1. Constantine Porphyrogenitus, De administrando imperio, 42, ed. Moravcsik-tr., Gy. Jenkins2, R.J.H. (Washington 1967) 182–185 Google Scholar; Theophanes Continuatus, ed. Bekker, I. (Bonn 1838) 122–123.Google Scholar

2. The influence of the standard studies which feature this date — for example, Bury, J.B., A History of the Eastern Roman Empire from the Fall of Irene to the Accession of Basil I (802–867) (London 1912) 417 Google Scholar; Vasiliev, A.A., The Goths in the Crimea (Cambridge Massachusetts 1936) 108 Google Scholar — is such that one often finds it quoted in recent publications, long after its foundation has been demolished. Thus the entry ‘Sarkel’ ( Pritsak, O.) in The Oxford Dictionary of Byzantium, III (Oxford 1991)Google Scholar 1894, still dates the building of the fortress ‘around 833’; curiously enough, the entry ‘Cherson’ (O.Pritsak and A.Cutler), ibid., I, p.419, dates the creation of the thema ‘ca. 832’, that is one year before the presumed date of Petronas’ mission to Sarkel.

3. Grumel, V., ‘Chronologie des patriarches iconoclastes du IX^e siècle,’ Echoes d’Orient 34 (1935) 162–166, see 164–165.Google Scholar

4. Sorlin, I..’Le problème des khazares et les historiens soviétiques dans les vingt dernières années’, TM 3 (1968) 423–455 Google Scholar, see 436, n. 51.

5. Theophanes Continuatus, ed. Bekker, p. 121.

6. Theophanes, Chronographia, ed. Boor, C. de (Leipzig 1883) 362.Google Scholar

7. Treadgold, W., The Byzantine Revival 780–842 (Stanford 1988) 297, n. 406 (p. 441)Google Scholar. One could add that the next two terms, Theodotos’ and Antonios’, are also clearly rounded. John’s six years and one month — the list’s only indication of months — show the way the list was compiled. The 37 years and one month between the death of Tarasios (February 18, 806) and Methodios’ consecration (March 11, 843), reduced by a mistake in calculation to 36 years and one month, were rather approximately divided between four patriarchs, the last on the list getting the spare month.

8. Treadgold, W., ‘The Chronological Accuracy of the Chronicle of Symeon the Logothete for the Years 813–845’, DOP 33 (1979) 157–197, see 178–179 Google Scholar. This recapitulation of Treadgold’s reasoning is called for by the inconclusive treatment of the question in Codoñer, J.Signes, El periodo del segundo iconoclasmo en Theophanes Continuatus (Amsterdam 1995) 535–541 Google Scholar, which does not show a clear grasp of the arguments at stake.

9. Treadgold, (cited n. 7) 313. Most recently, (Athens 1995) 295–296 Google Scholar, lists the three dates proposed (833, 838, 839) and picks up 838 as the ‘most convincing’, yet does not explain why this year carried her conviction more than the others. Signes Codoñer (cited n. 8) 543–548, observing that the connections like in Theophanes Continuatus are not always reliable, denies the chronological link between John VII’s appointment (for which he declines to fix a date) and the Khazar episode (which he would not date either). This deconstructionism for its own sake — hardly sustained by the analysis of hypothetical sources, pp.546–547 — is unwarranted (cf. infra, n. 10).

10. Treadgold (cited n. 7) 312–313. For Signes Codoñer (as ih n. 9), this is Theophilos’ fourth campaign. He claims mistakenly (p.543) that Treadgold dated it in 838, yet ventures no date of his own and makes no attempt to fit it in the general context of the hostilities of the 830s. In fact, the situation described in Theophanes Continuatus fits perfectly in 839, the year after both Byzantium (in 837) and the Arabs (in 838) have scored a major victory, the Arabs’ superior strength being handicapped, like in 837, by a rebel in their midst; see e.g. Bosworth, C.E., entry ‘al-Mu’tasim bi-Llah’, Encyclopédie de l’Islam² , 7 (1993) 777–778 Google Scholar, on the rebellion of Mazyar b. Karin b. Wanda(d)hurmuz in Tabaristan.

11. Treadgold (cited n. 7) 315, states that ‘Petronas Camaterus returned from building Sarkel to report that the archontate of Cherson was in danger, presumably from the Russians, and to recommend that Theophilus create a new theme there (…)just when the Persian revolt had come to an end’ (my italics, C.Z.). The end of the Persian revolt and the subsequent military reform — which allegedly involved the brand-new Crimean thema (cf. infra) — are dated in the fall and in the early winter of 839 but, in any case, before the end of that year (ibid., p.314 and especially n. 432 on p.448).

12. De administrando imperio, 42, ed. Moravcsik, p.184. This indication applies to the travel by water; riding across the steppe — if this could be done safely — would take half that long. On the way back, sailing down the Don reduces the duration of travel.

13. For this standard itinerary, cf. Protector, Menander, frag. 43 = 19,1 in Blockley, R.C., The History of Menander the Guardsman (Liverpool 1985) 170–174.Google Scholar

14. This is a purely theoretical calculation. Even if Petronas was to decide to leave Sarkel as soon as he reached it, he would not be able to undertake the return journey in November and December.

15. It should also be remembered that one of the functions of the Sarkel anecdote is to explain the nickname given to Petronas who became the eponymous founder of the Kamateroi family: . He was called so obviously not for being a transport ship but, at least in our writers’ perception, for being as hard on his people as the were exhausting for the rowers. Petronas the ‘slave-master’ would be a fitting translation for the nickname.

16. This purely technical problem escaped Wozniak, F.E., ‘Byzantine Policy on the Black Sea or Russian Steppe in the Late 830s’, Byzantine Studies/Etudes byzantines 2 (1975) 56–62 Google Scholar, see 61, who suggests that leaving the war vessels at Cherson (as a rear guard) ‘expressed an apparently long-standing Byzantine distrust of the Khazars.’

17. This initial name of the future thema of Cherson is explained below.

18. Oikonomidès, N., Les listes de préséance byzantines des IX^e et X^e siècles (Paris 1972) 49 Google Scholar. Recently, Baranov, I.A., ‘Administrativnoe ustrojstvo rannesrednevekovogo Hersona’, in Ajbabin, A. ed., Materialy po arheologii, istorii i etnografii Tavrii (hereafter MAIET) 3 (Simferopol 1993) 137–145 Google Scholar, contested the generally admitted localisation of the thema of Klimata in Crimea and suggested attaching it to the (hypothetical) town of the Anonymus Hase instead. The Crimean thema created under Theophilos was, according to Baranov, ‘short-lived and extraordinary in character,’ while the permanent thema of Cherson was only established under Constantine Porphyrogenitus, by the middle of the tenth century. However, the author’s reliance on the Anonymus Hase, a notorious forgery, and the lack of satisfactory explanation for the existence of numerous ninth-century seals of strategoi of Cherson undermine his case.

19. Treadgold, (cited n. 7) 282–283, 312–319; idem, Byzantium and Its Army 284–1081 (Stanford 1995) 118ff Google Scholar., with references to the author’s earlier studies on the subject.

20. Zajbt, N. and Zajbt, V., ‘Pecati stratigov vizantijskoj ferny Herson,’ in Vizantija i Srednevekovyj Krym (= Anticnaja drevnost’ i Srednie Veka 27), Simferopol 1995, 91–97, see 92.Google Scholar

21. Nesbitt, J. and Oikonomidès, N., Catalogue of Byzantine Seals at Dumbarton Oaks and in the Fogg Museum of Art, I (Washington 1991) 189,Google Scholar n° 82.19–20

22. Slovo na prenesenie mostem preslavnago Klimenta neboli Legenda Chersonska, ed. Vasica, J., Acta Academiae Velehradensis 19 (1948) 38–80 Google Scholar, see 74 (reprinted as Die Legende, Korsuner von der Überführung der Reliquien des hl. Clemens [Slavische Propyläen 8], Munich 1965)Google Scholar; cf. the editor’s somewhat smoothed translation, p.65: qui tunc urbis gubernacula bene et leniter tractans. The unclear description of Nikephoros’ position generated the notion, in the later Slavonic tradition, that the transfer of relics took place under the emperor Nikephoros.

23. Leon’s sources are studied and his text is edited by Meyvaert, P. and Devos, P., ‘Trois énigmes cyrillo-méthodiennes de la “Légende Italique” résolues grâce à un document inédit’, AB 73 (1955) 375–461 Google Scholar, see 457 for the passage quoted.

24. Zajbt, N. and Zajbt, V. (cited n. 20) 91–92.Google Scholar

25. Studita, Theodoras, Epistulae, 31, ed. Fatouros, G., I (Berlin-New York 1992) 88 Google Scholar; on this episode, cf. Alekséenko, N.A., ‘Un tourmarque de Gothie sur un sceau inédit de Cherson’, REB 54 (1996) 271–275.Google Scholar

26. On the localisation of the Crimean Gothia, see now Ajbabin, A.I., ‘Gli Alani, i Goti e gli Unni’, in Dal mille al mille. Tezori e popoli dal Mar Nero (Milan 1995) 156–170 Google Scholar. The best study of the term klimata, in the context of Crimea, is by Vasil’evskij, V.G., ‘Russko-vizantijskie otryvki IV: zapiska greceskogo toparha’, in idem, Trudy, II, 1 (St. Petersburg 1909) 136–212, see 195–200 Google Scholar. Nystazopoulou, M. G., ‘Note sur l’Anonyme de Hase improprement appelé Toparque de Gothie’, BCH 86 (1962) 319–326, see 324, n. 7 Google Scholar, promised a new study of the term which has never appeared; Bozilov, I., ‘Anonim Haze’: B’garia i Vizantija na dolni Dunav v kraja na X vek (Sofia 1979) 146–155 Google Scholar adds little to Vasil’evskij.

27. Theophanes, Chronographia, ed. Boor, de, 377–379 Google Scholar; Nikephoros Patriarch of Constantinople, Short History, 45, ed. tr. Mango, C. (Washington 1990) 106–108.Google Scholar

28. Le Synekdèmos d’Hiéroklès et l’opuscule géographique de Georges de Chypre, ed. Honigmann, E. (Brussels 1939) 62 (11. 854–857 Google Scholar: Isauria, see the editor’s notes ad loe), 66 (11. 990, 993, 996) and 68 (1. 1041). The editor’s note (p.69, ad 1. 1090) that districts are called klimata for being situated on mountain slopes puts the emphasis on the less pertinent meaning of the word klima.

29. graeco-latinae, Glossae, editae ex codice Harleiano 5792, in Corpus glossariorum latinorum, 2 (Leipzig 1888) 350.Google Scholar

30. On regio, see Feissel, D. and Worp, K.A., ‘La requête d’Appion, évêque de Syène, à Théodose II: Leid P. Zrevisé’, Oudheidkundige mededelingen uit het Rijksmuseum van Oudheden te Leiden 68 (1988) 97–111, see 101–103.Google Scholar

31. AASS Juni.V, 191; cf. Huxley, G., ‘On the Vita of St. John of Gotthia’, GRBS 19 (1978) 161–169.Google Scholar

32. I owe this important observation to Dr.Aibabin, Alexander (Simferopol), whose Histoire ethnique de la Crimée à la haute époque byzantine is in an advanced stage of preparation.Google Scholar

33. Ed. Papadopulos-Kerameus, A, Fontes Trapezuntini, I (repr. Amsterdam 1965) 117–118.Google Scholar

34. Theophanes, Chronographia, ed. Boor, de, pp.351 and 451.Google Scholar

35. Constantine Porphyrogenitus, De administrando imperio, ed. Moravesik, p. 168; 48; 64; 186.

36. Nesbitt and Oikonomides (cited n. 21) 182, n° 81.1.

37. Sokolova, I.V., Monety i pecati vizantijskogo Hersona (Leningrad 1983) 149–150, n° 14.Google Scholar

38. Nesbitt and Oikonomides (cited n. 21) 182.

39. Thus extending the existence of the thema of Klimata over the 830s-870s — so Sokolova, I.V., ‘Les sceaux byzantins de Cherson’, in Oikonomidès, N. ed., Studies in Byzantine Sigillography, 3 (Washington 1993) 99–111 Google Scholar, see 99 — is clearly unwarranted.

40. While this is not the place to elaborate on the relations between Khazaria and Byzantium, genuinely cordial before the Khazars’ conversion to Judaism in the 860s — cf. Zuckerman, C., ‘On the Date of the Khazars’ Conversion to Judaism and the Chronology of the Kings of the Rus Oleg and Igor’, REB 53 (1995) 237–270 Google Scholar — one recent misconception needs to be removed. Ahrweiler, A., ‘Les relations entre les Byzantins et les Russes au IX^e siècle’, Bulletin d’information et de coordination de l’Association Internationale des Etudes Byzantines 5 (1971) 44–70 Google Scholar, see 66, reprinted in eadem, , Byzance: le pays et les territoires (London Variorum 1976), n° VII Google Scholar, writes that according to ‘La Vie grecque de S. Jean le Psichaïte, confesseur sous le règne de Léon l’Arménien (813–820)’, ed. Gheyn, P. van den, Muséon 3 (1902) 97–125, see 118ff.Google Scholar, the Khazars threatened the Byzantine possessions in Crimea towards 820. This threat, which occasionally materialises into a Khazar raid against Cherson, features prominently in several recent Russian studies of Byzantine Crimea. However, the Life in question mentions neither a Khazar threat nor, for what it is worth, any Khazars at all.

41. See for more detail Zuckerman, C., ‘Les Hongrois au pays de Lébédia. Une nouvelle puissance aux confins de Byzance et de la Khazarie ca 836–889’, to appear in Oikonomidès, N. ed., Byzantium at War (Athens 1997).Google Scholar

42. Sazanov, A.V., ‘K hronologii citadeli Baklinskogo gorodisca IX-XI vv.’, in Mogaricev, Ju. M. ed., Problemy istorii i arheologii Kryma (Simferopol 1994) 42–57 Google Scholar, convincingly revises Bakla’s chronology and links its new fortifications to the creation of the thema. Sazanov, p.54, claims that Gercen, A., ‘Krepostnoj ansambl’ Mangupa’, MAlET 1 (1990) 88–166, see 137–138 Google Scholar, established that the ramparts of Doros (modern Mangup), the capital of Crimean Gothia, were repaired by the Khazars ca. 840. According to Gercen, however, the archaeological findings only allow for dating ‘the major repairs of the defence system’ of Mangup in the IX-X centuries, while the dates 843 and 903 quoted by this author originate in the marginalia ‘discovered’ in a Torah scroll by Firkovich, A.S., a notorious forger (in any case, the marginalia mention neither walls nor repairs). Ajbabin, A.I., ‘Osnovnye etapy istorii gorodisca Eski-Kermen’, MAIET 2 (1991) 43–51 Google Scholar, see 48, argues that the creation of the thema of Klimata marked the beginning of a distinct (third) period in the history of Eski-Kermen, another major fortress of Gothia. Yet Ajbabin only shows the continuous existence of the fortress through the ninth century and after; he provides no evidence of building activity on the site which could be conclusively dated in the middle of the ninth century.

43. Epistola Anastasii apostolicae sedis bibliothecarii ad Gaudericum episcopum, ed. Perels, E. and Laehr, G., in Monumenta Germaniae Historica, Epistolae VII (Berlin 1928) 435–438, see 436:… qui scillicet non tarn urbis cives quam esse carceris habitatores, cum non auderent extra earn progredi, viderentur. Google Scholar