ON HORACE’S PYRAMIDS

ON HORACE'S PYRAMIDS (C. 3.30.1–2)Author(s): Michael B. Sullivan Source: The Cambridge Classical Journal , Vol. 60 (2014), pp. 100-108 Published by: Cambridge University Press Stable URL: https://www.jstor.org/stable/10.2307/26430497 REFERENCES Linked references are available on JSTOR for this article: https://www.jstor.org/stable/10.2307/26430497?seq=1&cid=pdfreference#references_tab_contents You may need to log in to JSTOR to access the linked references. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to The Cambridge Classical Journal

The Cambridge Classical Journal (2014) 60, 100–108 doi:10.1017/S1750270514000050 © The Author(s) 2014. Published by Cambridge University Press Michael B. Sullivan Loeb Classical Library, Harvard University Press This paper contends that Horace’s comparison of his completed poetic monument to pyramids at the end of Odes 1–3 is both figurative and literal insofar as we possess ample art historical, literary and papyrological evidence from antiquity for the stacking of an appropriate number of book (sc)rolls in ‘pyramidal’ form. Most notable in this regard is the dedication to Delian Apollo of a triangular casket containing the ten books of Aristarchus’ edition of Alcaeus, whose resonances with the Pythagorean τετρακτύς, and implications for Horace’s own oeuvre, are duly explored. In EXEGI MONVMENTVM (C. 3.30), the final poem of his first three books of Odes, Horace famously boasts that he has completed a poetic monument more lasting than bronze and loftier than the pyramids.1 The comparison of his work to this markedly un-Roman architectural form has given pause to many readers, who have proposed interpretations as varied as the figurative triumph of Augustus over Cleopatra, the imitation of an Egyptian topos likening books to pyramids, and even an oblique reference to the political ambition and subsequent downfall of the elegiac poet and prefect of Egypt Cornelius Gallus.2 While Horace’s evocation of Egypt is clearly au courant in the wake of Actium, a second, complementary explanation for his pointed comparison can be found not at Giza but on Delos, where an inscription lists among wooden objects housed in the treasury of Andros ‘a triangular casket holding books of Alcaeus’.3 Most scholars agree that this casket must have contained the standard ten-book edition of Alcaeus compiled by Aristarchus, since ‘4, 3, 2, 1 fits a triangle nicely’.4 In mathematical, musical and mystical terms, this arrangement is ideally suited to a poetic monument insofar as it instantiates the cosmic Figure 3 Late Roman (4th century CE) marble sarcophagus, The Metropolitan Museum of Art, Gift of Mrs. Joseph Brummer and Ernest Brummer, in memory of Joseph Brummer, 1948 (48.76.1) is reproduced by permission of the Metropolitan Museum of Art, New York. 1 Hor. C. 3.30.1–2 Exegi monumentum aere perennius | regalique situ pyramidum altius. 2 3 4 Respectively, Galinsky (1996) 352; Fuchs (1962) 149–51; Gibson (1997). For further bibliography, see Nisbet and Rudd (2004) 364–6. Alc. T 10 Campbell ξύλινα· . . . τρίγωνον θήκην ἔχουσαν βυβλία Ἀλκαίου. The inscription postdates 166 BCE. I translate Greek θήκη with English ‘casket’ in an attempt to preserve its polyvalence. All translations are my own. Lyne (2005) 544. On Horace’s use of this edition generally, see Lyne (2005) 542–4, with bibliography.

Fig. 1. The τετρακτύς of the decad as the fourth triangular number (1 + 2 + 3 + 4 = 10) harmony of the Pythagorean τετρακτύς, which was widely revered in antiquity for its triangular form in two dimensions (Fig. 1) and its pyramidal form in three (Fig. 2).5 Indeed, according to the author of the Theology of Arithmetic formerly attributed to Iamblichus, Plato’s nephew Speusippus devoted no less than half of his booklet On Pythagorean Numbers to the natural, teleological and poetic qualities of the sacred decad: ὅτι καὶ Σπεύσιππος ὁ Πωτώνης μὲν υἱὸς τῆς τοῦ Πλάτωνος ἀδελwῆς, διάδοχος δὲ Ἀκαδημίας πρὸ Ξενοκράτου, ἐκ τῶν ἐξαιρέτως σπουδασθεισῶν ἀεὶ Πυθαγορικῶν ἀκροάσεων, μάλιστα δὲ τῶν Φιλολάου συγγραμμάτων, βιβλίδιόν τι συντάξας γλαwυρὸν ἐπέγραψε μὲν αὐτὸ Περὶ Πυθαγορικῶν ἀριθμῶν, ἀπ’ ἀρχῆς δὲ μέχρι ἡμίσους περὶ τῶν ἐν αὐτοῖς γραμμικῶν ἐμμελέστατα διεξελθὼν πολυγωνίων τε καὶ παντοίων τῶν ἐν ἀριθμοῖς ἐπιπέδων ἅμα καὶ στερεῶν περί τε τῶν πέντε σχημάτων, ἃ τοῖς κοσμικοῖς ἀποδίδοται στοιχείοις, ἰδιότητός <τε> αὐτῶν καὶ πρὸς ἄλληλα κοινότητος, <περὶ> ἀναλογίας τε καὶ ἀντακολουθίας, μετὰ ταῦτα λοιπὸν θάτερον τὸ τοῦ βιβλίου ἥμισυ περὶ δεκάδος ἄντικρυς ποιεῖται wυσικωτάτην αὐτὴν ἀποwαίνων καὶ τελεστικωτάτην τῶν ὄντων, οἷον εἶδός τι τοῖς κοσμικοῖς ἀποτελέσμασι τεχνικὸν ἀw’ 5 Speus. fr. 28 Tarán ap. [Iamb.] Theol. Ar. p. 62 Ast (p. 84 de Falco) ἔν τε ἐπιπέδοις καὶ στερεοῖς πρῶτά ἐστι ταῦτα· στιγμὴ γραμμὴ τρίγωνον πυραμίς· ἔχει δὲ ταῦτα τὸν τῶν δέκα ἀριθμὸν καὶ τέλος ἴσχει, by which he means that the limit of a point is 1, the limit of a line is 2, the limit of a triangle is 3 and the limit of a pyramid is 4, all of which add up to 10, thereby allowing the τετρακτύς to accommodate all these forms; see Tarán (1981) 281–4; Waterfield (1988) 113 n. 20, 114 n. 23. On the τετρακτύς in general, see Burkert (1972) 72– 3, 186–8, 427; Nolan (2002) 272–4. That Speusippus’ book (or at least the edition to which pseudo-Iamblichus had access) was entitled Περὶ Πυθαγορικῶν ἀριθμῶν demonstrates that the τετρακτύς of the decad was under discussion, even though Speusippus working in the Platonic tradition seems to prefer the less mystical term ὁ τῶν δέκα ἀριθμός; see Tarán (1981) 262–3. For a concise history of Pythagoreanism at Rome, see Kahn (2001).

Fig. 2. The τετρακτύς of the decad rearranged as the third pyramidal number (1 + 3 + 6 = 10) ἑαυτῆς (ἀλλ’ οὐχ ἡμῶν νομισάντων ἢ ὡς ἔτυχε) θεμέλιον ὑπάρχουσαν καὶ παράδειγμα παντελέστατον τῷ τοῦ παντὸς ποιητῇ θεῷ προεκκειμένην. Speusippus, the son of Plato’s sister Potone and head of the Academy before Xenocrates, assembled an elegant little book out of a choice selection of everhandy Pythagorean lectures (particularly from the works of Philolaus) and called it On Pythagorean Numbers, in the first half of which he expounds in most harmonious and orderly fashion the nature of linear numbers, polygonal numbers, and all sorts of plane numbers, together with solid numbers and the five forms that are assigned to the cosmic elements, their uniquenesses and their similarities, their proportionalities and their reciprocities. After that, in the remaining half of the book he proceeds straight to the decad, which he demonstrates is the most natural and fulfilling of all things both perceptible and actual insofar as it is inherently (and not by human contrivance or chance) conducive to making orderly finished products, and provides the foundation and perfect paradigm as preordained by God, the Maker of All [τῷ τοῦ παντὸς ποιητῇ θεῷ].6 Thus, from a Pythagorean perspective, the Andrian dedication of the Alcaic casket to Delian Apollo was triply appropriate, comprising the gift of a mortal poet to a divine poet in the most poetic form possible.7 As I shall argue, it is to this sacred Apolline form as instantiated by the Alcaic ‘pyramid’ of ten scrolls in Aristarchus’ edition that Horace at 6 7 [Iamb.] Theol. Ar. p. 61 Ast (p. 83 de Falco). For comparable resonances of the τετρακτύς at Delphi, see Burkert (1972) 187. On Pythagoras as ‘Hyperborean Apollo’ incarnate, see Burkert (1972) 91, 141–4. For an aniconic representation of Apollo as a pyramid at Megara, see Paus. 1.44.2 ἔστι δὲ ἐν τῷ γυμνασίῳ τῷ ἀρχαίῳ πλησίον πυλῶν καλουμένων Νυμwάδων λίθος παρεχόμενος πυραμίδος σχῆμα οὐ μεγάλης· τοῦτον Ἀπόλλωνα ὀνομάζουσι Καρινόν, καὶ Εἰλειθυιῶν ἐστιν ἐνταῦθα ἱερόν, on which see Gaifman (2012) 70–1.

least partially alludes when boasting over the superiority of his own ‘orderly finished product’ at the consummation of his first three books of Odes.8 For even if Horace’s personal copy of Alcaeus was not housed in a triangular casket such as the one offered to Apollo on Delos (presumably an honour reserved for deluxe editions, suitable for dedication), the studied polyvalence of Odes 3.30.1–2 suggests that the Roman poet both knew of the Aristarchean edition’s capacity for pyramidal arrangement and regarded it as a fitting comparandum for the presentation of his own monumental achievement.9 Like θήκη and its various synonyms in both Greek and Latin, monumentum simultaneously evokes two types of 8 While it may be objected that this arrangement of scrolls technically constitutes a triangular prism and not a pyramid, the presence of the τετρακτύς in both forms (see Fig. 1 and Fig. 2) seems to have elided the distinction between three-dimensional triangle and pyramid for the purposes of the Andrian dedication. Indeed, Speusippus has a compressed and somewhat obscure argument for the presence of the τετρακτύς in all triangles, on which see Tarán (1981) 283–4. In what follows I refer to all such triangular prisms of scrolls as ‘pyramids’, though I concede that in terms of strict Pythagorean numerology the τετρακτύς is only present in stacks of ten. Still, we know that Horace was hardly a strict Pythagorean; see e.g. Hor. Sat. 2.6.63–4: ‘o quando faba Pythagorae cognata simulque | uncta satis pingui ponentur holuscula lardo?’ Also note that a triangular prism more closely resembles an Egyptian pyramid (in mathematical terms, a ‘square pyramid’) than a tetrahedron, which is what the term πυραμίς regularly denotes in Greek mathematics without further specification; see Nicom. Ar. 2.13–14 Hoche; Speus. fr. 28 Tarán ap. [Iamb.] Theol. Ar. p. 63 Ast (p. 85 de Falco). On Speusippus’ contention that the τετρακτύς is present in all pyramids regardless of base shape, see Tarán (1981) 284–5, 289 n. 73; cf. Nicom. Ar. 2.14.4 Hoche καὶ ἐκ τούτου δῆλον γίνεται, ὅτι στοιχειωδέστερα τὰ τρίγωνα· πᾶσαι γὰρ ἁπλῶς αἱ δεικνύμεναι καὶ wαινόμεναι πυραμίδες ἀπὸ τῶν καθ’ ἑκάστην πολυγώνων βάσεων τριγώνοις μέχρι κορυwῆς περιέχονται. For numerous numerological consultations I am grateful to Joel Kalvesmaki, who saved me from several errors in approaching this complex material; needless to say, he bears no responsibility for any problems that remain or points where we differ. 9 Although no representation of a triangular casket survives from antiquity, art historical and papyrological evidence confirms that scrolls were routinely stacked in this form for storage. Closest in time to Horace is a carbonised pyramid of six accounting scrolls found inside the remains of a wooden armarium from the Casa del Sacello di Legno at Herculaneum (ins. V.31), on which see Pugliese Carratelli (1950) 274; Camodeca and Del Mastro (2002) 282–5; Meyer (2009) 574. An armarium holding a stack of scrolls in the Alcaic arrangement with one scroll missing and another in use is likewise depicted on the fourth-century CE sarcophagus of a Greek physician found at Ostia and now at the Metropolitan Museum of Art in New York (Accession Number 48.76.1; see Fig. 3), on which see Birt (1907) 262–3: ‘Mit solchen Schränken müssen wir uns also die großen Bibliotheken des klassischen Rollenbuchwesens bevölkert denken! Wir lernen daraus aber noch, daß man die Rollen im Armarium zwar auf einander-legte; aber sie lagen, wie unsre Zigarren in der Zigarrenkiste oder wie die Flaschen im Weinschrank, nicht überkreuz, sondern in der gleichen Richtung, Pyramiden bildend, so daß der Druck der oberen sich immer auf zwei untere verteilte’ (emphasis added). Thus, even if the casket on Delos was unique and Horace had no knowledge of it, the pyramidal arrangement of an appropriate number of scrolls (e.g., 3, 6, 10, 15, etc.) would have been a familiar sight to anyone with access to an armarium, and the ten-book division of Aristarchus’ Alcaeus made that work particularly susceptible to such arrangement, as the Delian inscription attests. Horace’s (plural) pyramids may even invite comparison with multiple Greek lyric poets: on the six books of Alcman, see Alcm. T 1 n. 7 Campbell; on the question of whether the Alexandrian edition of Sappho comprised nine or ten books, see Sappho T 1 n. 9 Campbell; Page (1955) 112–19. I am grateful to one of the anonymous referees of the Cambridge Classical Journal for pointing out that fifteen books of Ennius’ Annales (which opened with a dream of Pythagorean metempsychosis, and probably closed with the construction of another poetic monument, the Temple of Hercules Musarum) could likewise be stacked in the form of a pyramid, and that a possible pun on perennius would invite comparison with Horace’s Latin predecessor, as well; on fifteen versus eighteen books of Annales, see Zetzel (2007) 13–14. This observation in turn prompts me to note that the fifteenth book of Ovid’s Metamorphoses famously opens with a long speech by Pythagoras (Ov. Met. 15.75–478), and closes with a sphragis clearly modelled on Hor. C. 3.30 (Ov. Met. 15.871–9): characteristically, Ovid implies that he has built the bigger and better pyramid. The issue may even have some

Fig. 3. Sarcophagus with a Greek physician (Ostia, early 300s CE). commemorative vessel: funerary, for the memorialisation of the dead; and literary, for the preservation of knowledge through writing.10 The poet’s initial comparison of his work to bearing on the intended design of Lucan’s De bello civili, which breaks off in its tenth book with Caesar still in Egypt. 10 LSJ s.v. θήκη 1, 2; OLD s.v. monumentum 2, 4, 5. On the double meaning, see Fowler (2000) 197–8; Lowrie (2009) 117–22, with bibliography. For monumentum in the sense of a multi-volume literary work, see e.g. Cic. Phil. 2.20: ‘Nec vero tibi de versibus plura respondebo; tantum dicam breviter, te neque illos neque ullas omnino litteras nosse, me nec rei publicae nec amicis umquam defuisse et tamen omni genere monimentorum meorum perfecisse, ut meae vigiliae meaeque litterae et iuventuti utilitatis et nomini Romano laudis aliquid adferrent’; Liv. praef. 6: ‘Quae ante conditam condendamve urbem poeticis magis decora fabulis quam incorruptis rerum gestarum monumentis traduntur, ea nec adfirmare nec refellere in animo est.’ Note also that θήκη is the head noun of the Greek compound βιβλιοθήκη, equivalent to Latin monumentum at Ov. Pont. 1.1.5: ‘publica non audent [libelli] intra monimenta venire’; Sen. Dial. 6.1.3: ‘restituisti in publica monumenta libros quos vir ille fortissimus sanguine suo scripserat’. On the sustained analogy of physical and poetic corpus in Horace generally, see Farrell (2007). Cf. Hor. Epist. 2.1.264–70: ‘Nil moror officium quod me gravat, ac neque ficto | in peius voltu proponi cereus usquam | nec prave factis decorari versibus opto, | ne rubeam pingui donatus munere et una | cum scriptore meo capsa porrectus operta | deferar in vicum venentem tus et odores | et piper et quicquid chartis amicitur ineptis’, where capsa at Epist. 2.1.268 does similar work. Comparable synonyms include Greek τεῦχος and Latin armarium, each of which similarly designates both bookcase and sepulchre; see LSJ s.v. τεῦχος 2, 4; OLD s.v. armarium 1, 2; cf. Val. Max. 1.7.ext.3: ‘[Simonides] memor<iam> autem beneficii elegantissimo carmine aeternitati consecravit, melius illi et diuturnius in animis hominum sepulcrum constituens quam in desertis et ignotis harenis struxerat’. For the funereal connotations of the more regularly archival κιβωτός (not adduced by LSJ s.v. κιβωτός), see Paus. 6.9.6–8; cf. S. OC 1760–7. On Trajan’s

bronze similarly recalls both tablet and tomb.11 Horace’s unusual application of the masculine noun situs further reinforces this semantic doubling, since the term normally denotes physical position or even arrangement on a page (e.g., of the gods’ names in two verses of Ennius at Apul. Soc. 2), while the perfect passive participle from sinere occurs regularly in sepulchral contexts (e.g., in the epitaphic formula hic situs est).12 The adjective regali similarly calls to mind not just the royal tombs of Egypt, but also the highest grade of papyrus used in deluxe editions.13 Even the verb exegi contributes to the effect, since in architectural contexts exigere means ‘to set [straight]’, as in the erection of a funerary monument such as a column (e.g., Cic. Verr. 2.1.133), while elsewhere in Augustan poetry the word more commonly connotes ‘to perfect’, as in the polishing of a literary work (e.g., Hor. Epist. 2.1.71–2).14 The comparative adjective altius likewise pertains to both structural height and stylistic elevation (or even profundity, given the dual nature of altus/a/um).15 Situated among these polyvalent pointers, Horace’s pyramids invite contemporary bibliophiles to compare his first three books of Odes not just to the wondrous tombs of Pharaonic Egypt, but also to the poetic contents of Aristarchus’ edition of Alcaeus.16 In other words, if we grant pyramidum the same literary valence present in exegi, monumentum, aere, regali, situ and altius, we can understand Odes 3.30.1–2 to mean not only ‘I have completed a monument more lasting than bronze | and loftier than the kingly site of the column as a visual analogue of the same trope (i.e., a funerary monument designed like a scroll, situated in a library) see Birt (1907) 269–82. The metonymy is more pervasive than most of us realise; ‘reading Horace’ of course actually means ‘reading [the book of] Horace’. 11 Nisbet and Rudd (2004) 368. Cf. OLD s.v. aes 1, 8. 12 Indeed, Horace’s usage of situs here is otherwise unattested in Latin literature ( pace OLD s.v. situs2 2, Sil. 3.242 belongs under sense 1b). On the funereal connotations of sinere, see also Nisbet and Rudd (2004) 369, with bibliography. Apul. Soc. 2: ‘sunt illi duodecim numero situ nominum in duo versus ab Ennio coartati: Iuno Vesta Minerva Ceres Diana Venus Mars | Mercurius Iovis Neptunus Vulcanus Apollo’. Horace appears to have been at least partially anticipated in this wordplay by Terence, who exploits the semantic overlap of cistella, monumentum, riscus and sinere to humorous effect at Ter. Eun. 753–54: ‘TH. abi tu, cistellam, Pythias, domo ecfer cum monumentis. | CH. Viden tu illum, Thais? PY. Vbi sitast? TH. in risco: odiosa cessas.’ 13 Cat. 22.6–8 ‘cartae regiae, novi libri, | novi umbilici, lora rubra membranae, | derecta plumbo et pumice omnia aequata.’ Interestingly, this grade of papyrus was later called not ‘regal’ but ‘Augustan’. See Suet. ap. Isid. Orig. 6.10.2: ‘[Cartarum] prima et praecipua Augustea regia, maioris formae in honorem Octaviani Augusti appellata.’ Cf. Plin. HN 13.74. 14 Cic. Verr. 2.1.133: ‘ad perpendiculum columnas exigere’; Hor. Epist. 2.1.71–2: ‘sed emendata videri | pulchraque et exactis minimum distantia miror.’ On the Greek tradition of funerary columns, see McGowan (1995); on the reception of that tradition at Rome in the first centuries BCE and CE, see Frischer (1983) 68–73; Davies (2000) 32–3. Cf. esp. Suet. Iul. 85: ‘postea [plebs] solidam columnam prope viginti pedum lapidis Numidici in foro statuit [in]scripsitque PARENTI PATRIAE. Apud eam longo tempore sacrificare, vota suscipere, controversias quasdam interposito per Caesarem iure iurando distrahere perseveravit.’ For further literary usages of exigere, see Nisbet and Rudd (2004) 367–8. The attenuating force of the root (as in exiguus, exigue, exiguitas, etc.) is not insignificant given Horace’s appropriation of Archaic Greek lyric via Alexandria, and the Roman poet’s three books versus Alcaeus’ ten, on which see below. 15 OLD s.v. altus 1, 13, 14. 16 In this respect Odes 1–3 functions as a literary analogue of the Mausoleum of Augustus, a monument likewise inaugurated in 23 BCE upon the death of Marcellus, and one that similarly invites comparison with Egyptian, Greek and Greco-Egyptian models such as the Pyramids of Giza, the Mausoleum of Halicarnassus and the Sema of Alexander, respectively; see Davies (2000) 49–60.

Fig. 4. The exiguous Horatian ‘pyramid’ of Odes 1–3? pyramids’, but also ‘I have perfected a casket/collection [of scrolls] more lasting than bronze [tablets] | and more [stylistically] elevated/profound than the deluxe arrangement of [Alcaeus’] pyramids’.17 Thus, in the final ode of his three-scroll collection, Horace subtly and wittily asserts his superiority over Alcaeus through the application of the very Alexandrian bookishness through which he filters his aemulatio of the Greek lyric poet elsewhere.18 If correct, this interpretation prompts two speculations. First, let us imagine for a moment the day in 23 BCE when either Horace or his messenger arrived at the ‘kingly site’ of the turris Maecenatiana bearing the dedication copy of Odes 1–3 to Maecenas, ‘scion of Etruscan kings’.19 When the poet’s eminent friend removed the lid from its case, what did he see? In my view it is possible (though of course ultimately unverifiable) that he was confronted with three scrolls arranged in an exiguous pyramid on the Alcaic model (Fig. 4). Such a presentation would certainly befit a collection aspiring to Archaic lyric grandeur on an Alexandrian scale.20 Secondly (and even more speculatively), it is perhaps relevant that, when Horace put down his stylus for the last time a year or two before his death in 8 BCE, his complete oeuvre—two books of Satires, one book of Epodes, four 17 The choice of translations between ‘casket’ and ‘collection’ depends on whether the θήκη on Delos was unique; on this question, and on the possibility that Horace’s pyramids may evoke editions of other poets in addition to Alcaeus, see n. 9 above. For monumentum in the sense of a multi-volume work of literature, see n. 10 above. 18 Cf. e.g., Hor. Epist. 1.19.26–34, esp. 1.19.32–4: ‘hunc ego, non alio dictum prius ore, Latinus | vulgavi fidicen. iuvat immemorata ferentem | ingenuis oculisque legi manibusque teneri.’ On Horace’s refinement of Alcaeus and the latter’s status as the leading (though not necessarily dominant) model for the lyric persona of Odes 1–3, see e.g., Nisbet and Hubbard (1970) xii; Nisbet and Hubbard (1978) 205; Feeney (1993) 45–53; Strauss Clay (2010). 19 On the importance of dedication copies, see Starr (1987), esp. 214, 218. C. 1.1.1 ‘Maecenas atavis edite regibus’. 20 In such a case, altius would signify stylistic elevation or profundity; the irony of a casket stylistically ‘loftier’ yet physically smaller than Alcaeus’ would seem characteristically Horatian. Alternatively, even a cylindrical capsa could evoke the image of a funerary column as opposed to the Alcaic pyramids. See n. 14 above. For a comparable arrangement of three scrolls from Herculaneum, see Camodeca and Del Mastro (2002) 283 n. 11, 295.

books of Odes, two books of Epistles and the so-called Ars poetica—might have filled ten scrolls in their own right.21 Could the phrase exegi monumentum apply not just retrospectively to Odes 1–3, but also prospectively to a planned ten-book edition of the poet’s complete works, to be housed in a triangular casket resembling his Greek predecessor’s, and likewise instantiating the cosmic harmony of the τετρακτύς? In light of the numerous uncertainties, it is a riddle I am not prepared to answer, though one perhaps worthy of both Horace and the Sphinx. Works cited Ast, F. (ed.) (1817) Theologumena arithmeticae, Leipzig. Barchiesi, A. (2007) ‘Carmina: Odes and Carmen saeculare’, in S. J. Harrison (ed.) The Cambridge companion to Horace, Cambridge, 144–61. Birt, T. (1907) Die Buchrolle in der Kunst: archäologisch-antiquarische Untersuchungen zum antiken Buchwesen, Leipzig. Burkert, W. (1972) Lore and science in ancient Pythagoreanism, Cambridge, MA. Camodeca, G., and G. Del Mastro (2002) ‘I papiri documentari ercolanesi (P.Herc.MAN.): Relazione preliminare’, Cronache ercolanesi 32, 281–96. Campbell, D. A. (ed. and tr.) (1982–93) Greek lyric (5 vols), Cambridge, MA. Davies, P. J. E. (2000) Death and the emperor: Roman imperial funerary monuments, from Augustus to Marcus Aurelius, Cambridge. de Falco, V. (ed.) (1922) Iamblichi Theologoumena arithmeticae, Leipzig. Farrell, J. (2007) ‘Horace’s body, Horace’s books’, in S. J. Heyworth (ed.) Classical constructions: papers in memory of Don Fowler, Oxford, 174–93. Feeney, D. (1993) ‘Horace and the Greek lyric poets’, in N. Rudd (ed.) Horace 2000, a celebration: essays for the bimillennium, London, 41–63. Fowler, D. (2000) ‘The ruin of time: monuments and survival at Rome’, in (id.) Roman constructions: readings in postmodern Latin, Oxford, 193–217. Frischer, B. (1983) ‘Monumenta et arae honoris virtutisque causa: evidence of memorials for roman civic heroes’, Bullettino della Commissione archeologica del Governatorato di Roma 88, 51–86. Fuchs, H. (1962) ‘“Nun, o Unsterblichkeit, bist du ganz mein . . .” Zu zwei Gedichten des Horaz’, in E. von Beckerath et al. (eds.) ΑΝΤΙΔΩΡΟΝ: E. Salin zum 70. Geburtstag, Tübingen, 149–66. Gaifman, M. (2012) Aniconism in Greek antiquity, Oxford. Galinsky, K. (1996) Augustan culture: an interpretive introduction, Princeton. Gibson, B. J. (1997) ‘Horace, Carm. 3.30.1–5’, CQ 47, 312–14. Hoche, R. (1866) Nicomachi Geraseni Pythagorei introductionis arithmeticae libri II, Leipzig. Kahn, C. H. (2001) ‘The Pythagorean tradition in Rome’, in (id.) Pythagoras and the Pythagoreans: a brief history, Indianapolis, 86–93. Lowrie, M. (2009) Writing, performance, and authority in Augustan Rome, Oxford. Lyne, R. O. A. M. (2005) ‘Horace Odes Book 1 and the Alexandrian edition of Alcaeus’, CQ 55, 542–58. McGowan, E. P. (1995) ‘Tomb marker and turning post: funerary columns in the archaic period’, AJA 99, 615–32. 21 Here I assume that the Carmen saeculare became attached to Horace’s manuscript tradition sometime after the poet’s death, since the Carmen ‘was pointedly excluded from the self-edition’ of Odes 4 (Barchiesi (2007) 148), was for aesthetic reasons unlikely to have been appended to any other book, and was at seventy-six lines unlikely to have been issued as a scroll in its own right. I am grateful to one of the anonymous readers of CCJ for correcting an earlier hypothesis, and to Richard Thomas for his consultation on the problem.

Meyer, E. A. (2009) ‘Writing paraphernalia, tablets, and muses in Campanian wall painting’, AJA 113, Nisbet, R. G. M., and M. Hubbard (eds.) (1970) A commentary on Horace, Odes Book I, Oxford. Nisbet, R. G. M., and M. Hubbard (eds.) (1978) A commentary on Horace, Odes Book II, Oxford. Nisbet, R. G. M., and N. Rudd (eds.) (2004) A commentary on Horace, Odes Book III, Oxford. Nolan, C. (2002) ‘Music theory and mathematics’, in T. Christensen (ed.) The Cambridge history of Western music theory, Cambridge, 272–304. Page, D. (1955) Sappho and Alcaeus, Oxford. Pugliese Carratelli, G. (1950) ‘L’Instrumentum scriptorium nei monumenti pompeiani ed ercolanesi’, in A. Maiuri (ed.) Pompeiana, Naples, 266–78. Starr, R. J. (1987) ‘The circulation of literary texts in the Roman world’, CQ 37, 213–23. Strauss Clay, J. (2010) ‘Horace and Lesbian lyric’, in G. Davis (ed.) A companion to Horace, Chichester, 128–46. Tarán, L. (ed.) (1981) Speusippus of Athens: a critical study with a collection of the related texts and commentary, Leiden. Waterfield, R. (trans.) (1988) The theology of arithmetic: on the mystical, mathematical, and cosmological symbolism of the first ten numbers, attributed to Iamblichus, Grand Rapids, MI. Zetzel, J. E. G. (2007) ‘The influence of Cicero on Ennius’, in W. Fitzgerald and E. Gowers (eds.) Ennius perennis: The Annals and beyond, Cambridge, 1–16.