^{page 116 note 1} Cf. Heinze, , ad loc.Google Scholar; Fraenkel, , Horace, p. 296.Google Scholar

^{page 116 note 2} Nam. 2. I.Google Scholar

^{page 116 note 3} If we may take his ‘works’ to begin with Satires i. Maecenas appears at the beginning of Satires i, Epodes, Odes i, and Epistles I.

^{page 116 note 4} p. 227; cf. also Pöschl, V., Die groβe Maecenas-Ode des Horaz (Heidelberg, 1961).Google Scholar

^{page 116 note 5} p. 281; cf. also pp. 185, 230, etc.

^{page 116 note 6} Hermes lxxxv (1957), 344Google Scholar; I am indebted to Professor O. Skutsch for drawing my attention to this article.

^{page 117 note 1} Cf. Fraenkel, , p. 407.Google Scholar

^{page 117 note 2} Cf. Skutsch, , CP lviii (1963), 239 n. 6 and references there.Google Scholar

^{page 117 note 3} Consul suffectus in 23 B.C., when the book was published.

^{page 118 note 1} The fact that it is in a different metre from the other eleven odes in the block helps to set it apart as something special.

^{page 118 note 2} If this is significant it means that the notorious Cerberus stanza in 3. 11. 17–20 cannot, after all, be deleted. The calculation is slightly complicated by the unique metre of 3. 12, which, like the stichic 3. 30, has here been taken to consist of four stanzas (or sixteen lines).

^{page 118 note 3} Horace could not very well boast to Maecenas that the hopes he had expressed in 1. 1. 35–36 had now been fulfilled by him: such thoughts needed the Muse as their addressee. A similar arrangement is followed in Epistles i.

^{page 119 note 1} p. 418 n. I.

^{page 119 note 2} See, e.g., Quinn, K., Latin Explorations (1963). p. II.Google Scholar

^{page 119 note 3} If we remember that the poem may it self be read as an example of dulce est desipere in loco, to which Virgil is invited in its last line, just as the Bandusia ode (3. 13) is surely itself the poem that is promised in its last stanza, we need not perhaps feel with Fraenkel that the assumption that the ode is addressed to the poet Virgil ‘turns Horace, the most tactful of poets, into a monster of callousness’ . (If, as seems probable, Horace had Catullus 13 in mind, we may add that Fabullus was certainly Catullus' close friend and may, for all we know, have been a fellow poet.)

^{page 120 note 1} This position is unaffected by the deletion of lines 17 and 33 under Meineke's canon, though it could not survive Lach-mann's more drastic surgery. A reference to the ‘carmen proper’ may be thought special pleading in this context, but the carmen here seems structurally comparable to Juno’ s speech in 3. 3, on which see below.

^{page 120 note 2} On this stanza cf. Fraenkel, , p. 295.Google Scholar

^{page 120 note 3} As Quirinus (1.2.46, 3.3.15, 4–15–9) he shows no special liking for the centre, any more than his mother Ilia (1.2.17, 3.9.8, 4. 8. 22).

^{page 120 note 4} 1.37.6 (balanced by regia, 25), 3.24.45, 4. 3. 9.

^{page 121 note 1} Cf. Heinze's introduction to 4. 1.

^{page 121 note 2} 1.1, 18, 22, 29, 33; 2. 2, 3, 6, 10, 11, 14, 17; 3. 17, 29; 4. 2, 8, 14, 15. (In 4. 2 the second half of the name occurs in line 26, matching Caesarem eight lines later at the other side of the centre; cf. below, p. 124 n. 2 and pp. 128–9.

^{page 121 note 3} 1. 6, 20; 2. 7, 9.

^{page 121 note 4} 2. 16, where this may be due to the anaphora of otium, and 2. 20, on which see above.

^{page 121 note 5} 1. 24; 2. 12; 3. 8, 16; 4. 12.

^{page 123 note 1} That was bad enough for a Roman, ; cf. sub rege Medo Marsus et Apulus, 3. 5. 9.Google Scholar

^{page 123 note 2} Cf. here 1. 34, where there is a similar overlap, but this time into the central stanzas: cogor relictos (5), and valet ima summis (12). See also below, p. 125.

^{page 123 note 3} Fraenkel, , p. 161.Google Scholar

^{page 123 note 4} 39 D., Z 8 L.-P.

^{page 123 note 5} i.e. ‘borrowing of a motto’, cf. Fraenkel, , p. 159 and n. 2.Google Scholar

^{page 123 note 6} At least some Catullan poems are open to the same kind of analysis; see, e.g., Greece & Rome xvi (1966), 155–7.Google Scholar

^{page 124 note 1} e.g. Ovid, (F. 2. 132, etc.).Google Scholar

^{page 124 note 2} Even though this ode is exceptional also in referring to Caesar Augustus by both names: in the only other example of this (4. 2) the names are separated, echoing perhaps the separation of the addressee’ s two names (lulle, line 2, and Antoni, line 26). On the place of Augustus in 2. 9 cf. also Ludwig, W., Hermes lxxxv (1957), 340.Google Scholar

^{page 124 note 3} It is well known that this applies also to almost every metrical rule Horace made for himself in adapting Greek lyric metres to Latin.

^{page 125 note 1} Fraenkel, , p. 253.Google Scholar

^{page 126 note 1} Cf. Heinze, , ad loc.Google Scholar

^{page 126 note 2} Cf. above, p. 120.

^{page 126 note 3} The present bibit in Ψ must surely be wrong.

^{page 127 note 1} Fraenkel, , p. 291.Google Scholar

^{page 127 note 2} ‘The apparently obvious answer to the opening question Quern virion … ? is avoided or, at any rate, is given only in an indirect form. Had Horace been the courtier into whom many modern critics attempt to turn him, he would have felt obliged to write a very different poem' (Fraenkel, , p. 293).Google Scholar

^{page 128 note 1} ‘As if he were thinking of a minor and a maior magistrates’ (Fraenkel, , p. 297).Google Scholar

^{page 128 note 2} This alone shows laetum, the reading of Ψ, to be impossible.

^{page 128 note 3} pp. 291 ff.

^{page 128 note 4} On this statement cf. Fraenkel, , p. 364.Google Scholar

^{page 128 note 5} Formally this ode is a recusatio; but when, in 44 ff., Horace offers to make his own con tribution, this is probably also a poetic re ference to the four ‘political’ odes later in the book.

^{page 128 note 6} Here surely we have an echo of the Gigantomachia, but on the level of hero-kings rather than gods.

^{page 128 note 7} Note how the two halves of Iullus Antonius' name are used (lines 2 and 26) to mark beginning and end of the first half.

^{page 128 note 8} The phrase is foreshadowed by laurea donandus Apollinari (9).

^{page 128 note 9} ‘The sound-pattern sank into Horace's musical memory, and much later, in a poem in the same metre, he praised Augustus himself, quo nihil maius, etc’ (Fraenkel, , p. 293 n. 3, on 1. 12. 17).Google Scholar

^{page 129 note 1} See Marquardt, , Römische Staatsverwaltung, ii. 586Google Scholar; Privatleben der Römer, ii. 525.Google Scholar

^{page 129 note 2} To my feeling at least, the way in which Jupiter is here reintroduced into the ode is more reminiscent than anything else in Horace of the Theocritean ‘model’ ( 17. 135–7).

^{page 129 note 3} Bonus is so hackneyed a word that little attention tends to be paid to its powerful significance as a word of good omen. This is exemplified not only by its traditional association with faustus, felix, and fortunatus (see, e.g., Cic. de div. 1. 102Google Scholar), but also by its frequent occurrence in such poems as Catullus 61, the surviving parts of which contain it no less than thirteen times, often twice in the same line.

^{page 129 note 4} The Ennian echo in the opening lines (0 Romule, Romule die, ׀ qualem te patriae custodem di genuemnt, Ann. III V.) merely adds poignancy to this.Google Scholar

^{page 130 note 1} Even though the same style is used, by way of parody, in an address to a wine jar in 3. 21; cf. Norden, , Agnostos Theos, p. 152.Google Scholar

^{page 130 note 2} It can hardly be coincidence that Venus appears in the first line of the book as well as the last.