1. In an appendix I have collected together a range of comments, on Lavinia in general and our lines in particular. In an effort to keep my paper succinct and clear I have refrained from expressing agreement or disagreement with this or that aspect of this or that interpretation.

2. Cf. Aristotle, , Poetics 1450 a 37–b2Google Scholar.

3. This seems to me a reasonable description of Aeneas' killing of Turnus. His own understandable desire to avenge the death and dishonour of Pallas has been reinforced by the compelling and touching plea of Evander: 11.177–81. Evander, at any rate, feels Aeneas has a duty to him in this respect. Note too that while it is part of heroic and indeed Roman ethics to despoil a defeated enemy (cf. Pallas at 10.449, about to fight Turnus), the wearing of spoils breaks what one might call a taboo: cf. Reinach, S., Cultes, Mythes et Religions, tome III (Paris, 1908), pp. 223ff.Google Scholar; also Aen. 8.562, 11.5–11, 83–84, 193–96. Aeneas, we could say, has honour on his side—and piety, and perhaps other things. But once more piety conflicts with Piety. Turnus is subiectus and no longer a threat to pax; from the grand point of view of Anchises (6.851–53) he should have been spared.

4. See my note on Ciris 293–94. The sentiment belongs most naturally to a parent or equivalent who loses or has the prospect of losing an only child. Amata exaggerates emotionally. The death of Turnus would not signify for her an óρφ⋯νιον γ⋯ρας.

5. There are no other examples of an appeal per has lacrimas in Vergil.

6. How useful for the modest Lavinia, and for Vergil! The fact that Amata is a plausible character speaking with plausible psychology does not of course prevent her advancing other people's action.

7. Cf. Callimachus fr. 80.10 with Pfeiffer ad loc, Musaeus, Hero and Leander 160–61 with Kost ad loc. (Apoll. 3.297–98 (see below) is a little different.) Bomer collects many other blushes in his note on Met.. 3.423. It is interesting that Ovid depicts rubor on the face of the virginal Daphne { Met. 1.484), and that his Phaedra perceives it on the face of Hippolytus (Her. 4.72). Ovid may possibly mean to suggest that the two are virginal in reaction to stimuli they have at some time felt, or do in some way feel – they are not abhorring a completely unknown quantity. That would obviously be correct psychology; and the potential connotations of rubor would on this reading be fully exploited. I do not know whether Met. 1.469–71 works against such an interpretation; note too that Ovid does sometimes (as some of Bomer's passages show) simply use rubor for the ‘roses’ part of a natural, unstimulated ‘milk-and-roses’ complexion. (Jasper Griffin reminds me that blusing is not Homeric. Vergil colours the epic according to a later sensibility. But note Enn., Ann. 352Google Scholar, et simul erubuit ceu lacte et purpura mixta.)

8. Cf. Ovid, , Am. 1.2.32Google Scholar, Ars 1.608, etc. for pudor as an obstacle to love.

9. This is not an example of hypallage, pace Servius (Appendix (1)), who seems to have had influence on, among others, T. E. Page ad loc. Cf. Heyne on 64–69: ‘non est hypallage… utrumque dici et animo repraesentari potest; ut et ignis h. calor sanguinis ruborem faciat, et rubor, sanguine moto, calorem.’ A. J. Bell's interesting discussion of Vergilian hypallage (The Latin Dual and Poetic Diction, Oxford, 1923, esp. p. 320)Google Scholar does not, incidentally, include a subject-object interchange.

10. Rhet. Her 4.67: significatio est res quae plus in suspicione relinquit quam positum est in oratione (cf. Cic., Oral. 139Google Scholar, significatio saepe erit maior quam oratio); Quint. 8.3.83, amplior uirtus est ἓμϕασις, altiorem praebens intellectum quam quern uerba per se ipsa declarant.

11. Rieks, R., ‘Die Gleichnisse Vergils’ in Aufstieg und Niedergang der römischen Welt 31, 2 (Berlin/New York, 1981), p. 1087Google Scholar.

12. Knauer, G. N., Die Aeneis und Homer (Göttingen, 1964)Google Scholar.

13. Knauer, pp. 317–20.

14. A neutral sense ‘stain’, ‘dye’, is a figment of lexicographical imagination (this point is pressed upon me by D. P. Fowler). It is not illustrated by Heliodorus 10.15, pace LSJ; and in Iliad 4.141ff., μιάνη (141) is surely an example of ‘intrusion’ by a ‘tenor’ term (cf. 146, μιάνθην αἵματι μηροί) into the ‘vehicle’: cf. Silk, M. S., Interaction in Poetic Imagery (Cambridge, 1974), pp. 138–42CrossRefGoogle Scholar. For other uses of μΙαίνω in Homer, see Iliad 16.795, μΙάνθηααν δ⋯ ἕθειραι/αἵματι, 797, 17.439, 23.732: a sense ‘sully’ is clear in all of them.

15. See Lewis and Short sub voc; Aen. 11.277, 591, 848, 12.797.

16. sanguineus might also be interpreted as the intrusion of ‘tenor’ terminology (the blood in Lavinia's cheeks) into the ‘vehicle’ (cf. n. 14). But this certainly does not preclude, nor is it as important as, the interpretation offered above.

17. Cf. Otis, Brooks, Virgil (Oxford, 1963), pp. 70ffGoogle Scholar.

18. But note too the interesting passages cited by Enk on Prop. 2.3.11–12; also some of the passages listed by Bömer at Ov. Met. 3.423. It should be observed that such imagery sometimes applies to a stimulated effect (Lavinia, Catullus' uxor), sometimes to a natural ‘milk-and-roses’ complexion (Propertius' Cynthia). But I do not think the distinction affects the point I am making, and so I have not encumbered the text with it.

19. Plato, Symp. 179A and following.

20. Note the case of Achilles, Symp. 179E–180B.

21. My best thanks are due to D. P. Fowler and Jasper Griffin for helpful contributions and criticisms.