Topsoil, surface detritus and hill-wash

Trenches 1 (C001) and 2 (C501) were both covered by a c. 0.10 m-thick layer of finds-rich topsoil and surface detritus (4.10, 5.7, 5.26, 5.28, 5.34, 6.3 and 6.12). Topsoil deposits in Trenches 4 (C701) and 5 (C301) were more substantial (c. 0.20–0.60 m thick), reflecting their location at the base of a steep slope. These deposits were very mixed and rich in stone and portable finds (e.g. the Early Roman glass beaker Fig. 2.2a, and the marble column Fig. 2.2m), no doubt eroded from the acropolis or upper cavea given the presence of votives and pottery predating the theatre's construction (5.2, 5.16, 5.17, 5.27, 6.1, 6.5 and 6.6).

Beneath the surface layers, deposits of mixed hill-wash/tumbled material were noted in other trenches. Trench 2 is particularly noteworthy as it lies just beneath the acropolis. This deposit (C502) contained a wide range of finds including pre-theatre pottery (5.5, 5.9, 5.14), stamped tiles (6.7, 6.8, 6.9), an Early Roman glass bowl rim (Fig. 2.2b), and a Venetian coin.Footnote ² Together, C501, C502 and C503 contained around half of the metallurgical slag recovered in the 2008 campaign (16 of c. 30 pieces, with more in contexts downslope), strongly suggesting that ironworking took place on or near the acropolis. This is most likely Byzantine in date, noting that ironworking (with much slag) is documented in the settlement over the orchestra (see §4), although three pieces of slag in the packing of the theatre substructure (C906) indicate that metalworking somewhere in the area predated the theatre and was thus likely to be connected with the sanctuary on the acropolis.

Downslope, Trench 4 (C702, C703) (Fig. 2.3) contained few artefacts (noting the iron nail, Fig. 2.2e, typical in most mixed deposits, and the fragment of marble acanthus leaf from a Roman column capital, Fig. 2.2n). However, Trench 5 in the lower central cavea was somewhat richer: C302, C303 and C304 all contained slag; C302 produced worked bone (Fig. 2.2h) and an architectural member (Fig. 2.4); while 4.19 (from C304) exemplifies the Middle Byzantine (late twelfth to early thirteenth century) pottery that dominated these contexts.

Fig. 2.3.

Trench 4 post excavation, from the west.

Fig. 2.4.

Trench 5 post excavation, from the south.

Theatre construction

The substructure of the theatre was partially excavated in Trenches 1–3 in the upper cavea, giving insights into its formation. It was exposed but not excavated in Trenches 4 and 5 in the lower cavea.

Cavea walls

The southern elevation of the north retaining wall of the cavea was exposed in Trench 1 (C015) (Fig. 2.5). Roughly finished and built of randomly sorted, rounded stones in mortar (C042), the wall was surprisingly shallow, measuring only c. 0.60 m wide and 1 m high (six courses), with no indication that further courses had been destroyed (Fig. 2.6). Some 7 m to the south of this wall, a second curvilinear, mortared wall (C7/41) was noted in Trenches 1 and 2 (C506/518) but not explored in either. Both elevations of this wall were exposed in Trench 3 (C905): at 0.50 m wide and standing to a height of 2.1 m, it was much more substantial than C015. It comprises uncoursed, mortared (C916), unsorted, rounded stones and is roughly finished: it is in good repair, although, as would be expected, the top has been somewhat eroded. It is notable that the upper section seemed to contain much larger stones than the lower (Fig. 2.7).

Fig. 2.5.

Plan of Trench 1 pre-excavation.

Fig. 2.6.

Southern elevation of north retaining wall of the cavea, C15/41, in Trench 1.

Fig. 2.7.

Southern elevation of wall in the northern cavea, C905/916, in Trench 3.

Theatre subsoil packing, and tumble (lower cavea)

In Trench 1, the north retaining wall of the cavea sits on C035, a yellow-brown compact silty clay similar to the packing material between the two cavea walls (C005 and C008). Both deposits are similar in colour (yellow to mid-brown) and composition (compact silty/sandy clay with few inclusions). Downslope of C015, deposits C002, C012, C019, C034, C037 and C046 make up the theatre packing. They vary slightly in make-up and colour but are predominantly yellow/green–red/brown fine silty-clays with oxidised clay inclusions, and extremely compacted as though tamped into place. They contain little pottery (e.g. 5.18 from C002).

In Trench 2, C508 lies between the cavea walls, with C516, C517 and C519 making up the deposits downslope of wall C518. As in Trench 1, these deposits consist of very compacted, light coloured, silty clays. Finds were recovered only from C508 but include part of an Archaic terracotta sima from a public building (5.1). This discovery, as findings from Trench 3 (see also §5), confirm Woodward and Hobling's (Reference Woodward1923–5, 252) observation on the quantity of material from the Athena sanctuary contained in fill laid down for the theatre cavea.

In Trench 3, the packing material was more intensively explored: certain contexts previously excavated in ST92/93 IX were identified in section, and some lower deposits were further explored with rewarding results (Fig. 2.8). Ceramic dates are heavily concentrated in Classical–Early Hellenistic, consistent with the construction date for the theatre proposed by Waywell and Wilkes (Waywell et al. Reference Waywell, Wilkes, Powell, Fradgley, Hayes, Walker and Sanders1995, 438, 449–51) based on pottery from their trench ST92/93 IV. The south retaining wall of the cavea (C905) overlay deposit C908. Downslope (south) of this wall the theatre was packed by C906. Finds from these deposits included 5.3, 5.19, 5.21, 5.32, a bronze fibula (Fig. 2.2g), and three pieces of slag which, as noted above, indicate that metalworking predated the construction of the theatre. Upslope of the wall, C903, C907, C910, C911, C912, C913, C914 and C915 were all identified in section, and C909 was excavated. These deposits were largely comparable in colour (yellow/brownish red), with oxidised clay inclusions, but the composition of the upper contexts is more varied than the lower (C906, C908 and C909 are much more uniform in composition and colour). One deposit differed: C904, a mixture of rounded cobbles and angular sandstone (including one piece which had been shaped), may be rubble from the acropolis. Finds from these deposits include 5.4, 5.13, 5.24 and 6.11.

Fig. 2.8.

Western baulk of Trench 3.

Downslope, the more secure deposits in Trench 4 contained a mixture of stony tumble and spolia: C704 overlay a tile-rich deposit C705, which contained a variety of finds, including 4.21–4.23, 4.25–4.28, the marble basin Fig. 2.2o, the bronze fibula Fig. 2.2f, and metallurgical slag. In Trench 5, C305 is similar in colour and composition to the theatre packing: it could also have washed downslope (the late second/first century BC stamped tile 6.10 fits either interpretation), although the ceramic assemblage contains nothing post-Augustan. Beneath C305, a deposit of cobble rich, coarse soil (C307) contained a worked sandstone block (WS12): it overlay a deposit of light brown compacted silty clay (C308), which also contained a large architectural block (WS11). C305, C307 and C308 could not clearly be assigned as original or coherent post-Roman deposits but may instead be tumble from higher in the cavea.

Postholes

Close to the north retaining wall of the cavea in Trench 1 was a shallow, oval, round-bottomed pit-like feature (C029/30), 0.46–0.50 m in diameter and 0.10 m deep, containing ceramics and animal bone. A comparable feature in a similar position in Trench 2 (C510/511), 0.65 m in diameter and 0.09 m deep, contained pottery, tile, and metallurgical slag (Fig. 2.9). Given their respective locations in similar positions in relation to the cavea wall, these pits may relate to the theatre structure, perhaps as bases of eroded postholes and potentially part of an awning covering the upper sections of the cavea. Such an awning should be expected (Montilla Reference Montilla1969), but for its supports to extend beneath the stone seating (as, for example, in the large theatre at Argos or the theatre at Butrint) implies a significant installation (Sear Reference Sear2006, 386–7, 410–11).

Fig. 2.9.

Posthole C511 post excavation.

Terrace wall

Sections of a terrace wall which arc through the west and central areas of the mid cavea are similar in composition and construction. This wall was investigated where it crossed the southern parts of Trenches 1 and 2. Spolia visible outside the trenches were recorded in situ (Fig. 2.10).

Fig. 2.10.

The theatre from the east: the terrace wall is visible in the bottom of Trench 1 and in the central cavea area.

In the south-east part of Trench 1, a dry-stone terrace wall (C009) at least three courses (c. 0.50 m) high ran approximately east–west across the trench. It was built of rounded stone cobbles and tile with occasional architectural spolia. The wall was set into a cutting (C028) through the upper theatre packing (C002, C019), and sat in a loose, dark brown sandy clay (C013). Pottery from C013 (which includes 4.18 and 5.20) gives a terminus post quem of the first quarter of the thirteenth century for the construction of the wall. At the eastern limit of the trench, the wall was partially breached by a large architectural member from the theatre (WS18) which had tumbled downslope (Figs 2.11, 2.12). Upslope, two deposits of mixed tumbled material (C012, C016) had built up behind WS18; finds include 5.12, 5.15 and 5.25.

Fig. 2.11.

East-facing section showing terrace wall C009 in Trench 1 and truncation of the theatre packing.

Fig. 2.12.

Southern elevation of terrace wall C009.

In Trench 2, the wall (C513) was built into a steep vertical cut c. 2 m deep (C514) made through the theatre packing (C508, C516, C517 and C519). The wall is a dry-stone construction with some loose stones, perhaps tumble from C512 and C516, overlying it, and appears much more robust here than elsewhere to allow for substantial material upslope. It contains a sizeable, worked stone (WS13) and, in comparison with other areas, sits in a much deeper cut and is taller (at least five courses high) while roughly the same width (0.35 m). Only a short length (0.37 m) was present in the south-west corner of Trench 2, although it is visible just beyond the western limit of the trench. The wall finishes abruptly and neatly, suggesting a terminus (Fig. 2.13).

Fig. 2.13.

Northern elevation of terrace wall in C513 in Trench 2. Note the large block of worked stone in the centre left of the image.

Radial pathway

At the north limit of Trench 2, north of the cavea wall (C506/7) and south of the acropolis, lay a deposit of loose angular and rounded cobbles and much tile (6.2) in a matrix of compacted, silty clay (C503). This deposit was 1 m wide, 0.20 m deep, and extended for at least 5 m across the entire trench. It was rich in finds, with significant pottery (4.20, 5.22), glass (as the Early Roman aryballos, Fig. 2.2c), slag, and worked stone (including part of a fluted column drum, Fig. 2.14). While such material might suggest a midden, the location of the deposit favours a different explanation (see below). A deposit to its east (C504) with grey, mortared material may be related. Downslope of C503 but north of the cavea wall (C506), a crescent shaped, shallow-sided cut (C505) into the theatre packing (C508) may be a step cut into a rather steep climb. It measured 1.30 m long (east–west) and 0.80 m wide and was excavated to a depth of 0.14 m (Fig. 2.15).

Fig. 2.14.

East-facing section of C503, Trench 2.

Fig. 2.15.

‘Step’ C503 to the right of wall C505, with C507 to the left, from the east.

Immediately south of the cavea wall (C506) the underlying theatre packing (C519) had been removed to a depth of 0.40 m to create a flat level platform (C520), 0.65 m wide and extending across the entire 5 m of the trench. Subsequently, C507 was deposited on this ‘platform’, abutting C506 on its southern (downslope) side. C507 consisted of grey, loose cobbles with adhering mortar, in a silty clay matrix, with pottery: the cobbles largely lay at the base of the feature, with the clay and mortar placed above to bind them and make a smoother surface. Initially we suspected that C507 was tumble from the cavea wall; however, the modification of the theatre substructure to create the platform and the layering of stones with mortar on top indicate a deliberate construction (Fig. 2.16). Although the context pottery is very worn and contains a high proportion of residual material, the presence of two Middle Byzantine sherds is consistent with a late date.Footnote ³

Fig. 2.16.

Eastern section of C507 in Trench 2: note the flat surface (C520) and sequence of deposition of large stones with mortar on top.

These features all directly align along the length of Trench 2 and between the acropolis and the orchestra. Furthermore, the terrace wall is largely absent in this area and has what appears to be a terminus within Trench 2. Taken together, the wall terminus, step, and deposits suggest that a formalised, radial pathway gave access between these two significant areas of settlement after the abandonment of the theatre (Fig. 2.17).

Fig. 2.17.

‘Pathway’ in Trench 2, showing deposits C503, C507 and ‘step’ C505.

Funerary structureFootnote ⁴

South of the cavea retaining wall in Trench 1 (C015), a large pit (C033) with a flat-concave base and shallow sides was cut through the theatre packing (C005, C008). Its maximum dimensions were 4.5 m north–south and 4.8 m east–west, and it was 0.90 m deep.

A rectangular structure (C003) was built into this pit – the pit itself remaining visible to the west and north of it and south of the retaining wall (C015) (see Figs 2.5 and 2.18ab). The external dimensions of this structure were 2.56 m north–south and 2.06 m east–west, creating an internal space of 2.05 x 1.90 m. The walls are preserved to a maximum height of 1.08 m (nine courses). The southern wall was the most eroded, with both the eastern and western elevations eroded towards their southern ends in keeping with the slope of the theatre. The walls were built of randomly coursed rounded and angular stones (including some grey marble with evidence of tooling) bonded with a creamy white mortar (C041). The internal faces on all sides were well finished. The corners were bonded, indicating a one-time construction. The eastern wall is the thickest (0.45 m maximum) and longest. In the south-west corner the structure was truncated by a pit (0.50 x 0.19 m) which was not excavated but may have been a depression left by a large stone in the wall that had eroded downslope.

Fig. 2.18.

Structure C003 in Trench 1: a) pre-excavation; b) post-excavation.

An additional, short wall (C043) (Fig. 2.19a) bonds perpendicular to the western side: it is 0.78 m long and 0.48 m wide, preserved to a height of 0.76 m. It was built of mortared (C047) unsorted squared and rounded stones with both faces left unfinished, suggesting that they were not intended to be seen (Fig. 2.19b). This wall terminates abruptly, with no clear terminus or evidence of destruction/quarrying. Its function is unclear, but its unfinished state may suggest that it belongs to an incomplete structure or annexe.

Fig. 2.19.

Structure C003 in Trench 1: a) post-excavation plan; b) detail of spur wall (C043/047) to the west.

Tomb stratigraphy

Inside the structure, four depositional episodes were identified in three phases (with hints of a fourth). These reflect differing uses. The basal fill of the structure (C031) was sampled to a depth of 0.13 m against the north wall, where the bottom of C003 was exposed (Fig. 2.20). This fill of compact silty clay contained some human bone fragments and pottery.

Fig. 2.20.

Structure C003 post excavation showing test C031, from the west.

The structure was remodelled with the addition of a wall (C038) parallel to the west wall. This was 1.25 m long, 0.42 wide, and 0.55 high (preserved to a maximum of five courses). It was a dry-stone construction with random courses of rounded and angular dressed stone, including a large grey marble block (WS19): it abutted but did not bond with the north wall of C003. Some ephemeral crushed bone was recovered from underneath it. C038 created a 0.17 m-wide space into which a significant quantity of human bone (predominantly long bones and mandibles) was deposited (C032), suspended in a matrix of friable, reddish clay-silt which also contained plain and fine pottery of the first half of the thirteenth century, tile and shell (Fig. 2.21a–c). East of this wall, in the larger chamber (1.90 x 0.85 m), C050 comprised the deposition of damaged human crania deliberately arranged in the northern half of the context: other disarticulated human bones were also noted (Fig. 2.22ab). Near the southern wall was a further small group of long bones. The bones in C050 were suspended in a matrix of yellowish-brown soil which also contained a large Late Roman column capital against the southern wall of the structure (Fig. 2.23).Footnote ⁵ Although C050 was deposited after the construction of C038, the modification of the tomb and the deposition of bones in C032 and C050 plausibly belong to the same phase of activity.

Fig. 2.21.

Phase 1 episode 1a: a) wall partition (C038) and bone deposit (C032); b) photograph from the east; c) after the removal of wall C032, from the east.

Fig. 2.22.

Phase 1, episode 1b: a) deposit of crania and long bones (C050); b) photograph from the east.

Fig. 2.23.

Late Roman column capital (OF41) from C050.

A subsequent inhumation (C048) abuts wall C038. The body lay supine (measuring 1.30 x 0.35 m) oriented north-west–south-east with the right arm across the abdomen. It had been severely truncated (C049), with only the right arm, ribs, pelvis, and legs remaining. Its close proximity to the wall left space for additional burial(s) alongside it, which would have been disturbed (Fig. 2.24).

Fig. 2.24.

Phase 2, episode 3: a) supine inhumation, truncated (C048); b) C048 from the west.

The final phase of activity (C004) was a large deposition of disarticulated bone and some semi-articulated bone groups in very fragmentary condition. Associated pottery indicates that C004 is at least thirteenth century in date (4.29): iron slag likely derives from Byzantine metalworking somewhere close by (see §4). Earlier material in this context includes 5.10, 5.23, the loomweight Fig. 2.2i, and architectural spolia including lapis lacedaemonius cladding (Fig. 2.2l). This deposit covered a maximum area of 1.90 m (north–south) by 0.85 m (east–west), to a maximum depth of 0.25 m. It was overlain by C014, which consisted of dark brown, loose soil with some angular stone inclusions, and residual topsoil (Fig. 2.25ab).

Fig. 2.25.

Phase 3, episode 4: a) truncation of C048 (C049) and deposit of mixed human bone (C004); b) photograph from the south: note the column capital OF41 in the bottom left of the structure.

The phases of activity can be summarised as follows (analysis of the bone remains and further reconstruction of the funerary activities follows in §3):

• Phase 0

• Original construction and use of the tomb (C003 and C031?). The ephemeral bone in C031 may indicate that the bone deposited in Episodes 1a and 1b (below) originated in Phase 0.

• Phase 1: C038, C032, C050

• Episodes 1a: construction of the dividing wall (C038) and packing of long bones and mandibles behind it (C032).Footnote ⁶

• Episode 1b: stacking of crania and other loose bone in the main chamber (C050).

• Phase 2

• Episode 3: placement of an inhumation against the dividing wall (C048), possibly with further burials.

• Phase 3

• Episode 4+: truncation of the left side of the inhumation (C049) and deposition of mixed human bone in one or more events (C004).

Refuse/backfill deposits in Trench 1

The pit to the north and west of the ossuary (C003) and south of the cavea retaining wall (C015) was backfilled with a series of material-rich deposits, some identified only in section. These are described sequentially below (Figs 2.26, 2.27). The latest pottery in all is Early Byzantine, and it is clear that they collectively represent backfilling undertaken as a single act or within a short timeframe. The Early Byzantine ceramic assemblage is coherent and includes large well-preserved sherds with some joins between deposits (see §4 with Fig. 4.9).

Fig. 2.26.

West- and east-facing sections of deposits between the cavea wall (C015) and structure C003.

Fig. 2.27.

East-facing photograph of the deposits north of C003: note the ash rich layer C018.

C024/034 was a 0.15 m-deep deposit of fine, yellow, silty clay. Immediately above it lay C022: orange-brown, compacted silty clay, 0.30–0.50 m deep (5.30, 5.33). C021 abutted C022 on its west side: identified in the northern baulk, it comprised yellow-brown, ash rich silty clay. Above it, abutting the ossuary, lay C021, a small, 0.13 m-deep deposit of light brown, loose silt with few finds but including a coin of Constantine I.Footnote ⁷ C020 was a more substantial deposit (0.12–0.22 m deep) containing a larger quantity of artefacts, including 4.2, 4.3, 5.8, 5.31, and the Roman loomweight Fig. 2.2j. Above it, C018, a 0.14 m-deep deposit of grey, compacted, ashy/clay silt with much stone and charcoal, was very rich in artefacts (4.1, 4.9), many of which were heavily burnt (as the Roman tile in Fig. 2.28). C018 is very similar in composition to C023, identified in a further slot trench to the east of structure C003 and probably from the same context. C023 is a deposit of ash rich material with finds including a whetstone (Fig. 2.2k) and a Late Roman/Early Byzantine wineglass (Fig. 2.2d). The topmost deposit, C006 (0.06–0.16 m deep), contained dark-brown, moderately compact friable clay-silt (4.5, 4.6, 4.13, 4.15, 4.16, 5.6). Over these deposits lay a tumble of medium sized angular and rounded un-mortared stones in a coarse brown clay matrix (C017, 0.60–0.80 m in diameter, 0.16 m deep) which (despite the lack of mortar) probably originated from the cavea wall (C015) to the north-west and was repurposed for backfill (4.12).

Fig. 2.28.

Heavily burned Roman tile from C018.

Excavation of the spur wall of the ossuary revealed other backfill deposits in the north-west of Trench 1. These included: C027, brown sandy clay (4.8, 4.11, 4.14, 4.17, 5.29) and C025 (0.10 m deep). A small, narrow deposit C026 (0.95 m), wedged between C006 and the spur wall C043, comprised compact, yellow-brown, sandy clay, with few finds. To the eastern side of the structure lay C011, rich in rounded stone in a clay matrix, covering an area 1.38 x 0.56 m.

Preliminary osteological results

The key preliminary results of the osteological study are presented below. Since final data analysis is in progress, the full osteological record is not published here, and illustrations (i.e. photographs, graphs and tables) are kept to a minimum. The numbers given below are minimum estimations and subject to slight change upon final publication.

The bone assemblage consists of a large quantity of commingled and largely disarticulated cranial, dental, and post-cranial remains, including approximately 6000 identified bone fragments and 400 teeth. The fairly high level of fragmentation necessitated the reconstruction of many bones from several conjoining fragments. Nevertheless, the material is generally well preserved in terms of weathering and general completeness (Fig. 3.1ab).

Fig. 3.1.

Example of typical state of preservation in terms of weathering and bone completeness: a) humeri from C003; b) femora from C050.

The MNI of the entire assemblage is 29 based on the maximum occurrence of mandibles (25) informed by additional non-adult bones. All categories of skeletal element (cranial, dental, and post-cranial) are represented, including small elements such as hand and foot bones. No strong differentiations were observed in the representation of different skeletal elements or between parts or sides of the skeleton. Analysis of exact bone frequencies, and especially exploration of variation between the MNI estimations according to different bones, is highly significant for the reconstruction of funerary activities (Bello and Andrews Reference Bello, Andrews, Gowland and Knüsel2006). The key patterns observed will be discussed briefly in the next section, and fully explored between and within contexts in Moutafi et al. (Reference Moutafi, Tsempera, Loughlin and Valentein preparation).

The remains of at least 23 adults and six juveniles were recovered (Fig. 3.2). Non-adult material included an adolescent of around 16 years at death, an older child of around 13 years, three children of 10, eight, and five to six years respectively, and a two-year-old infant (Fig. 3.3). The adult remains span all age categories from very young adults (18–20 years old) to older individuals (>50 years old), with the majority falling into prime to mature adulthood (i.e. 30–40 years old).

Fig. 3.2.

Age distribution (N = 29; counts and percentages shown).

Fig. 3.3.

Non-adult bones found in various contexts.

Both sexes are attested in the sample, with females outnumbering males (Fig. 3.4). Based on cranial morphology, 14 females and five males were identified among the adults: a positive sex determination was not possible for the other remains due to a lack of sexually diagnostic elements. Morphological observations in the preserved os coxae and metric observations of the long bones confirm the cranial-based sex ratio.

Fig. 3.4.

Sex distribution (adults only, N = 23; counts and percentages shown).

Only the right part (arm, leg, pelvis and thorax) of the semi-articulated skeleton C048 was preserved in situ (Fig. 2.24b). The skull and the remainder of the skeleton were disturbed and probably mixed with the surrounding commingled remains (bones of the left upper limb were discovered in C004 and positively re-individuated). The skeleton is that of a 30–35-year-old woman, around 1.55 m in stature. Her bones are quite robust, indicating pronounced musculoskeletal activity, with no pathological evidence apart from mild traces of spinal osteoarthritis in the form of slight degenerative changes in her vertebrae (slight porosity, productive changes, and lipping on cervical, thoracic, and lumbar vertebral bodies).

Palaeopathological data for the entire sample will be presented analytically in Moutafi et al. (Reference Moutafi, Tsempera, Loughlin and Valentein preparation), together with further occupational and diet related evidence. In summary, the sample displays normal prevalence rates of common palaeopathological conditions, including dental pathologies (mostly caries and antemortem tooth loss with fewer cases of calculus and abscesses), indications of metabolic disease and diet deficiencies (e.g. cribra orbitalia, dental hypoplasias), trauma, and common degenerative pathologies (e.g. joint disease or osteoarthritis). Evidence for infectious diseases (in the form of non-severe periosteal lesions associated with non-specific bone infections) is very limited.

Reconstructing the burial activities

Clarification of the character of the funerary structure and reconstruction of the burial activities performed are the second major goal of this study. Methods pertaining to funerary taphonomy (analysis of preservation characteristics, dispersal patterns, MNI and bone frequencies) are applied in conjunction with stratigraphic evidence to assess the use of the structure and reconstruct as accurately as possible the nature, frequency and sequence of funerary episodes.

The first question is whether this was a tomb of collective, successive primary burials which subsequently received secondary treatment within the structure, or an ossuary that mostly received disarticulated remains from other graves. The presence of at least one primary burial in situ, partially intact, and a few other instances of semi-articulation noted within the mass of disarticulated bones is positive evidence that the structure was used at least on occasion for this purpose. Bone preservation also indicates that most fragmentation and dispersal took place inside the structure. Despite the high degree of fragmentation, it was possible to find many conjoining fragments especially within the same contexts. Some recent fractures had occurred during excavation, but there were also several old ones indicating fragmentation during the use of the grave. Many bones were reconstructed from several small fragments (up to 10 in some cases). Refitting analysis and re-individuation of the remains suggested the relative independence of the main contexts, but also some meaningful associations between them. No conjoining parts of long bones were found between different contexts, but some bone pairs have been identified, with matches found between C004 and C032, as well as C050 and C032. The relationship between C050 and C032 was also firmly established by the reconstruction of an infant skull dispersed in both contexts. However, no associated material was found between the two contexts of the main chamber (C004 and C050). Finally, bone frequencies also point to the presence of primary interments that subsequently received secondary treatment within the structure.

Bone representation analysis shows a generally consistent minimum number of around 20–25 adult individuals by all main bones, with no pronounced discrepancies. In addition, several small bones are present (such as hand and foot phalanges) which would not be expected in an ossuary of exclusively secondary burials. Such bone frequencies suggest the original presence of several primary burials which were eventually commingled and disturbed due to the continuous and successive use of the structure (Bello and Andrews Reference Bello, Andrews, Gowland and Knüsel2006; Duday Reference Duday, Gowland and Knüsel2006; for further details of criteria, see Moutafi and Voutsaki Reference Moutafi and Voutsaki2016; Moutafi Reference Moutafi2021, 58–61). A more nuanced picture is given when we compare bone frequencies between the different contexts. Long bones are almost equally represented in all three main contexts, though appear slightly fewer in the side compartment (C032). However, skull elements demonstrate significant discrepancies: C004 and C050 share an almost equal number of adult skulls (between 10 and 15 individuals each), while C032 contains only a few cranial fragments coming from one adult and two juveniles. Interestingly, the exact opposite is observed in the representation of mandibles: C004 includes fragments of only three mandibles and C050 of another six, while the side compartment, C032, holds the vast majority (16 in total, most fairly complete).

These patterns suggest that the structure served mostly as a collective tomb rather than an ossuary. The tomb received successive primary burials which then became disarticulated, displaced, and eventually commingled due to continuous, intensive, and prolonged use of the same space. It is possible that some charnel use took place as well, with the tomb receiving at some point secondary remains of burials originally placed elsewhere. This is suggested by subtle differentiation in patterns of bone representation (fully explored and discussed in Moutafi et al. Reference Moutafi, Tsempera, Loughlin and Valentein preparation). Finally, it is evident that the eventual placement of the bones is the result both of accidental disturbance during continuous accumulation of burials, but also of conscious re-arrangements and secondary placements designed to free space for new interments while keeping most bones from previous burials inside the structure. Specific selective practices are evident in this process: for example, crania were kept together in C004 and C050 (Fig. 3.5), but their associated mandibles were moved to the side compartment, C032.

Fig. 3.5.

Assemblage of crania at the back of the tomb in C050.

The next step is to use taphonomic analysis to make a more nuanced reconstruction of the major use phases of the tomb (as set out in §2). It should, however, be borne in mind that the events described below do not represent the full range of funerary episodes or single occurrences but are rather schematic outlines of the major phases of use, each representing the end result of several minor episodes that cannot fully be discerned.

Epilogue

The analytical field recording of the osteological material in C003 offered an unprecedented opportunity to work with principles of funerary taphonomy and archaeothanatology on a Byzantine burial context. The preliminary results presented here have already shed light on the demographic profile of these people, the character of the structure, and the specific acts performed in it. It was shown that the tomb was collective, used as a primary burial locus for successive burials of both sexes and all ages (note however the minimal presence of infants), with continuous internal re-arrangements in order to keep older remains while adding new ones. At some point(s) during its history, the structure probably also received some already disarticulated material from other graves. The forthcoming publication of the complete bioarchaeological study will fully present the osteological data, including biological and taphonomic details, to allow for a deeper contextual discussion of this interesting context in its temporal framework. In the future, we hope to enhance the current results with selective biomolecular analyses (C14 dating, a-DNA and stable isotopes) and cross-examination with other anthropological material from the Sparta acropolis discovered in old excavations (see §4).

Introduction

The Roman theatre was a monument continuously utilised from antiquity to Medieval times. Byzantine builders adapted and modified its form to meet the daily needs of an urban neighbourhood. The acropolis as a whole was enclosed by fortification walls perhaps during the fourth century, transforming the hilltop into the kastron of Lakedaimon. The theatre thus formed part of the intramural settlement of the Byzantine city. The Roman stoa, west of the theatre, was also repurposed with the construction in the Middle Byzantine period of the church and monastic complex once linked to Osios Nikon (Waywell and Wilkes Reference Waywell and Wilkes1994, 424–9; Kourinou-Pikoula Reference Kourinou-Pikoula1998; Armstrong Reference Armstrong, Gallou, Georgiadis and Muskett2008). These monuments document the continuum between the physical fabric of the Roman city of Sparta and the emerging Byzantine style of vernacular and religious urban architecture.

The 2008 excavation sought to investigate how urban needs were accommodated, allowing for the conformation of the theatre area, and to understand the development of the urban fabric here and in relation to the acropolis. In this context, it is worth recalling Robert Ousterhout's (Reference Ousterhout2019, 333) observation that

By focusing on the archaeological evidence brought to light in 2008, my aim in this section is to address exactly these concerns and to contribute to the investigation of daily lifeways in a Byzantine city. I begin by placing the 2008 excavations in the context of British School research on the theatre of Sparta.

Excavations in the Sparta theatre: a multi-period conception of archaeological stratification

Detailed documentation of Byzantine archaeological stratigraphy often seems to be an afterthought in Mediterranean archaeology. It was common practice in the last century to remove post-antique remains with almost no record of stratigraphy or architecture. The goal of bringing to light Classical monuments encouraged the removal of inscriptions, spolia, and architectural features incorporated into Byzantine buildings, with a general disregard for post-antique phases of occupation. The strategies adopted by successive campaigns of the British School at Athens (BSA) in the theatre varied according to the research methodologies of their time, yet from the beginning they stand out for the attention paid to post-antique evidence.

The British School began to excavate at Sparta on 10 February 1906 (Fig. 4.1). Reports published in the School's Annual, unpublished excavation notebooks kept in the School's archive, and the archive of the Byzantine Research Fund (BRF) officially established in 1908 (a unique collection of architectural drawings and photographs of monuments across the Byzantine world) together enable us to trace how, in the context of the Sparta excavations and the wider exploration of Laconia, archaeologists and architects mutually influenced each other's methodology in studying Byzantine heritage. From the late 1880s to the mid-1930s, the intellectual circles around the Arts and Crafts Movement in Britain included archaeologists, art historians and architects with a flowering interest in Byzantine art and architecture. This movement inspired the exploration and recording of Byzantine heritage, enhancing and heavily influencing the direction of Byzantine studies (Kakissis Reference Kakissis2023). Byzantine material culture in Sparta was not simply neglected and discarded. Rather, the British excavations can be seen as pivotal – a laboratory for the study of the diachronic phasing of architectural structures and the appreciation of archaeological stratigraphy as a sequence of phases of accumulation and spoliation.

Fig. 4.1.

Excavation Permit for Sparta, 10 February 1906 (BSA Archive, BSA Corporate Records-Athens: BSA Excavation Permits).

The archive of the British School at Athens is a rich source of data with which to address transformations in the methodology informing the excavation of the multiperiod urban site of Sparta over the past century. Excavation diaries, handwritten notes, photographs, negatives and drawings produced since the beginning of the British School excavations constitute published and unpublished legacy data which may deepen reconstruction of the topography of Lakedaimon.

Early excavators tended to prioritise the periods which were their major object of interest: nonetheless, and perhaps surprisingly, the ‘less crucial’ periods were also recorded. Reporting on the 1906 season, Guy Dickins, a student of the BSA briefly in charge of the Sparta excavations, describes the opening of trial pits along the Byzantine fortification (then standing and visible) in order to ‘discover its relation with the theatre-buildings, and to extract any inscriptions or marble fragments that might be built into it. In this last respect we were fortunate as we began to discover a great quantity of architectural, sculptured, and inscribed marbles’ (Dickins Reference Dickins1905–6, 396). Even though the main research questions guiding the investigation focused on Classical Sparta, the 1905–6 excavation notebook reveals the excavators’ appreciation of the diachronic composition of the archaeological stratification. Notebook 1 in the BSA Sparta archive shows how they distinguished strata by level, with attention to soil composition, brief characterisation of the matrix, and even identification of cuts and pits. Although the excavation methodology was not yet rigorous and scientific, the most significant architectural phases of post-Classical features were noted and illustrated. Moreover, the team clearly understood the complexity of Byzantine occupation of the Roman theatre and were aware that they were excavating examples of vernacular architecture.

Fig. 4.2 shows a notebook entry with the kind of observations which led Dickins to propose a periodisation of the difference phases of post-antique construction and occupation in the theatre area (Dickins Reference Dickins1905–6, 400). Here is a transcription of part of the entry:

Fig. 4.2.

BSA Sparta Excavation Notebook 1 (BSA Archive, Sparta Excavation Records, Sparta 1), p. 18.

The work of Ramsay Traquair, architectural student of the BSA, is equally remarkable. In 1906 he investigated Byzantine and Frankish architectural ‘remains’ in Laconia, recording churches and fortifications. He then continued to explore Greece until 1909 as a member of the BRF team of architects (Traquair Reference Traquair1905–6a; Kakissis Reference Kakissis, Llewellyn Smith, Kitromilides and Calligas2009, 134; Kourelis Reference Kourelis2023). Traquair spent one month in Sparta investigating the Byzantine defensive walls, leading to a report in the School's Annual (Traquair Reference Traquair1905–6b) and an unpublished text (Fig. 4.3). Traquair investigated the topography and construction of the walls. He also examined the use of spolia, highlighting not just their original character as was usual, but stressing the utilitarian scope of reuse and even its aesthetic and decorative value, and paying attention to the location of spolia within the monument (Fig. 4.4). Moreover, he analysed the stratigraphy of the defensive walls, identifying at least three phases of construction and development. He then tried to develop a historical discourse linking this relative sequence to the major events in post-classical Spartan history, in order to provide absolute dates for the various construction phases. He locates the first phase of the fortification in front of the Stoa and relates it to the aftermath of the Herulian invasion in 267 AD. According to Traquair's interpretation, the majority of the walls were built during his second phase, which he dated between the fourth and fifth centuries AD, identifying a terminus post quem in the sack of Sparta by Alaric in 396 AD. Finally, works including the addition of towers are dated between the ninth and tenth centuries, during the revival of the Middle Byzantine empire.

Fig. 4.3.

Traquair's report on the Sparta defensive walls, 11 April 1906 (BSA Archive, Sparta Excavation Records, Sparta 32, side a and b).

Fig. 4.4.

East end of Late Roman fortifications (BSA Archive, BSA SPHS Image Collection, Reference no. BSA SPHS 01/2626.6908; Traquair Reference Traquair1905–6b, 427, fig. 10).

Architecture was not the only object of investigation. In Sparta Excavation Notebook 1, the presence of Byzantine pottery is often recorded when describing the levels excavated, sometimes with its stratigraphic position and a very brief note of quantity. A portion of the Byzantine ceramics recovered in these excavations was kept and, together with material from trenches outside the fortification, informed the pioneering Reference Dawkins and Droop1910–11 article on Byzantine glazed and matt painted pottery by the School's Director R.M. Dawkins and School student J.P. Droop (an expert on Geometric–Classical pottery). As Joanita Vroom (Reference Vroom2003, 35–6) has noted, this study has the great merit of focusing not only on decorative motifs, which were divided into seven groups (five lumping together diverse sgraffito motifs and the final two categorising painted motifs on glazed and unglazed vessels), but also on the morphological characteristics of the pottery. Dawkins and Droop published profile drawings and paid attention to the technology of pottery production, describing the fabrics and firing conditions and presenting a brief characterisation of the clay. While admitting their limited knowledge of Byzantine matters and observing that ‘date and general relations are still matters of some doubt’, they suggested chronologies on the basis of the archaeological evidence from Sparta and comparative material from Cairo, Pergamon and Constantinople on display in the Kaiser Friedrich Museum, Berlin (Dawkins and Droop Reference Dawkins and Droop1910–11, quotation at 23; Kourelis Reference Kourelis2023, 167).

The first excavation campaign concluded in 1910. When investigations resumed in 1924, continuing until 1928 under the direction of A.M. Woodward, attention focused on the orchestra area (Figs 4.5, 4.6). Consistent with the methodologies of the time, Woodward sought to recover the monumental structures of the Roman theatre. He did, however, plan post-antique structures before removing them, as shown in Fig. 4.7. Excavation records show an appreciation of the accumulation of strata and the nature of the post-antique evidence. So, for example, the notebook record of excavation east of the stage in 1925 reports: ‘Byzantine walls forming rooms north of that with a cistern, of characteristic “jumble-style”, including several marbles and a poor granite column standing on a plain rectangular marble base. Floor level here showed nearly a foot-thick layer of broken tiles in very dusty earth, with rather hard clay above’ (BSA Archive, Sparta Excavation Records, Sparta 137, 95). The soil compositions of the various levels and their main contents were recorded, and the nature of domestic space was occasionally discussed. Excavations east of the parodos brought to light a ‘small square structure of rubble, containing an archaic Doric capital built upside down in the centre, associated with many traces of burning: this may actually have been hearth of a Byz. dwelling erected against the wall’ (BSA Archive, Sparta Excavation Records, Sparta 137, 113). This excavation notebook makes clear that most of the orchestra and lower cavea lay under three and in some areas four metres of Medieval occupation. Working in such cramped conditions proved very difficult, as Woodward reported to the press in 1924 (Fig. 4.8).

Fig. 4.5.

General view during excavation east of cavea, 1925 (BSA Archive, BSA SPHS Image Collection, Reference no. BSA SPHS 01/7469.C2591).

Fig. 4.6.

General view of the stage area of the theatre, looking south-east, 1925 (BSA Archive, BSA SPHS Image Collection, Reference no. BSA SPHS 01/7430.C2552; Woodward Reference Woodward and Hobling1923–5, 140, fig. 9).

Fig. 4.7.

Unpublished plan of the theatre showing the Byzantine buildings after the excavation seasons 1924–6 (BSA Archive, Sparta Excavation Records, Sparta 233).

Fig. 4.8.

Extract from an unknown newspaper (BSA Archive, Sparta Excavation Records, Sparta 137, detail p. 28).

Fig. 4.9.

Quantification of ceramics from the bothros in Trench 1 (N = 336).

The record shows that the excavators recognised two phases of post-antique occupation in the theatre area, describing closely packed structures and possible domestic and artisanal complexes including a Byzantine house with an olive press east of the central line of the cavea (BSA Archive, Sparta Excavation Records, Sparta 137, 11). The overall picture is of a settlement on terraces separated by narrow alleyways which in places followed the curvature of the theatre. Yet even though the excavators recorded substantial Byzantine remains, their published reports give little sense of the complexity of the urban topography they had revealed. This is also true of the portable finds. Woodward kept what he regarded as the most significant finds, including an enormous quantity of Byzantine pottery and coins (see e.g. Woodward Reference Woodward and Hobling1923–5, 157–8), but did not publish them.

Some 60 years later, School excavations resumed under the direction of John Wilkes and Geoffrey Waywell. They explored the area of the Roman Stoa (1988–91) and the theatre (1992–8), where reconstruction of the stratigraphy of two deep sections at the edges of the 1920s excavation further documented post-antique phases of activity. These excavations confirmed a prolonged period of Medieval occupation in several areas of the theatre, mainly in the form of small-roomed buildings in stone and mud-brick with tiled roofs, similar to the secular architecture documented in other Byzantine cities and villages in Greece (Waywell and Wilkes Reference Waywell and Wilkes1994; Reference Waywell and Wilkes1999; Waywell et al. Reference Waywell, Wilkes, Powell, Fradgley, Hayes, Walker and Sanders1995; Reference Waywell, Wilkes, Bland, Sidell, Wilkinson and Chandler1997).

This phase of research specifically investigated post-antique material culture, producing publications which have stimulated methodological debate about how reliance on textual, numismatic, or archaeological sources may lead, to paraphrase a famous publication, to ‘different sources, different histories’ (Brandes Reference Brandes, Broglio and Ward-Perkins1999). Sparta was long seen as deserted from the late sixth to the eighth century, devastated by earthquake, epidemics, and invasions (Armstrong Reference Armstrong, Cavanagh, Crouwel, Catling and Shipley2002, 350–2). Since there is a well-known gap in written sources for the Peloponnese between the sixth and tenth centuries, this view relies solely on the Chronicle of Monemvasia, which narrates the invasion of Sparta by Slavic tribes. According to this reconstruction, the city's inhabitants emigrated to the east coast of Laconia (where they founded Monemvasia in an inaccessible coastal location, settling there with their own bishop), as well as to Mount Parnon and to Sicily (Bon Reference Bon1951, 31–64). Sparta then remained more or less unoccupied until the ninth century and the Byzantine military recovery under Nikephoros I. The Chronicle also states that the population of Patras fled to Reggio Calabria, the inhabitants of Argos to the island of Orobe, and the Corinthians to Aigina.

More recent philological analysis of the Chronicle of Monemvasia by Ilias Anagnostakis and Anthony Kaldellis has emphasised structural, verbal and factual parallels with the work of Pausanias. Its tenth-century author, now generally identified as Arethas, based his narrative on Pausanias with the aim of bolstering the authority of the See of Patras to govern that of Lakedaimon. The Chronicle, it is suggested, could also have been used to address competing claims by the Corinthian bishopric. The See of Corinth governed regions with indigenous populations, including the Lakedaimonians who founded Monemvasia, while the See of Patras governed regions that included defeated populations as well as the mixed settlers who migrated to the Peloponnese in the first decade of the ninth century, during the reign of Nikephoros (Anagnostakis and Kaldellis Reference Anagnostakis and Kaldellis2014). This analysis has fostered debate about the validity of the Chronicle as a document of historical events (for a synthesis, see Valente Reference Valente2021).

Archaeological data to challenge the narrative of abandonment were revealed for the first time in Wilkes and Waywell's excavation in the orchestra, in the form of a stratified sequence of domestic occupation dated, thanks to Guy Sanders’ (Reference Sanders1995a) study of the Byzantine material, from the eighth to tenth century. Based on discoveries of so-called globular amphorae (Saraçhane types 5 and 31) in association with handmade cooking ware inconsistent with Roman manufacturing tradition, Sanders suggested that Sparta was not completely abandoned but that local communities coexisted with settlers documented in the archaeological record by so-called ‘Slavic’ pottery. The fact that a non-Byzantine cultural group was for the first time recorded archaeologically in an urban domestic context makes the Byzantine neighbourhood in the theatre of particular interest. However, setting aside assessments of the Chronicle and of the archaeological record, one cannot dismiss the view expressed by Pamela Armstrong (Reference Armstrong, Cavanagh, Crouwel, Catling and Shipley2002, 351) that ‘there was not enough population to support a basic level of civic structures’, although it is worth recalling that Sparta retained the title of bishopric of Lakedaimon during the ecumenical council of 680/1 (Sanders Reference Sanders1995a, 455–7).

As the latest chapter in this long history of research, the 2008 excavation provides further data to enrich the debate. I preface discussion of the ceramic evidence by noting that all material was kept and fully recorded, and that a relative chronology for the ceramic assemblages was defined on the basis of stratigraphy. In two instances, coins provide termini post quem for the pottery with which they are associated (see §2). Absolute chronologies for the pottery presented below have been established by comparison with well-dated contexts in other excavations.

A connected city during the so-called ‘Dark Ages’: the new evidence in context

The oldest Byzantine occupation strata documented in 2008 are found in Trench 1. Immediately to the north and west of the Middle Byzantine tomb, the pit dug to contain the structure was partially preserved (Figs 2.26, 2.27; see §2). This bothros had been backfilled in a series of deposits containing very large sherds, often with sufficient joins to reconstruct entire vessel profiles. In C018 in particular, the matrix consisted largely of ash, large chunks of oxidised clay, and much animal bone. We may therefore speculate that the pit was backfilled using domestic refuse which originated in the vicinity and was at least in secondary if not primary deposition. Fig. 4.9 presents a full quantification of the pottery from all stratigraphic units in the pit: 99 per cent of it consists of utilitarian vessels such as cooking, storage, and tablewares, completely unglazed and very often undecorated. The following catalogue illustrates the main types identified.

• 4.1 (P141). Fig. 4.10. Plate: sherd preserving one-sixth of the rounded rim and one-tenth of the slightly rounded, widely flaring body. A single groove at the base. Greenish yellow glaze inside and out.

• Fine, hard fabric: 2.5Y 8/2 (white). Wheelmade.

• Preserved height 0.06 m, rim diameter 0.16 m.

• Trench 1: C018.

• 4.2 (P164). Fig. 4.10. Beaker: two joining and two non-joining sherds preserving one-third of the flat base and ovoid body; concave neck and everted rim with squared lip (one-sixth preserved). Uneven inner and outer surface, undecorated.

• Fabric medium-hard with a few angular, tabular white and crystalline inclusions: 10R 6/8 (light red). Handmade.

• Base diameter 0.10 m, rim diameter 0.16 m, h. 0.16 m.

• Trench 1: C020.

• 4.3 (P159). Fig. 4.10. Beaker: three joining sherds preserve one-fifth of the rim and body; base lost. Ovoid body, convex shoulder, flaring rim slightly concave inside, and rounded lip. Undecorated.

• Fabric medium-hard with frequent voids and a hackly/irregular break; inclusions abundant, large to very large, angular, spherical, crystalline limestone or quartz, plus a few fine, angular, platy silver schist. 10R 5/8 (red). Handmade.

• Preserved height 0.12 m, rim diameter 0.06 m.

• Trench 1: C020/F010.

• 4.4 (P125). Fig. 4.10. Beaker with almost flat base and straight/slightly flaring lower body, curved profile at maximum diameter (just above median), concave shoulder and flaring rim (thickened on exterior with squared lip). Two non-joining sherds preserve one-fifth of the base and one-sixth of the body and rim. Two incised wavy lines on the shoulder.

• Fabric medium-hard with a few angular, tabular white and crystalline inclusions: 10R 5/6 (red). Handmade.

• Base diameter 0.12 m, rim diameter 0.28 m, h. 0.28 m.

• Trench 1: C006.

• 4.5 (P249). Fig. 4.10. Jar with insloping upper body, thickening towards flaring lip with squared edge. Vertical strap handle attached over lip. Single sherd preserving one-fifth of the rim, one-tenth of the body, and one-sixth of the handle. Undecorated.

• Fabric medium-hard with frequent voids and a hackly/irregular break; inclusions abundant, large to very large, angular, spherical, crystalline limestone or quartz, plus a few fine, angular, platy silver schist. Core 5YR 7/1 (light grey), margins 5YR 6/6 (reddish yellow). Handmade.

• Preserved height 0.064 m, rim diameter 0.14 m.

• Trench 1: C027.

• 4.6 (P139). Fig. 4.10. Beaker: a single sherd preserves one-quarter of the flat base and slightly flaring lower body; upper body and rim lost. Undecorated.

• Fabric medium-hard with frequent, elongate voids and a hackly/irregular break; inclusions abundant, large to very large, angular, spherical, crystalline limestone or quartz, plus a few fine, angular, platy silver schist. 2.5YR 5/3 (reddish brown). Handmade.

• Preserved height 0.049 m, base diameter 0.085 m.

• Trench 1: C006.

• 4.7 (P165). Fig. 4.10. Beaker with sloping shoulder, slightly flaring rim and rounded lip. Five joining sherds preserve one-sixth of the body and one-eighth of the rim. Surface decorated with an irregularly combed motif.

• Fabric medium-hard as 4.6: 2.5YR 5/1 (grey). Handmade.

• Preserved height 0.078 m, rim diameter 0.13 m.

• Trench 1: C006.

• 4.8 (P312). Fig. 4.11. Beaker with flat base, globular body with maximum diameter below neck, flaring rim with concave inner profile, tapered lip. Two non-joining sherds preserve one-sixth of the lower body, and one-fifth of the rim. Undecorated.

• Fabric medium-hard with a few angular, tabular white and crystalline inclusions: 5YR 2.5/1 (black). Wheelmade.

• Restored height 0.043 m, base diameter 0.09 m, rim diameter 0.10 m.

• Trench 1: C027.

• 4.9 (P145). Fig. 4.11. Cooking pot with sloping shoulder and convex flaring rim with squared lip. Single sherd preserves one-tenth of the body and rim and one handle (oval in section, attached over lip). Base and second handle lost. Undecorated. Cf. Sanders Reference Sanders1995a, 453, cat. no. 5.

• Fabric hard with a few angular, tabular white and crystalline inclusions. 10R 5/6 (red). Wheelmade.

• Preserved height 0.054 m, diameter 0.115 m.

• Trench 1: C018.

• 4.10 (P90). Fig. 4.11. Stewpot with sloping shoulder; flaring rim with concave inner profile and squared lip. Single sherd preserves one-seventh of the body and one-eighth of the rim. Grooved shoulder; single groove below rim on exterior.

• Fabric medium-hard with granular, even break; few medium to large, reddish-black to black, sub-rounded, spherical inclusions; rare large, subrounded, spherical red inclusions; few fine, angular, platy silver schist (?) inclusions; few small to medium irregular pores. 10R 4/4 (weak red). Wheelmade.

• Preserved height 0.038 m, rim diameter 0.13 m.

• Trench 1: C001.

• 4.11 (P171). Fig. 4.11. Small cooking pot with flat base and globular body, concave neck, and squared lip outwardly sloping. Thirteen joining sherds preserve one-eighth of the base and one-third of the rim. Grooved upper body and shoulder.

• Fabric medium-hard, similar to 4.10: 10R 5/8 (red). Wheelmade.

• Height 0.65 m, base diameter 0.11 m, rim diameter 0.115 m.

• Trench 1: C027.

• 4.12 (P129). Fig. 4.11. Cooking pot with globular body, concave neck, and flaring rim with squared lip. Single sherd preserves one-quarter of the body and rim; base lost. Wheel-ridged body.

• Fabric medium-hard, similar to 4.10: 2.5YR 5/6 to 6/6 (red to light red). Wheelmade.

• Preserved height 0.091 m, rim diameter 0.115 m.

• Trench 1: C017.

• 4.13 (P126). Fig. 4.11. Stewpot with insloping shoulder, rim with concave profile, and round lip. Single shallow groove around mid-rim. Single sherd preserves one-quarter of the rim and one-sixth of the body; base and handles lost. Smoothed body surface.

• Fabric hard, with smooth break. Few, medium to large, sub-rounded, spherical, cream-coloured lime inclusions; rare small to medium, angular, spherical, white quartz; rare small, angular, platy, abundant silvery sparkling, white mica. Few regularly shaped voids. 10R 3/1(dark reddish grey). Wheelmade.

• Preserved height 0.056 m, rim diameter 0.16 m.

• Trench 1: C006.

• 4.14 (P185). Fig. 4.11. Six non-joining sherds preserve one-third of a slightly rounded base of a stewpot. One-sixth of a tall, slightly flaring lower body. Wheel-made.

• Fabric hard, similar to 4.13: 10R 5/6 (red).

• Preserved height 0.079 m, maximum diameter 0.17 m.

• Trench 1: C027.

• 4.15 (P158). Fig. 4.11. Complete vertical strap handle.

• Fabric hard, similar to 4.13: 5YR 5/4 (light reddish brown).

• Length 0.09 m, thickness 0.009 m.

• Trench 1: C006 F15.

• 4.16 (P137). Fig. 4.12. Globular amphora: single sherd preserves one-third of the neck and rim. Cylindrical neck and thickened rim with concave interior and pointed lip. Oval handles attached from lower rim to lower neck. Wheel-ridged interior; exterior smoothed. Undecorated. Cf. Sanders Reference Sanders1995a, 452, cat. no. 2.

• Fabric very hard, 2.5YR 5/6 (red).

• Preserved height 0.078 m, rim diameter 0.04 m.

• Trench 1: C006.

• 4.17 (P184). Fig. 4.12. Amphora: two non-joining sherds preserve one-quarter of the ovoid body (maximum diameter below median) and half of the carinated shoulder and conical neck.

• Fabric very hard: 2.5YR 6/6 (light red).

• Maximum neck diameter 0.065 m, maximum body diameter 0.23 m, height preserved 0.295 m.

• Trench 1: C027.

Fig. 4.10.

Ceramics from the bothros in Trench 1 (with catalogue numbers).

Fig. 4.11.

Ceramics from the bothros in Trench 1 (with catalogue numbers).

Fig. 4.12.

Ceramics from the bothros in Trench 1 (4.16 and 4.17); Middle Byzantine pottery (4.18–4.22).

The finds from the bothros show close similarities with the stratified sequence revealed in the orchestra (Sanders Reference Sanders1995a). There the oldest, eighth-century contexts were characterised by handmade pottery and globular amphorae. Data from Trench 1 confirm this picture. The contents of the bothros include handmade beakers in a fabric containing organic inclusions (indicated by elongated voids), fired at a low temperature possibly for a short time, so hypothetically in a bonfire. Similarities in the clay paste used to produce handmade cooking ware suggest consistency in the manufacturing process and in clay preparation, despite variation in vessel shape. Wheelmade globular cooking vessels associated with this handmade ware may also have been locally produced, as indicated by fabric analysis in progress at the Fitch Laboratory of the BSA (Valente and Kiriatzi Reference Valente and Kiriatziin preparation). However, we do not yet know where in Sparta pottery workshops of this period were located. As observed at Corinth, Byzantine globular shaped cooking vessels appear to be produced in a clay paste already in use for the Late Roman type (Joyner Reference Joyner2007; Graybehl Reference Graybehl2010). Similarly, the shape continued from the Late Roman period until the late thirteenth century AD, with variations which mainly pertain to the rim profile (Valente Reference Valente2018).

Local wheelmade pots in the bothros are found in association with imported globular stewpots with concave rim in a fabric rich in white mica (4.13, 4.14 and 4.15, Fig. 4.13:1,2). This ware has also been identified in Corinth, where seriation suggests a date from the second half of the eighth to the early tenth century (Valente Reference Valente2018, 81–3). Stewpots in what macroscopically appears to be the same fabric as 4.13, 4.14 and 4.15 were found in the orchestra (Sanders Reference Sanders1995a, 453; material re-appraised by the author). Similar cooking ware was also found in the British survey on Melos (Guy Sanders pers. comm.) and in excavations at Otranto, where stratigraphic evidence suggests an eighth- to ninth-century date, narrowed by rehydroxylation analysis to the first half of the ninth century (Arthur, Buccolieri and Leo Imperiale Reference Arthur, Buccolieri and Leo Imperiale2017, 181–6, cat. no. 4). In 1995, Sanders proposed a similar date for the Spartan contexts in which this ware was found. Petrographic analysis conducted at the University of Salento suggests an Aegean–Asia Minor provenance,Footnote ⁸ while ongoing petrographic investigation of the Sparta theatre material confirms that the fabric recipe is not consistent with Peloponnesian geology, and that the area of origin should be sought within the so-called metamorphic Cycladic unit which spans Southern Euboea through the Cyclades to Samos and western Anatolia. Planned chemical analysis will provide further evidence for provenance, but in the current state of research, a western Anatolian origin is likely, possibly in the Ephesus area. If so, finds from Melos, Corinth, Sparta, and Otranto combine to suggest one of the sea routes that connected the eastern Aegean to the Italian peninsula in the eighth and ninth centuries.

Fig. 4.13.

Byzantine pottery fabrics: 1 and 2: imported cooking ware; 3: White Ware fabric; 4: Laconian Measles Ware; 5: Corinthian Measles Ware (photographs: Rossana Valente).

In Sparta, as in Corinth, this cooking ware is found in association with a very few examples of imported, glazed, so-called White Ware. The specific type found in the bothros is generally dated to the ninth century according to evidence from Saraçhane (Hayes Reference Hayes1992, context 34, no. 29 and context 36, no. 1) and Corinth (Sanders 1995b, cat. no. 69). Petrographic analysis places the Sparta examples in Harriet White's ‘Altered Feldspar’ fabric class (White Reference White2009, 94–8), which comes from the region of Constantinople (Fig. 4.13:3). Four White Ware fragments were found in 2008, and previous publications have also documented the presence of Constantinopolitan glazed table ware in Sparta (Armstrong Reference Armstrong2001; Katsara Reference Katsara and Yenişehirlioğlu2018; Reference Katsara, Petridis, Yangaki, Liaros and Bia2021). No unglazed Constantinopolitan White Ware was found in 2008, though this may reflect the paucity of the material rather than any failure to import this utilitarian ware to Sparta. Imported globular amphoras as 4.16 and 4.17, also found as residual in other contexts excavated in 2008, together with those documented by Sanders (Reference Sanders1995a, 452, cat. nos 2 and 4), further suggest that this inland region of Laconia was connected to maritime trade networks.

A fundamental step if we are to evaluate the diffusion of ceramics and the networks of distribution involved would be to quantify these data. This could also help to challenge the socio-economic picture which has so far dominated historical discourse on these centuries. The coexistence of Byzantine and non-Byzantine material, both locally produced and imported, in this sector of the city allows us to hypothesise that despite the economic and political crisis experienced by the empire and this region, there was a measurable sophistication of supply in urban locations like Sparta. The archaeological record of these centuries is only now starting to become visible in Aegean excavations (Poulou Reference Poulou2018; Reference Poulou, Cau Ontiveros and Florit2019; Reference Poulou, Castrorao Barba and Tanasi2023; Poulou-Papadimitriou Reference Poulou-Papadimitriou, Crow and Hill2018; Poulou and Leontsini Reference Poulou, Leontsini and Poulou2022; Valente et al. Reference Valente, Jackson, Crow and Turner2023), but future research should aim to create quantified datasets to enable systematic estimates of the nature and extent of network distributions in the Aegean, from Asia Minor to the Italian peninsula, in which Sparta too participated. This could also be a fruitful approach to evaluating combinations of, and intersections between, the multiple social variables behind production, consumption, and trade.

Middle Byzantine funerary practices: the 2008 evidence in context

On the basis of the stratigraphic evidence and associated pottery (§2), the Middle Byzantine tomb was built between the last quarter of the eighth century and the first quarter of the ninth and remained in use at least until the late thirteenth century. Within this period, there were four phases of deposition of human remains followed by one layer of anthropogenic material (§3). Consistent with widely documented practice in Byzantine and Medieval funerary contexts, the tomb contained no grave goods (Poulou-Papadimitriou, Tzavella and Ott Reference Poulou-Papadimitriou, Tzavella, Ott, Salamon, Wołoszyn, Musin and Špehar2012).

This is not the only case of post-antique inhumation in the acropolis area. In 1906, excavators noted ‘late tile-built graves in the stoa area’ (BSA Archive, Sparta Excavation Records, Sparta 1, 4) and a Byzantine grave in an imprecisely defined location within the theatre (BSA Archive, Sparta Excavation Records, Sparta 1, 16), and they excavated a burial ‘in a pit S.W. of theatre’, reporting as ‘found with skeleton an iron buckle: Byz. work, diam. 0.5’ (BSA Archive, Sparta Excavation Records, Sparta 1, 45). More graves were excavated in the 1920s, including various tile burials in the theatre area (BSA Archive, Sparta Excavation Records, Sparta 137, 27, 37), one west of the parodos (BSA Archive, Sparta Excavation Records, Sparta 137, 95), and one on the acropolis (‘close to the Byz. wall a byz. burial, feet to N, N-E was above the roman wall’) (BSA Archive, Sparta Excavation Records, Sparta 137, 209). One structure in particular recalls our tomb (imprecisely located, it is likely somewhere north-west of the skene). This is described as having a mortared wall just below the surface along two sides, and a small unfluted column lying across the third side 0.60 m below the surface. An ‘ossuary’ was found against the wall at a depth of 1.25 m, and another mortared wall (at 0.95 m) ran along the south side of the pit. A terminus post quem for the structure is provided by a road beneath it, at a depth of 1.40 m, with a coin of Constantine in its surface and one of Hadrian lower in its fabric. The comparison is tentative because it is unclear whether by ‘ossuary’ the excavator intended a bone receptacle or a bone deposit, although it is not a term generally used in connection with single burials. No illustration is provided (BSA Archive, Sparta Excavation Records, Sparta 137, 139–40).

It is difficult to date these burials precisely from the information recorded in the excavation notebooks. The single graves can be hypothetically assigned at least to the Early Byzantine period, according to the stratigraphy distinguished by the excavators. They appear to be so-called tile-built burials, ‘alla cappuccina’, a type also documented in the Sanctuary of Artemis Orthia (Dawkins Reference Dawkins1929, 48, fig. 27). The presence of single and multiple burials within the city walls would not be a unique occurrence in Byzantine urban settlements. Although intra-mural burial was repeatedly forbidden from the time of Justinian on (Theodosian Code IX 17.6), burials within the kastron populated by vernacular architecture were de facto already common in Byzantine cities. The practice was legalised only in the late ninth and early tenth centuries, during the reign of Leo VI (Novella LIII: Poulou-Papadimitriou, Tzavella and Ott Reference Poulou-Papadimitriou, Tzavella, Ott, Salamon, Wołoszyn, Musin and Špehar2012, 388). At Sparta, however, the relationship between the graves and the urban settlement remains unclear. If the single graves were Early Byzantine, the construction of overlying dwellings in the Middle Byzantine period would not be unexpected. In the forum area of Corinth, for example, the Middle Byzantine city developed over graves (containing single and multiple burials) which had occupied the area since the sixth century (Ivison Reference Ivison, Christie and Loseby1996; Sanders Reference Sanders and Christie2004). In addition to clarifying the date of the Sparta burials, future research might also consider whether the reburial of bones identified in phase 4 of the tomb excavated in 2008 (see §3) might have been occasioned by urban development in this area in the Middle Byzantine period.

Two further multiple burials excavated between 1989 and 1991 are clearly associated with the Middle Byzantine monastic complex in the area of the Roman Stoa. One was found in the northern area of trench RSC1 and the other in trench RSXII. Each contained one articulated skeleton and further disarticulated bones belonging to several individuals (males and females of all ages), carefully placed, with few grave goods (Waywell and Wilkes Reference Waywell and Wilkes1994, 386, fig. 3 [RSXII plan], 388–9 [RSXII], pl. 60).

Finally, excavations in the western part of the theatre cavea conducted by the Ephorate of Antiquities of Laconia in 2014 (ArchDelt 69 [Bʹ 1], 453) revealed three Byzantine tile graves a few centimetres below topsoil, oriented east–west, and apparently without grave goods. Two contained single inhumations (of an adult and a sub-adult). In tomb 3, which is the only one to preserve a section of the tile covering, two primary burials were revealed (of an adult and a child) plus one instance of a secondary burial. The excavators propose an Early Byzantine date, although a later date remains possible.

Middle Byzantine Sparta

The 2008 excavation did not provide clear evidence of Middle Byzantine vernacular or religious building. Yet the material evidence, despite being found in secondary deposition, still contributes to discussion of the economic growth experienced by Sparta at this time.

As noted in §2, the discovery of numerous iron slags provides further evidence of artisanal activity in the settlement occupying the theatre area, with 16 of the c. 30 slags discovered coming from Trench 2 (C501, 502 and 503). Slags of kiln-type (C503) and iron forging type (C302, Trench 5) were found, the latter with the typical flat base and concave upper surface, voids oriented towards the surface, and metal inclusions of different sizes. Chemical analysis will be required to determine whether this slag type relates to iron forging or refining.Footnote ⁹ These iron slags provide clear evidence of a nearby metal workshop, since the iron forging type in particular was too large and heavy to be transported far. A metal workshop (as well as a kiln) documented during excavation of buildings in the west orchestra and lower cavea in 1992 and 1993 (trench ST 92/93 I) relates to the ninth- and tenth-century first phase of occupation in this area (Waywell et al. Reference Waywell, Wilkes, Powell, Fradgley, Hayes, Walker and Sanders1995, 439, 446–7).

Ceramic evidence too supports the picture of an economically prosperous Sparta rich in public and private buildings and with a strong artisanal tradition especially during the twelfth and thirteenth centuries. So-called Measles Ware, a distinctive painted sgraffito ware dated to the second half of the twelfth century, provides a starting point. Local productions have been identified in Corinth on the basis of petrographic analysis (White Reference White2009), in Argos on the basis of macroscopic fabric analysis and kiln wasters (Vassiliou Reference Vassiliou and Yenişehirlioğlu2018; Reference Vassiliou2022), and in Sparta, supported by our ongoing petrographic study (Sanders Reference Sanders1993, 255, 267; Bakourou, Katsara and Kalamara Reference Bakourou, Katsara, Kalamara and Bakirtzis2003, 234; Valente and Kiriatzi Reference Valente and Kiriatziin preparation). Measles Ware of Corinthian production has also been petrographically documented among the 2008 finds in the form of plates manufactured in the so-called ‘clay-temper’ fabric recipe also used at Corinth for the production of Slip Plain Glazed, Green and Brown Painted I and II, Dark on Light Slip Painted Ware, and Fine Sgraffito Ware (Valente and Kiriatzi Reference Valente and Kiriatziin preparation) (Spartan fabric: Fig. 4.13:4; Corinthian fabric: Fig. 4.13:5). Whether these ceramics attest to a regular trading network between cities or more sporadic connections is a matter for future research, as it is not yet possible to quantify their distribution. However, debate about this ware is definitely reopened. At this stage in research, it is preferable not to assign Measles Ware found outside these cities to Corinth by default, since fabric analysis is required to shed light on the complex trading network linking the Peloponnese with the coasts of the Adriatic (so far this ware is identified mainly in Italy and Albania: see Vassiliou Reference Vassiliou2022 for an updated overview). Pamela Armstrong (Reference Armstrong, Cavanagh, Gallou and Georgiadis2009, 318 n. 50) has, for example, suggested a Spartan origin for the Measles Ware found in Otranto on the basis of its macroscopic description.

In addition, various examples of so-called ‘Middle Byzantine Production’ from Chalkis have been documented petrographically in Sparta (Waksman et al. Reference Waksman, Kontogiannis, Skartsis and Vaxevanis2014) and identified among finds from Trenches 1, 2, 4 and 5. These include examples of Green and Brown III and Incised Sgraffito tableware (Fig. 4.15:1). Günsenin type III amphorae are shown petrographically to have been imported from Chalkis (Waksman et al. Reference Waksman, Skartsis, Kontogiannis, Todorova and Vaxevanis2018) and are also found in late twelfth- and thirteenth-century contexts from the 2008 excavations (Fig. 4.14:2).

Fig. 4.14.

Byzantine pottery fabrics: 1: ‘Middle Byzantine Production’ from Chalkis; 2: Günsenin type III amphorae; 3: local Middle Byzantine glazed wares; 4: cooking wares (photographs: Rossana Valente).

Fig. 4.15.

Middle Byzantine pottery (with catalogue numbers).

Glazed tableware hypothetically of local production appears to span the second half of the twelfth to the thirteenth century. In addition to Measles Ware, there may have been a local production of Slip Plain Glazed Ware, Sgraffito, and Incised Sgraffito Wares, although no archaeological evidence of workshops has yet been found in Sparta (Fig. 4.14:3). Coarse and cooking wares further enriched the repertoire of utilitarian vessels, sharing traits with the Byzantine cooking wares documented elsewhere on the Greek mainland. Byzantine ceramic cooking pots were closed vessels: open shapes disappear from the repertoire and were perhaps produced in other materials (as suggested by other visual evidence: Bakirtzis Reference Bakirtzis2003). Ongoing petrographic research suggests that the Middle Byzantine stewpots presented here are made in fabric recipes which, while diverse, are consistent with local geology. We may therefore speculate about local production, even though to the author's knowledge cookware wasters have yet to be recorded in Sparta (Fig. 4.13:4).

Utilitarian vessels for storage purposes, as those with incised handles (Sanders Reference Sanders1993, cat. nos 35, 36, 38 and 46), were widely recorded in the 2008 excavations but in very fragmentary condition. The catalogue below includes two examples of coarsewares not previously documented in Sparta (4.29 and 4.30).

• 4.18 (P85). Fig. 4.12. Dish with straight flaring body. Single sherd preserves one-twentieth of the body; base and rim lost. White slip inside; a sgraffito eagle head; greenish yellow glaze.

• Fabric soft to medium-hard with granular, even break; frequent to common fine, sparkling, angular, platy inclusions (silver schist?); rare medium to large, angular, platy silver schist; rare voids. 2.5YR 6/6 (light-red)

• Length 0.026 m, width 0.03 m.

• Trench 1: C013.

• 4.19 (P212). Fig. 4.12. Jug with cylindrical neck, flaring rim, and rounded lip. Vertical strap handle over mid neck, oval in section. Three joining sherds preserve complete neck, one-third of the rim, and one-fifth of the handle. Missing body and base. Double ridges on the lower and mid neck. White slip, colourless overglaze, and sprinkles of brown glaze on the exterior extending inside over the rim.

• Medium-hard fabric, similar to 4.18: 2.5YR 6/8 (reddish yellow).

• Preserved height 0.055 m, rim diameter 0.05 m.

• Trench 5: C304.

• 4.20 (P127). Fig. 4.12. Bowl with concave, flaring body and T-shaped rim. Single sherd preserves one-tenth of the body and rim; base lost. Double groove on the exterior. White slip inside and out, green glaze over the entire vessel.

• Medium-hard fabric, similar to 4.18: 2.5YR 6/8 (reddish yellow).

• Preserved height 0.029 m, rim diameter 0.27 m.

• Trench 2: C503.

• 4.21 (P213). Fig. 4.12. Plate with slightly flaring ring foot, almost flat undersurface, and widely flaring, convex body and rim with rounded lip. Single sherd preserves a quarter of the base and one-fifth of the body and rim. Painted fine sgraffito ware. White slip inside extends over lip, dripping onto the exterior. Concentric circles in the centre define a medallion, double circles on lower body enriched with triangular and semi-circles decorated with cross-hatched decoration, vertical hatches on rim. Painted strokes of green glaze inside.

• Medium-hard fabric, similar to 4.18: 2.5YR 6/8 (light red).

• Preserved height 0.055 m, foot diameter 0.095 m, rim diameter 0.21 m.

• Trench 4: C705.

• 4.22 (P198). Fig. 4.12. Straight flaring rim with rounded lip. Single sherd preserves one-eighth of the rim; base and body lost. White slip inside extends over lip. Incised crossed-hatched decoration on rim, yellow glaze.

• Fabric medium-hard, similar to 4.18: 2.5YR 6/8 (light red).

• Preserved height 0.01 m, rim diameter 0.23 m.

• Trench 4: C705.

• 4.23 (P199). Fig. 4.15. Dish with shallow ring foot with rounded edges, flat underbase, shallow convex, widely flaring lower body. Single sherd preserves complete foot and one quarter of the body; rim lost. White slip inside, sgraffito rabbit in sitting position, band decorated with incised vertical lines on lower body. Greenish yellow glaze inside.

• Fabric medium-hard with smooth even break; few medium to large, sub-rounded, spherical, cream-coloured lime (?) inclusions; rare small to medium, angular, spherical, white quartz; rare small, angular, platy, silvery sparkling schist. 5YR 6/4 (light reddish brown).

• Preserved height 0.09 m, foot diameter 0.112 m.

• Trench 4: C705.

• 4.24 (P20). Fig. 4.15. Bowl with straight, flaring foot, almost flat underside, flaring lower body. Single sherd preserves one-quarter of the foot and one-tenth of the body; rim lost. White slip inside, incised medallion at centre decorated with wavy lines alternated with glaze painted strokes. Yellow glaze inside.

• Fabric medium-hard, as 4.23: 5YR 6/8 (reddish yellow).

• Preserved height 0.04 m, foot diameter 0.10 m.

• Trench 2: surface clearance.

• 4.25 (P192+195). Fig. 4.15. Bowl with flaring lower body curving up to vertical upper body, flaring rim with tapered lip. Two joining sherds preserve one-tenth of the body and one-sixth of the rim: foot and lower body lost. White slip on interior, thinner on exterior. Incised band around body decorated with hatched motif; incised wavy line on lip with dotted motif. Yellow glaze inside.

• Fabric medium-hard, as 4.23: 5YR 6/8 (reddish yellow).

• Preserved height 0.068 m, rim diameter 0.26 m.

• Trench 4: C705.

• 4.26 (P308). Fig. 4.15. Small cooking pot with almost globular body, sloping shoulder curving sharply up to vertical rim, outwardly thickened and triangular in section, inwardly sloping lip. Two joining sherds preserving one-tenth of the body and rim; base lost. Grooved shoulder.

• Fabric medium-hard, granular/even break; inclusions few medium, sub-rounded, tabular white quartz; few, medium to large, rounded, tabular to spherical mudstone; rare fine, angular, platy schist; few, irregularly shaped voids. 5YR 5/6 (yellowish red). Wheelmade.

• Preserved height 0.07 m, rim diameter 0.11 m.

• Second half 12th century AD.

• Trench 4: C705.

• 4.27 (P260). Fig. 4.15. Round, grooved shoulder curving up to straight, slightly flaring rim with pointed lip. Single sherd preserves one-eighth of body and rim: base lost.

• Fabric medium-hard, as 4.26: 5YR 5/6 (yellowish red). Wheelmade.

• Preserved height 0.04 m, rim diameter 0.11 m.

• Trench 4: C705.

• 4.28 (P259). Fig. 4.15. Stewpot with sloping shoulder curving up to vertical, thickened rim with squared lip. Single sherd preserving one-tenth of the body and rim; base lost. Single groove around mid-rim, grooved shoulder.

• Fabric medium-hard, as 4.26: 5YR 5/6 (yellowish red). Wheelmade.

• Preserved height 0.051 m, rim diameter 0.21 m.

• Trench 4: C705.

• 4.29 (P166). Fig. 4.15. Cooking ware stamnos with combed decoration. Numerous fragments from body and shoulder, missing base, rim, and handle. Convex body, sloping shoulder. Parallel stumps of horizontal handle, round in section, on shoulder. Rib with incised hatches on upper shoulder at neck attachment. Combed wavy line on maximum diameter and upper body, intersecting multiple bands of combed lines.

• Fabric very hard with frequent angular, tabular white and crystalline inclusions, rare subrounded-spherical black inclusions. Core 2.5YR 2.5/2 (very dusky red), edges 2.5YR 3/0 (very dark grey). Wheel-made.

• Height 0.134 m, width 0.114 m, thickness 0.04 m.

• From a context dated stratigraphically to the first half of 13th century. No comparative material is known to the author, although the condition of the vessel suggests that it is not residual. A Middle Byzantine date is therefore proposed, possibly second half 12th–13th century AD.

• Trench 1: C004.

• 4.30 (P186). Fig. 4.15. Three joining fragments preserve the almost complete profile of a pedestal plate, with flaring, high foot, widely flaring almost straight body, and rounded lip. Medallion impression with concave profile at centre.

• Fabric medium-hard, similar to 4.18: 2.5YR 6/8 (reddish yellow). Wheelmade.

• Preserved height 0.055 m, rim diameter 0.15 m.

• Pedestal plates are a typical shape in the Byzantine dining set. They are generally glazed, but unglazed examples are also documented which replicate the shape and fabric of the contemporary glazed examples.

• An 11th-century date is indicated by comparanda from Corinth (Valente Reference Valente2018, nos 321, 322), although 4.30 comes from a context dated to the first half of the 13th century.

• Trench 1: C004.

• 4.31 (P15). Fig. 4.15. Günsenin type I amphora. A single fragment preserves one-sixth of a vertical handle, triangular in section with a deep central groove, each side pressed down into a concavity. Neck hollowing preserved at the top. Surface compacted and smoothed, no additional treatment.

• Fabric medium-hard, fine with rare subangular-oval black and angular-rounded white and crystalline inclusions: 5YR 6/8 (reddish yellow).

• Preserved length 0.064 m, width 0.040 m, thickness 0.014 m.

• From a context dated to the modern period. Günsenin type I amphorae date from the 10th–12th century AD: Günsenin Reference Günsenin1990, fig. 40; Valente Reference Valente2023, 795 for a synthesis.

• Trench 2: C001

The neighbourhood which developed over the orchestra and lower cavea of the theatre in the Middle Byzantine period featured buildings separated by alleys which generally followed a radial plan dictated by the profile of the cavea. Despite some indications that buildings encroached into these alleys, the overall impression is of a planned settlement. By contrast, the upper cavea appears to have been left unoccupied. Research at the so-called church of Osios Nikon (a three-aisled basilica with a triconch sanctuary on the eastern hill of the acropolis) by the BSA and the Ephorate of Antiquities of Sparta identified an eleventh-century phase developed on a mid-sixth- to early seventh-century basilica (Sweetman and Katsara Reference Sweetman and Katsara2002; Katsara Reference Katsara, Cavanagh, Gallou and Georgiadis2009; Sweetman Reference Sweetman, Cavanagh, Gallou and Georgiadis2009). On the south slopes of the acropolis, a market area was defined by a two-storey colonnade.

Both combining evidence from the 2008 excavation with the results of over a century of research into post-antique activity in the theatre area and setting this into the larger context of research within and beyond the city walls reveal the complex urban topography of Byzantine Sparta. Recent work by Evi Katsara (Reference Katsara, Petridis, Yangaki, Liaros and Bia2021) has shed new light on topography outside the walls and the wealth of material culture associated with it, drawing on evidence from excavations conducted by the Ephorate of Antiquities of Laconia. The urban development of Sparta emphatically disproves the idea that Byzantine ‘building occurs without any obvious planning and, lacking any deliberate architectural scheme, materializes dynamically to meet needs as they arise’ (Bouras Reference Bouras1983, 22). Instead, it appears to indicate that a new fortified city developed as a centre of power overlying the ancient acropolis. The new urban topography saw the location of manufacturing activities (notably metalworking) within the settlement and the concomitant development of a vernacular architecture that could accommodate private daily life as well as craft activities. Burial intra muros appears as a regular phenomenon. Monumental construction centred on churches. This urban fabric was not a continuum but rather a series of nuclei within and outside the city walls.

This state of affairs did not end suddenly with the Frankish conquest of the Peloponnese, when control of Sparta passed to the Principality of Achaia. However, when Laconia was reconquered in 1262 and the Byzantine Despotate of the Morea created, the castle of Mystras (founded in 1249) became the new seat of administration for the entire Byzantine-controlled Peloponnese. By the last quarter of the thirteenth century, Mystras had become the political and economic heart of Laconia, the movement of the bishopric leaving the previously prosperous city socially and economically marginal. On present evidence, artisanal activities, including pottery production, continued up to the early fourteenth century (Sanders Reference Sanders1993; Dimopoulos Reference Dimopoulos, Böhlendorf-Arslan, Uysal and Witte-Orr2007). But this is beyond the scope of the 2008 excavation results.

Architecture

• 5.1 (OF46). Fig. 5.1. Two non-joining pieces of a raking sima. Vertical plaque with cut-out border of triangular teeth above a torus. Exterior monochrome with no evidence of polychromy or incision. Mantle attached just below the level of the torus; patches of paint preserved. The function of the hole pierced through one ‘tooth’ is obscure.

• Fabric fine, with many large black, red, and white stony inclusions, micaceous: 2.5YR 7/6 (light red). Paint dark purple-brown.

• A: height 0.107 m, width 0.162 m, thickness 0.027–0.05 m. B: height 0.088 m, width 0.07 m, thickness 0.026–0.033 m.

• Late 7th century: c. 630–620 BC.

• 5.1 recalls the Laconian Type I sima of the original roof of the early temple of Artemis Orthia (Winter Reference Winter1993, 98–100, 104–5, fig. 11; Skoog Reference Skoog1998, 46–7), but lacks the cavetto moulding between the teeth and the torus on the outer side or below the torus. It belongs on the transition between monochrome roof elements without a tooth border (among the earliest at Sparta) and polychrome with teeth, torus, and cavetto: it is dated accordingly (Nancy Winter pers. comm.). Similarly transitional traits are evident on the disc acroterion of Winter's Type I variant 2 (c. 620) from the sanctuary of Artemis Orthia (Winter Reference Winter1993, 102; George and Woodward Reference George and Woodward1929, 120, 137, cat. no. 16, fig. 90), which resembles 5.1 in its paint and forming.

• 5.1 must come from an Archaic public building on the acropolis. Its proposed date is consistent with roof elements previously recovered (notably the acroterion of Winter's Type II 2a, c. 625/620–580 BC: Woodward Reference Woodward1926–7b, 40–2, fig. 1; Skoog Reference Skoog1998, 51–2, 157, cat. no. 11), which may be associated with the temple of Athena Chalkioikos or more probably (on grounds of size) with the further temple of Athena excavated by Woodward (Reference Woodward1926–7b, 39–43) south of the Chalkioikos sanctuary (Spallino Reference Spallino, Longo, Di Cesare and Privitera2016, 703–6).

• Trench 2: C508.

Fig. 5.1.

Artefacts from the acropolis sanctuary (with catalogue numbers).

Votives

• 5.2 (OF35). Fig. 5.1. Bronze bird pendant with suspension loop on the back; square section shaft with ring above the lower break. The bird has a small, rounded body, near horizontal, with two diagonal grooves on the front marking wings; the neck and head are upstanding; the beak is sharply pinched; there is a groove around the back of the head and a small, raised disc on top.

• Height 0.035 m, depth 0.027 m.

• Early Iron Age.

• 5.2 recalls both the category of bird pendants with back loop described by Kilian-Dirlmeier (Reference Kilian-Dirlmeier1979, 128–49, 261) and hammer pendants (as her 49–50, no. 307, pl. 20 [from Tegea]); see also Woodward and Hobling (Reference Hobling1923–5, 270, 274, fig. 5, no. 11 from the Sparta acropolis). However, the combination of back loop, shaft, body form, and modelling is unparalleled.

• Trench 4: C701.

• 5.3 (OF11). Fig. 5.1. Bronze votive double axe with part of the haft.

• Width 0.022 m, height 0.021 m, maximum thickness 0.008 m, haft length 0.017 m, haft thickness 0.003 m.

• Protogeometric–6th century BC (not closely datable).

• As Kilian-Dirlmeier Reference Kilian-Dirlmeier1979, 248–9, 252–4, pl. 94:type B, cat. nos 1609, 1613 (from Geometric deposits at the sanctuary of Artemis Orthia).

• Trench 3: C908.

• 5.4 (OF6). Fig. 5.1. Lead wreath; reed/spike leaf type. Three fragments comprising near complete circle, outer edge lost.

• Diameter c. 0.018 m.

• c. 620–250 BC.

• As Boss Reference Boss2000, 112–13, motif 34, 1, fig. 82, type 1222 (but without ball attachments); Braun and Engstrom Reference Braun and Engstrom2022, 208–9, fig. 2 (as they note, spike wreaths without context should be dated broadly to Lead II–VI; cf. Boss Reference Boss2000, 148–75).

• Trench 3: C909.

• 5.5 (P86). Fig. 5.1. Miniature terracotta pedestalled amphoriskos with three handles. Top hollowed out; body solid; base lost. Surface compacted with no added colour.

• Fabric fine, with small–medium mainly white inclusions: 5YR 6/6 (reddish yellow).

• Height 0.055 m, maximum diameter 0.047 m, thickness rim 0.006 m.

• Archaic/Early Classical: 7th–5th century.

• As e.g. Droop Reference Droop1929, 106–7, fig. 82k (Artemis Orthia); Buschor and von Massow Reference Buschor and von Massow1927, pl. XV, nos 31, 32 (Amyklaion); cf. Catling Reference Catling and Sanders1992, 70, 72, fig. 20 (Menelaion).

• Trench 2: C502.

• 5.6 (OF24). Fig. 5.1. Terracotta figurine. Aphrodite type, half-draped: lower body draped (garment bunched at waist), upper body nude. Figure stands with weight on right (front) foot, left leg slightly bent to rear, left hand on hip, right in front of belly grasping drapery. Head and lower body lost. Hollow, with deep relief (unfinished interior roughly pushed into mould). Surface largely eroded: traces of white slip preserved, but no other colour.

• Fabric dense and gritty, a little silver mica and a few grey inclusions: 2.5YR 6/6 (light red).

• Height 0.069 m, maximum width 0.034 m, maximum depth 0.02 m.

• Hellenistic.

• Iconography as Bonias Reference Bonias1998, 188–9, cat. nos 400–2, pl. 49.

• Trench 1: C006.

Pottery

Archaic–Early Classical

• 5.10 (P162). Fig. 5.1. Small, closed vessel: mid-upper body sherd. Decorated with horizontal bands enclosing in two cases a wavy line and in the third (at the bottom), a row of dots.

• Fabric very hard (misfired), burnt, a few small black inclusions: now 10R 2.5/1 (reddish black) to 10YR 7/3 in less burnt patches, cream slip, thin grey paint.

• Height 0.029 m, width 0.031 m, thickness 0.002 m.

• Archaic/Early Classical.

• The tiers of ornament and use of dots and wavy lines generally recall schemes current in Elis and the central Peloponnese during the Archaic period. An unpublished kantharos from Megalopolis on display in Tripolis Museum has the same crudely executed motifs.

• Trench 1: C004 spit 1 F17.

• 5.11 (P12). Fig. 5.1. Pithos: overhanging rim and neck. Black-glazed except for the underside of the rim.

• Fabric hard, slightly coarse with small black and white inclusions: 10R 5/8 (red) with a grey core. Glaze grey/black, matt on the exterior, glossier on the interior and rim top.

• Height 0.029 m, estimated rim diameter 0.022 m, thickness 0.009 m.

• Archaic: 6th century.

• As Stibbe Reference Stibbe2000, 81–2, figs 185–90 (as 185).

• Trench 3: C902.

• 5.12 (P135). Fig. 5.2. Hydria or neck amphora; outturned rim with ridged outer lip and low ring on neck below. Black-glazed.

• Fabric fine, lacks visible inclusions: 5YR 6/6 (reddish yellow). Burnt; glaze now flaking.

• Height 0.055 m, estimated rim diameter 0.23 m, thickness 0.006 m.

• Archaic/Classical: 6th–5th century.

• The large rim diameter suggests a hydria rather than an amphora: Catling Reference Catling1996, 60–1, type 2, as 30f, fig. 14:8, nos 19, 22; Pelagatti Reference Pelagatti and Gnade1991, especially figs 4 and 5; Stibbe Reference Stibbe2000, 75–6.

• Trench 1: C012.

• 5.13 (P131). Fig. 5.2. Hydria, oinochoe, or amphora; broad neck and rounded shoulder, with a ridge at the junction of shoulder and neck, and in the neck above. Black-glazed.

• Fabric fine, very hard, a few small black and white inclusions: 7.5YR 7/4 (pink) to grey. Glaze matt, grey/black.

• Height 0.077 m, maximum preserved diameter 0.28 m, thickness 0.01 m.

• Archaic: (late 7th–)6th century.

• Hydria: Catling Reference Catling1996, 61, type 30g, fig. 14.8:23 (chronology based on neck proportions which cannot be restored here). Oinochoe: Stibbe Reference Stibbe2000, 149, C3, fig. 123.

• Trench 3: C909.

• 5.14 (P113). Fig. 5.2. Amphora, hydria, or oinochoe, with neck-ring. Top of a vertical strap handle with a deep central groove and rolled edges, attached to the neck at the ring (profile preserved in the section). Black-glazed.

• Fabric fine, lacks visible inclusions: 5YR 6/6 (reddish yellow), lightly burnt. Glaze unevenly fired (mottled red).

• Height 0.051 m, width 0.045 m, thickness 0.012 m.

• Archaic: 6th century.

• As Catling Reference Catling1996, 68–9, type 37e, fig. 14.10:15; Pelagatti Reference Pelagatti and Gnade1991, figs 2, 4, and especially 8.

• Trench 2: C502.

• 5.15 (P169). Fig. 5.2. Bell krater. Pointed rim with concavity in wall beneath; insloping profile. Monochrome in and out.

• Fabric very hard fired, a few small white inclusions: 2.5YR 2.5/1 (reddish black) with redder areas. Glaze grey/black out, redder in. Misfired.

• Height 0.036 m, estimated rim diameter 0.20 m, thickness 0.006 m.

• Archaic: 6th (second quarter)–early 5th century?

• Stibbe Reference Stibbe1994, 53, Group C, cat. no. C3, fig. 158; Catling Reference Catling1996, 40–1, type 5c, fig. 14.2:25 (plainware, c. 550–450?).

• Trench 1: C012.

• 5.16 (P245). Fig. 5.2. Large bowl with overhanging rim. Black-glazed.

• Fabric fine; 2.5YR 4/4 (reddish brown). Glaze black with a metallic sheen, thick on the upper face, washy on the underside.

• Height 0.018 m, width 0.049 m, thickness 0.018 m.

• Archaic: 6th century.

• Stibbe Reference Stibbe1994, 91, Group C, as C3, fig. 349.

• Trench 4: C701.

• 5.17 (P173). Fig. 5.2. Silhouette style skyphos, upper body. At left, grazing goat (?) to right; at right rear of another animal; ground line and one further band beneath; top of thick band or glazed zone at lower break.

• Fabric fine, lacks macroscopically visible inclusions: 5YR 7/6 (reddish yellow). Glaze black.

• Height 0.037 m, width 0.025 m, thickness 0.003 m.

• Archaic: probably Laconian II (c. 620–580).

• Silhouette animals occur earlier in the 7th century (Margreiter Reference Margreiter1988, 111–18, e.g. pl. 65:727,728; Droop Reference Droop1929, 66–9) but continue into the early 6th: Droop Reference Droop1929, 73, figs 46o, 58cd.

• Trench 4: C701.

• 5.18 (P136). Fig. 5.2. Kothon or mug; two joining sherds from the base and lower body. Rounded bottom, hollowed towards the centre, resting surface poorly defined. Black-glazed in and out, resting surface reserved.

• Fabric very hard (lightly burnt in firing), a few small white inclusions: 5YR 6/4 (light reddish brown).

• Height 0.067 m, width 0.055 m, thickness 0.007 m.

• Archaic: 6th century.

• Stibbe Reference Stibbe1994, 43–7, see Groups C and D, figs 97–109.

• Trench 1: C002.

• 5.19 (P97). Fig. 5.2. Cup: concave rim to junction with body. The rim diameter suggests a stemless cup, although 5.19 cannot be attributed to type. Black-glazed in and out.

• Fabric fine, a few macroscopically visible white inclusions: 5YR 6/8 (reddish yellow), lightly burnt in firing.

• Height 0.026 m, estimated rim diameter 0.12 m, thickness 0.003 m.

• Late Archaic/Early Classical: c. 550–450?

• As Catling Reference Catling1996, 47, type 15d, fig. 14.5:15.

• Trench 3: C906.

Fig. 5.2.

Artefacts from the acropolis sanctuary (with catalogue numbers).

Classical

• 5.20 (P133). Fig. 5.2. Large kothon: two joining sherds from mid body. Red-figure decoration: part of the drapery of a female figure (perhaps to right). Interior glazed.

• Fabric fine, with a few small white inclusions: 2.5YR 6/8 (light red), core redder, exterior slipped.

• Height 0.076 m, width 0.069 m, thickness 0.008 m.

• Classical: last quarter 5th or early 4th century.

• While 5.20 could be a bell krater, the interior treatment favours the slightly closed kothon (the most common Laconian red-figure shape at Sparta: McPhee Reference McPhee1986, 155–6). For the shape, see Stroszeck Reference Stroszeck, Schierup and Sabetai2014, 142–3, 150, cat. nos 1–2, figs 1–2.

• Trench 1: C013.

• 5.21 (P68). Fig. 5.2. Oinochoe shape 8 (Pheidian mug): foot ring and ribbed lower wall. Narrow base; small foot ring marked off from the wall with a groove; rounded body profile above.

• Fabric fine, lacks visible inclusions, lightly burnt in firing: 5YR 6/8 (reddish yellow). Glossy glaze.

• Height 0.03 m, estimated base diameter 0.037 m, thickness 0.003 m.

• Classical: c. 450–early 4th century.

• Sparkes and Talcott Reference Sparkes and Talcott1970, 73, cat. nos 216–18; for the small base, see CVA: Leiden 3 (5), pl. 57:4. The longevity of the shape in Laconia is unknown, but it continues into the early 4th century in Athens (Sparkes and Talcott Reference Sparkes and Talcott1970, 73 n. 28) and Apulia (CVA: Toledo 2 [20], pl. 94:2).

• Trench 3: C908.

• 5.22 (P124). Fig. 5.2. Small bowl with projecting rim. On rim top, a series of closely spaced (sometimes overlapping) stamped palmettes. Wide vertical ribs run from a groove just below the rim. Black-glazed in and out.

• Fabric fine with a few small white inclusions, burnt in firing: 5YR 7/4 (pink).

• Height 0.019 m, estimated rim diameter 0.07 m, thickness 0.002 m.

• Classical: late 5th(–4th) century.

• The date derives from the plain-walled shape in Attic black glaze (Sparkes and Talcott Reference Sparkes and Talcott1970, 135, cat. nos 879–81; Rotroff Reference Rotroff1977, 165, cat. nos 1045–9). The decoration of 5.22 is not attested in Attic and may be a Laconian innovation.

• Trench 2: C503.

• 5.23 (P130). Fig. 5.2. Stemless cup: profile of foot and edge of wall/tondo. In tondo, part of a motif in added white (perhaps animal fur). Stilt-like foot sliced diagonally to give a narrow resting surface; an angular groove scraped around the exterior junction with the body. Underside of foot and base half covered with a pool of glaze; the unaffected area suggests that the inner face of the foot may have been reserved and the base glazed, with a reserved band round the outer edge.

• Fabric hard, fine, lacks visible inclusions: 7.5YR 7/6 (reddish yellow).

• Height 0.013 m, width 0.046 m, thickness 0.002 m.

• Classical: late 5th–4th century.

• On the chronology of high ring feet on Laconian cups, see Catling Reference Catling1996, 55–6, especially type 27k.

• Trench 1: C004.

• 5.24 (P74). Fig. 5.2. Small one-handler; rim rolled out and hollowed underneath; rounded body turning in at lower break. Black-glazed in and out.

• Fabric fine, very hard (burnt in firing), a few small black inclusions: 5YR 6/2 (pinkish grey). Glaze matt black.

• Height 0.02 m, estimated rim diameter 0.08 m, thickness 0.002 m.

• Classical: second half 5th–4th century.

• Profile as Catling Reference Catling1996, 51, type 18, fig. 14.5:45; Lang Reference Lang1992, 54–6, fig. 8:1.

• Trench 3: C903.

• 5.25 (P116). Fig. 5.3. Bowl with incurving rim; four joining sherds preserve the profile from rim to centre base. Low disc foot. Glazed in and out.

• Fabric 5YR 7/8 (reddish yellow) with a few small white and black inclusions. Glaze matt grey/brown.

• Height 0.027 m, estimated rim diameter 0.66 m, estimated base diameter 0.043 m, thickness 0.005 m.

• Classical: late 5th–4th century.

• As Coldstream Reference Coldstream, Coldstream and Huxley1972, 160, cat. no. 19, fig. 48.

• Trench 1: C016.

Fig. 5.3.

Artefacts from the acropolis sanctuary (with catalogue numbers).

Stone

• 6.1 (OF33). Fig. 6.1. Part of an inscribed halter (jumping weight) of green lapis Lacedaemonius preserving the pointed end of the weight and part of the grip-hole, inscribed around the outer edges. Length 0.126 m; width 0.06 m; letter height 0.012–0.018 m.

• Trench 5: C301.

• Eds. Briefly reported in AR 55 (2008–9) 31 (SEG LIX 388); ed. pr. Pitt Reference Pitt2014–19 (Bull. ép. 2020, 157).

• Late 6th–early 5th century BC

Fig. 6.1.

Part of an inscribed halter (jumping weight) of green lapis Lacedaemonius preserving the pointed end of the weight and part of the grip-hole, inscribed around the outer edges. Drawing: Robert Pitt.

Brick and tile stamps

Ten stamped bricks and tiles add to the extensive collection discovered in earlier British excavations and surveys of the acropolis and wider region, and in numerous rescue excavations in recent decades. They date principally from the Late Hellenistic period (late second–first century BC), when a number of specific projects, notably the city walls and the theatre stage building, necessitated the creation of large amounts of such building materials. It seems that certain Spartan magistrates ordered the manufacture of tiles from workshops in this period for a variety of projects and may even have stored them up for future repairs. The stamps often preserve the names of the officials and contractors involved, but there are also several series of tiles marked simply ‘Public, of Athena’, which have been interpreted as originating from factories owned by the goddess for use in public constructions.Footnote ¹⁰

• 6.2 (P150). Fig. 6.2. Pan tile fragment with no preserved edges. Height 0.10 m, width 0.135 m, thickness 0.016 m; letter height 0.016 m. Cf. Wace type 13; IG V 1, 857.

• Trench 2: C503.

• 2nd century BC?

Fig. 6.2.

Pan tile fragment with no preserved edges. Drawing: Robert Pitt.

Fig. 6.3.

Fragment of a large, stamped brick, preserving top, bottom and one side with a retrograde inscription. Drawing: Robert Pitt.

Fig. 6.4.

Flat tile broken on all sides with mortar at the breaks from reuse. Drawing: Robert Pitt.

Fig. 6.5.

Cover tile broken on all sides, dark brown-red glaze on the convex. Drawing: Robert Pitt.

Fig. 6.6.

Pan tile broken on all sides. Drawing: Robert Pitt.

Fig. 6.7.

Small tile fragment broken on all sides. Drawing: Robert Pitt.

Fig. 6.8.

Pan tile with no preserved edges. Drawing: Robert Pitt.

Fig. 6.9.

Cover tile broken on all sides. Drawing: Robert Pitt.

Fig. 6.10.

Tile broken on all sides. Drawing: Robert Pitt.

Fig. 6.11.

Two joining tile fragments with red slip, broken all sides. Drawing: Robert Pitt.

Pottery

• 6.12 (P22). Fig 6.12. Graffito on a fragment of a large Classical Laconian closed vessel with black-glaze exterior. Height 0.065 m, width 0.055 m, thickness 0.009 m; letter height not preserved, scratched with a 0.002 m-thick instrument.

Fig. 6.12.

Graffito on a fragment of a large Classical Laconian closed vessel with black-glaze exterior. Drawing: Robert Pitt.

Trench 1: C001.