¹ Olster, D.M., ‘The Date of George of Pisidia's Hexaemeron’, Dumbarton Oaks Papers xlv (1991) 159–72CrossRefGoogle Scholar (hereafter ‘Olster’).

² Ed. Migne, J.-P., Patrologia Graeca xcii (Paris 1865) cols. 1161–1754Google Scholar; an improved text in Aelian, , Varia Historia, ed. Hercher, R. (Leipzig 1866) ii 601–62.Google Scholar ‘Panegyrical’ poems: Pertusi, A. ed. (with Italian tr.), Giorgio di Pisidia Poemi, I: panegirici epici, Studia Patristica et Byzantina vii (Ettal 1959).Google Scholar Other long poems: PG xcii. Epigrams and shorter poems: Sternbach, L. ed., ‘Georgii Pisidae carmina inedita’, WSt xiii (1891) 1–62Google Scholar, and xiv (1892) 51–68. Prose encomium of St Anastasius the Persian: Flusin, B. ed. (with French tr.), Saint Anastase le Perse et l'histoire de la Palestine au début du vii^e siècle (Paris 1992) i 189–259.Google Scholar

³ Olster 159–61; cf. Whitby, Mary, ‘A new image for a new age: George of Pisidia on the emperor Heraclius’ in Dabrowa, E. ed., The Roman and Byzantine army in the Near East (Cracow 1994) 197–225Google Scholar; Ludwig, C., ‘Kaiser Herakleios, Georgios Pisides und die Perserkriege’ in Speck, P. ed., Poikila Byzantina xi (Bonn 1991) 73–128 esp. 74, 104–28.Google Scholar

⁴ For the date, see the meticulous discussion of Flusin (n. 2) ii 293–309.

⁵ This is the approach of Ludwig (n. 3) who argues independently for association between the Hexaemeron and Heraclius' Persian wars on the basis of analysis of the prooemium and conclusion.

⁶ See Flusin (n. 2) ii 312–327 for the period around 630. The polemical Contra Severum belongs after 630, see Pertusi (n. 2) 16.

⁷ See Mary Whitby (n. 3), esp. 198 n. 4.

⁸ In support of the view that George was no theologian, Olster (n. 1) 160 cites Contra Severum 695–9 (PC xcii 1673) where George disclaims any ability to present dogma. But lack of expertise in technical doctrinal issues does not preclude religious conviction.

⁹ I argue below that the sequence of ideas developed in the following section of this poem (iii 404–61) provides an important parallel for the interpretation of the conclusion of the Hexaemeron.

¹⁰ Ludwich (n. 3) 104–14 argues that the prologue is addressed to Christ, David and Heraclius.

¹¹ No. cvii ed. Sternbach (n. 2) WSt xiv (1892) 66–8; 1.8 alludes to 1 Cor. 13.12. Cf. Pertusi, A., ‘Dei poemi perduti di Giorgio di Pisidia’, Aevum xxx (1956) 395–427 at 398–9Google Scholar, Gonnelli, F., ‘Le parole del cosmo: osservazioni sull' Esamerone di Giorgio Pisida’, BZ lxxxiii (1990) 411–22 at 411–12.Google Scholar Aspects of the scholarship and profundity of the Hexaemeron are illuminated by Gonnelli (loc.cit.) and Bianchi, G., ‘Note sulla cultura a Bisanzio all' inizio del VII secolo in rapporto all' Esamerone di Giorgio di Pisidia’, RSBN xii–xiii (1965–1966) 137–43Google Scholar; id., ‘Sulla cultura astronomica di Giorgio di Pisidia’, Aevum xl (1966) 35–52. But Olster correctly observes (159) that much work remains to be done on the Hex., not least the production of a modem critical edition: see Bianchi ‘Note’ at 137–8 for work in this field. Dr Joannis Vassis is currently preparing an edition.

¹² Olster 168–72.

¹³ Migne differs at the following points: 1841 1848 1853 1864 1865 1878 1886 1890 1892 1896 1898 omitted, (I adopt the reading suggested by the JHS referee).

¹⁴ I here follow Olster 169 n. 79 in preferring the MS reading to the conjecture (‘unenslaved’) printed in Migne and Hercher.

¹⁵ These are the dates suggested by Olster (169). But it is likely that Heraclius did not return to Constantinople at all until mid-631, after the restoration of the Cross to Jerusalem, cf. Nicephorus, Breviarium chs. 18–19 (ed. C. Mango, Washington DC 1990) with Mango's notes, pp. 185–6; Flusin (n. 2) ii 282–91. Heraclius had returned to the east to face the Arabs by 633, where he directed operations until the decisive Byzantine defeat at the Yarmuk in 636, see Kaegi, W.E., Byzantium and the early Islamic conquests (Cambridge 1992) 66–146.CrossRefGoogle Scholar

¹⁶ My interpretation of these lines differs substantially from that of Olster. At 1841 is used metaphorically (LSJ s.v. B.IV, Lampe s.v. B.3), and (1842) is imperative; there is no reason to postulate a change of subject from God to Heraclius at 1841.

The reference to the heavenly gates provides another link with the earlier part of the poem, since at Hex. 1749–65 (PG xcii 1568–9) George expresses the hope that, if he can understand the universe, he will aspire to approach the narrow gate of heaven which on opening terrifies those who enter. At Her. Red. 67–9 God is called opener of the gates to peace.

¹⁷ For Heraclius as the sun, cf. Exp. Pers. iii 1–12 (he eclipses the heavenly bodies worshipped by the Persians); Contra Sev. 673–4, 691, 706–16 (PG xcii 1671–6; Heraclius the sun of religious enlightenment). For the Persian and Roman emperors described respectively as sun and moon in the language of diplomacy, see Whitby, Michael, The emperor Maurice and his historian: Theophylact Simocatta on Persian and Balkan warfare (Oxford 1988) 205 n. 15.Google Scholar

¹⁸ At Her. i. 77 Khusro is called the ‘universe-destroying lion’ (kosmophthoron) in contrast to Heraclius the kosmorustes; the former term is also applied to Khusro by George's contemporary, the historian Theophylact Simocatta (viii. 15. 7); cf. Hex. 1851 and see also n. 37. On Contra Severum, see n. 6.

¹⁹ On the date of Heraclius' return, see n. 15. Descriptions in Theophanes (328. 2–10 de Boor) and Nicephorus (19.1–6 Mango) differ considerably, and may refer to distinct occasions (Mango on Nic. loc. cit.). George's references to charioteers (Her. i 207) and ‘the theatre of life’ (211) recall Nicephorus' account of a triumph. Neither of the historical sources refers to Heraclius holding the icon which he had carried with him on the Persian campaigns (Exp. Pers. i 139–51, with Pertusi's note, 142–3), but the detail is plausible, even though it is frequently uncertain how literally George should be understood.

²⁰ On the text, see n. 14.

²¹ For ‘your fear’ in the sense ‘fear of you’, cf. In Bon. 78.

²² Again I differ from Olster in details of translation. I would put greater emphasis on the adjective (1856): Heraclius' second victory will be a spiritual one achieved through faith. I interpret plus aorist participle (1858) as causal, giving the grounds for Heraclius' second victory (not future). Adjectival (1858) covers the range of meanings ‘unseen’, ‘unknown’, ‘secret’, ‘not evident to sense’, see LSJ s.v., and, for its use in George, cf. Hex. 44, 52, de Van. Vit. 119, Contra Sev. 364.

²³ On the importance of the end of Exp. Pers. for understanding the end of the Hexaemeron, see n. 9. The lines here discussed continue directly after the invocation to God at Exp. Pers. iii 385–403.

²⁴ A parallel may be intended with the purificatory shedding of Christ's blood to save sinful man. For analogy between Heraclius and Christ, see Trilling, J., ‘Myth and metaphor at the Byzantine court: a literary approach to the David plates’, Byzantion xlviii (1978) 249–63 at 259–60Google Scholar, Olster 161–4, Mary Whitby (n. 3) 214–15.

²⁵ On biblical imagery in George, see Mary Whitby (n. 3), and see further below p. 126 on the links between this passage and Hex. 1869–75.

²⁶ So Pertusi (p. 173), who sees the same allusion at Exp. Pers. iii 343–6, 407, Her. i 145–7 and Hex. 1853–8. But the sins are identified as Heraclius' own only in the poem to Bonus, and Ludwich (n. 3) 117–18 rightly doubts allusion to Heraclius' incest even there.

²⁷ Heraclius' premature ageing is also mentioned at In Bon. 144.

²⁸ For this idea, cf. In Christi Res. 39–41 (PG xcii 1377).

²⁹ Olster argues (161–7) that Heraclius' purification of sin is twofold, the elimination of Khusro who personifies the world's evil, and the purification of the Byzantines from the consequences of their own sins.

³⁰ For the same contrast between physical and spiritual barbarians, cf. Eus. Laus Const. 7, esp. secs. 1–2. (I am indebted to Professor Aldo Corcella for this reference.) For similar parallelism between barbarian invasion and the attacks of Satan in the context of fifth-century Gaul, see Roberts, M., ‘Barbarians in Gaul: the response of the poets’ in Drinkwater, J.F. and Elton, H. ed., Fifth-century Gaul: a crisis of identity? (Cambridge 1992) 97–106, esp. 104–6.Google Scholar

³¹ | (795) | | | | | (800) The passage is translated (somewhat differently) and discussed by Olster (165–6), who postulates a contrast between Satan (who attacks the mind) and a barbarian (who attacks the body), where I see an analogy.

³² This passage is cited by Olster (162), who later (165–6) posits two models of barbarian in order to reconcile it with Hex. 794–6 where he sees a contrast between Satan and a barbarian (see previous note); if, as I believe, Satan is likened to a barbarian, this explanation is unnecessary.

For metaphorical cf. also Bell. Avar. 142 where Sergius' armed tears are said to burn to dust the barbarian courage (discussed below, p. 126); In Christi Res. 18 (PG xcii 1376) (at the Resurrection).

³³ Cf. p. 119–20 on the contrasting tone of Heraclias i 192–218.

³⁴ For more detailed discussion of these passages in the context of George's use of Heracles imagery, see Mary Whitby (n. 3) 207–9.

³⁵ The clichés associated with imperial rule are identified as the repeated calls for victory against the barbarians, the role of the children as creators of peace, and their weighty anxieties.

³⁶ Cf. Olster(170 n. 83).

³⁷ No. xlviii (ed. Sternbach WSt xiv [1892] 56). Lines 9–10 describe Heraclius as ‘slayer of Chosroes’ and ‘universal slaughterer’ for the latter, cf. Hex. 1851 and n. 18. Olster mentions this poem and the In Christi Res. passage (170), but in order to stress the distinction between their praise of Heraclius Constantine alone and the two successors mentioned in Hex. Pertusi (n. 2) 16 n. 1 argues that In Christi Res. belongs about 630, but the reference to ‘the stings of the scorpions by the Danube’ (123) suggests the Avars, and hence a date in the mid 620s (I owe this observation to Dr Howard-Johnston.) For another reference to Heraclius acting in concert with his son Constantine on a building project, cf. A.P. ix 655.

³⁸ Menander Rhetor 377.28–30 (p. 94 Russell and Wilson) recommends such a prayer for the conclusion of the basilikos logos.

³⁹ Heraclius had a daughter Epiphania (born 611) and a son Heraclius Constantine (born 612) by his first marriage to Eudocia (Chron. Pasch. 702.10–703.2). The marriage to Martina produced many children, several of whom died in infancy, but their dates of birth and death, like the date of the marriage itself, are uncertain. Heraclonas was probably born in Lazica in 626 (Nic. 12.14–16) and Theodosius was married to Nike daughter of Shahrbaraz in 629/30 (Nic. 17.16–19). The problem is discussed by Speck, P., Das geteilte Dossier. Poikila Byzantina ix (Bonn 1988) 33–40Google Scholar, whose date for the birth of David (627) is, however, too early.

⁴⁰ See p. 120–1. As stressed above (n. 9), the Expeditio Persica parallel is important not only for its similarity of content, but for the analogous sequence of thought. The prayer that Heraclius be filled with warmth towards God like that of Moses and Elijah intervenes between the two passages under discussion, but this is an expansion of the earlier theme of Heraclius as God's faithful shield-bearer (iii 402–3), and is partially paralleled in the Hexaemeron in the comments made about Sergius (1869–75; see further p. 126).

⁴¹ | | | | (435) | | | | | (440) | |

⁴² So Pertusi (n. 2) 162; cf. n. 26.

⁴³ Cf. also In Bon. 116, Bell. Avar. 531–2.

⁴⁴ The adjective initiates the comparison between Sergius and Moses (made explicit at 1874), since it is the term used of Moses' speech impediment at Exodus 4.10 and 6.30. (I am indebted to Mr Peter Coxon, Mr Andrew Mein, Dr Jim Martin and Professor L.G. Whitby for drawing my attention to these passages.) In a linguistically similar but contrasting statement at Contra Sev. 589 (PG xcii 1665) fasting is regarded as having a beneficial effect on the mind of Ephraem the Syrian.

⁴⁵ A pun on two senses of which can mean both ‘provide with a mouth’ (LSJ s.v. II) and ‘harden’, ‘steel’ (LSJ s.v. III). At Hex. 1891 is used in the Euripidean sense ‘fence’ (LSJ s.v. IV). Lines 1872–3 allude to the biblical story of the burning bush (Exodus 3. 1–6), and thus continue the comparison between Sergius and Moses.

⁴⁶ Nic. 26. 1–2 with Mango's note (p. 190). His funeral is described in Const. Porph. de caer. ii. 30 (630.12–631.4 Reiske).

⁴⁷ For example, those of Basil of Caesarea and Ambrose of Milan.

⁴⁸ So God delegates Aaron to speak on behalf of the stammering Moses at Exodus 4.14–17 and 7.1–2 (cf. n.44; I am indebted to Dr Susanna Phillippo for this point). George's privileged vision perhaps suggests his own enhanced state of grace, achieved by his spiritual endeavours to reach God through understanding of the universe, cf. esp. Hex. 1760–5.

⁴⁹ Bell. Avar. 495–501, Laud. S. Anast. 2; cf. Sternbach, L., ‘Analecta Avarica’, Rozprawy Akademii Umiejetnosci, Wydzial Filologiczny ii 15 (Cracow 1900) 326 n. 1.Google Scholar For the image used of Heraclius, cf. (for example) Exp. Pers. iii 415–25 (mentioned above, p. 120), and see further Mary Whitby (n. 3) 213.

⁵⁰ Bell. Avar. 130–44 are discussed by Trilling (n. 24) 257–8. Lines 145–61 are elaborated with New Testament imagery of the fruitful vine.

⁵¹ Theod. Sync. 305.6–7 (Sternbach) The passage alludes to Numbers 10.33–4 which does not, however, explicitly mention Moses' silent voice. For this, Querci (on Hex. 1874) cited Exodus 14.15 as an example of Moses calling upon God silently. But see next note.

⁵² Theodore introduces the Moses image with a reference to the defeat of Amalek (Exodus 17.8–16): whereas Moses held out his arms against Amalek, Sergius held out the icon ‘not made with hands’, and ran round the walls weeping as he exhibited it to the barbarians (304.36–305.6). Similarly George's reference to Sergius’ stand (on the walls of Constantinople), which brings about the enemy's fall, recalls Moses' words to Joshua (Exod. 17.9) like Moses, Sergius stands high above the enemy, who are miraculously defeated. This suggests that it is Moses' silent prayer to God for the defeat of Amalek which inspired the reference to his ‘silent voice’ in the context of the Avar assault. (The link between George's reference to Moses' silent prayer and the story of Amalek was suggested to me by the late Dr Margaret Gibson.)

⁵³ They are mentioned together at Laud. Anast. 2 in connection with Sergius: cf. Flusin (n. 2) ii 382–4. For the crossing of the Red Sea, cf. also Exp. Pers. i 135–8 (of Heraclius) and Bell. Avar. 495–501 (of Sergius).

⁵⁴ The verb is used, as at Hex. 1878. For the expression with cf. Contra Sew 211, 303.

⁵⁵ The ceremony is described in Const. Porph. de caer, ii 27 (627.12–628.20 Reiske).

⁵⁶ See n. 38 and cf. also Menander Rhetor 376.28–9.

⁵⁷ Exp. Pers. iii 428–61, discussed above pp. 124–5; In Bon. 162–4 (longevity), Bell. Avar. 535–41.

⁵⁸ Cf. also Hex. 1900. The claim that Heraclius' victory over Persia has won him the right to world rule is unlikely to have been made after the decisive Byzantine defeat by the Arabs at the Yarmuk in 636, although Olster (169 n. 81, cf. 172) argues that omission of unpleasant events is characteristic of George's panegyric of Heraclius.

⁵⁹ See above pp. 124–5.

⁶⁰ Cf. Exp. Pers. iii. 404–6 where Heraclius is invoked to strike terror into the enemy, ‘and every foreigner who runs towards bloodshed will bend his neck in trembling before this man’ (405–6 see p. 120). The parallel reinforces the change of emphasis from the situation in 622: barbarian submission is now certain.

⁶¹ On the text, see n. 13.

⁶² The preface refers to George's dumbness in past times of adversity (5–7), which balances the epilogue's reference to Sergius' dumbness (1870–75). If the closing lines refer to Sergius, then the children (1884) are the clergy or people of Constantinople; the preface alludes (46) to George's poem as a metaphorical child of Sergius. I am indebted to Dr Paul Magdalino for discussion of the reference of Hex. 1881 ff. See further Ludwig (n. 3) 119–28.

⁶³ The metaphor is explored in detail by Olster (161–8).

⁶⁴ Theophanes 327.24–328. 2, which is probably based on a lost part of George's poetry, cf. Olster 161; Sternbach, L., ‘De Georgii Pisidae apud Theophanem aliosque historicos reliquiis’, Rozprawy Akademii Umiejetnosci, Wydzial Filologiczny ii 15 (Cracow 1900) 35–7Google Scholar; Pertusi (n. 2) 292, 307; Howard-Johnston, J., ‘The official history of Heraclius' Persian campaigns’ in Dabrowa, E. ed., The Roman and Byzantine army in the East (Cracow 1994) 57–87 at 74 n. 35, 83.Google Scholar

⁶⁵ Cf. esp. Theoph. Sim. Hist, v 15.3–7, with Mango, C., ‘Deux études sur Byzance et la Perse sassanide. II. Héraclius, Sahrvaraz et la Vraie Croix’, Travaux et Mémoires ix (1985) 105–17 at 117Google Scholar, and Whitby, Michael, ‘Greek historical writing after Procopius: variety and vitality’, in Cameron, Averil and Conrad, L.I. ed., The Byzantine and early Islamic Near East I: problems in the literary source material (Princeton 1992) 25–80 at 73.Google Scholar

⁶⁶ Mango (n. 65).

⁶⁷ Flusin (n. 2) ii 312–27, cf. Mango on Nicephorus 19, p. 186. I intend to explore these ideas in detail elsewhere.