¹ I wish to thank Dr R.C.S. Felsch for inviting me to study and publish this material. I am also grateful to him for much assistance conducting the study as well as his unfailing hospitality at Kalapodi. Andreas Nitsche also gave much help at Kalapodi. I am grateful to H.W. Catling and G.D.R. Sanders for reading and commenting on sections of this article. The drawings are my own and were inked by Elizabeth Catling, who also made the map. Many of the photos were taken by R. Felsch. All are warmly thanked. This work was supported by a Twenty-Seven Foundation Historical Award.

Information about the excavation can be found in the following reports: Felsch, R., Kienast, H.. AAA 8 (1975) 1–24Google Scholar; Felsch, R.C.S. et al. AA 1980, 38–123Google Scholar; 1987, 1–99.

² The pottery is stored in the apotheke of the German excavations in Kalapodi. I would like to thank the guard who collected much of the material published here for his help and cooperation when I have been working at Kalapodi.

³ All coarse wares collected are published.

⁴ Where the total body of material representing a site was small, it is usually all included. Two sites, Agios Georgios and Drisbeyi, are each represented by a single sherd, and Kryovrysi by two sherds. They are not isolated stray finds but simply what happened to have been collected. There are sites at these locations.

⁵ Koder, J., Hild, F., Soustal, P.Hellas und Thessalia, Tabula Imperii Byzantini Vol. 1 (Vienna 1976).Google Scholar

⁶ Loc. cit. 179.

⁷ For the route see Felsch, AA 1980, 41 fig. 1.Google Scholar The ancient and mediaeval routes were the same, having changed only in modern times.

⁸ Wheler, G.A Journey into Greece (London 1682) 463Google Scholar; Leake, W.M.Travels in Northern Greece Vol. 2 (London 1835) 163–171Google Scholar; Gell, W.The Itinerary of Greece (London 1827) 223–5.Google Scholar

⁹ Yorke, V.W.JHS 16 (1896) 291–312.CrossRefGoogle Scholar

¹⁰ Morgan, C.H.Corinth vol. 11, The Byzantine Pottery (Harvard 1942Google Scholar, hereafter Morgan; Megaw, A.H.S. ‘Zeuxippus Ware’ BSA 63 (1968) 67–88Google Scholar, hereafter Megaw 1968; Megaw, A.H.S. ‘An early thirteenth-century Aegean glazed ware’ in Studies in Memory of David Talbot Rice edd. Robertson, G. and Henderson, G. (Edinburgh 1975)Google Scholar, hereafter Megaw 1975.

¹¹ Frantz, A. ‘Turkish pottery from the Agora’, Hesperia 11 (1942) 1–28CrossRefGoogle Scholar, hereafter Frantz 1942.

¹² ‘White’ may be quartz, lime or shell; it is usually impossible to determine which without a microscope, though sometimes the shape of a shell is clear. ‘Red’ may be grog. The ‘purple’ grits which occur so frequently in our fabrics have not been identified.

¹³ Wheler, op. cit. 463. The track through the site was probably the old route out of Kalapodi westwards. Wheler shows that the centre of Kalapodi was then higher up than today's centre.

¹⁴ Leake op. cit. 170; Gell op. cit. 223.

¹⁵ Leake, op. cit. 163–4; Gell, op. cit. 225.

¹⁶ The date of Agios Athanasios is recorded in a painted foundation inscription in the church.

¹⁷ Bogdanos is from the Bulgaro-Macedonian branch of the Slavic languages and must post-date the christianization of Bulgaria.

¹⁸ Leake op. cit. 163, 166.

¹⁹ Yorke, op. cit. 303, his map, Pl. XIV, indicates only the upper church.

²⁰ Leake op. cit. 169; Gell op. cit. 222–3; Yorke op. cit. 304.

²¹ Fossey, J.M.The Ancient Topography of Eastern Phokis (Amsterdam 1986) 77.Google Scholar

²² Agios Vlasios is especially associated with the early christianisation of Bulgaria: for his Life see Acta Sanctorum Nov IV (1925) 657–669; for his influence in tenth-century Bulgaria see Dvornik, F.Byzantinoslavica 1 (1929) 35–39.Google Scholar

²³ Leake op. cit. 171, noted a village, known as Purnari, somewhere in this vicinity. The description of its position relative to his route corresponds with Drisbeyi. However this may be unknown site.

²⁴ The following abbreviations are used for the sites: P: Panagia, S: Sphaka, V: Valtesi, M: Michaelidina, B: Bogdanos, Sm: Smixi, Pa: Panagitsa.

²⁵ P 30, 31, V 21, M 11, Sm 27, 28, 33, 34.

²⁶ In general it is not wise to build chronological arguments upon fabric analysis, and especially upon nonscientific analysis, but the pattern suggested by the occurrence of fabric C has, in this particular instance, some supporting evidence.

²⁷ Morgan 141.

²⁸ Morgan 78 fig. 56 d, e.

²⁹ It is a simple decorative technique to apply.

³⁰ It is difficult to correlate our description of N with Morgan's of the Corinthian fabric: Morgan 95; on present evidence the only common feature is that they are both red. The comparative rarity of N and its micaceous quality, not common amongst our fabrics, must mean that it is not local.

³¹ Evidence for local pottery production is examined below p. 46. Evgenidou's standardized production at many centres during the Tourkokratia would support this; Evgenidou, D.Archaiologia 3 (May 1982) 64.Google Scholar

³² Painted incised: P 62, 64, V 43, 47, B 21, Pa 9; developed sgraffito: Pa 51, V 35, B 35, Sm 56, post-Byzantine: B 22, 23, 24.

³³ Hobson, R. L.A Guide to Islamic Pottery (London 1932) 92.Google Scholar

³⁴ See Charitonidou, A.Epeirotika Chronika 2nd. series 25 (1983) 288–289Google Scholar, for the exploitation of a clay source in Epeiros and the way the clay was made ready for production of majolica derivatives in the nineteenth century.

³⁵ Morgan 84–5.

³⁶ Morgan 85, cat. no. 570.

³⁷ For instance, slip-painted: V 2, S 1, B 2, 4; incised and painted incised: S 8, 10; monochrome of varying dates: V 62, 63, 70, S 13, B 32.

³⁸ B 33 and its chronological implications are discussed below, p. 47.

³⁹ A fuller description of the technique is given by Morgan 96–97. The earliest examples in Greece are eleventh century.

⁴⁰ The classification of the Byzantine types can be found in Morgan 97–103.

⁴¹ Morgan 96 notes ‘that green glazes are more common in this ware than any other’.

⁴² The form of B 4 makes is post-Byzantine.

⁴³ Morgan 97–100 cat. nos. 747–774, fig. 75, pl. XXXII.

⁴⁴ It is similar to V 64 and V 65.

⁴⁵ These categories are for vessels with painted decoration alone, when paint occurs in conjunction with other decorative techniques, such as sgraffito or incision, the wares are classified elsewhere.

⁴⁶ Twelfth century, first quarter: Morgan 75–77; twelfth century, second quarter: Morgan 77–80.

⁴⁷ 13, 14, 17–23 are first quarter of the twelfth century, 11, 15, 16, 24—27 are second quarter, following Morgan, loc. cit.

⁴⁸ For majolica in Italy see Berti, G., Cappelli, L., Francovitch, R. ‘La maiolica arcaica in Toscana’ in La ceramica medievale nel mediterraneo occidentale (Florence 1986) 480–510.Google Scholar For Kutayha ware see Hobson, R.L.A Guide to the Islamic Pottery of the Near East (London 1932), 91–3Google Scholar and Lane, A.Connoisseur 104 (1939) 246.Google Scholar

⁴⁹ Charitonidou, A.Epeirotika Chronika 2nd. Series 25 (1983) 287–294Google Scholar; Frantz 1942 group 9, 22, group 10, 4.

⁵⁰ For the Athens examples see Frantz 1942 group 8, 7 and 8, group 9, 1–3.

⁵¹ de Bock, W.Memoires de la Societe nationale des Antiquaires de France 66 (1897)Google Scholar; Talbot-Rice, D.Byzantine Glazed Pottery (Oxford 1930) 32–3, 82–94.Google Scholar

⁵² Morgan 117–123.

⁵³ For instance the various features of P 32: for the band, Morgan cat. no. 1463, pl. XLVII a, fig. 22 b; for the medallion, Morgan cat. no. 1391, pl. XLVII e; for the form, Morgan cat. no. 1143, fig. 103 g. See Kritzas, AAA 4 (1971)Google Scholar, fig. 1, for the form and band of P 34, from the shipwreck. Also the Catalogue for the Centenary of the Christian Archaeological Society (Athens 1984), no. 114, illustrates the motifs of both P 32 and P 34 on one bowl. By the middle of the twelfth century, the date of our pieces, most of this type of sgraffito was being imported into Corinth: Morgan 116.

⁵⁴ Megaw, and Jones, BSA 78 (1983) 261.Google Scholar

⁵⁵ Megaw 1968. P 40 is his class 1A, Sm 59 is 1B.

⁵⁶ Megaw, Report of the Department of Antiquities of Cyprus 1971, 127Google Scholar; Pringle's examples come from a site near Haifa, : Levant 16 (1984) 104Google Scholar, cat. nos. 58–63. Note the pink slip on his 58 and 60, which is also on our examples.

⁵⁷ From these fragments it is not possible to determine how far decoration does continue on the outside.

⁵⁸ A full bibliography covering Greece and the Balkans is given by Bakirtzis, Ch., Balkan Studies 21 (1980) 152Google Scholar, n. 18.

⁵⁹ Charitonidou, A.Archaiologia 4 (Aug 1982) 60–62Google Scholar: T 1836 with V 44, 47, Pa 8; the exterior of T 1827 with the exterior of V 42.

⁶⁰ Birtasevic, M.Mediaeval Ceramics (Belgrade 1970) cat. nos. 137 and 138, pls. 68 and 70Google Scholar; Nikolakopoulos, G.A.Archaiologia 28 (Sept 1988) 82–3Google Scholar, cat. no. 3 should be classed with our painted incised wares, not with the comparanda he cites (Archaiologia 17 (Nov 1985) 44–45, cat. no. 11).

⁶¹ Evgenidou, D.Archaiologia 3 (May 1982) 64.Google Scholar

⁶² It is impossible to develop any further chronological refinements of these wares from surface finds.

⁶³ Megaw 1975.

⁶⁴ Megaw illustrates these decorations on bowls from Paphos, Cherson, Antioch-on-the-Orontes and Pergamon: wavy lines: pl. 15. 2, 5, 6; pl. 16. 2, 6; pl. 17. 2; concentric circles: pl. 15. 3, 4, 5; pl. 15. 3, 4; pl. 17. 1, 3, 5, 6; ring of circles: pl. 16. 1, 2; pl. 17. 4, 6; cancelled motifs: pl. 15. 2; pl. 16. 1, 6; pl. 17. 2, 4.

⁶⁵ Megaw 1975 36–37. There are other examples with this treatment not included in the catalogue, some are illustrated in pl. 6, unpublished material from Panagia.

⁶⁶ Green and brown painted: Morgan 78 fig. 56, d, e; sgraffito: 128 fig. 103, j.

⁶⁷ Illustrated Megaw 1975 36 fig. 1.

⁶⁸ Megaw 1975 pl. 17. 6 may illustrate a complete version of which M 13 could be a fragment.

⁶⁹ Evgenidou, D.Anthropologika 3 (1982) 32–33Google Scholar, Pl. 6β. 2, fig. 5. 2.

⁷⁰ Examples: V 28, Sm 46.

⁷¹ Sm 45 has a reasonably high foot but is unusual for being well formed and for having carefully incised, though standard, decoration. It is in the usual coarse fabric.

⁷² Thessaloniki, : Archaiologia 3 (May 1982) 63 fig. 10Google Scholar; Anthropologika 3 (1982) 32 pl.6.β,γ. Chalkis, : AD 29 (1973–1974) pl. 327.α,β,στ.Google ScholarThebes, : AD 23 (1968) pl.160γ, 162α, 164γ.Google ScholarQuimun, Tall: IEJ 29 (1979) 77 fig. 5. 9–10.Google Scholaral-Qubaiba, : I monumenti di Emmaus el-Qubeibeh e dei dintorni (Jerusalem 1947) 146 pl.47. 7–8, 14.Google ScholarJaffa, : Levant 14 (1982) 109, 111 fig. 1.Google ScholarCaesarea, : Levant 17 (1985) 189–90 cat. nos. 56–58, fig. 11.Google Scholar

⁷³ For other firing stilts of the Ottoman period see Bakirtzis, Ch.Balkan Studies 21 (1980) 151, figs 24–5.Google Scholar

⁷⁴ Since it was made when the clay was wet it should indicate the potter rather than the owner.

⁷⁵ The liguatures for stigma and -ου- developed the forms shown here when Greek books began to be printed in the fifteenth century. By the sixteenth they were a standard feature of handwriting. The damaged -ο- at the edge of the inscription is the final part of one of the abbreviations for common prepositions and syllables which also entered common usage after the introduction of printed Greek books: see G.F. Ostermann and A.E. Giegengack Abbreviations in Early Greek Printed Books (Washington 1936). The possible readings for this abbreviation are: -γο-, -γϱο-, -μο-, and- ϱο-.