¹ Mynors, R. A. B., Virgil, Georgics (Oxford, 1990).Google Scholar

² Thomas, R. F., Virgil, Georgics (Cambridge, 2 vols., 1988).Google Scholar

³ Doig, G., ‘Vergil, Georgics, I, 491-2’, AJPh 86 (1965), 85–88. The idea of taking Haemias a reference to blood goes back at least to Paratore (1955), though Doig deserves the credit for adding Emathiamand working out the argument in full.Google Scholar

⁴ Nor would a site half-way between the two battles solve the problem. As Doig notes, Emathia is to the west and Mount Haemus well to the north of both battlefields.

⁵ The fact that Emathiamis not similarly glossed is no doubt one reason why Paratore and Thomas find the pun in Haemimore convincing.

⁶ V Poschl, Horazische Lyrik(Heidelberg, 1970), p. 89 n. 47.

⁷ Pointed pairing of the colours white and red (or purple) is of course endemic in Latin verse.

⁸ Latin etymologists derive the name of Haemonia either from Mount Haemus or from Haemon, one of the Spartoiand so a second-hand son of Ares. Details in Maltby, R., A Lexicon of Ancient Latin Etymologies (Liverpool, 1991).Google Scholar

⁹ Apollonius of Rhodes, Argonautica, Book III(Cambridge, 1989), ad 1086, and The Argonautica of Apollonius: Literary Studies(Cambridge, 1993), p. 68.Google Scholar

¹⁰ This was one of his editorial suggestions. In True Names: Vergil and the Alexandrian Tradition of Etymological Wordplay(Ann Arbor, 1996), p. 265, J. J. O′Hara registers 1.492, but not 2.484-8. The latter is the less likely of the two, since the two phrases are four lines apart, but the double connection–both cold and blood–seems significant even at that distance.

¹¹ The juxtaposition of fluminaand Haemonioin 25 suggests that Propertius may also intend a poetic contrast of blood and water.

¹² Since caedehere means ‘gore’, we have blood (in the pentameter) paired with sand (in the hexameter), just as in the Georgicspassage from which we began. Are we to take Patroclus as a dead gladiator?

¹³ As elsewhere, I must omit consideration of unrelated textual problems, though I will say that it seems to me unlikely that this is the beginning of a poem, still less a whole book, as Lachmann suggested.

¹⁴ Details in Smyth, W. R., Thesaurus Criticus adSexti Propertii Textum (Leiden, 1970).Google Scholar

¹⁵ Haemoniocannot be purely ornamental, since Helicon is of course in Boeotia, not Thessaly.

¹⁶ It may also be significant that his one mention of Antigone's Haemon (2.8.21-2) is in a context of suicide by sword-thrust: quid? non Antigonae tumulo Boeotius Hae mon I corruit ipse suo saucius ense latus,... ?In favour of such a pun is the Sophoclean parallel, against it the lack of any sanguinary language in the context, along with the fact that Haemon was not known for anything much except his suicide and that the sword was the standard tragic way of killing a man. Propertius nowhere mentions Haemus or Haemonia.

¹⁷ I will mention only two here. Haemonian puns are surprisingly infrequent in Ovid. I find only three or four sanguinary contexts for roughly 40 instances of Hae moniusand related words, and coincidence could easily account for that many. Exceptions are for the most part early, and occur in Ovid's reworkings of the Propertian passages examined above: examples are Amores2.1.31f. (Hector's corpse) and 2.9.7f. (Telephus). On the other hand, two of the three instances of Haemoniusin Seneca's tragedies seem significant. In the Medea, Haemonius... Athosis the source of the drugs that Medea cuts cruentafalce(720-22)–is she gathering mandrakes? In the Agamem non,Polyxena is Haemonio desponsa rogo(640). Note that it is the pyre, not the hero buried under it, that is Haemonian: it will be quite literally sanguineuswhen Polyxena is sacrificed.

¹⁸ The interpretation of 27f (barbarus exclusis agitat uestigia lumbis —let subitofelix nunc mea regna tenet!)is extremely controversial, but the first word at least seems secure.

¹⁹ If there is any scholar still living who thinks it profitable to try to identify this praetor with a historic personage, the poetic appropriateness of this particular step in his cursus honorummakes it and him that much more likely to be purely fictional.

²⁰ Of course, that is a conscious metaphor, with a known inventor (Robert Schumann), while Propertius' Illyriciswould be the opportunistic use of a convenient linguistic coincidence.

²¹ Although they do not refer to birthplaces, Catullus 44.1 (Ofunde noster, seu Sabine seu Tiburs)and Horace Ep.1.5.5 (inter Minturnas Sinuessanumque Petrinum)are quite similar. Petronius parodies the practice in Sat.48.2: in suburbano... quod ego adhuc non noui. dicitur confine esse Tarraciniensibus et Tarentinis

²² Whether the speaker of 1.21 is dying or already dead, whether he is the same as the poet's unnamed pro pinquusin 1.22, and what relation the latter is to the poet, are all very much disputed. However, none of these questions requires an answer here, since none of them makes any difference to my point. The latest and fullest treatment is by DuQuesnay, I. M. Le M., ‘In Memoriam Galli: Propertius 1.21’, in Woodman, T. and Powell, J. (edd.), Author and Audience in Latin Literature(Cambridge, 1992), pp. 52–83. He argues (pp. 75f.) that the two men should not be identified.Google Scholar

²³ Parker, H. N., ‘The Fertile Fields of Umbria: Prop. 1.22.10’, Mnemosyne45 (1992), 88-92.Google Scholar

²⁴ The OLDlists 1.1.2 under contingo§ 7.a (‘affect emotionally, move, touch’), but it seems better to take it as another instance of contingo§ 6, along with 1.22.9.

²⁵ An earlier version of this paper was presented at the biennial meeting of the Classical Association of the Midwest and South (Southern Section) in Chapel Hill in 1994.1 particularly wish to thank Jenny S. Clay for correcting a silly blunder, and Stephen Heyworth, whose editorial comments and suggestions were extremely helpful: note 10 is only the tip of the iceberg.