The room bounded by wall 366 and robbing trench 497 to the north (1006.00 N), wall 365 to the south (1002.10 N), and walls 332 to the east (262.07 E) and wall 306 to the west (258.01 E) was excavated between April 5 and April 20, 2012. This room is located approximately 10 meters south of the so-called Turkish House. Its earliest phases predate the 11th c CE when two walls (366 and 306) were built as part of a larger, unknown structure. Activity in the room continued through to the early modern period, when the north portion of wall 306 and the west portion of wall 366 were robbed out (497), and in the 18th century a bothros was placed in the middle of the room. In the intervening 700 years, layers of dumped fill were used to level out the space during the late 11th – early 12th c., and subsequently walls 332 and 365 were constructed to create the current space. This area was excavated under the supervision of Angele Rosenberg-Dimitracopoulou and Larkin Kennedy. We worked with a crew of workmen including Thanasis Notis (pickman, foreman), Vangelis Kollias (wheelbarrow man, screener), Thanos Kioseloglou (shovelman, replacement pickman), and occasionally Tasos Kakouros filled in for Thanasis Notis as pickman as well.

Pre-11th c CE

Wall 366 at 1007.00N, which extends from 260.40E to 262.20E, and wall 306 at 258E, which extends from 1004.78N to 1002.58N, were constructed prior to the 11th c CE. Wall 366 consists of fairly regular courses of large, squared blocks separated by tile. These courses continue unchanged throughout the visible elevations of the wall, and extend below the elevation reached at the end of session 1. Wall 306 is slightly smaller and is similarly constructed of fairly regular courses of large, squared blocks.

Byzantine period

During the Byzantine period, this area was leveled out through the deposition of successive layers of dump fill. Three layers of dump fill (718 Lot 2012-16, 716, 713) were thrown in along the northern wall 366 during the 11th c CE. In the early 12th c CE, another layer of fill (711) was tossed into the room’s northern portion. These deposits included the disposal of glass and ceramic vessels, a glass bracelet, bone needles and pins, iron slag, nails, ceramic tile, faunal material, charcoal, and plaster. In context 718 were also disposed two bronze bracelets (MF 2012-10, 2012-12), one with a hook and eye closure and one composed of wire flattened at one end and wrapped around the wire body to form a ring for closure, as well as a shell bead, possibly made of mother of pearl, and incised with a starburst pattern (MF 2012-2).

These earlier deposits form a bowl allowing later fills to spill down in successive lenses away from walls 306 and 366. Later dump fills (711, 708, 690 Lot 2012-14, 685, 683 Lot 2012-15, 681, 677, 670, 660 Lot 2012-13, 650, 646 Pottery Lot 2012-13 Bone Lot 2012-2, 645) were successively laid around the room, probably also during the early 12th c CE. These deposits similarly included the disposal of glass and ceramic vessels, a waster, a bone awl and a pin, a stone tessera, a millstone fragment, an iron implement, iron nails, slag, and bloom, bronze, plaster, ceramic tile, faunal material, architectural revetment, mainly of marble but also of terracotta, and charcoal. A stamped amphora handle was kept from context 685, though not inventoried. Of note were coin 2012-13 (931-944 CE) and coin 2012-18 (969-1030 CE) from context 646, coin 2012-35 (976-1030 CE) and coin 2012-36 (945-950 CE) from context 690, and a bronze bracelet with a hook clasp (MF 2012-3) which was also retrieved from context 690.

In the southwestern corner of context 690, spreading to the east from wall 306, a cluster of large faunal bones (mainly bovid) was accompanied by large fragments of stewpots. This and other subtle differences in the dumping activity were explored with soil samples taken from contexts 646 and 690. Although clear changes in soil indicate that these fills are the result of years of dumping activity and not a single deposit, it is not possible to archaeologically distinguish between these events and the finds do not reveal a more precise chronology either. Few ceramic sherds joined, and none between contexts, indicating a large number of individual pots were represented, rather than many fragments representing a few vessels as would be expected in a primary deposit. These successive fills therefore appear to represent individual loads of dirt that should ultimately be considered part of the same protracted depositional event which leveled the area.

Later in the 12th c CE a drain (426) extending from 1011.75N/259.26E to 1004.97N/260.30E was cut into the early 12th c CE fill. Slabs of stone lined the walls of the drain and its floor was made of packed soil. The drain may continue south into the adjoining room, although this needs to be further explored. It also extends north into the Blue team’s area, where it appears to cut through wall 366. In previous excavation seasons, fill for the drain has been dated on the basis of the pottery to the 11th century, though it has also been considered to date to at least the Frankish period on the basis of stratigraphy (590, 591). The portion of the drain south of wall 366 still preserved coverstones, while that north of the wall only retained a few coverstones and displayed other signs that the northern portion of the drain had been exposed outside for a period of time prior to being covered by later depositional events. Stones from the drain’s walls were robbed out (between approximately 1007.00 N and 1006.00) as part of the robbing activity (c 497) that also makes determination of the exact relationship between drain 426 and wall 366 problematic, as stones were also removed from the northern portion of wall 306 and the portion of wall 366 close to where wall 366 and drain 426 would have met in antiquity. This robbing has been dated to the early modern period on the basis of its stratigraphic relationship with robbing trench 596.

Still later in the 12th c CE, wall 332, which extends from 1006.37N/262.38E to 999.90N/262.07E was built in order to create a smaller, interior space. This wall is constructed of roughly squared blocks with some tile demarcating the irregular courses. It rests on a foundation of spherical boulders that begin at elevation 86.35 at the north end and elevation 86.72 at the south end, and extend down to a depth of 86.23. These non-corbelled courses bell out wider than the finished wall face. At the time the wall was built, a foundation trench 654 was dug into the surrounding deposits of leveling fill. This trench contained the disposal of glass and ceramic vessels, a glass tube, bronze, marble, and iron. As the material contained in this trench dates to roughly the same time period as the deposits it is laid against, however, this foundation trench can be considered to have been filled by the same material as that which was dug up in order to provide a trench for wall 332’s foundation courses, and not by extraneous material.

Even later in the 12th c CE, wall 365, extending from 1002.58N/257.89E to 1002.44N/261.99E, was constructed in order to subdivide this room and create two smaller interior spaces in place of one larger interior space. This wall was possibly constructed on top of drain 426, though the boulders partially exposed at elevation 86.70 require further attention in order to better understand this relationship, and whether these boulders can indeed be considered as part of the southern extend of the drain. In wall 365, large, roughly squared blocks rest on a foundation made up of a single course of rounded boulders extending to a maximum depth of 86.67. Threshold 664, consisting of a break in wall 365, joined this room to the room to the south and contained an iron needle (MF 2012-8) as well as an iron nail and body sherds from a glass vessel. Coin 2012-14 from the Late Byzantine period (1070-1075 CE) was found just below the two large stones preserved in the south-most portion of threshold 664. Threshold 536, made up of a single layer of crushed tile, provided access to the room from the east. This threshold was laid on the edge of existing stone courses of wall 332 and bridged the gap between this wall and wall 365. To the west of threshold 536, there was a concentration of carbon and bone with some sherds from glass vessels which could be associated with this doorway (641). An iron bloom was also saved from context 641, though not inventoried. A flotation sample was taken from both of these contexts as they could represent deposition associated with the use of the doorways, and which could be compared at a later date with the contents of flotation samples taken elsewhere in the room.

Early Modern

In the 18th c CE, a bothros (346/671) was dug into the middle of room, cutting through all of the earlier leveling fill. It stretches from 1005.06N to 1003.84 N and 259.79E to 261.38 E. Four large, tabular stones capped the bothros on its northern side (346) to a depth of 86.81. Under the capstones, the interior of the bothros was lined with regular courses of rounded boulders until an elevation of 85.35 m (671). Under the stone courses is a thick white clay foundation layer which was not excavated. When it was dug, the bothros cut through the drain 426. Built into the regular courses of the structure was a large amount of tile that appears to have used glass slag as a temper. A sample was saved for further testing, as it may indicate the presence of glass making in the vicinity. A Byzantine coin (coin 48) from Thessalonika (1143-1180) was also built into the bothros wall (671). After the stone courses end at 85.35, the bothros belled out into a cavity wider at the bottom than at its mouth, and continued to an elevation of 83.90 m. It appeared to Thanasis Notis (the team’s pickman) that the bothros terminated at this point.

When it went out of use, the bothros (346/671) was filled with a deposit taken from elsewhere around the site which contains a cohesive collection of Frankish material dating to the first half of the 13th c CE (726, 733, 739, 743) including glass and ceramic vessels, iron nails, iron slag and an antler. In addition to the relatively high volume of cooking and tableware, a significant concentration of faunal and fish bone and charcoal were included in the deposit. Two Byzantine coins were also included in the deposit, coin 2012-50 (1112- 1137 CE) and coin 2012-52 (1143- 1180 CE), a bronze token (originally assigned coin no. 2012-46), and two illegible coins (coin 2012-44 and 2012-49). This deposit appears, therefore, to be a secondary deposit of Frankish material that was thrown into the bothros, possibly when it was no longer necessary for storage.
Sometime after the filling of the bothros, a pit (c431) was dug into it, cutting through both the southern end of the structure to a depth of 86.30. It stretches from 1004.35 N to 1002.43 N and 259.71E to 261.70E. It was filled with Frankish material (336 and 338) similar to that filling the bothros 671/346 (726, 733, 739, 743).

Further excavation is necessary to address four remaining questions. First, further excavation is necessary within bothros 346/671 in order to ensure that it does terminate. Thanasis Notis suggested that it might continue laterally to the north rather than down. Next, investigation of the room to the south would indicate whether drain 426 does in fact continue. Threshold 536 could also be removed in order to better determine the association of walls 332 and 365. Finally, more soil should be removed from within the room in order to determine a more precise chronology for walls 306 and 366.