Hostname: page-component-7b7567c8c-jswmv Total loading time: 0 Render date: 2026-11-15T15:21:29.261Z Has data issue: false hasContentIssue false
  • English
  • Français

KYTHERA FORTY YEARS ON: THE POTTERY FROM HISTORICAL KASTRI REVISITED1

Published online by Cambridge University Press:  22 August 2014

Alan Johnston,
Kathleen Slane and
Joanita Vroom
Alan Johnston*
Affiliation:
Institute of Archaeology, University College London; Institute of Classical Studies, University of London
Kathleen Slane
Affiliation:
Department of Art History and Archaeology, University of Missouri, Columbia
Joanita Vroom
Affiliation:
Faculty of Archaeology, University of Leiden
*
alan.johnston@ucl.ac.uk
Rights & Permissions [Opens in a new window]

Abstract

We present a review of the post Bronze Age material excavated by Coldstream and Huxley at Kastri on Kythera, as part of the ongoing Kythera Island Project. In particular we refer to material not published in Coldstream and Huxley 1972. Greek material is largely confined to one deposit and dates to the period c.500 to 380 bc. From the Roman period material is more plentiful; the two relevant strata are third and late sixth to seventh centuries, with some earlier Julio-Claudian pieces. While later Medieval pottery also is largely from one deposit, the entirety of it is presented, and a mis en scène with respect to Kytheran modern history is offered.

Kythera. Σαράντα χρόνια μετά. Επανεξετάζοντας την κεραμεική των ιστορικών χρόνων από το Καστρί

Παρουσιάζουμε μία ανασκόπηση των ευρημάτων που χρονολογούνται μετά την Εποχή του Xαλκού, από τις ανασκαφές των Coldstream και Huxley στο Καστρί των Κυθήρων. H εκ νέου μελέτη τους αποτελεί μέρος του εν εξελίξει προγράμματος για την Nήσο Κύθηρα (Kythera Island Project). Πιο συγκεκριμένα, επικεντρωνόμαστε στο υλικό που δεν έχει δημοσιευθεί στο Coldstream και Huxley 1972. Tα αρχαία ελληνικά ευρήματα περιορίζονται ως επί το πλείστον σε έναν αποθέτη, και χρονολογούνται περίπου στο 500–380 π.X. Mεγαλύτερη αφθονία παρατηρείται στα ευρήματα των ρωμαϊκών χρόνων, τα οποία προέρχονται από δύο στρώματα του 3ου και του ύστερου 6ου–7ου αιώνα, αντιστοίχως. Περιορισμένος αριθμός πρωιμότερων ευρημάτων έχει χρονολογηθεί στην εποχή των αυτοκρατόρων Iουλίων–Κλαυδίων. H μεταγενέστερη μεσαιωνική κεραμεική προέρχεται επίσης από έναν μόνο αποθέτη, και παρουσιάζεται στο σύνολο της, προσφέροντας έτσι μία συγκριτική επισκόπηση αναφορικά με την σύγχρονη ιστορία των Κυθήρων.

Information

Type
Articles
Information
Annual of the British School at Athens , Volume 109 , November 2014, pp. 3 - 64
Copyright
Copyright © The Council, British School at Athens 2014 

INTRODUCTION

The publication of J.N. Coldstream and G.L. Huxley, Kythera: Excavations and Studies conducted by the University of Pennsylvania and the British School at Athens (Coldstream and Huxley Reference Coldstream and Huxley1972) was a milestone in the study of relations between Crete and the Greek mainland in the Bronze Age. It is perhaps remembered particularly for that; however, the excavations at Kastri, ancient Skandeia, were also of significance for later periods of the history of the island, and were duly set into that history by Nicolas Coldstream and George Huxley some forty years ago. The current survey work in the area of Kastri, directed by Cyprian Broodbank and Evangelia Kyriatzi in the Kythera Island Project (KIP), owes much to their achievement, and has naturally involved restudy of the material then excavated.

In this article we severally review the original publication, and in various ways relate it to more recent research, including indeed the KIP survey. The three sections are perforce differently focused; the Roman deals with several (mostly) stratified deposits, the other two with only one. All, however, seek to publish more material than was included in Coldstream and Huxley Reference Coldstream and Huxley1972 in order to give as full a picture as possible of these important deposits for the study of southern Peloponnesian historical archaeology.

We review the published material, and add further pieces that were not in the original publication. New drawings and photographs are added where appropriate. A number of sherds have been sampled by Evangelia Kyriatzi and the authors for scientific analysis and are accordingly briefly listed as such below. Users of the online version will find the relevant pages and plates (S1–S31) of Coldstream and Huxley Reference Coldstream and Huxley1972 in the Appendix online at http://journals.cambridge.org/ath.

The map at Fig. 1 shows the location of the Kastri trenches.

Fig. 1.

Map of Kastri trenches.

1. First Millennium BC (Alan Johnston)

This section reviews the evidence of pottery import and use at the presumed site of ancient Skandeia in the first millennium bc (in the event the first two-thirds of it) which was originally presented in Coldstream and Huxley Reference Coldstream and Huxley1972, produced soon after the joint British School at Athens and Philadelphia excavations in 1963–5.

Only a selection of the excavated pottery was published there, and a fuller treatment will obviously give a better picture of the totality of the material retained after the excavations, which is almost wholly confined to that from the deposit in Trench IV, termed σ in the publication and hereafter; there are merely some fifteen unstratified finds from elsewhere – ‘Deposit’ ω, 302–22 and two trial trenches. The picture regarding the history of the site overall will be further broadened by the results of the survey, which are merely adumbrated below, especially with respect to the finds in the area around the trench in question.

Published material

A few additional comments can be made with regard to the pieces published in Coldstream and Huxley Reference Coldstream and Huxley1972, 159–65, pls. 44–6 [S1–S6, S9–S11], from some of which depend broader questions, notably of dating and provenance.

At the outset it is useful to set out the problem of distinguishing Laconian from Attic material: there is a substantial overlap in the appearance of the two fine clays, in the range of pink to orange, which may only be fully resolved either by clay analysis (not a readily available field technique) or alternatively by the presence of supplementary features such as shape or the appearance of the glaze,Footnote 2 while any obvious inclusions in the clay point very strongly to a Laconian origin (McPhee Reference McPhee1986, 154–5; Catling Reference Catling, Cavanagh, Crouwel, Catling and Shipley1996, 35–6). Judgements are not made easier by post-depositional changes in surface appearance. Consequently one should be cautious in attributing this range of material if a fine clay is the only usable criterion. The light clay of Laconian sixth century decorated ware is rare in later periods.

Comment below often refers directly to the relevant text and illustrations in Coldstream and Huxley Reference Coldstream and Huxley1972 [S1–S11].

Attic

Among the pieces taken as Attic in Coldstream and Huxley Reference Coldstream and Huxley1972, 159–65, σ 9, σ 13, σ 17, σ 20 and σ 28 [S1, S7, S9] fall into the uncertain category, whether Laconian or Attic, the last (as already noted there) also in part because the underfoot is glazed as far as preserved, though that is not in itself a reason for denying an Attic origin.

Regarding other aspects of the Attic material, chronologically, the skyphos σ 13 [S1, S7, S9] has a horseshoe handle, indicating that it is among the later pieces from the deposit, and another skyphos, σ 15 [S1, S9], has a lower wall that is concave, not straight as in the drawing in Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 48 [S7]. On the other hand the preserved handle of the Laconian cup σ 29 [S2, S7, S10] is oval, an earlier piece, as is the Attic black-figure cup σ 7 [S9]; the remains of decoration on the amphora σ 22 [S2, S7, S9] (if it is not a stand) also seem black-figure, probably palmettes; the lip has an unusually abrupt, cut-off top (Fig. 2a ).

Fig. 2.

(a) σ 22; (b) σ 54; (c) σ 9; (d) σ 50.

Other minor decorative details are that the remains of the red-figure decoration on σ 9 [S1, S7, S9] are indeed minimal (Fig. 2c ); the impressed ‘c's on the bolsal σ 11 [S1, S9] are not clearly intentional; the skyphos σ 14 [S1, S9] has miltos underneath, and there is no reserved band on the bowl σ 16 [S1, S7, S9]; the dribble of glaze inside the oinochoe σ 23 [S 2, S9] is an informative technical detail.

With respect to shape, the bell-krater σ 3 [S7, S9] would seem to be mid fifth century; σ 9 [S1, S9] (Fig. 2c ) is not a skyphos but some form of bowl, and, as noted, is Laconian; and σ 27 [S2, S9] should be a mug, to judge from the middling quality of the glaze on the inside.

With regard to the graffiti, σ 15a [S1, S7, S9] could perhaps be the remains of an alphabet row with chi (which is clear from autopsy), psi, omega written as lambda, and then alpha. It cannot be dated closely as a sherd, but if it is of the later fifth century it brings to mind the writing exercise found by Tsaravopoulos (Reference Tsaravopoulos1999, 264–6) in the cave at Diakofti on Kythera. Alternatively, the first two signs could be in the Laconian alphabet – xi, chi – then lambda and alpha, from which it is difficult to discern any sense. On the cup σ 33 [S3, S7, S10] the graffito is not simple epsilon but a ligature of tau and epsilon.

Laconian

  • σ 32 [S3, S7, S10]. The wall of this cup is very thin, perhaps therefore an earlier piece; there is some probably accidental glaze underneath. The three-bar sigma of the graffito is rare in Laconia, but we do not necessarily have to resort to an Athenian writer to explain it.

  • σ 33 [S3, S7, S10]. This cup also has glaze underneath, though worn; it is fired nearer black than ‘brown’.

  • σ 36 [S3, S10], a mug with a well-everted rim. The traces of the inscribed letter to the right of the published drawing (Coldstream and Huxley Reference Coldstream and Huxley1972, 162 no. 36, fig. 48:36 [= S7:36]) are more likely to belong to a mu than to a retrograde sigma.

  • σ 37 [S3, S10], another mug with a flaring rim. Regarding the graffito, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 48:37 [= S7:37] does not include the remains of an initial letter, perhaps delta.

  • σ 40 [S3, S7, S10] is glazed under.

  • σ 41 [S3, S7, S10] does not seem Attic but an imitation; there is a reserved band on the lower wall.

  • σ 43 [S3, S7, S10]. Plate; there is no reserved line, though there is a stacking ring on the floor inside.

  • σ 45 [S4, S7, S10]. Bowl; the lower wall, foot and underside are reserved. It is questionably Laconian, but no other origin seems plausible.

  • σ 50 [S4, S10]. Strainer vase; ‘trickle of glaze’ is an understatement. The inside of the spout is glazed, as is part of its floor; glaze passed through four holes on to the inside of the wall. It is large for a strainer vase. Fig. 2d , where the spout is presented at an angle consonant with the wheel marks on the inner surface.

  • σ 51 [S4, S10]. Large closed vase; it is reserved under.

  • σ 54 [S4, S10]. The diameter of this large lid is c.18 cm. It has a small, low knob, and there are sharp wheel ridges under. Fig. 2b .

Corinthian

  • σ 57 [S4, S10]. The rim of this skyphos is more everted than in Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 48 [= S7:57]. The handle is still oval, not horseshoe. Fig. 3a .

  • σ 58 [S5, S10] has good vertical curvature, and so is a cup not a skyphos.

    Fig. 3.

    (a) σ 57; (b) σ 67; (c) σ 64.

  • σ 59 [S5, S7, S10]. Sampled (no. 98/258).

Semi-glazed, plain and coarse

  • σ 63 [S5, S8, S11]. It is worth noting that if this had only been a small fragment it could easily have been taken for the edge of a tile.

  • σ 64 [S5, S8, S11]. Profile of mortar. Fig. 3c . Sample no. 98/263.

  • σ 66 [S5, S8, S11]. Ledge rim of mortar; fairly small grits on inside; cf. Catling Reference Catling, Cavanagh, Crouwel, Catling and Shipley1996, 73. Sample no. 98/962.

  • σ 67 [S5, S11]. Trefoil oinochoe; five joining fragments. Clay rather more red than σ 60; surface orange-buff. Very hard fired and rather micaceous. Fig. 3b .

Amphorae

  • σ 68 [S5–S6, S8, S11]. The red paint is an intentional band.

  • σ 69 [S6, S11]. Chian new-style amphora, with only a little mica. Slight, triangular everted rim. Handle 3.8 × 2.2 cm. Impressed circle on top of handle. Fig. 4a .

  • σ 70 [S6, S11]. Clay close to σ 69. Handle, 4.5 × 2.9 cm.

  • σ 71 [S6, S11]. Much small mica. Clay squeezed and twisted into a pointed shape for the stem. Fig. 4b .

  • σ 72 [S6, S8, S11]. Diameter 16 cm.

  • σ 73 [S6, S8, S11]. Diameter 17 cm.

  • σ 74 [S6, S11]. Narrow ring-foot. Buff-ochre clay; cream surface; many small inclusions and some mica. Now three joining fragments. Fig. 4c .

  • σ 75 [S6, S8, S11]. Two joining fragments. Diameter 19 cm. Not flat-topped as in Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 49:75. Sharp offset of lip. Clay a little more buff-brown than orange.

  • σ 76 [S6, S8, S11]. Diameter 18 cm. Handle scar preserved. Orange-buff.

  • σ 77 [S6, S11]. Handle 3.7 × 2.2 cm; near amygdaloid shape. Much very small mica. Perhaps Samian.

Fig. 4.

(a) σ 69; (b) σ 71; (c) σ 74.

Cooking pottery

  • σ 78 [S6, S8]. The possible spout is puzzling; as can be seen in Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 46:78 [= S11:78], it is situated directly above the left root of the lost horizontal handle and so an implausible spout; but there certainly appears to be a deliberate thickening and downturn of the rim here.

  • σ 79 [S6, S8, S11] is nastily warped; same fabric as G23, below.

  • σ 80 [S6, S8, S11]. Sample no. 98/267.

‘Deposit’ ω (Coldstream and Huxley Reference Coldstream and Huxley1972, 201–3, pl. 58)

ω 310, ω 310a, ω 321 have not been relocated. I give additional notes on some of the others:

  • ω 308. The clay is more ochre than orange. 0.9 cm thick. A 7th century date is possible.

  • ω 309. The clay is buff. Decoration seems to be sigmas, or longer squiggles.

  • ω 311. The fabric – fine light buff – and the style seem to be Cretan. It is a large closed vase with a diameter of c.35 cm. There would therefore have been room for more than one figure on each side; were there two facing lions? Fig. 5a .

  • ω 313. Glazed inside; the wall is 0.8 cm thick, probably from a krater. The leg, seemingly equine, is large, and raises the question whether it may not belong to a Trojan Horse.

  • ω 316. From a krater, c.450–425 bc.

  • ω 317. Cup-skyphos. Fig. 5b .

  • ω 318. Mug; it seems to be Cretan and earlier rather than late Archaic. Preserved height 3.8 cm; thin wall. Fig. 5d .

  • ω 319. Amphora or hydria rim; concave inner surface. Cretan. Fig. 5e .

  • ω 320. Oinochoe? Thick dark glaze and thick fabric; it seems later than the 5th century. The neck is glazed inside, with substantial dribbling onto the shoulder.

  • ω 322. Amphora. Nine fragments, mostly joining. A small finger mark pushes up clay at the shoulder join. Diameter 34 cm. Handle 4.6 × 2.6 cm. Micaceous brown-buff. The extent of the red band around the neck is uncertain. North Greek, though the precise provenance is not easy to determine. Fig. 5c .

Fig. 5.

(a) ω 311; (b) ω 317; (c) ω 322; (d) ω 318; (e) ω 319.

Unpublished material

Notaras trench

Two Classical sherds are among the mainly Bronze Age pieces from trial Trench I in the Notaras property, briefly mentioned in Coldstream and Huxley Reference Coldstream and Huxley1972, 75:

  • Notaras 1.1. Body of Attic open vase; maximum preserved dimension 3.1 cm; 0.4 cm thick. Black-glazed with two red bands, each 0.25 cm thick, preserved. Perhaps from a black-figured pot.

  • Notaras 1.4. Profile, save inner edge of foot, of probably Laconian ‘salt cellar’. Grey-buff clay. Height 2.7 cm, diameter 6 cm. Fig. 6a .

Fig. 6.

(a) Notaras 1.4, ‘salt cellar’; (b) Classical House, krater.

Classical House

Only one sherd was seemingly retained from the trials noted in Coldstream and Huxley Reference Coldstream and Huxley1972, 74–5 (and not mentioned there); it is the foot and lower wall of a Classical krater, wholly glazed save perhaps for the worn stand-ring. Diameter of foot 12 cm. Perhaps Laconian but the clay is a rather pale ochre. It is marked ‘H.1.4’. Fig. 6b .

Deposit σ

A considerable amount of material was not included in the publication, and it is not intended to give a complete account here, but rather an overview, with emphasis on types not represented in the published material. Specific sherds, the great majority illustrated, are given catalogue numbers below, using ‘G’ to distinguish them from the material in the later sections of this article. The sub-categories below largely follow the categories in Coldstream and Huxley Reference Coldstream and Huxley1972, and the separate bags in which the pieces were, and largely still are, stored. This approach does result in some overlaps between semi-decorated and plain pieces, but their number is few.

Tiles

Only a few tile fragments were kept from the excavation and none were published, perhaps for the reason that they present no significant architectural details; indeed, illustrations add little to description. However, they do contribute much to the overall interpretation and dating of Classical period roofing material on the island. All are of Laconian type, save two fragments from the KIP survey which appear to be Corinthian, if they are not part of other shapes, e.g. bins.

  • G 1) Fragment with strong curvature, diameter c.19 cm. Thickness (of the main body, not the edges) 1.0 cm. Glazed on the convex side. The glaze is somewhat streaky but basically dark and lustrous. Sample no. 10/1.

    This must be a cover tile, and the KIP survey material clearly demonstrates not only the rarity of material with such curvature, but also that those pieces which have been found are all of a similar appearance. The rarity of cover tiles has been particularly noted in the Laconia survey (Catling Reference Catling, Cavanagh, Crouwel, Catling and Shipley1996, 85), and is borne out by the results of the KIP survey, where cover tiles are substantially less than 1% of the total.

  • G 2) Fragment with a straight-cut lower edge. Thickness 1.2 cm. Diameter c.47 cm. Medium-brown biscuit, purplish surface. Dark paint on concave side. Sample no. 98/269.

    Clearly a pantile, to be compared with:

  • G 3–4) two examples, one of two joining fragments, of moulded upper edge, with substantial ‘toe’. The larger piece, G 3 (Fig. 7 showing rear face and side) is preserved to 19 cm wide with a diameter of c.55 cm; 2.0 cm thick. The other is 1.4 cm thick. Glazed on the concave side.

    These must be the pantiles which accompany the rarer cover tiles like G 1. Their straight-cut edge therefore will be the lower, the moulded the upper, overlapped by the tile above. The material is not well enough preserved to judge any taper of the tile up or down.

  • G 5) Also kept, a fragment without any edge, 1.1 cm thick, with strikingly good dark glaze on the concave side.

    This type of material is very well represented in the survey, enabling us to place such material broadly in the Classical period. There is no survey evidence to suggest that they could stretch back to the Archaic period, but the possibility should perhaps not be ruled out.

Fig. 7.

Pantile, G 3.

Attic

There are ample plain black body fragments – the majority small and from open vases; one very fine piece is from a closed pot. A scrap of a squat lekythos has red-figure palmette decoration. One body fragment of a small closed pot may have minimal remains of black-figure decoration. Eleven sherds have gadroons or vertical grooving, all from open vases; three are relatively early and careful, another is probably from a kantharos, and one seems later, probably fourth century (and not surely Attic).

  • • Two small red-figured krater fragments.

  • • Two lamp fragments, one with vertical handle.

  • • Two skyphos feet, one large, of Attic type.

  • • Foot of a fluted squat lekythos, with miltos on underside.

  • • Two krater feet, one from a bell-krater, the other probably a column-krater.

  • • More than fifty rims of smaller pots, almost all open, ranging from simple bowl rims, via everted, to ledge rims. Two are from heavy cup-skyphoi or cup-kantharoi. Larger rims are from a red-figured krater, a lekanis and a probable hydria.

  • • Some thirty handles, both Attic and Laconian, about two-thirds of them round, the rest mostly strap handles.

  • G 6) Foot fragment of a kantharos. Slight groove on resting surface. Probably once wholly glazed; stamped palmettes and rouletting inside. Preserved height 1.7 cm. Fig. 8a . Close to Sparkes and Talcott Reference Sparkes and Talcott1970, no. 664, dated by them c.350 bc.

  • G 7) Foot of a large closed vase. Diameter 11 cm, maximum preserved dimension 10.8 cm. Two joining fragments. Fig. 8b .

  • G 8) Probably Attic, lower parts. Preserved height 5.3 cm. A curious piece; seemingly a narrow floor and tall pedestal foot, its outer profile running uninterrupted up to lower wall. Glazed save for underside. Fig. 8c .

Fig. 8.

(a) G 6; (b) G 7; (c) G 8.

Corinthian

Fifteen varied pieces in an ‘unpublished’ bag. Most are from lekanai or bowls – six body sherds, three rims, three varied feet. One rim is smaller, rounded; one handle may be from a skyphos. A flaring glazed rim could be Corinthian or Laconian. A neck fragment has been sampled, no. 98/257: it is rather unevenly glazed inside, with a neck diameter of c.15 cm; probably from a Classical krater.

  • G 9) Rim of a lekane, diameter c.35 cm. The downcurving ledge rim is closer to the mortar, σ 64 [S8:64]. Fig. 9a .

  • G 10) Foot of amphora or hydria, Fig. 9b . Cream-buff clay. Diameter c.14 cm, maximum preserved dimension 9.4 cm. Disc foot. Glaze on lower wall and outside of foot.

  • G 11) Foot of amphora or hydria, Fig. 9c . Cream-buff clay. Diameter 13 cm; maximum preserved dimension 8.2 cm. Low ring-foot. Similarly glazed.

Fig. 9.

(a) G 9; (b) G 10; (c) G 11.

A foot similar to G 10, but of pink-buff clay, is probably Laconian, diameter 15 cm, maximum preserved dimension 13.5 cm. Cf. Catling Reference Catling, Cavanagh, Crouwel, Catling and Shipley1996, 66 fig. 14.11:4. No disc feet were recorded in material published in Coldstream and Huxley Reference Coldstream and Huxley1972.

Laconian open vases

This group contains some thirty body sherds, with roughly equal numbers from larger and smaller pieces. One is red-figured, preserving part of an arm and another object; it has dull, streaky glaze inside, perhaps therefore coming from a mug, and uses relief line (McPhee Reference McPhee1986, 155). Another body sherd from a large vase has been sampled, no. 98/259.

There are eight lips of cups, all more or less everted (one may be Attic); there is also a fairly early cup with a light offset at the bottom of the lip outside. A horizontal bowl rim has two grooves on top with a reserved band between.

Four ring-feet are of substantial diameter; underneath, two are glazed, while another has miltos and at least one glaze ring, in imitation of Attic; the fourth has no floor preserved. One tall ring-foot with light groove on the stand surface, diameter 8 cm, has dull, dark glaze, and two incised rings on the floor; it is probably Laconian. There is also a krater foot with miltos on a reserved band at the join of foot and body.

There are some ten flat or disc feet; two are flat cup bases (another has the slightest ring-foot) and two larger flat feet are 6 cm and 11 cm in diameter.

The rest include one skyphos foot, diameter 8 cm, the full profile of a small bowl, one Attic or imitation Attic floor, foot lost, with carefully painted rings and band under, and handles, as noted above (under ‘Attic’).

  • G 12) Cup rim with graffito ]T[ . Fig. 10a .

  • G 13) Body, rim and part of vertical handle of Laconian kantharos. Wholly glazed. Maximum preserved dimension 5.1 cm. Near miniature in size. It is possible that this could be Argive. Fig. 10c .

  • G 14) Handle of cup, seemingly Laconian, with ‘advanced’ shape, akin to later fourth century skyphoi. Maximum preserved dimension 4.4 cm. Fig. 10a–b .

Fig. 10.

(a) G 12 and G 14; (b) G 14; (c) G 13; (d) G 15; (e) G 16.

Laconian closed

A restricted number of sherds, mostly fairly thick body fragments (16), many from oinochoai. Of the latter, two are neck/shoulder pieces, one with a double fillet at the turn, the other with probably a single fillet above a row of single impressed dots. One fragment is from a slim, lekythos-like body. One big flat foot is composed of two joining fragments.

There are a variety of handles: from larger jugs, one flattened 3.5 × 1.9 cm, one ridged 3.3 × 1.3 cm, one grooved, one sub-round; one from a smaller piece, 1.8 × 0.7 cm.

  • G 15) Two joining, heavily angled, shoulder to neck fragments, maximum preserved dimension 13.5 cm, appear to be from a stamnos. Fig. 10d .

  • G 16) A curving rim of a hydria or amphora may perhaps be Cretan; fine light brown with some white inclusions. Glazed save for a band near the break on the inside. Maximum preserved dimension 7.6 cm. Fig. 10e .

Amphorae

Not published, but retained, were five feet, two rims and six handles; of the last, one is Chian, 4.5 × 2.9 cm, with red paint traces at the top, and one probably North Greek, 4.2 × 2.8 cm, brown-buff, with much small mica and a red stripe across the neck join.

  • G 17) Foot, diameter 6 cm, maximum preserved dimension 5 cm. Pink-buff clay with little mica. The rather shallow floor indicates a full body. Fig. 11a .

  • G 18) Foot, diameter 5.8 cm, maximum preserved dimension 6.2 cm. Pink-buff clay with few inclusions. Only a slight depression under foot. Hollow stem. Fig. 11b .

  • G 19) Foot, diameter 4.1 cm, maximum preserved dimension 4.9 cm. Pink-buff clay with little mica. Fig. 11c .

Fig. 11.

Amphorae: (a) G 17; (b) G 18; (c) G 19; (d) G 20; (e) G 21; (f) G 22.

Two other feet are similar to G 17 (Fig. 11a ), diameters 6.6 cm and 7.0 cm.

  • G 20) Rim and part of handle. Maximum preserved dimension 5.8 cm. Slight thickened triangular lip. Handle 4.1 × 2.2 cm. Brown to purplish brown with some inclusions and mica. North Greek. Fig. 11d .

  • G 21) Rim. Maximum preserved dimension 5.8 cm. Height 1.5 cm. Brownish clay, with small inclusions. Groove on upper neck. Some remains of a red dipinto. Fig. 11e .

  • G 22) Handle, 4.3 × 2.4 cm, with central rib. Brick red with much small mica. Fig. 11f . Related are two handles from the late Archaic levels at the temple of Aphaea on Aegina (Johnston Reference Johnston1990, 53 nos. 132–3).

Cooking pots

  • G 23) Rim of small jug; micaceous light brown, burnished. Maximum preserved dimension 4.4 cm. Fig. 12a .

  • G 24) Rim and handle; micaceous light yellow-brown; height of lip 1.7 cm; near-vertical handle. Rather slight ledge for lid. Fig. 12b . Sample no. 10/3.

  • G 25) Chytra rim and body; handle well out of vertical. Micaceous brown. Maximum preserved dimension 12.5 cm. Fig. 12c . Sample no. 10/2.

  • G 26) Rim and wall fragment; hard-fired orange fabric. Diameter c.18 cm. Very thin wall. Fig. 12d .

Fig. 12.

Cooking pots: (a) G 23; (b) G 24; (c) G 25; (d) G 26.

There are two lid fragments in non-micaceous clay, and two in micaceous clay.

Most rim fragments curve, as σ 78–9, but one is fully straight. A bowl, preserved height 5.2 cm, diameter c.24 cm, has a plain ledge rim, 1.1 cm wide; it is in non-micaceous cooking ware.

Silver mica cooking pots are represented by three types of handle – vertical round, vertical grooved strap, and heavy, round, attached to lip. A slightly concave strap handle, 3.1 × 1.3 cm, has a shallow finger mark at the join with the body. One nearly flat lug handle may be from a tray.

Several body sherds have also been sampled.

Plain and domestic

A variety of shapes and fabrics are represented in this section, which reflects both the category as published in Coldstream and Huxley Reference Coldstream and Huxley1972 and (scarcely unconnected with that) the broad range of ‘semi-decorated utilitarian’ ware in circulation in the period.

  • • Substantial lower wall fragment of a large glazed amphora; regularly and heavily grooved inside.

  • • Oinochoe shoulder with spring of strap handle, 2.9 × 1 cm; rather steep shoulder; worn glaze, down to handle level, inside.

  • • Jug rim, diameter of neck c.6 cm; everted cut-off rim; worn glaze.

  • • Plain light-make bowl with ledge rim, 1.1 cm wide; diameter c.12 cm; wall 0.3–0.4 cm thick. Pink-buff clay with white inclusions. This represents a simple bowl type, seen widely in the KIP survey.

  • G 27) Lidded basin, maximum preserved dimension 6.8 cm. Pink-buff with some mica. Red-chocolate glaze, scumbled on the outside. Fig. 13a .

  • G 28) Basin. Maximum preserved dimension 12.3 cm. Salmon clay. Red-brown paint inside; a dribble over the rim outside. Fig. 13c .

  • G 29) Bowl? Maximum preserved dimension 13.5 cm. Red-brown paint inside, on top of rim and, sloppily, at top of wall. More likely a heavy duty bowl than a pedestal. Fig. 13d .

  • G 30) Bowl. Diameter c.23 cm. Two grooves near rim; nipple (rather than worn handle stump) to right. Buff clay. Traces of dull glaze inside. Corinthian? Fig. 13b .

  • G 31) Hydria or amphora, lip lost. Maximum preserved dimension 12.5 cm, neck diameter c.18 cm. Sharp shoulder angle; narrow plastic band on neck; wheel-ridging inside. Light buff clay with much small mica. Painted inside only. Fig. 14a .

  • G 32) Neck, shoulder and handle root of small hydria or oinochoe. Diameter c.17 cm. Fine buff clay. Band at turn of neck and three lines below handle; outside of handle painted. Laconian? Fig. 14b .

  • G 33) Flaring, but comparatively slight, ring-base of largish closed vase; fine clay, light grey in core. Diameter 6.5 cm. Reserved as far as preserved above a glaze band at the foot. Glazed under foot.

  • G 34) Very micaceous fine light brown; shoulder fragment, flattish with trace of handle attachment at top. Maximum preserved dimension 9.1 cm. Banded with part of handle swag. Cycladic or East Greek hydria – probably still 6th century. Fig. 14d .

  • G 35) One very difficult fragment because of curvatures. Maximum preserved dimension 7.8 cm; 1 to 1.1 cm thick; sui generis if a tile; least worst if taken as a rim, rather than a lid, though it is illustrated in Fig. 14 in another, different way. Painted under and on edge; top accreted. Fig. 14c and d .

  • G 36) One pithos wall fragment horizontally rilled. Maximum preserved dimension 12 cm; 2 cm thick. Rather small angular mudstone temper. Dull, dark glaze inside and out. Fig. 14e . Sample no. 10/4.

  • G 37) Wall and bit of disc foot of small closed vase. Wheel-ridged inside, shallow grooves out. Diameter foot 6 cm. Much small silver mica in red-brown clay with buff-beige surface; disc irregularly potted.

Fig. 13.

(a) G 27; (b) G 30; (c) G 28; (d) G 29.

Lekanai and mortars duplicate the types already published. Of the latter there are three rims and three feet, some included in material sampled for clay analysis:

  • G 38) Rim of lekane, similar to σ 60 (Coldstream and Huxley Reference Coldstream and Huxley1972, 164, fig. 49:60 [= S8:60]). Maximum preserved dimension 15 cm. Orange-buff with small white inclusions. Sample no. 98/261. Fig. 14f .

  • G 39) Floor of mortar. Maximum preserved dimension 10 cm. Sample no. 98/264.

  • G 40) Foot of mortar, maximum preserved dimension 12.5 cm, diameter c.20 cm. Painted out. Rather thinly gritted in, but worn; the floor is (now) rather thin for a mortar, 0.7 cm. Sample no. 98/260.

Fig. 14.

(a) G 31; (b) G 32; (c) G 35; (d) G 35 and G 34; (e) G 36; (f) G 38.

Commentary

While this fuller presentation does not alter the overall picture of ceramic usage at Kastri in the first millennium bc, it does offer some nuances which may be stressed here. In the introduction to this section I used the phrase ‘pottery imports’; that phrase should be set against the fact that it is not yet demonstrable that any pottery was being produced on the island during this period, though the KIP survey detected one or more kiln sites in the Mitata valley which were producing tiles at some point in the later Classical or Hellenistic period.

The context of deposit σ

The KIP survey catalogued 42 sherds from the square in which the deposit was located. Some tile fragments could belong to the fifth century, but no sherd is clearly of that period except perhaps for a body fragment of a lekane with dull, dark paint inside. Concentrations of Classical pottery were all picked up further inland than this coastal area. The conclusion in Coldstream and Huxley Reference Coldstream and Huxley1972, 73 that the deposit was a rubbish dump is therefore fully warranted. The very full range of material found, excluding any specifically votive material, points strongly to a domestic origin; lack of burning is perhaps not a sufficient reason to reject clean-up after a disaster; many pieces are preserved in only small fragments, and even if the pit was not fully excavated, it seems implausible that it is a primary dump. It is likely that pithoi were not found, save for the one unusual fragment, G 36 (Fig. 14e ), noted above.Footnote 3

Chronology

A few sherds of the eighth to seventh century were found in the course of the excavation and published in ‘deposit’ ω. They are indeed few, and the results of the survey reinforce the view that presence on the site was sporadic and thin, or both. Recent work on the secure site of Palaiokastro, rising high behind Kastri, has demonstrated more significant levels of this earlier period there.

Deposit σ has some material of a date substantially earlier than the terminal date noted below. σ 7 [S9], σ 22 [S2, S7, S9] and G 44 were singled out above, while one Laconian krater rim (σ 46: Coldstream and Huxley Reference Coldstream and Huxley1972, 163 [= S4, S10]) is probably of the earlier fifth century, and also well before the end of the fifth century are a Chian amphora, σ 68 (Coldstream and Huxley Reference Coldstream and Huxley1972, 164–5, fig. 49:68 [= S8:68]), and G 22 (Fig. 11f ).

As for the terminal date of σ, it was placed in the late fifth century in Coldstream and Huxley Reference Coldstream and Huxley1972, 73. As noted above, there is some considerably earlier material in the deposit, but it is clear that there are also sherds of the fourth century, even if in small quantity. For most of the plainer wares no solid conclusions can be drawn, but four black-glazed pots would appear to date after 400 bc, possibly by a generation or more. σ 15 [S1, S7, S9] is a skyphos of perhaps 400–375, akin to Sparkes and Talcott Reference Sparkes and Talcott1970 no. 349; G 14 (Fig. 10b ) is sui generis, but the form of handle is of the fourth century; G 6 (Fig. 8a ) is very clearly no earlier than c.375, and a second piece noted above (under ‘Laconian open vases’, third paragraph) with a grooved underside to the stand-ring of the foot should also be of the fourth century. It is more difficult to place the amphora ω 322 (Fig. 5c ), which has a fourth century feel to it, and the scraps of Laconian red-figured pots. The KIP survey has not produced parallel fourth century pieces from Kastri, where material belonging to the broad ‘later Classical–early Hellenistic’ period is largely confined to banded lekanai or the like and amphorae, a very few of which appear to date between c.350 and 200.

Coldstream and Huxley wisely refrained from relating the pottery from σ to political events of the fifth and fourth century; objective dating of the finer wares could just be set against the historical record, but the great majority of material cannot be said to be a sufficiently fine dating tool. The terminal date for the deposit suggested above would not readily fit any known historical episode; one could argue that the King's Peace in 387 bc, in which Sparta strengthened its grip, might have resulted in some displacement of inhabitants, while any local events following the battle between Boeotians and Spartans at Leuktra in 371 bc are likely to be too late with respect to the lowest possible dating of the material.

Area of import

A large percentage of the material retained consists of domestic pottery – cooking, storage and service. While it cannot be excluded that in such a secondary deposit some sanctuary material could infiltrate, it is generally assumed that the two did not mix; therefore all the material in the deposit comes from domestic contexts. The increasing amount of figured pottery – a little black-figure, then rather more red-figure – could of course be purely a reflection of the closing date of the fill, but does also seem to reflect a general trend in the use of Attic red-figured pots in the home in the later fifth century, not to mention the vast amount of high quality black-glazed ware, whether Attic or imitation, being consumed during the period.

As regards other wares, Laconian covers a large percentage of black-glazed and semi-decorated wares, with Corinthian also present in good amounts in the latter category. Corinthian transport amphorae, however, are not present, and only two possible fragments of them come from the survey of the site. The amphorae from σ, where identifiable, are from North Greece and Ionia.

2. The Roman Deposits τ, υ, φ, and χ (Kathleen Slane)

The Roman pottery reported in Coldstream and Huxley Reference Coldstream and Huxley1972, 166–75 exists today as a study collection that also contains unpublished sherds labelled with their findspots.Footnote 4 From the excavation records it is clear that only diagnostic sherds were retained, and only a selection was published. Deposit τ was dated to the third century ad, and the remaining deposits were assigned to the sixth–early seventh century. This material is very important. Questions centred on the seventh century, concerning the twilight of the Roman economy and society, the end of late antiquity and the beginning of the Middle Ages, and the breakdown of Mediterranean-wide trade routes are intensely debated today; as imports disappear from the archaeological record and a monetary economy with them, even the chronology remains in flux. Kastri is one of only a handful of eastern Mediterranean sites where such material was found stratified, identified and published. Nevertheless, it is rarely cited, except by John Hayes, and one purpose of this article is to make it more readily available to a wide audience. The material listed below serves to emphasise the importance of this late sixth–seventh century phase, found both on the Akroterion and in the trenches on the narrow saddle to its north between Kastri and Kastraki (Neck). Furthermore, while much of the unpublished pottery duplicates that published in Coldstream and Huxley Reference Coldstream and Huxley1972, some of it is Early Roman (a phase that was missing from the publication) or Middle Roman, third or fourth century.

This pottery also merits re-examination because, in the half century intervening since the excavations, the identification and dating of Roman pottery have been considerably refined and have become much more secure.Footnote 5 Hayes types rather than Antioch forms are now standard for fine wares, and amphorae and lamps can be much more precisely placed; in most cases the fabrics are now so well known that detailed description is superfluous. Some drawings have been improved and some sherds are being analysed. As a result it seems best to republish deposits τ to χ completely, adding the unpublished material, as well as some additional context pottery stored with ω. The order of presentation within each unit is fine wares, early to late; lamps; amphorae; cooking pots; and plain wares. Rather than assigning new catalogue numbers to material already published, I have simply listed the published material with additional references, interpolating a list of unpublished pieces. I have also investigated the possibility of subdividing the Late Roman strata into an earlier and a later phase.

Almost all the Roman pottery is imported. There are also a few imitations of imported fine wares, one or two of cooking pots (but one imitation is an import), and apparently some imitations of imported amphorae. The basins identified both here and in the KIP survey are mostly of one fabric and therefore are likely to be local or regional Laconian or Cretan products.Footnote 6

All of the pieces numbered in Coldstream and Huxley Reference Coldstream and Huxley1972, 166–75 [= S12–S21] were illustrated both by profiles and by photographs in the original report. Pieces marked * have new drawings. Illustrations from Coldstream and Huxley Reference Coldstream and Huxley1972, figs. 50–2, pls. 47–9 [= S22–S27], are available to online readers in the Appendix.

Second half of the third century: Deposit τ

Deposit τ was recovered from Neck Trench II (Coldstream and Huxley Reference Coldstream and Huxley1972, 69–70, fig. 31, layers 5 and 7) and comes from both sides of a Roman wall α. It is the earlier of two Roman phases. The third century date originally suggested is still a reasonable date for the five published pieces. A Çandarli ware basin marked NK 2.4 (υ 7) comes from a level below deposit υ north of wall ε and probably belongs to the same phase as τ. The published description also refers to several fragments (ω 328–9, i.e. African Red Slip Hayes [1972] forms 48–50) that would date to the second half of the third century and might suggest a date late in the century.Footnote 7 τ 1, τ 3, the casserole in bag 4, and a small casserole rim from AKR VIIA/B.2 are all of the same imported cooking fabric, while τ 4 is an imitation.

In addition to υ 7 and ω 328–9, 20 vessels are in a bag labelled deposit τ.Footnote 8 Two are nearly complete, τ 4 and τ 5; the others were divided among several bags.Footnote 9 None of the four fine-ware sherds in one bag (unfortunately without trench markings) is datable as early as the third century: they belong to the Late Roman phase of the site and are very likely from deposit υ, the upper level in the Neck trenches. Two imitations of North African lamps (marked NK 3.4) in another bag are certainly from that deposit. Accordingly, all six pieces are listed below with υ rather than here with τ.

Published material

  • υ 7 [S13], Çandarli rim, basin of Hayes form 1, first half of 3rd century? Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:7, pl. 47 [= S22:7, S25]. Marked NK 2.4.

  • τ 2 [S12, S25], keel-rim bowl like Robinson Reference Robinson1959, 62 K 29, pl. 12, probably Attic, first two-thirds of 3rd century; Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:2 [= S22:2].

    Additional fine wares, unpublished: keel-rim bowl, unidentified ware, probably mid 3rd century; carinated bowl, unidentified ware, 2nd or 3rd century; basin with grooved rim, slip-coated on the interior, 1st–2nd century.

  • τ 1 [S12], rim, non-joining base, and two other sherds of a thin-walled, globular mug, 2nd–mid 3rd century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:1, pl. 47 [= S22:1, S25]. Marked NK 2.5.

  • τ 3 [S12], stewpot or casserole rim, imported, later 2nd–mid 3rd century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:3, pl. 47 [= S22:3, S25]. Marked NK 2.3.

    Additional example, unpublished: casserole, probably 3rd century. The fabric is corrugated cooking fabric, the same as τ 1, τ 3, and the small casserole rim marked AKR 7A/B.2.

  • *τ 4 [S12], imitation of a corrugated cooking ware stewpot, nearly complete, late 2nd–late third century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:4 [= S22:4],Footnote 10 pl. 47 [= S25]. Marked 3318. Fig. 15a . Sample no. 10/26.

  • *τ 5 [S12], plain-ware basin with broad disc-base, vertical wall and out-turned, slightly drooping rim, probably local, 3rd–4th century?, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:5 [= S22:5],Footnote 11 pl. 47 [= S25]. This type was not slipped. Fig. 15b . Sample no. 10/31. Cf. Lawson Reference Lawson1996, 32 (ST 123) and/or 34b (ST 21), 115, fig. 16.3:3 and 8, dated Late Roman and undated, respectively.

Fig. 15.

(a) τ 4; (b) τ 5; (c) υ 4; (d) υ 15.

Unpublished lamps

  • Unglazed mould-made lamp with rays on shoulder, probably Hellenistic. Marked SF 366.

    Handle of an unglazed Corinthian lamp (Broneer Reference Broneer1930 type XXVII), second century ad. Marked SF 337.

Late sixth- and seventh-century phase across the whole site: Deposits υ, φ and χ

Deposit υ, mostly late sixth–early seventh century, is the latest phase on the Neck, described as coming from a fairly substantial building (γδζ) in Trench II and from Trench IV, associated with wall α; lamps labelled NK 3.3 and 3.4 must also be from the same levels. The material is as late as deposit φ on Kastri, but deposit υ contains relatively more sherds of the earlier sixth century, including some datable to the years before and after 500. In addition to pottery, five bronze coins were published from this deposit (υ 18–22: Coldstream and Huxley Reference Coldstream and Huxley1972, 211), of which one is early fourth century and the remainder were dated between 580 and 615. There were also two glass drinking cups (υ 28–29) and a third foot (υ 27) (Coldstream and Huxley Reference Coldstream and Huxley1972, 212).

Nineteen vessels are labelled as coming from these levels, 17 of which were published, and six additional, unpublished sherds were stored with material from other units. Of the 23 total, υ 4 (listed with Late Roman C in Coldstream and Huxley Reference Coldstream and Huxley1972, 167 [= S13], but noted as probably African) can be confirmed as African Red Slip. υ 7 was published as African Red Slip (Coldstream and Huxley Reference Coldstream and Huxley1972, 167 [= S13]), but it is Çandarli and comes from a stratum underlying deposit υ, at the same level as deposit τ but not physically connected with it. Considering their likely dates, it is possible that the flanged pan and the stamped beehive, υ 15 and υ 14 respectively (Coldstream and Huxley Reference Coldstream and Huxley1972, 168 [= S14, S22, S25]), also belong with deposit τ.

Published material

  • υ 8 [S13], African Red Slip form 67 rim, 350–430,Footnote 12 Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:8, pl. 47 [= S22:8, S25].

  • υ 11 [S13], African Red Slip form 99A rim, diameter 0.19–0.20 m, first half of 6th century,Footnote 13 Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 47 [= S25].

  • υ 12 [S13], African Red Slip form 99A rim, diameter 0.19–0.20 m, first half of 6th century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:12, pl. 47 [= S22:12, S25].

  • *υ 4 [S13], base of African Red Slip form 99A–B, diameter 0.08–0.10 m, 6th century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:4, pl. 47 [= S22:4, S25]. Fig. 15c .

  • Additional unpublished bowl (bag 1): African Red Slip form 91C, estimated diameter 0.18 m, 530–600 + , although this date has been much disputed.Footnote 14 Marked NK 3.4.

  • υ 9 [S13], rim of African Red Slip form 104B–C, no grooves and not recurved, 550–625,Footnote 15 Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:9, pl. 47 [= S22:9, S25]. Cf. Lawson Reference Lawson1996, no. 2 (ST 114), 112, fig. 16.1:5.

  • υ 13 [S13], foot of African Red Slip form 105/106, with chamfer on floor, late 6th but mainly 7th century,Footnote 16 Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:13, pl. 47 [=S22:13, S25].

  • υ 10 [S13], thin and gritty African Red Slip rim of form 105, half-round outside and flat on the interior, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:10, pl. 47 [=S22:10, S25]. Marked NK 2.6.

  • υ 3 [S13], rim of Late Roman C Hayes 3B, mid 5th century,Footnote 17 Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:3, pl. 47 [=S22:3, S25].

  • υ 1, υ 2 and υ 5 [S13] represent only two vessels, rims of Late Roman C Hayes 3E–3G, first half of the 6th century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:1–2,5, pl. 47 [=S22:1–2,5, S25].

  • υ 6 [S13], rim of Late Roman C 10C, mid 7th century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:6, pl. 47 [= S22:6, S25].

  • Additional fine wares, unpublished (with deposit τ): African Red Slip 103, foot, 6th century; African Red Slip 104/105, foot, 6th–7th century; Late Roman C form 3, 450–550; Cypriot Red Slip form 9, 550–650 or later.

  • υ 16 [S14], rim of Late Roman Amphora 2, Argolid fabric?, 5th–6th century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:16, pl. 47 [= S22:16, S25].

  • *υ 15 [S14], rim of wheel-ridged, flanged pan, micaceous cooking fabric, early 4th century (cf. Slane Reference Slane1990, 79 no. 167, fig. 16), Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50:15, pl. 47 [= S22:15, S25]. Fig. 15d . Sample no. 10/17.

  • υ 14 [S14], lower wall and edge of base of a beehive, hole pierced through wall before firing and stamp above hole, same fabric as the basin τ 5, probably 2nd–4th century, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 47 (upside down) [=S25].

  • υ 17 [S14], complete stopper for jug or amphora with basket handle, apparently handmade, soft calcareous fabric, date unknown, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 47 [= S25].

Unpublished lamps

  • R 1. Handle and shoulder of an African Red Slip lamp. Edge of the discus relief visible below the handle knob; small segmented motif on shoulder, perhaps a volute-diamond or quatrefoil. The discus is slightly pointed where it meets the handle and the outer profile is also oval rather than following the circular line of the shoulder. This may be Bonifay Reference Bonifay2004, 408–10 lamp type 67.13 (characterised by its oval discus), of the late 6th and early 7th century from northern Tunisia, or it could be the earlier type 55.18 (Bonifay Reference Bonifay2004, 382–6), of the second half of the 5th century from central Tunisia. Marked NK 3.3. Fig. 16a .

  • R 2. Red-slipped imitation of a North African lamp with rounded shoulder, 4th–5th century, cf. φ 33 (below). Marked NK 3.4.Footnote 18 Fig. 16b .

  • R 3. Nozzle of an imitation of a North African lamp, 6th century. Marked NK 3.4.Footnote 19 Fig. 16c .

Fig. 16.

Lamps: (a) R 1; (b) R 2; (c) R 3. Stamps: (d) φ 12; (e) R 20a.

Deposit φ

Deposit φ was the name given to all the ‘early Byzantine’ levels in the trenches on the Akroterion, including those from Trench VI inside the fortification wall (Coldstream and Huxley Reference Coldstream and Huxley1972, 55, 64–6; for the walls see fig. 28). From the sections (Coldstream and Huxley Reference Coldstream and Huxley1972, figs. 13, 18–19, 23–5), it appears that deposit φ was the level below topsoil (2) in the adjoining Trenches I, II, VIIA and VIIB, covering walls κ and μ; the underlying layer 3, filled with tumbled stones and immediately overlying the Late Minoan floor deposit λ, is also labelled deposit φ. It seems to have been found only west of the modern terrace wall that formed the west scarp of Trench VIII, over the two western houses identified in Trenches VII and VIII. The stratigraphic reports specifically mention the following levels: VII C 2 (covering wall σ) and the floor VII C 11, possibly contemporary with each other and earlier than the other buildings; VII A 4 (the robbing trench of wall κ) and VII A/B 4 (the pottery from wall ν). φ 72 [S19, S24, S27] comes from an oval pit in Trench II layer 2. Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 16 shows that in Trench VI, within the walls of the fort, levels 2 and 3 (labelled deposit φ) are the destruction debris of the heavy wall α, with coin ω 353 found on a stone paving. ω 365, a plastic lamp, shows that VI 4 and 5, the levels cut by wall α, were also of Late Roman date.Footnote 20 The ‘deposit’ contained three bronze coins of the last quarter of the sixth century (φ 74–76) and a glass lid (φ 82) (Coldstream and Huxley Reference Coldstream and Huxley1972, 212–3).

Interesting pieces earlier than the bulk of the material

  • φ 29 [S17], black-glazed(?) skyphos or kantharos handle, 1st century bc–1st century ad, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 48 [= S26]. The vertical ring of the handle is missing but the preserved thumb plate has a mould-made branch. Such mould-made thumb plates appear on Pergamene pottery in the later 2nd and 1st centuries bc and are common on lead-glazed skyphoi in the Augustan period and 1st century ad.Footnote 21 This seems likely to be a local or regional product.

  • φ 26 [S16], rim of a Cypriot sigillata rouletted crater, Hayes form P40, end of 1st–mid 2nd century,Footnote 22 Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26].

  • φ 28 [S17], Knidian bowl or patera fragment with inward-thickened rim and mould-made tongue pattern below rim on interior,Footnote 23 2nd century, perhaps a few years earlier or later, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26].

  • φ 7 [S15], Attic bowl foot, with two grooves on the floor, completely slipped, 2nd–3rd century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50, pl. 47 [= S22, S25]. This is either a keel-rim bowl like τ 2 or (more likely) an earlier shape. Marked 8B2.

  • *φ 9 [S15], large fragment of a casserole, African Red Slip (African black-top) Hayes form 23B, estimated diameter 0.325 m, end 2nd–mid 3rd century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 47 [= S23, S25]. Marked 7a.2. Fig. 17a .

  • φ 71 [S19], trefoil mouth and neck of a cooking ware pitcher, 2nd–3rd century, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27]. Cf. Robinson Reference Robinson1959, pls. 14:K 106, 23:M 101; Hayes Reference Hayes1983, 106–7 nos. 76–7, 122, fig. 6.

  • Additional early pieces, unpublished: Three joining fragments of rim of a local red-slipped hemispherical bowl, diameter 0.26 m. Two non-joining fragments of a rim of Çandarli Hayes form 2, 2nd–3rd century. Marked 5.5. Base of large bowl Çandarli Hayes form 3 or a basin with curving wall, estimated diameter 0.12 m; the narrow foot is relatively early, 2nd century. Marked 7/8.1.

Fig. 17.

Fine wares: (a) φ 9; (b) φ 24; (c) R 5; (d) R 4; (e) φ 25; (f) R 6; (g) R 7; (h) φ 13; (i) φ 12.

Late Roman material

  • Three rims of African Red Slip Hayes form 67, 360–450+. *φ 24 [S16] is Bonifay's 67C, dated second half of the 5th century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51:24, pl. 48 [= S23:24, S26]. The other two rims are more delicately formed and probably earlier. Fig. 17b . Sample no. 10/28.

  • Three rims of African Red Slip Hayes form 91D, including φ 20 and φ 21 [S16], 600–650, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26].

  • Eight rims probably from different vessels of African Red Slip Hayes form 99, including φ 3 [S15], 510–620, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50, pl. 47[= S22, S25].

  • Three different bases of African Red Slip Hayes form 99, including φ 5 and φ 6 [S15], 510–620, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50, pl. 47 [=S22, S25]. φ 5 marked 7A4; another marked AKR 7D.2.

  • φ 4 [S15, S22, S25], rim of African Red Slip Hayes form 104A, 530–580.

  • *φ 25 [S16, S23, S26], rim of African Red Slip Hayes 105, 580/600–660; cf. Hayes form 105.3. Fig. 17e .

  • Two bases of African Red Slip Hayes form 105, including φ 8 [S15], 580/600–660 + , Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50, pl. 47 [= S22, S25].

  • Two rims of African Red Slip Hayes form 80B/99, including φ 1 and probably φ 2 [S14], 7th century?Footnote 24 Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 50, pl. 47 [= S22, S25].

  • φ 10 [S15], two non-joining rim fragments of African Red Slip Hayes form 31 (if diameter is 0.32 m) or 14–17,Footnote 25 Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 47 [= S23, S25]. Marked 7A4.

  • Additional African Red Slip, unpublished:

    Complete profile and a second rim of African Red Slip Hayes form 48B, 240–300/325.

    Rim of African Red Slip Hayes form 49, 240–300/325.

    Rim of African Red Slip Hayes form 45A with rouletting on upper surface, 240–300/325 (likely 3rd century). Marked II 3.5?

    Rim of African Red Slip Hayes form 45 C, 240–300/325 (likely later).

    Rim of African Red Slip Hayes form 59, 320–380/400.

    Rim of African Red Slip Hayes form 99, estimated diameter 0.20 m, 510–620.

    Rim of African Red Slip Hayes form 91C, 530–600 + .

    *R 4. Foot of African Red Slip Hayes form 93? or similar, slip on interior only, diameter 0.086 m. 6th century or later. Burned. Marked 8B2. Fig. 17d .

    *R 5. Rim of African Red Slip Hayes form 102, diameter 0.10 m, late 6th–7th century.Footnote 26 The form is rare. Fig. 17c .

    Rim of African Red Slip Hayes form 103, diameter 0.24–0.26 m, 500–575. Marked 8B2.

    Rim of African Red Slip Hayes form 104C, 550–625.

    Rim of African Red Slip Hayes form 104B, 570–600 + .

    *R 6. Rim of African Red Slip Hayes form 105?, 7th century. Fig. 17f .

    *R 7. Rim of African Red Slip Hayes form 107, diameter 0.30 m, 600–650. Fig. 17g .

  • φ 22 [S16], rim of Late Roman C Hayes form 3B, 450–500, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26]. Marked 835.

  • φ 17 [S16], Late Roman C Hayes form 3C rim, diameter 0.20 m, 450–500, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26].

  • φ 23 [S16], rim of Late Roman C Hayes form 3C, 450–500, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26]. Marked 7.A2.

  • φ 27 [S16], rim of Late Roman C Hayes form 3F with rouletting, diameter 0.30 m, 500–550, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26].

  • Three rims of Late Roman C Hayes form 10A, including φ 18 and φ 19 [S16], 570–600 + , Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26]. φ 19 is marked 5.5; another 7a.10.Footnote 27

  • *φ 13 + φ 15 [S15–S16], rim of Late Roman C Hayes form 10C, 600–660+. φ 13, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51 (drawn together with φ 14), pl. 48 [= S23, S26]. Marked VIIIB2. Fig. 17h . φ 13 is sample no. 10/29.

  • Fourteen rims of Late Roman C Hayes form 10C, probably seven large and three smaller vessels, 600–660 + , Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51 (drawn together with φ 16), pl. 48 [= S23, S26].

  • Eight different bases of Late Roman C dishes, including φ 14 (no stamp) and φ 16 (with stamp [S16], Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 48 [= S26]), 400–660+. The numbers of the latter are reversed from the published drawings, and the bases are not from the same vessels as the rims, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26].

  • *φ 12 [S15], Late Roman C new form with a row of stamps around the centre;Footnote 28 four joining and one non-joining fragments preserve about one-third of the vessel. The stamps around the floor (an ovolo?) belong to Hayes' Group II of the second half of the 5th century; they have the thick outline characteristic of Late Roman C. The offset on the interior wall, which resembles that on some late 5th and 6th century African forms, suggests that this is an imitation of African Red Slip. Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 47 [= S23, S25]. Marked 5.2. Fig. 17i and Fig. 16d .

  • φ 11 [S15], probably an imitation of Late Roman C Hayes form 1, complete profile, about one-quarter preserved; although not slipped, the outside of the rim is fired a lighter colour than the wall, 400/425–470, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 47 [= S23, S25]. Marked 5.5. Cf. Quercia et al. Reference Quercia, Johnston, Bevan, Conolly and Tsaravopoulos2011, 87 no. 88, fig. 11, which may be the same fabric.

  • Additional Late Roman C, unpublished:

    Rim of Late Roman C Hayes form 3D–F, diameter 0.32 m, 500–550.

    Two different rims of Late Roman C, Hayes form 3F, small, 500–550. Marked AKR 7/8.2.

  • φ 30 [S17], back half of an African Red Slip lamp Hayes form IIA with a jewelled cross in the discus and three alternating motifs on the rim, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26]. See Bonifay Reference Bonifay2004, 385–8, lamp type 56C or less likely 57C, 6th and 7th century, respectively.

  • φ 31 [S17], handle and rim of an African Red Slip lamp Hayes form IIa with a row of quatrefoils on the rim, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26]. Cf. Bonifay Reference Bonifay2004, 408–10 lamp type 67, dated end of the 6th and first half of the 7th century. Marked 431.

  • φ 32 [S17], pierced handle and rim of a lamp with a jewelled chi-rho monogram (rho reversed) in the discus and parallel twisting strands on the rim, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26]. While the lamp appears to be African, the pierced handle would be remarkable, nor do the twisting strands on the rim appear on African lamps, although there is a similar motif on the unglazed circular lamps of Broneer's (Reference Broneer1930) type XXXII. Marked 416.

  • φ 33 [S17], back half of an imitation of an Asia Minor lamp. The lamp was red-slipped, with raised palm branch (parallel crescents) on the shoulder, raised ovolos at intervals on the discus, and a ring around the filling hole, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26]. Marked 422.

  • φ 33a [S17], back half of an imitation of an African Red Slip lamp, either Corinthian or local, with a jewelled cross in discus, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26]. Cf. Slane and Sanders Reference Slane and Sanders2005, 266 nos. 3-1, 3-2, 282-3, fig. 10, end of the 6th century. Marked 711.

  • Additional unpublished lamp: Lamp with vine-and-ray(?) rim, solid grooved handle and plain discus, Attic postglazing, 5th century. Similar glazed lamps usually have rays on the discus, but the type with plain discus enjoyed a vogue in Athens in the second half of the 5th and perhaps the 6th century; see Perlzweig Reference Perlzweig1961, 184–5 nos. 2595–620, pl. 41; Karivieri Reference Karivieri1996, 56–7 nos. 121–4, 197–8, pl. 10; Böttger Reference Böttger2002, 78 nos. 4599–606 and table 12, 286, pls. 78–9 (giving a date after c.425). Marked 1.2.

Amphorae

  • *φ 47 [S18], Late Roman Amphora 1 rim, neck and handles, probably Syrian, early 5th–early 7th century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 52, pl. 49 [= S24, S27]. Marked 7/8.2. Fig. 18a . Sample Kastri no. 10/11.

  • *φ 46 [S18], Late Roman Amphora 1 rim and one handle, probably Syrian, early 5th–early 7th century, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27]. Tiny subangular black, grey, and orange (infrequent) inclusions in an orange body with orange surface and light-brown core. Marked VIIIB2. Fig. 18b . Sample Kastri no. 10/10.

  • φ 60 [S19], body sherd of Late Roman Amphora 1, same fabric and inclusions as φ 46 except that the exterior surface is cream coloured, early 5th–early 7th century, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27]. Marked VIIIB2. Sample Kastri no. 10/12.

  • *φ 54 [S18, S24, S27], see end of list. Fig. 18d . Sample Kastri no. 10/8.

  • *φ 55 [S18], Late Roman Amphora 2 neck, handles and shoulder, 5th–late 6th century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 52, pl. 49 [= S24, S27]. Fig. 18e . Sample Kastri no. 10/7.

  • φ 52 [S18], body sherd of Late Roman Amphora 2, Argolid fabric, 5th–late 6th century, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27].

  • φ 50 [S18], body sherd of Late Roman Amphora 2 with a graffito on the shoulder, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27]. It appears to be the same fabric as φ 52 and φ 53, but it is softer and lacks the lime inclusions pockmarking the interior.

  • *φ 48 [S18], amphora rim, Samos-cistern type, 6th century, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27]. Highly micaceous, light orange fabric with small dark and white inclusions. The rim, which is simply the end of the wall, is slightly thickened, the shoulder is hollow between the neck and the carination, and the handle is very roughly attached at both ends, distorting the line of the wall; wheel-ridging on shoulder. Fig. 18f .

  • *φ 57 [S19], rim and neck of a small African spatheion, 7th century,Footnote 29 Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 52, pl. 49 [= S24, S27]. Fired light red with a white surface. Three handle scars preserved. Fig. 19a . Sample Kastri no. 10/5.

  • *φ 58 [S19], rim and neck of a second small African spatheion, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27]. Scars for upper attachments of handles; regular ‘drips’ of clay at widest part of rim. Fired yellow all the way through. Marked VIIIB2. Fig. 19c . Sample Kastri no. 10/6.

  • φ 62 [S19], toe of a third small African spatheion, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27]. Marked VIIIB2.

  • φ 63 [S19], toe of small African spatheion, Bonifay Reference Bonifay2004 variant 3D. Marked VIIIB2.

  • *φ 64 [S19], most of a reconstructed small African spatheion, same fabric as the others, with vertical paring on the body, Bonifay Reference Bonifay2004 variant 3C, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27]. Fig. 19b . Sample Kastri no. 10/34.

  • φ 61 [S19, S27], toe of a small amphora said to be like φ 62. The orange fabric appears to be African with a buff surface but the toe is broader and the body more ovoid than the other spatheia.

  • *φ 72 [S19, S27], many joining fragments of the upper body, neck, rim and handles of a large African amphora, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 52 [= S24]. Two narrow bands of combing on the neck and two on the shoulder. Fig. 20. Cf. Bonifay Reference Bonifay2004, 135–7 amphore type 42, second half of the 5th century; the bands of combed decoration seem to be characteristic. It comes from a pit in Akroterion Trench II, level 2 (Coldstream and Huxley Reference Coldstream and Huxley1972, 55).

  • φ 49 [S18], amphora with tightly curved, deeply ridged shoulder like a Gaza amphora, but the fabric is very fine, very hard, brown, with mica and white inclusions, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27].

  • φ 51 [S18], fragment of flat shoulder of an amphora; very fine, micaceous brownish-grey ware, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27]. Marked 7/8 2.

  • φ 59 [S19, S27], amphora shoulder with spaced incised grooves. Thick, crumbly, light-brown fabric with moderate tiny lime. Probably 7th century? Marking no longer legible.

Fig. 18.

Amphorae: (a) φ 47; (b) φ 46; (c) R 21; (d) φ 54; (e) φ 55; (f) φ 48.

Fig. 19.

African spatheia: (a) φ 57; (b) φ 64; (c) φ 58.

Fig. 20.

African amphora, φ 72.

Cooking pots

  • φ 65 [S19], rim and handle of a casserole, Late Roman micaceous Aegean cooking ware, diameter 0.180 m, mid 5th–early 6th century or later (the offset is unusually smooth), Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 52 (with φ 66), pl. 49 [= S24, S27]. Marked VIIIB2. See Slane Reference Slane2000, 309–10 with fig. 12 for the types and distribution of this ware; cf. Slane and Sanders Reference Slane and Sanders2005, 255 1-28 and 264 2-35, figs. 3, 6; in Slane and Sanders Reference Slane and Sanders2005, 287 I argued that it is probably a product of the Asia Minor coast south of Pergamon.

  • *φ 66 [S19], rim and upper body of a casserole (probably the same vessel as φ 65), Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 52, pl. 49 [= S24, S27]. Marked VIIIB2. Fig. 21a . Sample no. 10/21.

  • φ 67 [S19], rim and handle of cooking pot with short neck and oblique rim; lip turned up and with a groove on the outer face; handle, triangular in section, has a finger smear upward onto the rim, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 52, pl. 49 [= S24, S27]. All four handles of this type from φ have the finger smear from the handle onto the rim. Cf. Yangaki Reference Yangaki2005, 173 type XI.11, figs. 33b, 38e, rare in these excavations but more common in others, and characterised by a slight groove inside the lip and by two vertical handles, oval in section. The ware is her fabric 9, described as local. Possibly also Ricci Reference Ricci and Saguì1998, fig. 3.1, second half of the 7th century. Marked VIIIB2.

  • φ 68 [S19], cooking pot with low neck and oblique rim; upward-turned lip with groove on outer face (same shape as φ 67), Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27]. Coarse sandy, light orangey-brown fabric with abundant angular white inclusions and a matt black surface on the exterior. This is definitely not the fabric of φ 65–φ 66 because it is not micaceous and has much more angular inclusions.

  • φ 69 [S19], cooking pot with half-round moulding to support a lid on the interior of the rim, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27]. Saraçhane type B stewpot (and not the same fabric as φ 65–φ 66 or φ 67–φ 68). See Hayes Reference Hayes and Yon1980, 378 with figs. 7–9.1 (a map noting the Kythera examples as well as many from Cyprus and one from the mouth of the Danube), who identifies this as the local type of Constantinople in the 7th and 8th century; see also Martin Reference Martin, Di Vita and Martin1997, 355–6 no. 42, pl. CXXVIII:4, citing as well those from the 7th century Yassi Ada shipwreck and from Naples in the 6th to 7th centuries.

  • *φ 70 [S19], oblique rim of cooking pot with half-round moulding to support a lid on the interior of the rim, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 52, pl. 49 [= S24, S27]. Saraçhane type B stewpot. Fig. 21d . Sample no. 10/22.

  • Additional unpublished cooking pots:

    *R 8. Everted rim with two grooves near outer edge on top. Micaceous brown with rounded to subrounded quartz(?) and tiny round voids; 4th century? Fig. 21b . Sample no. 10/15.

    *R 9. Baggy stewpot with oblique rim, thickened at outer edge and one or two finishing grooves on top. Reddish-brown fabric with moderate angular quartz and feldspar and tiny white inclusions; surface fired grey. Fig. 21g . Sample no. 10/16.

    *R 10. Stewpot with low neck and out-turned rim with broad seat for lid on top, 4th–5th century? Very fine, very hard fabric, black with a tinge of dark reddish brown; no visible inclusions in rim, a few angular quartz or feldspar at shoulder. Fig. 21c . Sample no. 10/20.

    A thin-walled body sherd with curving bottom and offset vertical wall. Micaceous cooking fabric similar to the Istanbul pieces but much thinner and slightly finer.

    Two rims of another casserole in Late Roman micaceous Aegean fabric, diameter 0.16 m.

    *R 11. Four rims of stewpots similar to φ 67–φ 68 but with a more angular carination on the interior so the neck is very short or non-existent. Could these be earlier? One marked AKR 7/8.2. Fig. 21h .

    *R 12. Coarse sandy stewpot with tall neck and oblique rim; upward-turned lip with groove on outer face. Same type as φ 67–68. Hard-fired, very gritty light-grey fabric with moderate tiny rounded red grits and abundant angular chert and quartz; surface burned but not particularly black. Fig. 21i . Sample Kastri no. 10/19.

    *R 13. Another similar piece but the neck is shorter; handle, triangular in section, attached to rim. Fig. 21j . Sample no. 10/18. There are two additional rims of this type from φ.

    *R 14. Saraçhane type B stewpot rim with half-round moulding on inside (like φ 69 and φ 70). Muddy brown fabric with moderate tiny white inclusions and possible quartz; surface fired matt black. Fig. 21e . Sample no. 10/23.

    *R 15. Another similar rim in the same fabric. Fig. 21f . Sample no. 10/24. There is a third unpublished example.

Fig. 21.

Cooking pots: (a) φ 66; (b) R 8; (c) R 10; (d) φ 70; (e) R 14; (f) R 15; (g) R 9; (h) R 11; (i) R 12; (j) R 13; (k) χ 10; (l) φ 34.

Plain and domestic wares

  • *φ 34 [S17, S26], one-quarter of a plain-ware bowl, probably an imitation of African Red Slip 91D, diameter 0.124 m, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51 [= S23]. Light-brown fabric with a very wide range of inclusions including mica, other clear, white, black and red; self-slip preserved on interior. Marked VIIIB2. Fig. 21l . Sample no. 10/30.

  • φ 35 [S17], jug or cup with string-cut disk base and ovoid body, diameter 0.044 m, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 48 [= S26]. Very fine, light-red fabric with very fine scattered lime and exterior surface fired buff; it may be the same fabric as Robinson Reference Robinson1959, M355. Marked 8 B 2.

  • φ 36 [S17], same shape and fabric but less carefully finished on interior and some wheel-ridging on lower exterior, diameter 0.042 m, Coldstream and Huxley Reference Coldstream and Huxley1972 pl. 48 [= S26]. Marked VIIIB2.

  • φ 37 [S17], pitcher with everted rim and traces of slip on the neck; fine orange fabric, red slip, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 48 [= S26]. The fabric and finishing technique recall Slane and Sanders Reference Slane and Sanders2005, 270 nos. 3-17, 3-18, fig. 8.

  • *φ 43 [S18], pitcher with out-turned lip and long handle with deep groove on back, diameter 0.068–0.070 m, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 [= S27]. Very fine orange fabric with self-slip; could be the same as φ 35 and φ 36. Marked VIIIB2. Fig. 22a .

  • φ 38 [S18] (not marked on sherd), several joining fragments of the thick shoulder of a pitcher(?) with combing limited to shoulder: combed wave pattern between two combed bands. Very soft brown fabric. 6th century or later. Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 48 [= S26]. Marked 7/8. Cf. Lawson Reference Lawson1996, 17 no. 49 b G155/4, of which no description is given (this seems to be ST 125, of which there were three examples).

  • *φ 39 [S18], pitcher(?) with vertical branches and double axes incised on the belly, partly slipped, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 48 [= S26]. Fig. 22c .

  • *φ 40 [S18], a second pitcher with a double axe incised on the belly, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 48 [= S26]. Fig. 22b .

  • *φ 56 [S19], a small closed vessel with narrow neck, two vertical handles, and an oval mouth, gritty buff fabric; white-slipped, with dark stripes painted on the neck and handles, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 52, pl. 49 [= S24, S27]. Encrusted with barnacles. Fig. 23.

  • φ 41 [S18], plain-ware base with indented centre and central button, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 48 [= S26]. Marked VIIIB2.

  • *φ 42 [S18], oblique spout from plain-ware vessel with convex rim; groove at maximum diameter of rim and probable wheel-ridging below spout, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 51, pl. 48 [= S23, S26]. Fig. 22d . Sample no. 10/32. Cf. Ballance et al. Reference Ballance, Boardman, Corbett and Hood1989, 105 spouted jugs nos. 208 and 209, pl. 23; Slane and Sanders Reference Slane and Sanders2005, 274 no. 4-6, fig. 11; Ricci Reference Ricci and Saguì1998, 375–7, fig. 15:1–4 from the 7th century context of the Crypta Balbi in Rome. Similar spouts also appear on cooking pots at this time; see Slane and Sanders Reference Slane and Sanders2005, 272 no. 3-32 and 279 no. 4-28, figs. 8, 12. Marked VIIIB2.

  • φ 44 [S18, S23, S27], sombrero lid. Marked VIIIB2.

  • φ 45 [S18, S23, S27], sombrero lid. Marked 7/8.2.

  • φ 53 [S18], stopper recut from the body of a Late Roman Amphora 2, Argolid fabric, 5th–late 6th century, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49 (upside down) [= S27]. Marked VIIIB3.

  • *φ 54 [S18], rim, neck and handle of a Late Roman Amphora 2, Argolid fabric, 5th–late 6th century, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 52, pl. 49 [= S24, S27]. A heavy lime deposit on the interior shows the piece was reused as a water-pipe. Fig. 18d . Sample Kastri no. 10/8.

  • Additional unpublished material: Flat base of a thin-walled pitcher, diameter 0.063 m, probably cut down to use as a lid like φ 53.

Fig. 22.

Plain wares: (a) φ 43; (b) φ 40; (c) φ 39; (d) φ 42.

Fig. 23.

φ 56.

Deposit χ

Deposit χ was a small deposit from Trench XII levels 2 and 3, the northernmost and lowest trench on the slope of the Akroterion. All but two of the 16 sherds are marked AKR 12.3, a level that was partly sealed with white plaster (see Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 12); 11 of the published sherds can be identified. The date range is very consistent, late sixth and seventh century, and the group may mark the furthest extent of this phase on Kastri. χ 9 [S20, S27–S28], published as Late Roman C, has been re-identified as one of the latest forms of African Red Slip. The group also includes three pieces of vessel glass (χ 13–χ 15: Coldstream and Huxley Reference Coldstream and Huxley1972, 213), which I have not seen.

  • χ 4 [S20], base of African Red Slip form 99B–C, c.530–620, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 53:4, pl. 49 [= S28:4, S27].

  • χ 5 [S20], rim of African Red Slip form 104A, 530–580, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 53:5, pl. 49 [= S28:5, S27].

  • χ 1–2 [S20], four joining fragments of African Red Slip form 105 rim, 580–660 + , Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 53:1, pl. 49:1–2 [= S28:1, S27:1–2]. Three pieces are badly burned.

  • χ 9 [S20], rim of African Red Slip form 107, 600–650,Footnote 30 Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 53:9, pl. 49 [= S28:9, S27].

  • χ 3 [S20], rim of African Red Slip form 109, diameter 0.20 m, 580/600–mid 7th century,Footnote 31 Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 53:3, pl. 49 [= S28:3, S27].

  • Additional African Red Slip, unpublished:

    Body sherd of African Red Slip D, late 4th–7th century.

    Rim of African Red Slip form 105, diameter 0.36 m, 580–600/660 + .

    Base of African Red Slip form 105, diameter 0.16 m, 580–600/660 + .

  • χ 6 [S20], rim of Late Roman C Hayes form 10A, rough ridge on the inside, 570–600 + , Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 53:6, pl. 49 [= S28:6, S27].

  • χ 7 [S20], rim of Late Roman C Hayes form 10C, 600–660 + , Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 53:7, pl. 49 [= S28:7, S27].

  • χ 8 [S20], rim of Late Roman C Hayes form 10C, 600–660 + , Coldstream and Huxley fig. 53:8, pl. 49 [= S28:8, S27].

  • Additional Late Roman C, unpublished:

    Rim of Late Roman C Hayes form 10A, diameter 0.30 m, 570–600+. Marked AKR 12.2.

    Rim of Late Roman C Hayes form 10C, diameter 0.26 m, 600–660+. Marked AKR 12.2.

  • χ 12 [S20], Late Roman unguentarium toe with stamp, standard fabric, after 525?, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49:12 [= S27:12]. Published in Hayes Reference Hayes1971, 243, pl. 37a and mentioned again in Hayes Reference Hayes2008, 116 n.10, where such unguentaria are dated to the 6th and 7th centuries. Stamps like these (with multiple animals and an inscription) are not found in the Athenian Agora, however.

  • *χ 10 [S20], upper wall and rim of cooking-fabric bowl with everted rim; a pair of narrow grooves at the maximum diameter of the body and two grooves on the upper face of the rim, Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 53:10, pl. 49 [=S28:10, S27]. Fig. 21k . Sample no. 10/25.

  • χ 11 [S20], large flat lid with thickened rim, two grooves and a wave pattern incised along the edge, soft micaceous brown fabric, Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 49:11 [= S27:11].

Category ω

ω was used for unstratified sherds of interest in themselves.Footnote 32 More precise identifications are offered here. Only seven of the eight vessels have been identified. Seven bronze coins were also assigned to ω (ω 350–ω 356: Coldstream and Huxley Reference Coldstream and Huxley1972, 215) of which one or two are second century, two are probably fourth century, and the remainder seem to be sixth century.

  • ω 324, reflex handle of a carinated cup, of the fabric and form known as Knidian grey ware. It dates from the end of the 2nd century bc through to the Tiberian period in both Corinth and Athens and was widely dispersed around the Aegean;Footnote 33 kilns have been identified on the Knidian peninsula (Kassab Tezgör Reference Kassab Tezgör and Abadie-Reynal2003, 36–43, especially 38). Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 58:324, pl. 58. Marked NK 4.2.

  • ω 325, South Gaulish sigillata, foot and lower wall of a mould-made relief bowl, Dragendorff (Reference Dragendorff1895) type 37, Claudian–Trajanic. The decoration on the lower wall consists of a series of medallions separated by sets of six vertical wavy lines (‘palisade’); one medallion has a sitting hare right, the second the hindquarters of another animal. The lower edge of the relief field is not demarcated by a ridge or a moulding. The inside of the foot is not glazed. La Graufesenque product? Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 58:325, pl. 58. Marked NK 7C.8, which must be an error.Footnote 34

  • *ω 326, Italian sigillata rim, Ettlinger et al. Reference Ettlinger, Hedinger, Hoffmann, Kenrick, Pucci, Roth-Rubi, Schneider, Schnurbein, Wells and Zabehlicky-Scheffenegger1990, nos. 22.1 or 22.5; the body is plain, except for a groove near the top of the wall on the interior; the exterior mouldings are rouletted and there is a concave band flanked by grooves below the lip on the interior. Late Augustan or Tiberian. Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 58:326, pl. 58. Not marked. Fig. 24a .

  • ω 328, three joining fragments of African Red Slip Hayes form 50A, complete profile, mid 3rd–early 4th century. Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 58:328, pl. 58. Marked 727, AKR 7A/B.2.

  • ω 329, rim of African Red Slip Hayes form 48B, c.240–300/325. Marked II.3.

  • *ω 330, upper body of an amphora like Slane and Sanders Reference Slane and Sanders2005, 4-26, mid 7th century–later 8th/9th century. Published as being from the wall of a house in Trench AKR VII (Coldstream and Huxley Reference Coldstream and Huxley1972, 204), possibly reused as a flue. Fig. 24f .

  • [ω 353, an Ostrogothic coin of c.500, comes from the upper destruction levels within the fortification on Kastri (Coldstream and Huxley Reference Coldstream and Huxley1972, 55, Trench VI, levels 2 and 3). In a lower destruction level (Trench VI, levels 4 and 5) was found a terracotta lamp in the shape of a Negro boy, ω 365 (Coldstream and Huxley Reference Coldstream and Huxley1972, 216), one of three such Late Roman figurines.]

Fig. 24.

(a) ω 326; (b) R 19; (c) R 18; (d) R 16; (e) R 17; (f) ω 330; (g) R 22.

Additional unpublished material from Akrotiri

Sherds from AKR VIIB/2

There are 17 unpublished sherds marked AKR VIIB/2 Roman, and two unmarked pieces stored with them. AKR VIIB/2 was the level below topsoil in the southern extension of Trench VII, over the Minoan West House; it was one of the levels incorporated into the so-called deposit φ. The sherds so marked had been divided into three classes: group 1 is imported fine ware, group 2 is various cooking fabrics, and group 3 is amphorae. In all groups most of the sherds are second–third century, distinctly earlier than the late occupation on the Akroterion recorded elsewhere.Footnote 35 It is possible that these sherds are the fill of the tile drain noted in Coldstream and Huxley Reference Coldstream and Huxley1972, figs. 24 and 28. Note the apparently contemporary casserole now bagged with deposit τ, marked AKR VII A/B.2.Footnote 36 φ 7 and φ 9 (see above) are also contemporary with this group.

Group 1

  • Black-glazed kantharos handle, Hellenistic?

  • Unidentified open form, rouletted on the exterior, unidentified red-slip-coated ware, 1st–3rd century.

  • Rim of Eastern Sigillata B Hayes form 60B, probably 2nd century.

  • Rim of small Late Roman C Hayes form 3, probably 6th century.

Group 2

  • Globular stewpot with triangular rim, corrugated cooking fabric, 2nd century.

  • Fragment of globular stewpot, corrugated cooking fabric, 2nd–3rd century.

  • *R 16. Fragment of a globular stewpot, variant of corrugated cooking fabric, 2nd–3rd century. Fig. 24d . Sample no. 10/13.

  • Two rims of a small stewpot or casserole, corrugated cooking fabric, one earlier, one later, 2nd and 3rd centuries.

  • Rim of a small casserole, later 2nd–mid third century. Marked AKR 7A/B.2.Footnote 37

  • *R 17. Casserole bottom, corrugated cooking fabric, 2nd–3rd century. Fig. 24e . Sample no. 10/14.

  • Base of a trefoil-mouth pitcher, corrugated cooking fabric, 2nd or 3rd century.

  • Small globular stewpot, local imitation of corrugated cooking fabric, 2nd–3rd century.

  • Baggy stewpot, imitation of corrugated cooking ware, 3rd–4th century.

  • Flat base of a flanged pan or late frying pan, micaceous black fabric, early 4th or 5th century.

Group 3

Only the amphorae marked AKR VII B/2 are Roman.

  • Two-handled flask, mud-tempered ware, prehistoric?

  • Plain-ware shallow basin with everted rim, 1st century bc–1st century ad or earlier. Not marked.

  • Amphora toe, African fabric?, 2nd–4th century.

  • Amphora rim as Robinson Reference Robinson1959, G 197, probably Cretan, late 1st–beginning of 4th century. Cf. Robinson Reference Robinson1959, pl. 8; Riley Reference Riley and Lloyd1979, 180–3, Mid Roman Amphora 2.

Roman 1965

‘Roman 1965’ designates relatively early sherds subdivided by class like those from deposits τ and AKR VIIB/2. All are fine wares or lamps marked with their findspots from both the Neck and the Akroterion, probably similar to the sherds published as category ω. They are valuable because they reflect the same periods and wares that were identified in the KIP survey: groups 1 and 2 are Early Roman, first and early second century, while bag 4 is closely contemporary with deposit τ, second half of third century. Bag 3 is slightly later, with sherds of the fourth–fifth centuries, a period not otherwise separately distinguished in the excavations.

Group 1

Group 2

Group 3

  • Base and rim of African Red Slip Hayes form 48?, late 3rd or early 4th century. Marked 11.2, NK4.2.

  • Rim and base of African Red Slip Hayes form 50B, 4th century. Base marked 7A/B.

  • Rim of African Red Slip Hayes form 56, 4th century.

  • Two fragments of a small Late Roman C dish foot, Hayes form 3, 5th century.

  • Red-glazed lamp base, Attic, 4th century. Marked NK3.5.

  • Unpierced handle of unglazed lamp, Corinthian?, early 4th century? Marked NK 3.5.

Group 4

  • Rim of Çandarli Hayes form 2, 3rd century. Marked 6.2.

  • *R 18. Fragment of African Red Slip-A?, one of the predecessors of Hayes 50 such as 30–33, mid 3rd century. Fig. 24c . Sample no. 10/33.

  • *R 19. Fragment of African Red Slip-C, rim of Hayes form 48, non-joining fragment of ω 329 from Kastri (Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 58:329), mid 3rd century. Fig. 24b . Sample no.10/27.

  • Two joining fragments of a rim, African Red Slip form 45, mid 3rd–early 4th century. Marked NK 2.1.

  • Thin fragment of a rim, African Red Slip form 48, mid 3rd–early 4th century.

  • Nine fragments of African Red Slip form 50A rims, mid 3rd–early 4th century. Marked NK 3.5, 6.2, 8C.2, and NK 4.2.

  • Burned base of a globule-and-volute lamp, Attic?, 3rd century. Marked NK 3.5.

Bags in which the sherds seem to be stored by trench

Trench VIII B extension, unit 2 (unstratified)

Trench VIII B extension, unit 2 (unstratified) is marked on eight sherds and a stamped(?) tile; this material is either from the late wall γ or the lowest Roman level above the Late Minoan I floors.

  • Rim of African Red Slip Hayes form 99, 510–620.

  • Rim and base of African Red Slip Hayes form 105, 580/600–660 + .

  • Rim of Late Roman C Hayes form 3C, second half of 5th century.

  • *R 20a,b. Two non-joining fragments of a Late Roman C dish base, both with stamps. One is a hare, to the right, perhaps from the centre, and the other lies outside a groove. The style is Late Roman C group II, of the middle and second half of the 5th century (on form 3B–C). Fig. 16e (hare).

  • Rim of Late Roman C Hayes form 10A, 570–600 + .

  • Two non-joining fragments of Late Roman C Hayes form 10C, 600–660 + .

  • *R 21. Two joining fragments of the rim and neck of a Late Roman Amphora 1 with upper edge of handle attachment, early 5th–early 7th century. Light-red fabric with abundant tiny round solution voids and sparse to moderate tiny round black and white inclusions; surface fired yellow. Fig. 18c . Sample no. Kastri 10/9.

Numerous cooking pots and plain wares stored with deposit φ are from this trench.

Trench IX, 25 sherds

  • Rim of Eastern Sigillata B, Hayes form 32, c.25–50.

  • Rim of Çandarli, Hayes form 3, c.150–250.

  • Curving sherd from floor and wall of a vessel of African Red Slip-C, mid 3rd–early 4th century.

  • Two joining rim fragments of African Red Slip Hayes form 71 or 72, mid 5th century.

  • Five rims and a base of African Red Slip Hayes form 99, 510–620.

  • Rim and base of African Red Slip Hayes form 105, 580/600–660 + .

  • Two rims of Late Roman C Hayes form 3, 450–500.

  • Two rims of Late Roman C Hayes form 10A, 570–600 + .

  • Two rims of Late Roman C Hayes form 10C, 600–660 + .

  • Attic postglazing lamp handle, 5th century.

  • Local basin rim.

  • Handle of a trefoil-mouth pitcher, corrugated cooking fabric, 2nd–3rd century.

  • Rim of flanged pan, micaceous cooking ware, early 4th century. Cf. υ 15 above, and Slane Reference Slane1990, 79 no. 167, fig. 16.

  • Combed amphora body sherd.

  • Lid recut from an amphora toe.

Trench XI + VI (from within the late fortification wall), 25 sherds

  • Base of African Red Slip Hayes form 59A with gouges on the wall, 320–380/400.

  • Rim of African Red Slip Hayes form 67, diameter 0.38–0.40 m, 450 + .

  • Two rims and a base of African Red Slip Hayes form 99, 510–620.

  • Rim of African Red Slip Hayes form 103, 500–575.

  • African Red Slip, ridged cooking pot base to carination.

  • Rim of Late Roman C Hayes form 1, 400/425–470. Cf. Lawson Reference Lawson1996, 114 no. 22 (ST 57), fig. 16.2:7 (not identified).

  • Rim of Late Roman C Hayes form 10C, 600–660 + .

  • Red-slipped mould-made lamp with rosette discus, 1st century. Marked 6.4.Footnote 39

  • Unglazed Corinthian lamp with rosette discus (Broneer Reference Broneer1930 type XXVII), probably 3rd century. Marked 6.4.Footnote 40

  • Handle and neck, probably a pitcher.

  • Sombrero lid, Late Roman.

  • Rim with incised wave pattern and notches cut in the side, Late Roman?

  • Basin with thin, everted rim.

  • Everted rim of basin with possible groove on upper surface, Early Roman?

  • Rim of basin like τ 5, Middle Roman.

  • *R 22. Rims of two different basins (three fragments) with everted rims ending in a hook. Fig. 24g . Drawn sherd marked 6.5.

Discussion

Temporal distribution

For the KIP survey the Roman phase is defined as the Augustan period into the seventh century after Christ, and it is isolated from the earlier Hellenistic and later Byzantine periods by breaks in the pottery sequence. In the report of the excavations at Kastri, the gap between deposits σ and τ was seven or eight centuries, but a small amount of first century bc through second century ad material is preserved among the unpublished sherds. Because seventh century strata, in which earlier Roman sherds were residual, rested directly on Late Minoan deposits along the slopes of the Akroterion and on the Neck, it is probable that the intervening phases had been cut away by extensive terracing in the latest Roman phase. Third or second century bc material has not been identified from the 1960s excavations, and pottery and lamps of the first century bc were listed above with the Roman pottery, although they are dated ‘Hellenistic’.

Most of the pottery retained from the 1960s excavations belongs to the extensive Late Roman occupation phase of the sixth and seventh centuries.Footnote 41 This phase was found on Kastri, in the Akroterion trenches on its slopes, and in the Neck trenches; it was also the period of a large building on Kastraki, although no material can now be identified from that trench. The characteristic forms are African Red Slip 91D, 105, 107 and 109 and Late Roman C form 10C, noted by Hayes as characterising the latest phase of production of both wares, found in Cyrenaica and at Antioch, and perhaps best represented in publication of the Byzantine fort at Emporio on Chios. More recently the same material has been published from the Saraçhane (deposit 30), as assemblage 4 in Late Roman Corinth, and from Argos, as well as further afield, in Italy and Marseilles, for example. The ending date in the middle or third quarter of the seventh century given in Hayes Reference Hayes1972 reflects the perception that these were the forms prevalent at the time of the Arab raids of the 670s in the eastern Mediterranean, but in Tunisia it is increasingly conceivable that production and distribution of African Red Slip wares continued at least to the end of the seventh century.Footnote 42 Deposit χ, from Akroterion Trench XII, contains only four sherds (out of 15) that are datable earlier than c.580, and all but the Late Roman unguentarium χ 12 could possibly be dated after 600. In view of the large number of pieces and their wide distribution over the site, it seems likely that this phase lasted half a century or more.Footnote 43 Extensive burning and numerous joining fragments among the rims suggest destruction accompanied by burning and no later disturbance.

The coherence of deposit χ suggests the very late sixth–seventh century phase could be distinct from earlier Late Roman levels rather than continuous with them. While there is no stratigraphic support for this suggestion, the topographic distribution of types would allow it. There may also be a gap between the Middle Roman occupation and the late period. Few fourth century sherds are necessarily datable later than the early fourth century: fragments of African Red Slip 50B from the baulk VIIA/B, African Red Slip 56 (not marked) and African Red Slip 59A from within the fortification wall, as well as an Attic glazed lamp from NK 3.5. The middle of the fifth century is marked by two pieces, rims of African Red Slip 67 and Late Roman C 3B, from deposit υ on the Neck and similarly early fifth century sherds come from within the fortification wall and from Trench IX (on the slope between Trenches VII and XII), as well as from φ (African Red Slip 67, 71 or 72, Late Roman C 1). Fine wares of the second half of the fifth century are limited to Trenches VII and VIII, however (φ has Late Roman C 3C and Attic postglazing lamps, and the two contemporary Late Roman C stamps are from just east of the lower Late Roman terrace wall), and it is difficult to judge whether the earlier occupation is continuous with the sixth century. Evidence of the earlier sixth century is seen most clearly in the Neck trenches, where deposit υ has multiple examples of African Red Slip form 99A–B and Late Roman C form 3E–F of the first half of the sixth century; African Red Slip form 91C, 103, 104A, and the African Red Slip lamps could be contemporary. Most of the same forms also come from group φ (including an unpublished Late Roman C form 3 from VIIB/2). Finally, an African Red Slip 103 of the sixth century was found in Trenches XI + VI within the fortification wall.

Earlier Roman occupation of the site may also have been episodic. Early Roman (more or less first and early second century ad) is marked by Italian sigillata and Pergamene ware (from the Neck), Pontic sigillata from both the Neck and Trench VI (within the late fortification wall), and Eastern Sigillata B and Çandarli from several trenches on the Akroterion including Trench IX, as well as a single example of Cypriot sigillata (from φ). All were found mixed with Late Roman material. The only stratified material seems to be the unpublished sherds marked VIIB/2 Roman, which could come from the tile drain over the West House, and deposit τ from the Neck. Both are primarily second to third century, a Middle Roman phase. Whether this is the same phase as one attested by late third/early fourth century sherds is doubtful to me because they come from different places: the examples of African Red Slip 44, 48B, 49, 50A published from ω are matched by unpublished pieces from φ, from the Neck trenches, from within the fortification wall (11.2 and 6.2) and from Trench IX.

Whether the temporal variations suggested above reflect variations in the occupation of the area of Kastri or variations in the availability of imported fine wares is a critical question, because most of the excavated material retained and listed above was unstratified and imported.

Geographical distribution

For the Roman period, the distribution patterns of amphorae and of fine wares primarily reflect seaborne shipping networks. The dominance of African Red Slip beginning with forms 45, 48–9 and 50A in the second half of the third century is a pattern familiar around the eastern Mediterranean, although the presence of an earlier Severan sherd and of an African cooking pot (φ 9) (Fig. 17a ) of the same date is unexpected. Types dating c.325–400/425 (African Red Slip 58–61 and 53), on the other hand, appear to be less common here than elsewhere, although form 67 (all fifth century examples?) was found within the fortification, in deposit φ, and on the Neck. The fifth century forms 70–91B are missing as they are on many eastern sites, and on Kythera as elsewhere the supply resumes with the sixth century forms 99, 103 and 104.

The number of sites on which the types that characterise the seventh century, African Red Slip 105, 107 and 109 with late examples of other forms like 91D and 99, are found is much more limited than those receiving the sixth century African imports, and the predominance of these late western forms on Kythera is remarkable. The main eastern fine ware of the Late Roman period, Late Roman C, on many eastern sites eclipses the import of African Red Slip around the middle of the fifth and throughout the sixth century (Hayes Reference Hayes1972, 417 and see maps 8–16). On Kythera there are relatively few Late Roman C form 3 of the mid fifth–mid sixth century, or the earlier Late Roman C form 1, and an unusually high number of Late Roman C form 10A and, especially, 10C of the mid seventh century.

Except in the last phase this pattern favouring western sources is not matched among the amphorae or coarse wares, which are not well documented from the excavations and which come from markedly different sources than the fine wares. Amphorae are confined to the Late Roman period, when three types appear in quantity. There are many imports of Late Roman Amphora 1 from the general region of the Bay of Iskenderun and coastal Syria (probably not Cyprus) as well as possible imitations.Footnote 44 The necks of Late Roman Amphora 2 in deposit φ (φ 50–3) seem to be from the Argolid rather than the eastern Aegean, and one neck may be a local imitation of the type; note that two pieces from the context (φ 53 and φ 54) (Fig. 18d ) were found in secondary use, which supports a date closer to 650 than 580 for the deposit. The Samos-cistern amphora φ 48 (Fig. 18f ) is probably also from the eastern Aegean. African amphorae are represented by six small ‘spatheia’ in deposit φ (φ 57–φ 58, φ 61–φ 64) (Fig. 19a –c) and these were cited by Riley in his investigation of Late Roman Amphora 8a. More recently the small ‘spatheion’ has been characterised as a seventh century type by Bonifay (Reference Bonifay2004), who distinguishes variants from four sources: those from Kythera appear to belong to his variant C with a vertically pared body and bevelled rim, from Nabeul, and the toes most closely resemble variants B and D. φ 72 (Fig. 20) is also a Late Roman African amphora, probably of the second half of the fifth century. Such a quantity of African amphorae is very unusual on Greek sites.

The Early Roman period also has western rather than eastern imports. In the Augustan and Julio-Claudian period Italian sigillata (ω 326 and two rims from NK III) and a piece of South Gaulish relief ware (ω 325) were recorded, but no Eastern Sigillata A from Syria was identified. The later first–second century imports are from the Aegean (Pontic sigillata, Pergamene, Eastern Sigillata B, Knidian) with a single piece of Cypriot sigillata from further east. The presence of the Pontic material is a reminder that a trade route from the Black Sea still passes east and south of the Peloponnese on its way into the Mediterranean. In general the period from the early or mid second to mid third century is not well represented by fine wares (four pieces of Çandarli and no more than one Eastern Sigillata B), but it is not clear whether this is because the Neck and Akroterion/Kastri were not occupied, because the wares are quite nondescript and apt to be overlooked, or because Eastern Sigillata B and Çandarli are not common on Kythera (as they are in the Aegean). On the other hand, corrugated cooking ware of this period from the eastern Aegean, both cooking pots and pitchers, are present and were imitated (τ 4) (Fig. 15a ).

Local or regional production centres are not prominent among the pottery sherds except the series of basins, which seem to run from the early period to the Late Roman. φ 14, a stamped beehive, has a similar fabric to that of the basins. There are few parallels with Sparta except the brown bowl, φ 11; perhaps it is Laconian. Parallels with the Laconia survey include possible parallels for the Middle Roman basins and a Late Roman combed pitcher. φ 7 is an Attic bowl of the third century or earlier (Attic pottery was hardly exported beyond the Corinthian isthmus in the Roman period). As mentioned above, there are cooking pots imitating the widely exported Aegean type of the second–third century and they could have been made on Kythera. There are also imitations, in a different fabric, of the Late Roman Amphora 1.

The distribution patterns of Roman lamps do not reflect those of fine wares or of amphorae. The excavated material contains some unglazed Corinthian lamps of the second or third century, a few Attic lamps of the third, fourth and fifth centuries, and several African imports, apparently of the sixth and seventh centuries, which are very unusual in Greece, where local imitations usually replace the African Red Slip lamps sometime in the fifth or sixth century.

3. The Medieval pottery (Joanita Vroom)

Introduction

The Medieval ceramics excavated at Kastri come from one domestic rubbish deposit, published as ψ, ‘Psi’, which was the upper fill overlying Roman and Minoan burial levels in Tomb H (Coldstream and Huxley Reference Coldstream and Huxley1972, 175–7 [= S21, S29–30], 308, fig. 53, pl. 50 [= S28, S31]).Footnote 45 I present here a more detailed description and interpretation of the pottery finds from the deposit in an updated catalogue, followed by a discussion of the material.Footnote 46

The finds fall into two main categories: first the fine glazed wares of the assemblage are discussed (cat. nos. M 1–20), then the unglazed coarse wares and the Medieval amphorae (cat. nos. M 21–42). The catalogue fills out information concerning the size, the surface treatment, the decoration and the shape of the published material, and it is related to similar-looking finds from more recent publications. The published finds are placed first in each section of the catalogue (cat. nos. M 1–15, M 21–35), followed by 12 unpublished fragments, all with new descriptions and new drawings in order to give the full picture of the pottery assemblage (cat. nos. M 16–20; M 36–42).

The catalogue is accompanied by 23 new black-and-white drawings of published pieces (accompanying the photographs from Coldstream and Huxley Reference Coldstream and Huxley1972 in Figs. 25–9), and three drawings of unpublished fine wares (Fig. 27h–j ) and five of coarse wares and amphorae (Fig. 30a–e ). In order to complete the picture, some black-and-white photographs from Coldstream and Huxley Reference Coldstream and Huxley1972 are included in these figures, 10 of fine wares in various decoration styles, two of coarse wares and two of complete amphorae in Figs. 25 to 29 (see Coldstream and Huxley Reference Coldstream and Huxley1972 , pl. 50:1–30 [= S31:1–30] and pl. 87:Q18–19).

Fig. 25.

(a)–(b) M 1; (c)–(d) M 2; (e)–(f) M 3; (g)–(h) M 4.

Fig. 26.

(a)–(b) M 5; (c)–(d) M 6; (e)–(f) M 7; (g)–(h) M 8; (i) M 9.

Fig. 27.

(a)–(b) M 10; (c)–(d) M 11; (e)–(f) M 12; (g) M 13; (h) M 18; (i) M 19; (j) M 20.

Fig. 28.

(a) M 21; (b) M 22; (c) M 23; (d) –(e) M 24; (f)–(g) M 27; (h) M 29.

Fig. 29.

(a) M 30; (b) M 32; (c) Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 87:Q 18; (d) M 31; (e) M33; (f) Coldstream and Huxley Reference Coldstream and Huxley1972, pl. 87:Q 19.

Most of the sherds have been sampled for petrographical and chemical analyses by Evangelia Kiriatzi in the Fitch Laboratory of the British School at Athens. The results will be more fully described in a forthcoming publication.

Fine Glazed Wares

Published material

The fine wares from deposit ψ published in 1972 included nine fragments of Green and Brown Painted Ware (cat. nos. M 1–9, Figs. 25 and 26), two fragments of Slip-painted Ware (cat. nos. M 10–11, Fig. 27a–d ) and two fragments of Fine Sgraffito Ware (cat. nos. M 12–13, Fig. 27e–g ). Unfortunately, two body fragments of Monochrome Glazed Wares have not been seen by me (cat. nos. M 14–15)

Green and Brown Painted Ware

The majority of these fragments are of a Byzantine pottery type, which is known in the literature as ‘Green and Brown Painted Ware’ (see in general, Vroom Reference Vroom2003, 151–2 with further literature; 2005, 82–3). This term was introduced by Charles Morgan in his 1942 publication of the Byzantine ceramic finds from Corinth. In addition, he divided the painted decoration styles of this pottery type into five groups (Morgan Reference Morgan1942, 72–83). The vessels were covered on the inside with a thick white slip (also known in French as engobe), on which a decoration was painted under a lead glaze in green and brown colours. The painted motifs on the Kastri fragments belong mostly to Morgan's Group III with simple linear, spiral and cross-hatched combinations (Morgan Reference Morgan1942, 77–80 and pls. XXII–XXIII).

Shapes mainly consist of open vessels, such as shallow dishes and large bowls with a low ring-foot, a straight divergent upper body and a small everted rim, or smaller bowls with a convex divergent body, a straight rim and rounded lip. Similar-looking shapes are also common in the pottery repertoire of Corinth and of other sites on the Greek Mainland (such as in the Peloponnese and in Boeotia).Footnote 47

  • M 1. Bowl, profile (Coldstream and Huxley Reference Coldstream and Huxley1972, 175 no. 1, fig. 53 and pl. 50:1 [= S21, S28, S31:1]). Fig. 25a–b .

    Height: 0.080 m; width: 0.137 m; estimated diameter of base: 0.098 m; estimated diameter of rim: 0.280 m; thickness: 0.005–0.010 m.

    Fabric sampled for analyses (no. 08/01). Surface treatment: white slip on interior and rim and splashes of white slip on upper part of exterior; transparent lead glaze on interior and on rim; smoothed exterior.

    Decoration: green and brown painted decoration of spirals on interior. Shape: low ring-foot; convex divergent lower body; straight divergent upper part; small everted rim.

    See for shape, Morgan Reference Morgan1942, fig. 51B (Corinth); Hayes Reference Hayes1992, fig. 17 no. 12 (Istanbul); Vroom Reference Vroom2003, fig. 6.20:W10.11 (Boeotia); Poulou-Papadimitriou Reference Poulou-Papadimitriou and Bakirtzis2003, figs. 37 and 39 (Herakleion).

  • M 2. Bowl, base fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 175 no. 2, fig. 53 and pl. 50:2 [= S21, S28, S31:2]). Fig. 25c–d .

    Height: 0.056 m; width: 0.172 m; estimated diameter of base: 0.089 m; thickness: 0.005–0.009 m.

    Fabric sampled for analyses (no. 08/02). Surface treatment: white slip on interior and splashes of white slip on upper part of exterior; transparent lead glaze on interior; smoothed exterior. Decoration: green and brown painted spirals and cross-hatched circle on interior. Shape: low ring-foot; convex divergent lower body.

    See for similar-looking decoration, Dawkins and Droop Reference Dawkins and Droop1910–11, pl. XVIII no. 75 (Sparta).

  • M 3. Bowl, rim fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 175 no. 3, pl. 50:3 left fragment [= S21, S31:3]). Fig. 25e–f .

    Length: 0.044 m; width: 0.076 m; estimated diameter of rim: 0.280 m; thickness: 0.006 m.

    Fabric sampled for analyses (no. 08/09). Surface treatment: white slip on interior and on rim; transparent lead glaze on interior; smoothed exterior? Decoration: green and brown painted stripes on interior. Shape: straight divergent upper part; small everted rim.

    See for similar-looking decoration, Morgan Reference Morgan1942, pl. XXIII B (Corinth); Vroom Reference Vroom2003, fig. 6.20:W10.7 (Boeotia); Vroom Reference Vroom, Hodges, Bowden and Lako2004, 280, figs. 15.6–7 (Butrint); see for shape, Morgan Reference Morgan1942, fig. 51B (Corinth); Hayes Reference Hayes1992, fig. 17 no. 12 (Istanbul); Vroom Reference Vroom2003, figs. 6.20–21:W 10.11–12 (Boeotia).

  • M 4. Bowl, profile (Coldstream and Huxley Reference Coldstream and Huxley1972, 175 no. 4, fig. 53 and pl. 50:4 [= S21, S28, S31:4]). Fig. 25g–h .

    Height: 0.070 m; width: 0.145 m; diameter of base: 0.066 m; diameter of rim: 0.170 m; thickness: 0.006 m.

    Fabric sampled for analyses (no. 08/03). Surface treatment: white slip on interior and upper part of exterior; transparent lead glaze on interior (almost vanished); smoothed exterior. Decoration: green and brown painted spirals and triangles(?) on interior. Shape: low ring-foot; convex divergent body; straight rim with rounded lip.

    See for shape, Morgan Reference Morgan1942, fig. 56C and fig. 59D (Corinth).

  • M 5. Bowl, rim fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 176 no. 5, fig. 53 and pl. 50:5 [= S29, S28, S31:5]). Fig. 26a–b .

    Length: 0.101 m; width: 0.107 m; estimated diameter of rim: 0.300 m; thickness: 0.006–0.008 m.

    Fabric sampled for analyses (no. 08/04). Surface treatment: white slip on interior, on rim and on upper part of exterior; transparent lead glaze on interior (vanished); smoothed exterior. Decoration: (light) green and brown painted spirals and loops on interior. Shape: convex divergent lower body; straight divergent upper part; small everted rim.

    See for similar-looking painted decoration, Vroom Reference Vroom and Andrikou2006, fig. 13 upper right (Thebes); see for shape, Morgan Reference Morgan1942, fig. 51B (Corinth); Vroom Reference Vroom2003, figs. 6.20–21:W 10.11–12 (Boeotia).

  • M 6. Dish, rim fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 176 no. 6, fig. 53 and pl. 50:6 [= S29, S28, S31:6]). Fig. 26c–d .

    Length: 0.076 m; width: 0.093 m; estimated diameter of rim: 0.240 m; thickness: 0.005 m.

    Fabric sampled for analyses (no. 08/05). Surface treatment: white slip on interior and on rim; transparent lead glaze on interior; smoothed exterior. Decoration: green and brown painted spirals and splashes on interior. Shape: convex divergent body; small reverted rim.

    See in general, Morgan Reference Morgan1942, pl. XXII (Corinth); Vroom Reference Vroom2003, fig. 6.19:W10.6 (Boeotia).

  • M 7. Bowl, body fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 176 no. 7, pl. 50:7 [= S29, S31:7]). Fig. 26e–f .

    Length: 0.095 m; width: 0.061 m; thickness: 0.008–0.009 m.

    Fabric sampled for analyses (no. 08/11). Surface treatment: white slip on the interior and exterior; transparent lead glaze on the interior. Decoration: brown painted cross-hatched. Shape: convex divergent lower body; straight divergent upper part.

    See for similar-looking decoration, Dawkins and Droop Reference Dawkins and Droop1910–11, pl. XVIII nos. 66 and 73 (Sparta); Vroom Reference Vroom2003, fig. 6.21:W10.16 (Boeotia); Waksman and Wartburg Reference Waksman and Wartburg2006, fig. 2 no. 6 (Kouklia–Paphos); Yangaki Reference Yangaki2008, 602, fig. 15 (Nauplion).

  • M 8. Dish, rim fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 176 no. 8, fig. 53 and pl. 50:8 [= S29, S28, S31:8]). Fig. 26g–h .

    Length: 0.048 m; width: 0.095 m; diameter of rim: 0.220 m; thickness: 0.005 m.

    Fabric sampled for analyses (no. 08/06). Surface treatment: white slip and transparent lead glaze on interior; smoothed exterior? Decoration: green and painted stripes and connected half circles on interior. Shape: convex divergent lower body; straight symmetrical upper part; straight rim with rounded lip.

    See for similar-looking decoration, Poulou-Papadimitriou Reference Poulou-Papadimitriou and Bakirtzis2003, figs. 36–37 (Herakleion); see for shape, Morgan Reference Morgan1942, fig. 56A (Corinth); Hayes Reference Hayes1992, fig. 17 nos. 10–11 (Istanbul).

  • M 9. Dish, body fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 176 no. 13, fig. 53 and pl. 50:13 [= S29, S28, S31:13]). Fig. 26i .

    Length: 0.077 m; width: 0.082 m; estimated diameter of rim: 0.210 m; thickness: 0.005–0.006 m.

    Fabric sampled for analyses (no. 08/08). Surface treatment: white slip on interior and exterior; transparent lead glaze on interior (almost vanished). Decoration: green painted surface on interior and on rim; splashes of brown on interior. Shape: convex divergent body; small everted rim.

Slip-painted Ware

The surface of most vessels within this category is not completely covered with a white slip (or engobe), but only decorated with it as a kind of paint (see in general, Vroom Reference Vroom2005, 80–1). The painted motifs are then afterwards coated with a lead glaze, which is often transparent and which enhances the contrast between the dark fabric and the pale designs. Slip-painted Ware of the Middle Byzantine period is in Corinth also known as ‘Light on Dark Slip-painted Ware I–II’ (Sanders Reference Sanders1995, 241–2).

The shapes of this type of painted pottery from Kastri include a large dish with a low ring-foot and a small cup with a round grooved handle (cat. nos. M 10–11, Fig. 27a–d ). Both shapes have previously been found on sites in the Peloponnese, such as Corinth and Kenchreai. The decoration of the two Kastri fragments belongs to Morgan's group of the so-called ‘Late Linear Style’ with painted rectilinear and spiral motifs (Morgan Reference Morgan1942, 102–3).

  • M 10. Cup, rim-handle fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 176 no. 9, fig. 53 and pl. 50:9 [= S29, S28, S31:9]). Fig. 27a–b .

    Height: 0.037 m; width: 0.057 m; diameter of rim: 0.100 m; thickness: 0.003 m.

    Fabric sampled for analyses (no. 08/14). Surface treatment: white slip with (light) green paint on interior and exterior upper parts and on top of round handle; transparent lead glaze on interior and exterior. Shape: thin-walled convex symmetrical body; slightly everted rim with rounded lip; small round handle with a groove on top.

    See for shape, Morgan Reference Morgan1942, figs. 63E and 71D–E (Corinth); Adamsheck Reference Adamscheck1979, pl. 25 nos. LRB 36, LRB 48 (Kenchreai).

  • M 11. Bowl, base fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 176 no. 10, pl. 50:10 [= S29, S31:10]). Fig. 27c–d .

    Height: 0.023 m; length: 0.05 m; width: 0.092 m; estimated diameter of base: 0.100 m; thickness: 0.005–0.007 m.

    Fabric sampled for analyses (no. 08/12). Surface treatment: transparent lead glaze on interior; smoothed exterior?

    Decoration: painted spirals in white slip under the glaze on the interior. Shape: low ring-foot; recessed bottom; straight divergent lower wall.

    See for similar-looking decoration, Morgan Reference Morgan1942, figs. 75A and 79A, pl. XXXII:A (Corinth); Papanikola–Bakirtzi, Mavrikiou and Bakirtzis Reference Papanikola-Bakirtzi, Mavrikiou and Bakirtzis1999, 56–7 nos. 87 and 89 (Benaki Museum); Vroom Reference Vroom2003, fig. 6.17:W 9.4 and fig. 6.41:W 9.14 (Boeotia); Poulou-Papadimitriou Reference Poulou-Papadimitriou and Bakirtzis2003, figs. 24 and 29 (Herakleion); Yangaki Reference Yangaki2008, 601, fig. 13 (Nauplion).

Fine Sgraffito Ware

The third group of Medieval fine wares within the Kastri assemblage includes two small pieces of Fine Sgraffito Ware (cat. nos. M 12–13, Fig. 27e–g ; see in general, Vroom Reference Vroom2005, 84–5). The interior of this type of decorated ware is covered with a thick white slip (or engobe). The vessels are then engraved with a fine sharp tool through the white slip, thus showing delicate incised motifs, and finally covered with a transparent lead glaze.

Shapes include shallow dishes and large bowls with a straight symmetrical upper body and a straight rim with rounded lip, which are common in this ware. The two fragments from Kastri belong to Morgan's Group IV of the so-called ‘Developed Style’ with geometric designs or animal and human fugures represented on a free field (Morgan Reference Morgan1942, 127–35, pls. XLII–XLIV; see also Vroom Reference Vroom2003, 152 with further literature).

  • M 12. Bowl, body fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 176 no. 11, pl. 50:11 [= S29, S31:11]). Fig. 27e–f .

    Length: 0.030 m; width: 0.032 m; thickness: 0.004–0.009 m.

    Fabric sampled for analyses (no. 08/15). Surface treatment: white slip and transparent lead glaze on the interior; smoothed exterior. Decoration: incised curved lines under the glaze. Shape: convex divergent body.

    See for similar-looking decoration, Morgan Reference Morgan1942, fig. 104b and pl. XLIII:B,D (Corinth); Sanders Reference Sanders, Williams and Bookidis2003, fig. 23.2 no. 16 (Corinth).

  • M 13. Bowl, rim fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 176 no. 12, fig. 53 and pl. 50:12 [= S29, S28, S31:12]). Fig. 27g .

    Length: 0.037 m; width: 0.038 m; estimated diameter of rim: 0.210 m; thickness: 0.005 m.

    Fabric sampled for analyses (no. 08/16). Surface treatment: white slip and transparent lead glaze on interior and upper part of exterior. Decoration: no incised motifs survived. Shape: convex divergent body; straight symmetrical upper part; straight rim with rounded lip.

    See for shape, Morgan Reference Morgan1942, figs. 91B and 95A–B (Corinth).

Monochrome Glazed Wares

Unpublished material

The unpublished fine wares from deposit ψ include four fragments of Green and Brown Painted Ware (cat. nos. M 16–19, Fig. 27h–i ) and one fragment of Slip-painted Ware (cat no. M 20, Fig. 27j ). They all belong in the same decorated groups as the published ones described above.

Green and Brown Painted Ware

  • M 16. Bowl, body fragment.

    Length: 0.074 m; width: 0.037 m; thickness: 0.006–0.008 m.

    Fabric sampled for analyses (no. 98/285). Surface treatment: white slip and transparent lead glaze on interior; smoothed exterior. Decoration: green and brown painted stripes and abstract motifs on interior.

  • M 17. Bowl, body fragment.

    Length: 0.057 m; width: 0.032 m; thickness: 0.006–0.007 m.

    Fabric sampled for analyses (no. 98/284). Surface treatment: white slip on interior and upper part of exterior; transparent lead glaze on interior; smoothed exterior. Decoration: green and brown painted zone on interior.

  • M 18. Bowl, rim fragment. Fig. 27h .

    Length: 0.068 m; width: 0.082 m; estimated diameter of rim: 0.200 m; thickness: 0.006–0.007 m.

    Fabric sampled for analyses (no. 08/07). Surface treatment: white slip on interior and exterior; transparent lead glaze on interior. Decoration: brown painted stripes on interior. Shape: convex divergent body; straight rim with rounded lip.

    See for shape, Morgan Reference Morgan1942, fig. 55C (Corinth); Vroom Reference Vroom2003, fig. 6.21:W10.14–15 (Boeotia); Yangaki Reference Yangaki2008, 602, fig. 17 (Nauplion).

  • M 19. Bowl, rim fragment. Fig. 27i .

    Length: 0.032 m; width: 0.024 m; thickness: 0.005 m.

    Fabric sampled for analyses (no. 08/10). Surface treatment: white slip on interior and upper part of exterior; transparent lead glaze on interior. Decoration: green and brown painted stripes on interior. Shape: straight symmetrical upper part; straight rim with rounded lip.

    See for shape, Morgan Reference Morgan1942, fig. 56A (Corinth); Hayes Reference Hayes1992, fig. 17 nos. 10–11 (Istanbul); Vroom Reference Vroom2003, fig. 6.20:W10.9 (Boeotia).

Slip-painted Ware

  • M 20. Bowl, rim fragment. Fig. 27j .

    Length: 0.045 m; width: 0.048 m; estimated diameter of rim: 0.280 m; thickness: 0.006–0.007 m.

    Fabric sampled for analyses (no. 08/13). Surface treatment: transparent lead glaze on interior. Decoration: arcades painted in white slip under the glaze on the interior. Shape: convex divergent body; straight symmetrical upper part; straight carinated rim with rounded lip.

    See for shape, Morgan Reference Morgan1942, fig. 75A (Corinth); Vroom Reference Vroom2003, fig. 6.18:W 9.12 (Boeotia).

Unglazed coarse wares and Medieval amphorae

Published material

The rest of the published material from deposit ψ includes nine fragments of unglazed coarse wares (cat. nos. M 21–9, Fig. 28), as well as six fragments of Medieval amphorae (cat. nos. M 30–5).

Unglazed coarse wares

The fabrics of the coarse wares are often very gritty, containing large chunks of silver mica. They appear to be local as seen through a hand lens, but this will be established by future petrographical analyses. The shapes of the Kastri fragments include three larger cooking pots/jars, in modern Greek also known as tsoukalia (cat. nos. M 21–23, Fig. 28a–c ). The simple shapes with straight walls of two of these cooking pots/jars (cat. nos. M 21–22, Fig. 28a–b ) seem to be adjusted to the coarseness of the gritty fabrics. It looks as if the potter did not have much choice for making other cooking pot shapes. The handle fragment of a third jar (cat. no. M 23, Fig. 28c ) shows similarities with Middle Byzantine cooking pots found in Corinth, Sparta and Nichoria.

Apart from cooking pots/jars we can also distinguish fragments of basins (or lekanai) with flaring sides in the Kastri pottery assemblage (cat. nos. M 24, M 27, M 29, Fig. 28d–h ), although one of these could also be a high-walled mortar (cat. no. M 29, Fig. 28h ). The other two examples are decorated with incised wavy lines or a painted streak on the exterior surface (cat. nos. M 24, M 27, Fig. 28d–g ).

  • M 21. Jar, rim–handle–body fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 176 no. 16, pl. 50:16 [= S29, S31:16]). Fig. 28a .

    Length: 0.123 m; width: 0.076 m; estimated diameter of rim: 0.400 m; thickness: 0.007–0.008 m.

    Fabric sampled for analyses (no. 98/289). Surface treatment: smoothed exterior. Shape: straight convergent upper body; straight rim with direct lip; oval plain handle.

  • M 22. Jar, rim–handle–body fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 176 no. 17, pl. 50:17 [= S29, S31:17]). Fig. 28b .

    Length: 0.120 m; width: 0.054 m; thickness: 0.006–0.007 m.

    Fabric sampled for analyses (no. 08/25). Surface treatment: smoothed on exterior. Shape: straight convergent upper body; straight rim with direct lip; oval grooved handle.

  • M 23. Jar, rim–handle fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 176–7 no. 18, pl. 50:18 [= S29–S30, S31:18]). KK 98/290. Fig. 28c .

    Length: 0.086 m; width: 0.062 m; estimated diameter of rim: 0.160 m; thickness: 0.010 m.

    Fabric sampled for analyses (no. 08/24). Surface treatment: smoothed on exterior. Shape: convex convergent upper body; concave symmetrical neck; straight rim; curving broad ribbon handle.

  • M 24. Basin, rim fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 177 no. 19, fig. 53, pl. 50:19 [= S30, S28, S31:19]). Fig. 28d–e .

    Height: 0.089 m; width: 0.102 m; estimated diameter of rim: 0.240 m; thickness: 0.0068 m.

    Fabric sampled for analyses (no. 98/296). Surface treatment: smoothed on the interior and exterior. Decoration: gouged incised wavy line under rim; brown painted streak under exterior rim and on interior. Shape: straight divergent lower body; straight symmetrical upper body; slightly everted rim with direct lip.

  • M 25. Jar, shoulder fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 177 no. 20, pl. 50:20 [= S30, S31:20]).

  • M 26. Closed vessel, body fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 177 no. 21, pl. 50:21 [= S30, S31:21]).

    Length: 0.092 m; width: 0.070 m; thickness: 0.008 m.

    Fabric sampled for analyses (no. 98/288). Surface treatment: smoothed on interior and exterior. Decoration: group of incised wavy lines on upper part of exterior.

  • M 27. Basin, two rim fragments and two adjoining body fragments (Coldstream and Huxley Reference Coldstream and Huxley1972, 177 no. 22, pl. 50:22 [= S30, S31:22]). Fig. 28f–g .

    Height: 0.121 m; width: 0.254 m; estimated diameter of rim: 0.400 m; thickness: 0.011 m.

    Fabric sampled for analyses (no. 98/294). Surface treatment: smoothed on exterior. Decoration: group of incised wavy lines under exterior rim and on exterior shoulder; band of finger impressions on shoulder. Shape: convex divergent upper body; everted rim.

  • M 28. Closed vessel, body fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 177 no. 23, fig. 53, pl. 50:23 [= S30, S28, S31:23]).

    Length: 0.091 m; width: 0.107 m; estimated diameter of rim: 0.240 m; thickness: 0.007–0.008 m.

    Fabric sampled for analyses (no. 98/295). Surface treatment: smoothed exterior and interior. Decoration: two groups of incised wavy lines on body.

  • M 29. Basin, profile (Coldstream and Huxley Reference Coldstream and Huxley1972, 177 no. 24, fig. 53, pl. 50:24 [= S30, S28, S31:24]). Fig. 28h .

    Height: 0.173 m; width: 0.114 m; estimated diameter of rim: 0.360 m; thickness: 0.016–0.017 m.

    Fabric sampled for analyses (no. 98/287). Surface treatment: smoothed on exterior. Shape: straight divergent body; flat base with rounded transition; slightly reverted rim.

    See for similar shape (but in a different fabric), Wartburg and Violaris Reference Wartburg, Violaris, Zozaya, Retuerce, Hervás and Juan2009, fig. 3 no. 9 (Nicosia).

Medieval amphorae

The Kastri pottery assemblage includes eight fragments of Medieval amphorae, or long-distance transport jars for wine and oil. Of these, five sherds are already published in Coldstream and Huxley Reference Coldstream and Huxley1972, 177 [= S30] (cat. nos. M 30–5, Fig. 29a,b,d,e ), but three remain unpublished (cat. nos. M 40–2).

The published sherds in this catalogue can be divided into three types of amphorae. One group is known as the so-called ‘Günsenin type 3 amphorae’, named after the Turkish underwater archaeologist Nergis Günsenin who first recognised them as an amphora type (cat. nos. M 33–5, Fig. 29e ) (Günsenin Reference Günsenin, Déroche and Spieser1989). In Istanbul they are classed by John Hayes as his ‘Saraçhane type 61 amphorae’ (Hayes Reference Hayes1992, 76, fig. 26.10). They have a tall pear-shaped body with two heavy high-slung oval handles, which are very characteristic of these amphorae (see also Vroom Reference Vroom2003, 153–5, fig. 6.7; 2005, 98–9 MBYZ 15.1–3). Their fabrics are often tempered with organics such as chaff, especially on the handles. Their exterior upper body has a closely set horizontal combed surface (cat. nos. M 33–5). The combed upper body of one Kastri sherd within this amphora group is further incised with a graffito in the shape of a wavy line (cat. no. M 33, Fig. 29e ). A complete example of this type of amphora also came from the Kastri excavations (Fig. 29f ; Coldstream and Huxley Reference Coldstream and Huxley1972, 308, pl. 87:Q 19, although I was not able to see this complete amphora).

The other two types of amphorae found at Kastri are still unknown elsewhere (respectively cat. no. M 30, Fig. 29a , and cat. nos. M 31–2, Fig. 29 d and b), so hopefully petrographical and chemical analyses will give some more information about their fabrics in the near future. The handle shape of one of the Kastri fragments seems similar to an amphora found in a late eleventh/early twelfth century context at excavations in Argos (cat. no. M 31, Fig. 29d ) (Piérart and Thalmann Reference Piérart, Thalmann, Déroche and Spieser1980, pl. V:B1). A complete example of this type of amphora was recovered in the Kastri excavations (Fig. 29c : Coldstream and Huxley Reference Coldstream and Huxley1972, 308, pl. 87:Q 18, although I was not able to see this complete amphora). It is possible that the Kastri sherds (cat. nos. M 31–2, Fig. 29d and b ), made in a gritty red fabric, belonged to a Medieval amphora type, which is also known as a ‘Bjelajac amphora II’. This type was first recognised by Ljiljana Bjelajac on Serbian sites in the Danube region (Bjelajac Reference Bjelajac, Déroche and Spieser1989, 113–15, fig. 3/1:2.). The amphora seems to appear between the eleventh and thirteenth centuries, but its provenance is still unknown.

  • M 30. Amphora, rim–neck–handle fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 177 no. 25, fig. 53, pl. 50:25 [= S30, S28, S31:50]). Fig. 29a .

    Height: 0.074 m; width: 0.073 m; estimated diameter of rim: 0.059 m; thickness: 0.008–0.020 m.

    Fabric sampled for analyses (no. 08/18). Surface treatment: smoothed exterior. Shape: straight convergent upper body; straight symmetrical neck; everted rim; plain oval handle.

  • M 31. Amphora, neck–shoulder–handle fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 177 no. 26, pl. 50:26 [= S30, S31:26]). Fig. 29d .

    Length: 0.135 m; width: 0.092 m; thickness: 0.011–0.015 m.

    Fabric sampled for analyses (08/20). Surface treatment: smoothed exterior with buff slip. Shape: convex convergent upper body; straight symmetrical neck; heavy oval handle, ribbed on top.

  • M 32. Amphora, neck–handle fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 177 no. 27, pl. 50:27 [= S30, S31:27]). Fig. 29b .

    Length: 0.099 m; width: 0.050 m; thickness: 0.010 m.

    Fabric sampled for analyses (no. 08/21). Surface treatment: smoothed exterior. Shape: ribbed and grooved concave convergent neck; heavy oval handle.

  • M 33. Amphora, shoulder–handle fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 177 no. 28, pl. 50:28 [= S30, S31:28]). Fig. 29e .

    Length: 0.187 m; width: 0.015 m; thickness: 0.011–0.014 m.

    Fabric sampled for analyses (no. 08/19). Surface treatment: combed exterior. Decoration: incised wavy line on upper body. Shape: convex convergent upper body; heavy oval handle.

    See for shape, Sanders Reference Sanders1993, 282–3, fig. 15 (Sparta); Vroom Reference Vroom2005, 97–9 MBYZ 15.1–3.

  • M 34. Amphora, body fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 177 no. 29, pl. 50:29 [= S30, S31:29]).

    Fabric sampled for analyses? Surface treatment: combed exterior.

  • M 35. Amphora, body fragment (Coldstream and Huxley Reference Coldstream and Huxley1972, 177 no. 30, pl. 50:30 [= S30, S31:30]).

    Length: 0.182 m; width: 0.095 m; thickness: 0.015 m.

    Fabric sampled for analyses (no. 98/286). Surface treatment: combed exterior.

Unpublished material

The unpublished sherds from deposit ψ include four fragments of unglazed coarse wares, especially of closed vessels (cat. nos. M 36–9, Fig. 30a–b ), as well as three fragments of Medieval amphorae (cat. nos. M 40–2, Fig. 30c–e ).

Fig. 30.

(a) M 36; (b) M 37; (c) M 40; (d) M 41; (e) M 42.

Unglazed coarse wares

Two rim fragments belong to smaller cooking pots/jars with similarities in shape with vessels from Corinth, Argos and Nichoria (cat. nos. M 36–37, Fig. 30a–b ).

  • M 36. Jar, rim–neck–shoulder fragment. Fig. 30a .

    Fabric sampled for analyses (no. 98/292). Surface treatment: two grooves on upper exterior body. Shape: straight convergent upper body; heavy everted rim for receiving a lid on the inside.

  • M 37. Jar, rim–neck fragment. Fig. 30b .

    Fabric sampled for analyses (no. 98/290). Surface treatment: smoothed exterior? Shape: convex symmetrical body; concave convergent upper body or neck; straight rim with rounded lip.

  • M 38. Closed vessel, body fragment.

    Fabric sampled for analyses (no. 98/293). Surface treatment: smoothed exterior. Decoration: incised straight and wavy lines on exterior.

  • M 39. Closed vessel, body fragment.

    Fabric sampled for analyses (no. 98/291).

Medieval amphorae

Two fragments of so-called table amphorae were not published in the 1972 volume (cat. nos. M 40–1, Fig. 30c–d ). These are the upper parts of large jars with a concave neck, made in either a micaceous or a calcite fabric, which show similarities in shape with vessels found in the Peloponnese (in Argos) and in southern Italy.

Furthermore, among the unpublished sherds is a large oval handle fragment of an amphora, decorated with a painted broad wavy stripe in a dark brown-reddish colour on top (cat. no. M 42, Fig. 30e ). One can, for instance, distinguish such painted amphorae or jars in shape and decoration at Argos.

  • M 40. Amphora, rim–neck–handle fragment. Fig. 30c .

    Length: 0.122 m; width: 0.112 m; thickness: 0.008–0.009 m.

    Fabric sampled for analyses (no. 08/23). Surface treatment: smoothed exterior and interior. Shape: convex convergent upper body; straight symmetrical neck; everted, flattened rim; oval handle.

    See for shape, Laganara Fabiano Reference Laganara Fabiano2004, 50–1 nos. 1 and 3 (Castel Fiorentino).

  • M 41. Amphora, neck–shoulder fragment. Fig. 30d .

    Length: 0.080 m; width: 0.134 m; thickness: 0.006–0.007 m.

    Fabric sampled for analyses (no. 08/17). Surface treatment: smoothed exterior and interior. Decoration: three incised lines on shoulder. Shape: convex convergent upper body; straight symmetrical neck.

  • M 42. Amphora, handle fragment. Fig. 30e .

    Length: 0.165 m; width: 0.065 m; thickness: 0.016–0.017 m.

    Fabric sampled for analyses (no. 08/22). Surface treatment: smoothed exterior. Decoration: dark red-brown painted wavy stripe on top of handle. Shape: oval broad handle with slight ribbing on top.

    See for similar-looking painted decoration, Piérart and Thalmann Reference Piérart, Thalmann, Déroche and Spieser1980, pl. X nos. D9–D10 (Argos).

Discussion

There are four groups of Medieval fine wares with a glazed surface in the Kastri pottery assemblage: Green and Brown Painted Ware, Slip-painted Ware, Fine Sgraffito Ware and Monochrome Glazed Wares. The shapes are mostly bowls and dishes, suitable for the purpose of food serving and consumption. Only one cup fragment (cat. no. M 10, Fig. 27a–b ) could have been used for beverage consumption and distribution.

Recent research has disputed Charles Morgan's chronology of dating Green and Brown Painted Ware in the late eleventh to twelfth centuries, and Guy Sanders now attributes this type of pottery to around 50 years later in Corinth (Sanders Reference Sanders and Papanikola-Bakirtzi1999; Reference Sanders, Koder and Hild2000; Reference Sanders, Williams and Bookidis2003). The decorated fragments of Green and Brown Painted Ware from Kastri correspond mostly to Morgan's Group III, and should therefore be dated to the second half of the twelfth century (and, in particular, to the third quarter of the twelfth century) (cat. nos. M 1–9, 18–19, Figs. 25a–h , 26a–i and 27h–i ).

Green and Brown Painted Ware is widely distributed in central Greece (e.g. Athens, Boeotia, eastern Phokis, Aetolia) and in the Peloponnese: one can find fragments in large quantities in Corinth, but also in Kenchreai, Nemea, Asea, Sparta, Nichoria, Argos, Mystras, Messene and Crete (see Fig. 31; cf. Vroom Reference Vroom2003, 151 with further literature; Yangaki Reference Yangaki2006, figs. 2–3). Outside Greece, this type of pottery has been recovered in Istanbul, western and southern Turkey, Cyprus (e.g. Kouklia, Paphos, Nicosia), Israel (e.g. Acre), the Crimea (e.g. Chersonessos), the Balkans (e.g. Belgrade, Varna) and Italy (e.g. Venice, Otranto) (Vroom Reference Vroom2003, 151 with further literature; 2005, 83; Poulou-Papadimitriou Reference Poulou-Papadimitriou and Bakirtzis2003, figs. 30–9 for Petra and Herakleion; Violaris Reference Violaris2004, fig. 6; Wartburg and Violaris Reference Wartburg, Violaris, Zozaya, Retuerce, Hervás and Juan2009, fig. 5 nos. 1–4 for Nicosia). Its origins are as yet unknown, although production centres on Mainland Greece have been suggested: among them, Corinth, Nemea Valley, Sparta and Boeotia, of which the last suggestion is the most reliable (Dawkins and Droop Reference Dawkins and Droop1910–11, 27 no. 52, pl. XVII:52 for Sparta; Morgan Reference Morgan1942, 72 for Corinth; Sutton Reference Sutton, Wright, Cherry, Davis, Mantzourani, Sutton and Sutton1990, 655–8, fig. 27c–h, pl. 96e for the Nemea Valley; Armstrong Reference Armstrong and Felsch1996, 126 for Sparta; Vroom Reference Vroom2003, 151 for Boeotia).Footnote 48

Fig. 31.

(a) Distribution map of Green and Brown Painted Ware in the Eastern Mediterranean (dots = sites).

The sherds of Slip-painted Ware found at Kastri belong to the so-called group ‘Light on Dark Slip-painted II’ (cat. nos. M 10–11, 20, Figs. 27a–d,j ; cf. Sanders Reference Sanders1995, 241–2). The earliest examples in Corinth were first dated by Charles Morgan to the eleventh century, but the floruit of this ware appears to be in the twelfth century (Morgan Reference Morgan1942, 95–103 versus Sanders Reference Sanders1995, 240–2). The decorated motifs of the Kastri fragments are similar to the ‘Later Linear Style’ from Corinth, and should be dated to the second half (and, in particular, the third quarter) of the twelfth century.

Wasters of Slip-painted Ware have been found at Corinth, but recent archaeometrical research has shown that not all decoration types were produced here (Megaw and Jones Reference Megaw and Jones1983, 238–9, pl. 25:4 and note 44 there).Footnote 49 Another main production centre is situated in Chalkis, and from there the ware was widely distributed in the Aegean and throughout the Peloponnese: for example, in Nauplion, Isthmia, Sicyon, Asea, Laconia, Sparta, Argos, Mystras and Crete. Outside Greece, Slip-painted Ware of the Middle Byzantine period has been recovered in Istanbul, western Turkey, Cyprus (e.g. Kouklia, Paphos, Nichoria), Israel (e.g. Acre), Bulgaria (e.g. Varna) and Italy (e.g. Venice, Mantova, Otranto) (Sanders Reference Sanders1995, 241–2; Vroom Reference Vroom2003, 150–1; Reference Vroom2005, 80–1).

The origin of Fine Sgraffito Ware was much debated, but now an important production centre has been established in Chalkis (Vroom Reference Vroom2003, 152 with further literature on this debate).Footnote 50 The decorated pieces from Kastri are similar to Morgan's group of ‘Developed Style’, and can be dated between the second half of the twelfth century and the early thirteenth century (cat. nos. M 12–13, Figs. 27e–g ). It is widespread in the Aegean, in the Peloponnese (e.g. Corinth, Argos, Sparta, Mystras and Laconia) and beyond. It has, for example, also been found on sites in Turkey, Lebanon, Syria, Cyprus, Romania, Bulgaria, Russia, Albania, Serbia, Italy and even in Sweden (Vroom Reference Vroom2003, 152 with further literature).

The amphora sherds found at Kastri, published in Coldstream and Huxley Reference Coldstream and Huxley1972, include three groups of Medieval amphorae, among them some fragments and a complete example of the so-called ‘Günsenin type 3 amphora’ or ‘Saraçhane type 61 amphora’ (cat. no. M 33, Fig. 29e ). This last amphora type was widespread all over the eastern Mediterranean, extending from southern France to Syria and from Cyprus to southern Russia (Vroom Reference Vroom2003, 153–5 with further literature; Reference Vroom2005, 98–9 MBYZ 15,1–3). Fragments were even found on some sites in Sweden (e.g. Lund, Sigtuna) (Roslund Reference Roslund, Andersson, Carelli and Ersgard1997, 273–4, fig. 21.3). In general, the amphora can be dated from the (late) twelfth to the (early) thirteenth centuries (in particular the first quarter of the thirteenth century), but its origin of production is still unknown (Vroom Reference Vroom2003, 155 with further literature).Footnote 51

Most of the amphora finds from Kastri have narrow necks, which make them excellent containers for storing and transporting liquids over long distances (cat. nos. M 30–3, Fig. 29a,b,d,e ). Only one of the amphora fragments is decorated with a painted wavy line on the outside (cat. no. M 42, Fig. 30e ). This piece, a large handle fragment of a painted amphora, shows similarities to excavated jars from Argos, which were dated from the first half to the late twelfth century.

The group of unglazed coarse wares from Kastri can be divided between cooking pots/jars (tsoukalia) and basins (lekanai). Most of their forms appear to belong to the Middle Byzantine repertoire, and can be roughly dated in the twelfth and thirteenth centuries. It is interesting to see that there are no fragments of jugs (for pouring and serving) or large jars (for storage, also known as pithoi) in the Kastri pottery assemblage.

The fabrics and shapes of the unglazed coarse wares from Kastri are substantially related to food processing and preparation. The larger cooking pots or jars (cat. nos. M 21–3, Fig. 28a–c ) were surely used with heat, because their gritty fabrics make them resistant to thermal shock. Their open and deep shapes (with wide rims) show easy access to the contents.Footnote 52 Because of their relatively large sizes and open character, these vessels were perhaps used for the slow cooking of liquids or semi-liquids on an open fire or on a brazier, such as the boiling and stewing of large pieces of staple food (Vroom Reference Vroom2008, fig. 13 and table 1).

The wide-mouthed basins (cat. nos. M 24, 27, 29, Fig. 28d–h ), on the other hand, though easily accessible, were used without heating being involved. They were probably multifunctional, with uses varying from the kneading of bread to the chopping, mixing and grinding of ingredients.Footnote 53 As a result of this, the exterior walls of these basins were more finely decorated within the repertoire of cooking utensils with incised and painted wavy lines (cat. nos. M 24, 27, Fig. 28d–g ).

Summarising, one can conclude that the dates of the pottery finds from deposit ψ in Kastri are to be put in the Middle Byzantine period. The fine glazed wares with decoration can be dated in the second half of the twelfth century, maybe also going into the early thirteenth century. However, no fragments of characteristic fine wares of the very late twelfth century and thirteenth century (such as Incised Sgraffito Ware, Champlevé Ware or Zeuxippus Ware) were found in pit ψ, which seems to support a date in the second half of the twelfth century.Footnote 54 The amphorae and coarse wares usually have a longer period of use and should therefore roughly be dated in the twelfth and thirteenth centuries.

At the moment it is difficult to say much about the distribution patterns of the Medieval pottery finds from Kastri, as the origins of most wares are still much in debate. We can assume that the glazed fine wares and the amphorae are probably imported, and that the majority of the unglazed coarse wares were manufactured on the island of Kythera. The imports, on the other hand, reflect networks between Kastri and sites on the Greek Mainland and in the Peloponnese, as well as possible contacts with other parts of the eastern Mediterranean (probably by ship) (see, for the distribution of Green and Brown Painted Ware, Figs. 31a–b ). This happened at a period of time when the settlement of Mitata (further inland from Kastri) came under the control of the powerful Eudaimonoianni family of Monemvasia and when the connections of Kythera with the Greek Mainland and with Monemvasia grew stronger. Furthermore, the merchant fleet of this influential noble family must have facilitated the transport of goods and ceramics between Kythera, the Peloponnese and other coasts of the Aegean (Herrin in Coldstream and Huxley Reference Coldstream and Huxley1972, 46–7 with further references; see also Maltezou Reference Maltezou1982, 206–8; Koumanoudi Reference Koumanoudi, Koumanoudi and Maltezou2003, 87–8).

Supplementary Material

To view supplementary material for this article, please visit http://dx.doi.org/10.1017/S0068245413000117

Footnotes

1

The authors of individual sections are given at the start of each section. The drawings are all by Denitsa Nenova. It has been possible to restudy the material stored on Kythera and we are grateful for all the assistance given to us by Aris Tsaravopoulos, Epimelete of the 26th Ephorate of Prehistoric and Classical Antiquities, and the Phylax of the Kythera Museum, Nikos Kominos. The online version of this article replicates material published in Coldstream and Huxley 1972 and we are grateful to †Nicolas Coldstream and George Huxley, as well Faber and Faber and the Pennsylvania Museum for permission to include it. Dimensions in the first section are in centimetres, in sections two and three in metres.

2 The term ‘glaze’, whether noun or verb, is retained in this section to denote what is alternatively cited as ‘gloss’, though the noun ‘paint’ will be used to relieve monotony, without, one hopes, causing confusion. The verb ‘paint’ reflects the action, not the material, of decoration.

3 The dump from the excavations was revealed after severe flooding of the river in 2000; from what I observed of the material therein, discards for this period may have consisted only of amphora sherds.

4 Four trenches were excavated on the Neck and 12 at Akroterion, the west slope of the Kastri promontory; NK means Neck (trench), AKR is Akroteri, and SF is Small Find. It appears that the trenches were originally designated with Arabic numbers (on the sherds) that were changed to Roman numerals in the final publication. The last (Arabic) number on the sherd indicates the excavation stratum.

5 Although John Hayes had examined the fine wares, Hayes Reference Hayes1972 had not yet been published when Coldstream and Huxley Reference Coldstream and Huxley1972 appeared. I have updated the published fine-ware sherds by substituting Hayes-type numbers for Antioch types, and adding references to Bonifay Reference Bonifay2004, an important contribution to the identification of production centres of African fine wares and amphorae and for the identification of forms at the end of the industry that rarely circulated beyond Carthage. I have limited the references for amphorae to Riley Reference Riley and Lloyd1979 (Benghazi), Panella Reference Panella, Carandini and Panella1973 (Ostia) or Slane and Sanders Reference Slane and Sanders2005 (late Corinth), where fuller lists of comparanda can be found; for a revision of Broneer's Reference Broneer1930 typology of lamps, see Slane Reference Slane1990 and Slane and Sanders Reference Slane and Sanders2005.

6 The results of a petrographic study that will define the main varieties of cooking wares, amphorae, and plain wares using the excavated material as well as finds from the KIP survey will appear in the final report of the survey; samples designated for analysis are indicated here.

7 It is likely that these sherds are those marked Roman, group 4 below.

8 Some marked sherds were found in bags other than those with which they belonged; I am reporting them with their marked deposit and noting the bag to which they had been displaced in a footnote. Because there were only four trenches on the Neck, I am reporting two lamps, of the 1st and probably of the 3rd century, marked 6.4 with AKR Trench VI, which was within the fortification wall on Kastri, and a casserole rim marked AKR 7 A/B.2 with AKR VIIB. These pieces prove that there were Early Roman strata on Kastri as well as on the Neck.

9 They have clearly been divided by ware and do not reflect stratigraphic units. Group 1 is imported fine ware, group 2 is unidentified slip-coated wares, group 3 is lamps, and group 4 is vessels made in corrugated cooking fabric.

10 The scale is incorrect: height 0.25 m is 0.10 m at 2:5.

11 The scale is incorrect: height 0.228 m is 0.091 m at 2:5.

12 Alternative dates for form 67: Reynolds Reference Reynolds1995, 44 (350–425); Bonifay Reference Bonifay2004, 171–3 sigillée type 41A (second half of 4th century), 41B (c.400 and first half of 5th century), and 41C (after 450, second half of 5th century).

13 Alternative dates: Reynolds Reference Reynolds1995, 146 99A–B (475–550); Bonifay Reference Bonifay2004, 181 sigillée type 55A (first quarter of 6th century), 55B (middle decades of 6th century into the early 7th century), 55C (essentially a 7th century form).

14 Alternative dates for form 91C: Reynolds Reference Reynolds1995, 146 (550–575 + ?); Bonifay Reference Bonifay2004, 179 sigillée type 52 (middle decades of 6th century).

15 Alternative dates for form 104A–C: Reynolds Reference Reynolds1995, 146 (525–600 or later); Bonifay Reference Bonifay2004, 181–3 sigillée type 56A1 (late 5th–first third of 6th), 56A2 [=Hayes Reference Hayes1972 104A] (c.525–550), 56A3 (late 6th–middle of 7th century), 56B [=Hayes 104B] (middle and second half of 6th century), 56C [=Hayes 104C] (mid 6th–mid 7th century).

16 Alternative dates for forms 105 and 106: Reynolds Reference Reynolds1995, 146 form 105 (575–625), 106 (525–600); Bonifay Reference Bonifay2004, 183–5 and 208–210 sigillée type 57A (end 6th–first half of 7th century), 57B (middle of 7th century), 57C (second half of 7th century). (In Bonifay's typology form 106 does not exist and late variants of 105C sigillée type 88, and 106 sigillée type 91 are placed at the end of African Red Slip production.)

17 Reynolds Reference Reynolds1995, 147 shows c.450–500.

18 Stored with other lamps in bag 3 of τ.

19 Stored with other lamps in bag 3 of τ.

20 These are the only levels identified between the early Byzantine and Late Minoan I levels except the Late Helladic III debris in Trenches III and VIII.

21 Hübner Reference Hübner1993, 188–9 nos. 45–65 mostly on form 4, pls. 6–7 (dated 170/160 to the middle of the 1st century bc); Rotroff Reference Rotroff1997, 423 nos. 1765–6, pl. 138 (lacking context, similar to Pergamene); Hayes Reference Hayes2008, 206–7 nos. 854–61, fig. 27, pl. 46, mainly first century ad. An as-yet-undated Knidian version of the Pergamene type has also been identified: Kögler Reference Kögler, Briese and Vaag2005, 56, fig. 18.

22 Hayes Reference Hayes1985, 88, pl. XXI:4 (with parallels); Hayes Reference Hayes1991, 45, especially fig. 69:7 from debris of the Hadrianic earthquake.

23 The definitive discussion remains Kenrick Reference Kenrick1985, 327–37, see especially B498.

24 Alternative date for form 80B/99: Reynolds Reference Reynolds1995, 146, very tentatively the whole 6th century; Bonifay Reference Bonifay2004, 180–1 renames the form 99D and dates it to the second half of the 7th century.

25 Bonifay Reference Bonifay2004 (157–9, fig. 89) has reclassified these forms as ‘late A’ types 11 (plate) and 5–10 (bowls); he proposes dates in the second half of the 3rd and 4th centuries rather than 2nd century as Hayes had originally suggested.

26 Reynolds (Reference Reynolds1995) prefers a date in the first half of the 7th century.

27 VIIA.10 is below wall κ.

28 According to Bonifay (Reference Bonifay2004, 166–7, fig. 92), this form has sometimes been regarded as related to Hayes form 58 of c.300, but the example he illustrates was found in a context of c.525–550 at Saint Propice. That is close to the likely date suggested by the decoration and context of φ 12.

29 See Riley Reference Riley and Lloyd1979, 226–7 (citing these amphorae) and Bonifay Reference Bonifay2004, 127–9 spatheion 3, dating almost all examples to the 7th century and distinguishing variants from four sources including Nabeul (like φ 57–φ 58 and φ 64) and Moknine (like the toe φ 63); for examples from Gortyna, where African amphorae account for 12% of the late amphorae, see Martin Reference Martin, Di Vita and Martin1997, 375.

30 Alternative date for form 107: Reynolds Reference Reynolds1995, 146 (575–600 and perhaps later); only late variants are listed in Bonifay Reference Bonifay2004, 210 (burnished examples together with other late African products of the last third of the 7th century). At Corinth, as elsewhere, 107 appears regularly with 109 although in lesser numbers, and Reynolds' date therefore appears to be too early.

31 Alternative dates for form 109: Reynolds Reference Reynolds1995 (580–650); Bonifay Reference Bonifay2004, 186–8 sigillée type 60A–C, of which 60B is the canonical type with rounded lip and burnished interior (middle and second half of the 7th century, perhaps into the 8th century), and 60C with a more triangular lip and similar burnishing on the interior is dated second half of the 7th century.

32 ω 323, the nozzle of an unglazed, wheelmade lamp, probably 2nd century bc, which is bagged with the following sherds, has been omitted. Marked 3385.

33 Chemical analyses that link Knidian grey ware with later Knidian products were reported by Kenrick (Reference Kenrick1985, 502–3, 509, table II). Hayes has recently supplied a comprehensive list of finds from the Knidian peninsula and nearby (Hayes Reference Hayes2008, 62–3).

34 As pointed out above, trench numbers on the Neck go no higher than IV; but AKR VIIC.8 appears to be below the Late Minoan Ib floor in deposit λ.

35 One complete vessel is probably prehistoric.

36 See Coldstream and Huxley Reference Coldstream and Huxley1972, fig. 11 for a plan of Akroterion Trenches VII and VIII. VIIA/B is the baulk that separated VIIA from VIIB.

37 From deposit τ, bag 4.

38 This findspot is suspect. Trench VI, within the fortification wall, was only excavated in 1964 and there is no 6A/B.2A.

39 From deposit τ, bag 3.

40 From deposit τ, bag 3.

41 This period was referred to as early Byzantine in Coldstream and Huxley Reference Coldstream and Huxley1972, but ‘Late Roman’ has become preferred since the publication of Hayes Reference Hayes1972. It emphasises both the cultural and economic continuity of this period with those that precede it and the discontinuity with succeeding periods: from the end of this phase there is also a hiatus in material culture from both the excavations and the survey until the 12th century. See Herrin in Coldstream and Huxley Reference Coldstream and Huxley1972, 42–5.

42 In most instances the coins that accompany these forms are early 7th century as on Kythera, but the Crypta Balbi in Rome and some deposits at Carthage produced coins of the second half of the 7th century.

43 Armstrong (Reference Armstrong and Mango2009) has argued that Cypriot Red Slip and Late Roman Amphora 1 continued to be made and traded until the early 9th century and that 8th century contexts in the Aegean are largely unrecognised because of a dearth of coins following the Arab conquests. While it is possible that the date of the Kythera material might be extended later than the mid 7th century, there is no question that the forms involved already existed by the beginning of the 7th century or even the last quarter of the 6th century.

44 φ 41, φ 55, φ 46 and φ 60 as well as an unpublished example from AKR VIII, layer 2. The type is known from at least the early 5th century to as late as the Yassi Ada wreck, c.625.

45 In addition, an historical introduction to Byzantine Kythera is presented by Judith Herrin in Coldstream and Huxley Reference Coldstream and Huxley1972, 41–51.

46 It is the intention that this new interpretation and discussion of the excavated Medieval ceramics from Kastri will make the identification of the pottery fragments from the ‘Kythera Island Project’ easier.

47 It has been suggested by Ince and Ballantyne (Reference Ince and Ballantyne2007, 28–9, fig. 57a,c) that two fragments of Green and Brown Painted Ware were also found at Paliochora on Kythera, but this is not sure as the painted decoration on these two pieces looks later.

48 See also Armstrong Reference Armstrong1989, 304 and 307 for the suggestion of eastern Phokis as a possible production area. Recent chemical analyses (Waksman and Wartburg Reference Waksman and Wartburg2006) have shown that Green and Brown Painted Ware and other wares, among them Slip-painted Ware and Fine Sgraffito Ware, were produced in a main workshop or group of workshops exploiting very similar clays.

49 Recent archaeometrical research in Corinth by Harriet White (University of Sheffield) has shown, though, that not all her samples come from Corinth.

50 See also Waksman and Wartburg Reference Waksman and Wartburg2006 with new results for a major production area of Fine Sgraffito Ware and other Byzantine glazed ceramics.

51 Recently around eight shipwrecks with Günsenin type 3/Saraçhane type 61 amphorae were found in a survey and excavation area along the south-western coast of the Pagasitic Gulf in Magnesia; cf. Demesticha and Spondylis Reference Demesticha and Spondylis2011, 37–8, fig. 12. Chalkis can now be named as a possible production centre.

52 Their open shapes show some similarities to the later Siphnian cooking pots/jars (tsoukalia), which according to the written sources were manufactured on Siphnos around ad 1700; cf. Tournefort Reference Tournefort1717.

53 We can also see this multifunctional use of basins (lekanai) in 19th and 20th century Greece, where these vessels are used not only for the kneading of bread but also for personal hygiene and the washing of clothes.

54 For a similar composition of Middle Byzantine fine wares in a 12th century pit from Nicosia, cf. Wartburg and Violaris Reference Wartburg, Violaris, Zozaya, Retuerce, Hervás and Juan2009, figs. 5–6. For the distribution of 13th century wares in the Peloponnese see Vroom Reference Vroom2011.

References

REFERENCES

Adamscheck, B. 1979. Kenchreai. Eastern Port of Corinth, vol. IV: The Pottery (Leiden).Google Scholar
Armstrong, P. 1989. ‘Some Byzantine and later settlements in eastern Phokis’, Annual of the British School at Athens 84, 1–47.CrossRefGoogle Scholar
Armstrong, P. 1996. ‘The Byzantine and later pottery’, in Felsch, R.S. (ed.), Kalapodi. Ergebnisse der Ausgrabungen im Heiligtum der Artemis und des Apollon von Hyampolis in der antiken Phokis (Mainz), 334–63.Google Scholar
Armstrong, P. 2009. ‘Trade in the east Mediterranean in the 8th century,’ in Mango, M.M. (ed.), Byzantine Trade, 4th–12th Centuries (Farnham).Google Scholar
Ballance, M., Boardman, J., Corbett, S. and Hood, S. 1989. Excavations in Chios, 1952–1955: Byzantine Emporio (British School at Athens Supp. Vol. 20; London).Google Scholar
Bjelajac, L. 1989. ‘Byzantine amphorae in the Serbian Danubian area in the 11th–12th centuries’, in Déroche, V. and Spieser, J.-M. (eds.), Recherches sur la céramique byzantine (Bulletin de Correspondance Hellénique Supplément 18; Athens and Paris), 109–18.Google Scholar
Bonifay, M. 2004. Etudes sur la céramique romaine tardive d'Afrique (British Archaeological Reports International Series 1301; Oxford).CrossRefGoogle Scholar
Böttger, B. 2002. Kerameikos XVI, Die kaiserzeitlichen Lampen vom Kerameikos (Munich).Google Scholar
Broneer, O. 1930. Corinth 4. Part 2. Terracotta Lamps (Cambridge, Mass.).Google Scholar
Catling, R. 1996. ‘Archaic and Classical pottery’ in Cavanagh, W.G., Crouwel, J., Catling, R.W.V. and Shipley, G.G.J. (eds.), The Laconia Survey II. Archaeological Data (British School at Athens Supp. Vol. 27; London), 33–89.Google Scholar
Coldstream, J.N. and Huxley, G.L. 1972. Kythera: Excavations and Studies conducted by the University of Pennsylvania and the British School at Athens (London).Google Scholar
Dawkins, R.M. and Droop, J.P. 1910–11. ‘Byzantine pottery from Sparta’, Annual of the British School at Athens 17, 23–8.Google Scholar
Demesticha, S. and Spondylis, E. 2011. ‘Late Roman and Byzantine trade in the Aegean. Evidence from the HIMA Survey Project at Pagasitikos Gulf, Greece’, Skyllis 11, 34–40.Google Scholar
Dragendorff, H. 1895. ‘Ein Beitrag zur Geschichte der griechischen und römischen Keramik’, Bonner Jahrbücher 96, 18–155.Google Scholar
Ettlinger, E., Hedinger, B., Hoffmann, B., Kenrick, P.M., Pucci, G., Roth-Rubi, K., Schneider, G., Schnurbein, S. von, Wells, C.M. and Zabehlicky-Scheffenegger, S. 1990. Conspectus formarum terrae sigillatae italico modo confecto (Materialien zur römisch-germanischen Keramik 10; Bonn).Google Scholar
Günsenin, N. 1989. ‘Recherches sur les amphores byzantines dans les musées turcs’, in Déroche, V. and Spieser, J.-M. (eds.), Recherches sur la céramique byzantine (Bulletin de Correspondance Hellénique Supplément 18; Athens and Paris), 267–76.Google Scholar
Hayes, J.W. 1971. ‘A new type of early Christian ampulla’, Annual of the British School at Athens 66, 249–74.CrossRefGoogle Scholar
Hayes, J.W. 1972. Late Roman Pottery (London).Google Scholar
Hayes, J.W. 1980. ‘Problèmes de la céramique des VIIème-IXème siècles à Salamine et à Chypre,’ in Yon, M. (ed.), Salamine de Chypre, Histoire et Archéologie: État des recherches, Lyon, 13–17 mars 1978 (Colloques internationaux du CNRS, no. 578; Paris), 375–87.Google Scholar
Hayes, J.W. 1983. ‘The Villa Dionysos excavations, Knossos: The pottery,’ Annual of the British School at Athens 78, 97–169.Google Scholar
Hayes, J.W. 1985. ‘Sigillate Orientali’, in Enciclopedia dell'Arte Classica e Orientale. Atlante delle Forme Ceramiche II: Ceramica Fine Romana nel Bacino Mediterraneo (Tardo Ellenismo e Primo Imperio) (Rome), 1–96.Google Scholar
Hayes, J.W. 1991. Paphos, vol. 3. The Hellenistic and Roman Pottery (Nicosia).Google Scholar
Hayes, J.W. 1992. Excavations at Saraçhane in Istanbul, vol. II: The Pottery (Princeton).Google Scholar
Hayes, J.W. 2008. Athenian Agora XXXII, Roman Pottery: Fine-ware Imports (Princeton).Google Scholar
Hübner, G. 1993. Die Applikenkeramik von Pergamon (Pergamenische Forschungen 7; Berlin).Google Scholar
Ince, G. and Ballantyne, A. 2007. Paliochora on Kythera: Survey and Interpretation. Studies in Medieval and Post-Medieval Settlements (British Archaeological Reports International Series 1704; Oxford).Google Scholar
Johnston, A.W. 1990. ‘Aegina, Aphaia-Tempel. XIII. The storage amphorae’, Archäologischer Anzeiger 1990, 37–64.Google Scholar
Karivieri, A. 1996. The Athenian Lamp Industry in Late Antiquity (Papers and Monographs of the Finnish Institute at Athens V; Helsinki).Google Scholar
Kassab Tezgör, D. 2003. ‘La céramique fine de l'atelier A1 de Cnide,’ in Abadie-Reynal, C. (ed.), Les Céramiques en Anatolie aux époques hellénistique et romaine (Varia Anatolica 15), 35–43.Google Scholar
Kenrick, P.M. 1985. Excavations at Sidi Khrebish, Benghazi (Berenice). Vol. 3, Part 1. The Fine Pottery (Tripoli).Google Scholar
Kögler, P. 2005. ‘Import, export, imitation. Trade and the economic power of late Hellenistic and early Imperial Knidos,’ in Briese, M.B. and Vaag, L.E. (eds.), Trade Relations (Halicarnassian Studies 3; Odense), 50–62.Google Scholar
Koumanoudi, M. 2003. ‘Illi de Ca'Venier: The first Venetian lords of Kythera’, in Koumanoudi, M. and Maltezou, Ch. (eds.), Venezia e Cerigo. Atti del Simposio Internazionale Venezia, 6–7 dicembre 2002 (Venice), 87–106.Google Scholar
Laganara Fabiano, C. 2004. La ceramica medievale di Castel Fiorentino. Dallo scavo al museo (Bari).Google Scholar
Lawson, J. 1996. ‘Chapter 16. The Roman pottery,’ in Continuity and Change in a Greek Rural Landscape: The Laconia Survey, II. Archaeological Data (British School at Athens Supp. Vol. 27; London), 111–23.Google Scholar
Maltezou, Ch. A. 1982. ‘Le famiglie degli Eudemonoiannis e Venier a Cerigo dal XII al XIV secolo. Problemi di cronologia e prosografia’, Miscellanea Agostino Pertusi 2 (Rivista di Studi Bizantini e Slavi 2; Bologna), 204–16.Google Scholar
Martin, A. 1997. ‘Ceramica comune: vasi da fuoco,’ and ‘Anfore,’ in Di Vita, A. and Martin, A. (eds.), Gortina II, Pretorio, il materiale degli scavi Colini 1970–1977 (Monografie della Scuola Archeologica di Atene e delle missioni italiane in oriente 7; Padua), 346–65, 371–89.Google Scholar
McPhee, I. 1986. ‘Laconian red-figure from the British excavations in Sparta’, Annual of the British School at Athens 81, 153–66.Google Scholar
Megaw, A.H.S. and Jones, R.E. 1983. ‘Byzantine and allied pottery: A contribution by chemical analysis to problems of origin and distribution’, Annual of the British School at Athens 78, 235–63.CrossRefGoogle Scholar
Morgan, C.H., 1942. Corinth vol. XI: The Byzantine Pottery (Cambridge, Mass.).Google Scholar
Panella, C. 1973. ‘Appunti su un gruppo di anfore della prima, media e tarda età Imperiale’, in Carandini, A. and Panella, C. (eds.), Ostia III: Le terme del Nuotatore: scavo dell'ambiente V et di un saggio dell'arem (Studi miscellanei 21; Rome), 460–633.Google Scholar
Papanikola-Bakirtzi, D., Mavrikiou, F.N. and Bakirtzis, Ch. 1999. Byzantine Glazed Pottery in the Benaki Museum (Athens).Google Scholar
Perlzweig, J. 1961. The Athenian Agora VII, Lamps of the Roman Period, First to Seventh Century After Christ (Princeton).Google Scholar
Piérart, M. and Thalmann, J.-P. 1980. ‘Céramique romaine et médiévale’, in Déroche, V. and Spieser, J.-M. (eds.), Études Argiennes (Athens), 463–82.Google Scholar
Poulou-Papadimitriou, N. 2003. ‘Mέσοβυζαντινή κεραμική από την Κρήτην: 9ος-12ος αιώνας’, in Bakirtzis, Ch. (ed.), VIIe Congrès International sur la Céramique Médiévale en Méditerranée, Thessaloniki, 11–16 Octobre 1999 (Athens), 211–26.Google Scholar
Quercia, A., Johnston, A., Bevan, A., Conolly, J. and Tsaravopoulos, A. 2011. ‘Roman pottery from an intensive survey of Antikythera, Greece,’ Annual of the British School at Athens 106, 47–98.Google Scholar
Reynolds, P. 1995. Trade in the Western Mediterranean, ad 400–700 (British Archaeological Reports International Series 604; Oxford).Google Scholar
Ricci, M. 1998. ‘La ceramica comune dal contesto di VII secolo Crypta Balbi,’ in Saguì, L. (ed.), Ceramica in Italia: VI–VII secolo. Atti del Convegno in onore di John W. Hayes, Roma, 11–13 maggio 1995 (Florence), 351–82.Google Scholar
Riley, J. 1979. ‘Coarse pottery’, in Lloyd, J. (ed.), Excavations at Sidi Khrebish, Benghazi (Berenike), vol. 3, part 1 (Libya Antiqua Supplement 5; Tripoli), 91–497.Google Scholar
Robinson, H.S. 1959. Athenian Agora V, Roman Pottery: Chronology (Princeton).Google Scholar
Roslund, M. 1997. ‘Crumbs from the rich man's table: Byzantine finds in Lund and Sigtuna, c. 980–1250’, in Andersson, H., Carelli, P. and Ersgard, L. (eds.), Visions of the Past. Trends and Traditions in Swedish Medieval Archaeology (Lund and Stockholm), 239–97.Google Scholar
Rotroff, S. 1997. Athenian Agora XXIX, Hellenistic Pottery: Athenian and Imported Wheelmade Table Ware and Related Material (Princeton).Google Scholar
Sanders, G.D.R. 1993. ‘Excavations at Sparta: The Roman Stoa, 1988–91. Preliminary report, part 1(c): Medieval pottery’, Annual of the British School at Athens 88, 251–86.Google Scholar
Sanders, G.D.R. 1995. ‘Byzantine glazed pottery at Corinth to c.1125’ (PhD thesis, University of Birmingham).Google Scholar
Sanders, G.D.R. 1999. ‘Corinth workshop production’, in Papanikola-Bakirtzi, D. (ed.), Byzantine Glazed Ceramics: The Art of Sgraffito (Athens), 159–64.Google Scholar
Sanders, G.D.R. 2000. ‘New relative and absolute chronologies for 9th to 13th century glazed wares at Corinth: Methodology and social conclusions’, in Koder, J. and Hild, F. (eds.), Byzanz als Raum: Zu Methoden und Inhalten der historischen Geographie des östlichen Mittelmeerraumes im Mittelalter (Vienna), 153–73.Google Scholar
Sanders, G.D.R. 2003. ‘Recent developments in the chronology of Byzantine Corinth’, in Williams, C.K. II and Bookidis, N. (eds.), Corinth vol. XX, The Centenary 1896–1996 (Princeton), 385–99.Google Scholar
Slane, K.W. 1990. Corinth XVIII, 2, The Sanctuary of Demeter and Kore: the Roman Pottery and Roman Lamps (Princeton).Google Scholar
Slane, K.W. 2000. ‘East–west trade in fine wares and commodities: The view from Corinth,’ Rei Cretariae Romanae Fautorum Acta 36, 299–312.Google Scholar
Slane, K.W. and Sanders, G.D.R. 2005. ‘Corinth: Late Roman horizons’, Hesperia 74, 243–97.CrossRefGoogle Scholar
Sparkes, B.A. and Talcott, L. 1970. The Athenian Agora xii. Black and Plain Pottery of the 6th, 5th, and 4th Centuries (Princeton).Google Scholar
Sutton, R.F. Jr 1990. ‘Appendix: Ceramics of the historic period’, in Wright, J.C., Cherry, J.F., Davis, J.L., Mantzourani, E., Sutton, S.B. and Sutton, R.F. Jr, ‘The Nemea Valley Archaeological Project: A preliminary report’, Hesperia 59, 655–8.Google Scholar
Tournefort, J.P. de 1717. Relation d'un voyage du Levant, fait par ordre du Roy: contenant l'histoire ancienne et moderne de plusieurs isles de l'archipel, de Constantinople, des côtes de la Mer Noire, de l'Armenie, de la Georgie, des frontières de Perse et de l'Asie Mineure ... enrichie de descriptions et de figures d'un grand nombre de plantes rares, de divers animaux; et de plusieurs observations touchant l'histoire naturelle (Paris).Google Scholar
Tsaravopoulos, A. 1999. ‘Graffiti από τα Κύθερα’, Horos 13, 261–6.Google Scholar
Violaris, Y. 2004. ‘Excavations at the site of Palaion Demarcheion, Lefkosia’, Cahier du Centre d'Études Chypriotes 34, 69–80.Google Scholar
Vroom, J. 2003. After Antiquity. Ceramics and Society in the Aegean from the 7th to the 20th Century A.C. A Case Study from Boeotia, Greece (Leiden).Google Scholar
Vroom, J. 2004. ‘The Medieval and Post-Medieval finewares and cooking wares from the Triconch Palace and the Baptistery’, in Hodges, R., Bowden, W. and Lako, K. (eds.), Byzantine Butrint: Excavations and Surveys 1994–99 (Oxford), 278–92.Google Scholar
Vroom, J. 2005. Byzantine to Modern Pottery in the Aegean. An Introduction and Field Guide (Utrecht).Google Scholar
Vroom, J. 2006. ‘Byzantine garbage and Ottoman waste’, in Andrikou, E., Thèbes. Fouilles de la Cadmée II.2. Les tablettes en linéaire B de la Odos Pelopidou. Le contexte archéologique (Pisa and Rome), 181–233.Google Scholar
Vroom, J. 2008. ‘Dishing up history: Early Medieval ceramic finds from the Triconch Palace in Butrint’, Mélanges de l'Ecole française de Rome – Moyen Âge 120–2, 291–305.Google Scholar
Vroom, J. 2011, ‘The Morea and its links with southern Italy after ad 1204: Ceramics and identity’, Archeologia Medievale 38, 351–72.Google Scholar
Waksman, S.Y. and Wartburg, M.-L. von 2006. ‘“Fine Sgraffito Ware”, “Aegean Ware”, and other wares: New evidence for a major production of Byzantine ceramics’, Report of the Department of Antiquities, Cyprus 2006, 369–88.Google Scholar
Wartburg, M.-L. von and Violaris, Y. 2009. ‘Pottery of a 12th century pit from the Palaion Demarcheion site in Nicosia: A typological and analytical approach to a closed assemblage’, in Zozaya, J., Retuerce, M., Hervás, M.A. and Juan, A. de (eds.), Actas del VIII Congreso Internacional de Cerámica Medieval en el Mediterráneo, Ciudad Real-Almagro del 27 de febrero al 3 de marzo de 2006, vol. I (Ciudad Real), 249–64.Google Scholar
Yangaki, A.G. 2005. Le céramique des IVe–VIIIe siècles ap. J.-C. d'Eleutherna: sa place en Crète et dans le bassin égéen (Athens).Google Scholar
Yangaki, A.G. 2006. “Γραπτή εφυαλωμένη κεραμική από την ανασκαφή της αρχαίας Μεσσήνης”, Δελτίον της Χριστιανικής Aρχαιολογικής Eταιρείας 8, 435–44.Google Scholar
Yangaki, A.G. 2008. ‘Céramique glaçurée provenant de Nauplie et d'Argos (XIIe–XIIIe siècles): observations préliminaires’, Bulletin de Correspondance Hellénique 132, 587–616.Google Scholar
Figure 0

Fig. 1.Map of Kastri trenches.

Figure 1

Fig. 2.(a) σ 22; (b) σ 54; (c) σ 9; (d) σ 50.

Figure 2

Fig. 3.(a) σ 57; (b) σ 67; (c) σ 64.

Figure 3

Fig. 4.(a) σ 69; (b) σ 71; (c) σ 74.

Figure 4

Fig. 5.(a) ω 311; (b) ω 317; (c) ω 322; (d) ω 318; (e) ω 319.

Figure 5

Fig. 6.(a) Notaras 1.4, ‘salt cellar’; (b) Classical House, krater.

Figure 6

Fig. 7.Pantile, G 3.

Figure 7

Fig. 8.(a) G 6; (b) G 7; (c) G 8.

Figure 8

Fig. 9.(a) G 9; (b) G 10; (c) G 11.

Figure 9

Fig. 10.(a) G 12 and G 14; (b) G 14; (c) G 13; (d) G 15; (e) G 16.

Figure 10

Fig. 11.Amphorae: (a) G 17; (b) G 18; (c) G 19; (d) G 20; (e) G 21; (f) G 22.

Figure 11

Fig. 12.Cooking pots: (a) G 23; (b) G 24; (c) G 25; (d) G 26.

Figure 12

Fig. 13.(a) G 27; (b) G 30; (c) G 28; (d) G 29.

Figure 13

Fig. 14.(a) G 31; (b) G 32; (c) G 35; (d) G 35 and G 34; (e) G 36; (f) G 38.

Figure 14

Fig. 15.(a) τ 4; (b) τ 5; (c) υ 4; (d) υ 15.

Figure 15

Fig. 16.Lamps: (a) R 1; (b) R 2; (c) R 3. Stamps: (d) φ 12; (e) R 20a.

Figure 16

Fig. 17.Fine wares: (a) φ 9; (b) φ 24; (c) R 5; (d) R 4; (e) φ 25; (f) R 6; (g) R 7; (h) φ 13; (i) φ 12.

Figure 17

Fig. 18.Amphorae: (a) φ 47; (b) φ 46; (c) R 21; (d) φ 54; (e) φ 55; (f) φ 48.

Figure 18

Fig. 19.African spatheia: (a) φ 57; (b) φ 64; (c) φ 58.

Figure 19

Fig. 20.African amphora, φ 72.

Figure 20

Fig. 21.Cooking pots: (a) φ 66; (b) R 8; (c) R 10; (d) φ 70; (e) R 14; (f) R 15; (g) R 9; (h) R 11; (i) R 12; (j) R 13; (k) χ 10; (l) φ 34.

Figure 21

Fig. 22.Plain wares: (a) φ 43; (b) φ 40; (c) φ 39; (d) φ 42.

Figure 22

Fig. 23.φ 56.

Figure 23

Fig. 24.(a) ω 326; (b) R 19; (c) R 18; (d) R 16; (e) R 17; (f) ω 330; (g) R 22.

Figure 24

Fig. 25.(a)–(b) M 1; (c)–(d) M 2; (e)–(f) M 3; (g)–(h) M 4.

Figure 25

Fig. 26.(a)–(b) M 5; (c)–(d) M 6; (e)–(f) M 7; (g)–(h) M 8; (i) M 9.

Figure 26

Fig. 27.(a)–(b) M 10; (c)–(d) M 11; (e)–(f) M 12; (g) M 13; (h) M 18; (i) M 19; (j) M 20.

Figure 27

Fig. 28.(a) M 21; (b) M 22; (c) M 23; (d) –(e) M 24; (f)–(g) M 27; (h) M 29.

Figure 28

Fig. 29.(a) M 30; (b) M 32; (c) Coldstream and Huxley 1972, pl. 87:Q 18; (d) M 31; (e) M33; (f) Coldstream and Huxley 1972, pl. 87:Q 19.

Figure 29

Fig. 30.(a) M 36; (b) M 37; (c) M 40; (d) M 41; (e) M 42.

Figure 30

Fig. 31.(a) Distribution map of Green and Brown Painted Ware in the Eastern Mediterranean (dots = sites).

Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 1

Download Johnston Supplementary Material(Image)
Image 1.9 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 2

Download Johnston Supplementary Material(Image)
Image 2 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 3

Download Johnston Supplementary Material(Image)
Image 2.1 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 4

Download Johnston Supplementary Material(Image)
Image 2 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 5

Download Johnston Supplementary Material(Image)
Image 2 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 6

Download Johnston Supplementary Material(Image)
Image 2 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 7

Download Johnston Supplementary Material(Image)
Image 1 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 8

Download Johnston Supplementary Material(Image)
Image 1 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 9

Download Johnston Supplementary Material(Image)
Image 3.5 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 10

Download Johnston Supplementary Material(Image)
Image 2.7 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 11

Download Johnston Supplementary Material(Image)
Image 4.3 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 12

Download Johnston Supplementary Material(Image)
Image 1.7 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 13

Download Johnston Supplementary Material(Image)
Image 1.8 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 14

Download Johnston Supplementary Material(Image)
Image 1.5 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 15

Download Johnston Supplementary Material(Image)
Image 1.8 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 16

Download Johnston Supplementary Material(Image)
Image 1.8 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 17

Download Johnston Supplementary Material(Image)
Image 1.9 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 18

Download Johnston Supplementary Material(Image)
Image 2 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 19

Download Johnston Supplementary Material(Image)
Image 1.9 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 20

Download Johnston Supplementary Material(Image)
Image 1.8 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 21

Download Johnston Supplementary Material(Image)
Image 2 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 22

Download Johnston Supplementary Material(Image)
Image 832.9 KB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 23

Download Johnston Supplementary Material(Image)
Image 861.6 KB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 24

Download Johnston Supplementary Material(Image)
Image 546.1 KB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 25

Download Johnston Supplementary Material(Image)
Image 3.3 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 26

Download Johnston Supplementary Material(Image)
Image 3.7 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 27

Download Johnston Supplementary Material(Image)
Image 2.9 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 28

Download Johnston Supplementary Material(Image)
Image 2.4 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 29

Download Johnston Supplementary Material(Image)
Image 2 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 30

Download Johnston Supplementary Material(Image)
Image 1.9 MB
Supplementary material: Image

Johnston Supplementary Material

Supplementary Material 31

Download Johnston Supplementary Material(Image)
Image 4.8 MB
Supplementary material: PDF

Johnston Supplementary Material

Appendix

Download Johnston Supplementary Material(PDF)
PDF 84.9 KB