^{page 311 note 1} C.R., xiv (1900), 153Google Scholar f. He was followed by Norden, E. in his edition of Book VI (Berlin, 1926), 348.Google Scholar

^{page 311 note 2} T.A.P.A. xc (1959), 173 ff. For a brief bibliography, and reference to other bibliographies, see Otis, op. cit. 173–4.Google Scholar

^{page 311 note 3} Der Traum in der Aeneis (Noctes Romanae, 1952), 94.Google Scholar

^{page 311 note 4} De Div. 2. 146.

^{page 311 note 5} N.H. 10. 211. Macrobius, in Som. Scip. 1. 3. 4 elaborates on the theme. Similarly nowadays, if a person has nightmares after eating cheese, people in general regard this as less significant than if he experiences them without such a physical cause.

^{page 312 note 1} As Virgil tells us: 2. 270.

^{page 312 note 2} The words are those of Otis, B., loc. cit. See also his book Virgil, A Study in Civilized Poetry (Oxford, 1963), 304.Google Scholar

^{page 313 note 1} By Duebner, F. in his edition of the Aeneid (Paris, 1866) ad loc.Google Scholar

^{page 313 note 2} In Aen. 6. 893.

^{page 313 note 3} Op. cit. 90, following Heyne.

^{page 313 note 4} This hypothesis would then explain the passage where Pliny mentions that the eating of beans is eschewed by the Pythagoreans, either because it causes dreams, ‘aut ut alii tradidere, quoniam mortuorum animae sint in ea (sc. faba), qua de causa parentando utique adsumitur’ (N.H. 18. 118). Presumably by eating beans one would dream more, and thus disturb a larger number of spirits from their abode in Hades.

^{page 313 note 5} 6. 283–4. Virgil has adapted this image from the description of Sleep in Homer (Il. 14. 286 ff.), where hides high up in the branches of a fir-tree, looking like a bird.

^{page 313 note 6} Cf. Steiner, op. cit. 87–8.

^{page 313 note 7} Cf. e.g. Hdt. 7. 16

^{page 314 note 1} This was seen by Servius, in Aen. 5. 840, but the point has not been taken up since: ‘Bene autem discernit ista Vergilius, ut “Somnum” ipsum deum dicat, “somnium” quod dormimus, “insomnium” quod videmus in somnis.’

^{page 314 note 2} Cf. e.g. 492–3; 510; 700–2.

^{page 314 note 3} One can suspect a similar meaning for falsus in the falsi genitoris of I. 716. This may mean either ‘of a deceived father’ or ‘of his supposed father—the one he called father’. The latter is much more likely, in which case it provides a second parallel.

^{page 315 note 1} Lucian faced a similar problem at Nec. 22, and also contrived an easy exit.

^{page 315 note 2} I am grateful to Professor R. G. Austin and Dr. J. Briscoe for criticism of an earlier version of this article; neither is responsible for the views expressed.