Jahrb. 41 (1972), 327–64, esp. 336–8Google Scholar; Chantraine, P., Dictionnaire étymologique de la langue grecque (Paris, 1968), p. 930Google Scholar; Irwin, E., Colour Terms in Greek Poetry (Toronto, 1974), pp. 18–19, esp. n. 31Google Scholar; Fowler, B. H., ‘The Archaic Aesthetic’, AJP 105 (1984), 119–49, esp. 128–9.Google Scholar

Porphyreos is a formulaic epithet of death (thanatos) in battle–therefore, youthful (Iliad 5.82–3Google Scholar; 16.333–4; 20.476–7). The stress on blood suggests that this colour is meant (5.82; 16.333; 20.476).

⁹ KO. ‘Ιμ⋯ρα ⋯νθ’ ὓδατος ε⋯τω γ⋯λα, κα⋯ τὺ δ⋯, Κρâθι,

οἴνῳ пορΦ⋯ροις, τ⋯ δ⋯ τοι σ⋯α καρп⋯ν ⋯νε⋯και. (5.124–5)

Gow, A. S. F., Theocritus i (Cambridge, 1950), pp. 48–9Google Scholar, translates пορΦ⋯ροις (line 125) as ‘reddens’; see commentary (ii.114) on line 125, referring to Iliad 14.16, where the sea turns ‘purple’ (i.e. dark) with a swelling wave. The archaic meaning might be ‘having sheen or iridescence’, e.g. at Iliad 17.547 (rainbow) and 16.391 (sea). See Edgeworth, R. J., ‘“Off-Color” Allusions in Roman Poetry’, Glotta 65 (1987), 134–7Google Scholar; 135–6–citing (134 n. 1; 135nn.4and5) his own article, ‘Does “Purpureus” Mean “Bright”?’, Glotta 57 (1979), 281–91Google Scholar; Irwin, , 18, 103–10, 201Google Scholar; Gippur; and Dürbeck, H., Zur Charakteristik der griechischen Farbenbezeichnungen (Bonn, 1977), pp. 129–37, 175–7.Google Scholar

Austin, citing parallels in Greek poetry, would understand ‘uer purpureum’ (Eclogues 9.40) as ‘bright’ or ‘dazzling’ (p. 204, on Aeneid 6.641).

¹⁰ Quinn, K., Catullus, the Poems (London, 1970), p. 129Google Scholar and Syndikus, H. P., Catull i (Darmstadt, 1984), p. 126Google Scholar, only find one remote Greek parallel–Sappho, fr. 105c LP = LGS 225Google Scholar, the hyacinth in the mountain trampled by the herdsmen.

¹¹ Text of Richardson, N. J., The Homeric Hymn to Demeter (Oxford, 1974)Google Scholar; see his comment, pp. 291–2, and that of Graf, F., Eleusis und die orphische Dichtung Athens in vorhellenistischer Zeit (Berlin, 1974), pp. 83–5.Google Scholar

¹² Dunand, F., Le culte d'Isis dans le bassin oriental de la Méditerranée, I: Le culte d'Isis et les Ptolémées (Leiden, 1973). p. 224Google Scholar; Griffiths, J. G., Apuleius of Madauros. The Isis-Book (Metamorphoses, Book XI) (Leiden, 1975), pp. 159–61.Google Scholar

¹³ For rosettes on Attic tombstones, see, for example, Riemann, H., Kerameikos. Ergebnisse der Ausgrabungen, ii: Die Skulpturen vom 5. Jahrhundert bis in römische Zeit (Berlin, 1940), p. 40, fig. 36Google Scholar (c. 350 b.c.), p. 41, fig. 37 (c. 340/330 b.c.), p. 54, fig. 52 (end of 1st cent. a.d.); Brueckner, A., Der Friedhof am Eridanos bei der Magia Triada zu Athen (Berlin, 1909), p. 71, fig. 43Google Scholar (4th cent. b.c.), p. 105, fig. 60 (4th cent. b.c.) 108, fig. 68 and pp. 110–11, fig. 71 (4th cent. b.c.). The first is reproduced as well in Karo, G., An Attic Cemetery. Excavations in the Kerameikos at Athens (Philadelphia, 1943), p. 32 and plate 32Google Scholar. All these stelai are of males. See also Vedder, U., Untersuchungen zur plastischen Ausstattung attischer Grabanlagen des 4. Jhs. v. Chr. (Frankfurt, 1985), p. 332, figs. 25, 26Google Scholar. Reviewing Vedder, 's book in Gnomon 59 (1987), 345–9Google Scholar, B. Schmaltz remarks (348) that the rosettes were first predominant in the second half of the 4th cent. b.c., and apparently were in competition with Sirens as a symbol for youthful death.

¹⁴ The authenticity is defended by J. Hani in the Budé edition of Defradas, J., Hani, J., Klaerr, R., eds., Plutarque. Oeuvres Morales ii (Paris, 1985), pp. 3–12Google Scholar; but rejected by Sandbach, F. H., ‘Rhythm and Authenticity in Plutarch's Moralia’, CQ 33 (1939), 194–203CrossRefGoogle Scholar–see esp. 197–8, Ziegler, K., Plutarchos von Chaironeia (Stuttgart, 1964), pp. 158–61Google Scholar ( = RE xxi [1951] cols. 636–962)Google Scholar, and Flacelière, R., rev. of Hani, Plútarque. Consolation à Apollonios (Paris, 1972)Google Scholar, AC 42 (1973), 251–3.Google Scholar

For the Pindaric threnos in Plutarch see Einarson, B. and De Lacy, P. H., Plutarch's Moralia xiv (Cambridge, MA, 1967), pp. 338–9Google Scholar; and more generally, Lloyd-Jones, H., ‘Pindar and the After-Life’, in Hurst, A., ed., Pindare. Entretiens sur l'Antiquité Classique 31 (Vandoeu-vres-Genève, 1985), pp. 245–79; pp. 255–6Google Scholar; and Brillante, C., ‘La rappresentazione del sogno nel frammento di un threnos pindarico’, QUUC 25 (1987), 35–51.Google Scholar

¹⁵ Text of Snell, B., rev. Maehler, H., Pindari Carmina cum Fragmentis ii (Leipzig, 1989), p. 118Google Scholar. Ps.-Plutarch, , Moralia 120c hasGoogle Scholar: Φοινικορ⋯δια⋯ τε λειμνες; Plutarch, , Moralia 1130cGoogle Scholar: Φοινικορ⋯δοις ⋯ν λειμώνεσσι.

¹⁶ In verses on Hyakinthos by Euphorien, well-known to Roman poets, the ‘purple’ hyakinthos ‘wailing letters’ springs up in the sand where Aias had fallen (fr. 40 Powell):

ПορΦυρ⋯η ὑ⋯κινθε, σ⋯ μ⋯ν μ⋯α Φμις ⋯οιδν

‘Ροιτε⋯ῃς ⋯μ⋯θοισι δεδουп⋯τος Αἰακ⋯δαο

εἴαρος ⋯ντ⋯λλειν γεγραμμ⋯να κωκ⋯ουσαν.

No doubt Euphorien then related the aetiological myth of the flower's origin from the blood of the boy slain inadvertently by Apollo. The author is grateful to the Editors for this reference.

¹⁷ Thilo, G. and Hagen, H., eds., Servii Grammatici qui Feruntur in Vergila Carmina Commentarii ii (Leipzig, 1884), p. 122: 884Google Scholar. PVRPVREOS FLORES ut saepe 〈V 79〉 diximus, propter sanguinis similitudinem, quia aut anima est, aut animae sedes.

¹⁸ See Rocha-Pereira, M. H., Pausanias. Graeciae Descriptio ii (Leipzig, 1977), p. 139.Google Scholar

¹⁹ The support consists of a small chest decorated with rose-like blossoms and a uraeus snake twisting round a vase. See Andreae, B. et al. , eds., Helbig, W., Führer durch die öffentlichen Sammlungen klassischer Altertümer in Rom. ii (4th edn, Tübingen, 1966), pp. 304–5 no. 1484Google Scholar; Charbonneaux, J., La sculpture grecque classique ii (Paris, 1954), p. 69, pl. 45Google Scholar; Richter, G. M. A., Ancient Italy (Ann Arbor, 1955), pp. 49–50, figs. 168–71Google Scholar. Richter dates the original to c. 460 b.c.; Charbonneaux to c. 440 b.c.; ‘Helbig’ sees a reworking of a 5th cent. b.c. statue under late Hellenistic influence to suggest Aphrodite-Isis, and notes the relationship to Venus Genetrix.

²⁰ The writer is grateful to Professor M. C. J. Putnam of Brown University for reading the manuscript and making several helpful suggestions.