1) The first collection of Odes (books I-II-III) is thought to have been published in -23 (perhaps -22), but there are signs that 5 poems (out of 88) were added in a second edition post mortem Vergili. These signs are of a numerical nature (for instance, if one takes away these 5 poems, the first book will count 800 lines instead of 876, the second book 500 instead of 572) and of an architectural nature (pairings, general organisation of books, etc).
2) Od. I, 3 : This poem wishes Virgil (or his boat) a bon voyage to Greece. Is it the fatal journey of -19 ? As biographers do not mention another one, some scholars have supposed so, but in all events nobody explains why two thirds of this poem consist of a tremendous imprecation against those criminals who dare attack "heaven itself".
3) Od. I, 28 is one of the most mysterious ("indeniably bizarre in conception", say Nisbet and Hubbard). Is it a monologue or a dialogue ? Who speaks to whom ? For which purpose ? Only one thing is clear : the poet is masking his true intentions. In other words, the text is cryptic, and we cannot understand it if we do not perceive the violence and the malignity of its first half. If we do perceive it, we will ask ourselves why the speaker (Nauta, as a reference, perhaps, to I, 3) hates so much the great Archytas, and why Horace chose to insult (or rather to make his speaker to insult) a man who died three centuries before. These questions are solved if one supposes that "Archytas" is only a mask for another great Pythagorean, the poet Virgil (whose name appears in anagram).
4) Od. II, 6 : here Horace invites a certain Septimius to accompany him as far as Tarente where he will shed tears over "the still warm ashes of the poet, the friend", i.e. Horace himself, according to comments, but rather someone else if Horace's modesty is to be preserved. Virgil comes automatically to mind. The city of Tarente was tightly associated with him (as well as with Archytas), and some report that he died there and not in Brindisi.
5) Od. II, 9 : This poem attempts to console Valgius, an intimate friend of Virgil, in his grief for Mystes' death. Mystes is ordinarily thought to be a young slave, but : 1) Mystes is evocated through strong references to Virgil ; 2) Virgil (Vergilius) appears in anagram at line 9 : tu sempER VRGES fLEbILIbVS ; 3) Virgil was certainly a mystes, or initiate ; 4) The final allusion to the recent successes achieved by Augustus in Orient suits well the year -19, when the emperor returned to Rome from Orient, and when Rome celebrated the restitution of the standards by the Parthians.
6) In Od. II, 20, Horace seems to glorify himself, imagining his transformation in a transcendent swan, as a symbol of his postumous and universal fame. Horace is usually more modest, but before taxing him with bad taste, critics should notice that the proportion of Vergilian echoes is anormally high in this poem. Could Horace conjure up Virgil, precisely known as "the Swan", almost at each line, and proclaim at the same time : "I am the Swan" ? Is it not better to consider that the actual speaker is Virgil?
7) In the 4th book of the Odes, published a few years after Virgil's death, there is one poem, n° 12, which traditionally puzzles its readers. It appears to be an invitation addressed to Virgil, but the light-hearted tone has led many critics to suppose that the Virgil meant is not the poet. However, the third stance in particular is very Vergilian in tone and style. So Virgil is certainly Virgil, but since he is dead he cannot be invited to dinner !
A grammatical trick could settle the question, for if Vergili at line 13 were analysed as a genitive disguised in a vocative (sitim Vergili, and not sitim, Vergili), then the ode could well be addressed to Maecenas, for instance (as IV, 11). But this grammatical ruse should immediately awaken us to suspicion about Virgil's death, especially as the two first stances evoke a heinous crime committed by a tyran.
8) Epist. I, 5 : written in a vulgar and unpleasant tone, this epistle invites to dinner a mysterious Torquatus on the occasion of Augustus' birthday, which practically coincided, as we have seen, with the date of Virgil's death. If the poem was composed after this event, as it is likely to have been the case, or even just before (M. J. McGann, in his 1969 edition, p. 87, n. 2, and D. H. Porter, in Latomus (1972), p. 79-91, tend to assign it to the year -19 : which means that he was published after Virgil's death), Horace's indelicacy cannot be excused, except if we put these lines on the lips of Augustus ordering the burying of his crime.
TIBULLUS (cf. Bibliog. items n° 28, 37)
No informations can be found in Tibullus's work about Virgil's death, for the good reason that both of them disappeared at the same time. However, this very coincidence, once more, should intrigue us. Was it just bad luck, or did somebody plan it ? Every indication that Tibullus could have been murdered will of course reinforce the evidence about Virgil's case.
But had Augustus any reasons to hate Tibullus ? Apparently no. Admittedly Tibullus was a coward (cf. e.g. I, 10, 11-14), but Augustus too, and this was probably not a sufficient motive to be killed. But appearances sometimes lie. For sure, Horace did not think Tibullus was a coward, or else he would not have compared him (in Epist. I, 4) to Cassius Parmensis, the famous anti-Octavian lampooner, and one of the conspirators of the Ides of March, whom Augustus put to death. The Horatian comparison will remain incomprehensible as long as one continues to read Tibullus's elegies at their face value. Actually, he managed to build a double meaning, just like his fellow poets, and using like them the cacozelia latens system, based specially on a secret shift of speakers (Ego vs anti-Ego). Mathematical evidence supports philological analysis to guide the reader through the labyrinth, for Tibullus, a true perfectionist, assigns the same number of lines to Ego and anti-Ego, 624 if I am not mistaken (books I-II). So, under the "anemic" Tibullus (so R. Syme in The Roman Revolution), we find a virulent and irreconciliable opponent to Augustus, a new Cassius Parmensis !
There is some indication that he too was murdered, in particular :
a) Hor. Epist. I, 4 (cf. Bibliog. items n° 33, 35) addresses the deceased Tibullus as if he were alive and, under gently teasing appearances, makes fun of him. The speaker cannot plausibly be Horace, but the Enemy.
b) Hor. Od. I, 3 (cf. Bibliog. item n° 28) : in this first ode denouncing Virgil's murder, Horace makes important use of Tibullus' s references, but his borrowings are carefully veiled, as if it were taboo to associate Virgil and Tibullus in the circumstance.
c) Ovid. Amor. III, 9 (cf. Bibliog. items n° 34, 36)
d) Pseudo-Tibullus : cf. infra.
As well as Tibullus, he uses with a mathematical exactness the occult change of speakers to show Augustus's real face. Threats fell thick and fast on poets, Propertius himself, but also Horace and Virgil, who appear as condemned men. Virgil's murder seems to be alluded to in pieces II, 34, III, 21 (cf. Bibliog. item n° 37), II, 26-27-28 et III, 7 (cf. Bibliog. item n° 19).
OVID (cf. Bibliog. item n° 34 ; for Amor. II, 11-12, item n° 19 ; for Fast., item n° 20).
Amor. II, 6 (note at lines 32-34 the acrostic PVM : Publius Vergilius Maro) ; II, 11-12 ; Met. III, 511-733 (Virgil under Acoetes) ; V, 642-661 (Virgil under Triptolemus ; VI, 412-674 (Virgil under Philomele) ; VIII, 159-168 et 183-230 (Virgil under Daedalus) ; Fast. II, 79-118 (Virgil under Arion) ; Tr. II, 539-40 ; IV, 10, 51-54 ; V, 3, 27-30.
PSEUDO-MOSCHOS, Lamentation for Bion (cf. Bibliog. item n° 10)
The central argument is that the great number of echoes between this anonymous piece and Virgil's work would be better explained by the influence of the second on the first than the reverse. And some ambiguities in the Greek could lead to the conclusion that Bion was killed.
AUGUSTUS AS A FORGER
He practised poetry during his whole life. At 20, during the Modena war, he studied and declaimed every day (Suet. Div. Aug. 84) ; later on, he tried to compose an Ajax and wrote a collection of Epigrams as well as a poem in hexameters intitled Sicilia (ibid. 85) ; he enjoyed parodying Maecenas' style (ibid. 86 ; Macr. Sat. II, 4) ; a few days before his death, he improvised Greek lines and asked his friends to find out who was the author (ibid. 98). Therefore, it was not enough for him to eliminate the poets, he wanted furthermore to refine his vengeance by tempting to attribute them some lines of his own invention. So he could vie with them in their own field, while ridiculing them, settling accounts with them, and even secretly justifying their assassination. All that by means of a "double writing" roughly imitated from his victims' cacozelia. He conterfeited :
- Virgil : the Culex in particular (cf. Bibliog. item n°23), as well as a few lines of the Aeneid (cf. Bibliog. items n° 7, 8, 12, 17, 21 : recapitulated in item 27, p. 200 (V, 295-6 must be added: cf. RBPh 72  51 *; New : cf. also, on this site, my analysis of books I-V). Aen. IX, 774-7 (cf. item n° 12, p. 41-42), joining Horace and Virgil in violent death ( (under the masks of Clytius and Cretheus respectively : cf. H. Morland, SO XLIII  57-67, 102-112), can be considered as a signature of his crime by the criminal himself, the equivalent of Poem 116 for the Libellus of Catullus. The infelicitous prologue beginning by the words Ille ego… should also be considered as added by the emperor post mortem Vergili with malignant, and sarcastic, intentions (Bibliog. Item n° 31).
- Tibullus : Tib. III and Messalla's Panegyric (cf. Bibliog. item n°30 and 30 bis).
- Propertius : book IV (cf. Bibliog. item n°29)
- Ovid : specially in Fastes (cf. Bibliog. item n°20), but also in Metamorphoses ( VIII, 231-259, calumnies about Daedalus ; X, 64-85, calumnies about Orpheus : cf. Bibliog. item n°31 ; II, 367-389, calumnies about Cygnus : cf. Bibliog. item n° 28).
* added Feb. 17, 2001 :
Is Aen. 10. 547-549 another trace left behind by Virgil's murderer ?
Aen. 10. 545-9 reads as follows :
Anxuris ense sinistram / et totum clipei ferro deiecerat orbem; / dixerat ille aliquid magnum uimque adfore uerbo / crediderat caeloque animum fortasse ferebat / canitiemque sibi et longos promiserat annos.
1. The succession of pluperfects is very clumsy, due essentially to the fact that the posteriority of the first one (deiecerat) does not show.
2. The meaning of aliquid magnum dicere is blurred and uncertain: is Anxur a braggart or a kind of wizard ?
3. The irony against the mutilated Anxur is heavy, gratuitous (for Anxur has lost his hand, not his life) and terribly chocking in Virgil's mouth. In the long series of bloody "exploits" accomplished by Aeneas in this tenth book, it is the only time the poet mocks a victim of this violence. On the other hand, Aeneas (to our, and the author's horror) is very fond of insults and sarcasms against the men he defeats.
It is as if Aeneas had suddenly snatched the calamus from Virgil's hand (as well as hacked Anxur's left arm) to write instead of him and force him to take side with him. We assist to a revolt of the creature against its creator, of the written emperor (Aeneas-Augustus) against the writing poet.
Moreover, the phrases aliquid magnum dicere, caelo animum ferre and uim adfore uerbo can easily refer to a poet : cf. Hor., Sat. 1. 4. 43 s.: cui mens diuinior; magna sonaturum; uis ...uerbis ... inest. So, the present case is analogous to IX, 774-7 : cf. item 12, p. 41. And compare Hor. Carm. 1. 28, 1 ff. to Archytas-Virgil (the murderer is the speaker): it is the same tone (cf. item n° 35, 2d ed., p. 74).