The following is a summary of excavations in three areas in North of Nezi: the courtyard of a Byzantine house, the room north of the courtyard, a room south west of the courtyard and a suite of rooms south of the room south west of the courtyard.
The courtyard of the Byzantine house uncovered in the 1961 season in Agora SW-G by Steven Lattimore (Room 12 NB 230; NB 235) and Agora SW-F by Willam Berg (NB 229) was further excavated all three sessions of 2008. Excavation of the courtyard was overseen in the first session by Josh Geiske and Laurie Kilker from April 11th to the 23rd, with Cleomenes Didaskalou picking and Vasilis Kollias as barrowman. In the second session, from May 15th to 23rd,, Jody Cundy and Nate Andrade took over excavation of this area, with Thanasis Notis as pickman and Andreas Oikounomou as barrowman and siever. In the third session Megan Thomsen replaced Nate Andrade and excavation of the courtyard continued from May 26th to June 13th. The courtyard extends from 1027.85 to 1035.70 N and from 265.20 to 273.5 E. The room North of the courtyard was excavated by Nathaniel Andrade and Jody Cundy in the second session. It extends from 1035.4 to 1039.4 N and from 265.5 to 270.6 E and was included in Room 12 by Lattimore (NB 230 p.170).
The room southwest of the courtyard and rooms south of it were excavated in 1963 as Agora SW-J under the supervision of Daniel Geagan (NB 254; 262). They were further excavated in the first session of 2008 by Matthew Baumann and Nathaniel Andrade, working with Thanasis Notis as pickman,, Vasilis (Bill) Papanikolao as shovelman and Andreas Oikonomou as barrowman. Excavation in this area continued in the second session with Jody Cundy replacing Matthew Baumann. The area extends from 1020.00 to 1030.50 N and from 262.00 to 266.00 E.
The goal of excavation in these areas is to clarify the relationships between the various walls in North of Nezi, in particular their phasing, and to isolate the walls and spaces of the Byzantine house for later consolidation and presentation to the public.
This report will be divided both thematically and chronologically.

COURTYARD OF THE 1961 BYZANTINE HOUSE:
Post-Byzantine:
The post-Byzantine levels of the courtyard were excavated by Steven Lattimore in 1961 (NB 230; 235). Though it is not profitable to summarize all the levels previously excavated here, the features and deposit that have direct bearing on those excavated this season will be briefly discussed. A marble wellhead, capping the built well 61-9 (NB 230 well #2), was identified on the 15th of May, 1961 at an elevation of 86.85 (NB 230 p.103, 151, 159-60, 170, 176, 180, 193-4, 196; NB 235 p. 16, 27-8, 33-44; neg. 61-6-17; 61-6-19; 61-6-20, photo log 18, p.30) . The well-head was subsequently removed on May 27th, 1961 (NB 230, p.167). Lattimore’s excavation of the fill of well 61-9 (renamed structure context 5864) produced Turkish material, including a coin (61-507). He interpreted the resting surface of the well-head as a Turkish courtyard (strosis 85.246, NB 230 p. 170). While it is clear that well #2 was in use in the Turkish period, the date of the construction of the earliest phase of well 61-9 (5864), and the tunnels and vaulted chamber associated with it are earlier. Lattimore detected and excavated the fill of a circular cut around well 5864 (61-9), which produced mid-thirteenth century material lotted as 827 (NB 230 p.176).
Excavation in 2008 began with the removal of two segments of a pi-shaped terracotta drain (structure 5213) and the vertical stones that lined the exterior of the drain wall. According to the excavation notes of William Berg (NB 229, p.192), the drain, which extended from 1034.16 to 1034.72 N and 271.66 to 272.84 E, with a top elevation of 85.16 and bottom 85.01, was associated with well 61-19 (5864). The terracotta segments are individually 0.58m in length, 0.26 in width and 0.12 in height. Though not evident at the onset of the 2008 season, the Travlos plan of the area (1963) shows the course of the drain extending eastward beyond the north-south rubble foundations (wall 5649) to a length equal to its westward extent. A photograph of the newly exposed drain confirms this (61-15-4, vol.18 p.25) and shows the drain overlying the rubble foundations 5649. The foundations for the terracotta drain were removed as contexts 5780 and 5847. Although no pottery was collected from the removal of the drain segments and the excavation of the foundations produced no precisely datable sherds, the association between the terracotta drain and wall 5649 indicates that the drain was the latest feature in the courtyard left from the 1961 excavation campaigns.
Wall 5649 is the rubble foundations of a NS wall with three blocks from the first course preserved at a top elevation of (@@) and bottom of (@@), extending from 1033.75 to 1035.75 N and 273.31 to 274.20 E. Wall 5649 abuts the earlier EW wall 5741 (NB 230 wall 59) to the north and the NS wall 34 (NB 229) to the south. Wall 5649 is interpreted as a blocking off of the entrance corridor to the courtyard from the East. Permission to dismantle wall 5649 is awaited and it is expected that Frankish or later material will be recovered from that operation.
In the first session, a martyr left from the 1961 excavation campaign was excavated. It extended northward from wall 61 (structures 5883 and 5882) from 1030.68 to 1035.19 N and 271.25 to 273.53 E. The preservation of this strip of unexcavated layers is likely due to the later wall 57 (NB 230, p.166), which overlaid them. Lattimore established the leveling point GG at 85.515 on a stone in this wall (NB 230, p.166). William Berg describes the excavation west of wall 34 (NB 229, p.156), he encountered a hard surface at 84.46 and changed baskets to continue down to a clay surface at 84.06 and then to earlier levels (NB 229, p.167). From this it can be conjectured that the martyr was produced as a result of the pedestalling of wall 57-G and the preservation of leveling point GG. This can be clearly seen in a post-season photograph of sections F and G (61-27-2, vol.18 p. 35). Wall 57-G, however was not present at the opening of the 2008 season. Excavation of the martyr began with a cleaning pass (5191, top elevation 85.25), and concluded with deposit 5264 (bottom elevation 84.91). Of the thirteen contexts excavated in the martyr (...), the combination of root action and the truncation of the deposits made interpretation of the layers problematic. Although these deposit produced mostly 12th century pottery, they are likely best understood as part of Lattimore’s strosis 85.246 (NB 230 p.164; 173), which produced the coins 61-794, -795, -795, and was lotted as 1961-827. Strosis 85.246, was assigned a Frankish date.
Likely contemporary with the rubble wall 5649 that closes off the East entrance to the Byzantine courtyard are two pier rubble foundations identified by Lattimore as part of the EW wall 61. The West pier foundations (structure 5784) have a top elevation of 85.14 and bottom of 84.59 and occupy the space from 1030.13 to 1030.72 N and 271.49 to 272.19 E. The pier foundations 5784 abut the eastern edge of the north wall of the staircase 5783 (NB 230 61). The east pier foundations abut wall 34 (NB 229). They are comparable in terms of elevation and dimensions with the west ones (5784). Extending between these pier foundations and overlying them was a rubble wall, interpreted by Lattimore as part of wall 61-G, but it has eroded away since it was exposed in 1961.
Further enclosure of the Byzantine courtyard in the Frankish (or later) period is evidenced by the EW rubble wall 5508 (NB 230 wall 61). Wall 5508 has a top elevation of 85.13 and bottom elevation of 84.80 and extends from 1030.20 to 1030.85 N and 265.85 to 269.65 E, and is composed of two random courses of rough-hewn blocks in reuse and fieldstone. Like the pier foundations 5784, wall 5508 abuts the north wall of the staircase 5783. Permission to dismantle wall 5508 is awaited and, like wall 5649, expected to produce Frankish or later material.
Lattimore uncovered the remaining blocks of a NS wall 71 (renamed 5473) and the fill of the robbing trench for the rest of the wall (NB 235 p.19). Wall 5473 (NB 235 71) has a top elevation of 85.73 and bottom elevation of 84.72 and extends from 1038.90 to 1037.56 N and 266.10 to 265.31 E. The robbing trench extends from 1037.56 to 1027.90 N and 265.55 to 266.70 E. The robbing trench, which produced Frankish material, also lotted as 837, is clearly visible in a 1961 post-season photograph (61-26-6, vol 18, p.36). More of the fill of this robbing trench was detected in the stretch south of wall 5508 (wall 61-G) and was excavated as deposit 5510 with a top elevation of 84.86 and bottom elevation of 84.67. Consistent with lot 837, it produced Frankish material and was lotted as 2008-52. It is imagined that wall 71 (5473) formed the west boundary of the level identified as strosis 85.246 by Lattimore.
Just west of well 61-9 (5864) is another EW wall (5443) of unknown function. It extends from 1033.63 to 1032.95 N and 265.68 to 267.01 E. Wall 5443, composed one large ashlar block in reuse and random coursed fieldstone, has a top elevation of 85.03 and bottom elevation of 84.57. It seem that the fill of the foundation trench for wall 5443 was partially excavated in the 1961 campaign, and although no reference to the wall or this trench has been recovered from the excavation notes, the cut of the trench is discernable in the 1961 post-season photograph 61-27-4 (vol. 18, p.36). The fill of the foundation trench was further excavated as deposit contexts 5550 and 5824 this season. Though the highest level at which the cut was made for the construction of this wall is unknown, it was at least as high as the fill over highest of a series of pebble surfaces associated with the late Byzantine phase of the courtyard (5298, el. 84.99), to be discused below. The fill of the robbing trench for wall 71 presumably overlaid the western portion of wall 5443 and the foundation trench associated with it or truncated both. Excavation of the fill of the foundation trench for wall 5443 revealed that this wall is built on top of an earlier structure (wall 5906) that also will be discussed below.
In the portion of the courtyard that extends southward from the later curtain wall 5508, further enclosure of the space is evidenced by the NS wall 5360. It extended from 1027.91 to 1030.29 N and 263.94 to 265.59 E with a top elevation of 85.26 and bottom elevation of 84.92. Wall 5360, which consisted of a single course of rough hewn stones abutted the earlier EW wall 5411 to the S and EW wall @@ to the N. The dismantling of wall 5360 produced Frankish pottery, lotted as 2008-51 including a Brundisi proto-majolica bowl with a blue chevron pattern. The fill of an EW robbing trench 5506, that extends from @@ to @@ and produced Frankish material, demarcates the southern extent of the courtyard of the Byzantine house. Of the Frankish levels associated with wall 5508 and 5360, only a single soft lens abutting wall @@, context 5527 (lot 2008-53) remained at the onset of the 2008 excavations. William Berg reports the excavation of hard fill in the area enclosed by wall 5360 (NB 229 wall 39), 5508 (61-G) to wall 34-F (NB 229, p.190) as basket 18 of the 8th of June (lot 61-680). The removal of basket 18 revealed a surface at an elevation of 84.97 that is associated with a series of stairs rising from W to E.

Late Byzantine:
The late Byzantine courtyard of the 1961 house extends from the EW walls 5463 and 5741 in the N to walls 5411 and 5519 at it southern limit, from a NS wall indicated by the robbing trench @@ at the western edge, presumably to wall 34-F as the eastern boundary forming a regular rectangle. The intrusion of later features, such as well 5864 and walls 5508, 5649 and 5543, discussed above as well as earlier excavation, in particular the removal of the deposits that abutted wall 34-F in the SE corner of the courtyard, complicate an understanding of the relationships between the deposits in the courtyard. Nevertheless, it is possible to suggest the following: the courtyard is characterized by succession of layers of pebbly and tile cobbled surfaces with evidence of leveling repairs, suggesting continuous use throughout the 12th century. In this summary, I hope to be able to relate the various surface deposits to one another, to the features of the courtyard (wall 4442, wall 5906 and the EW drain 5863) and to points of communication with the adjacent rooms of the house (thresholds 5462, 5463 and @@)..
It is possible to phase the features associated with the courtyard of the 1961 Byzantine house in the following chronological sequence:
1. Walls 5463 and 5741, together with the threshold 586 are the earliest features
2. The courtyard installation (possible staircase foundations) wall 5906
3. Associated with the same stratum:
a. The N wall of the staircase 5783
b. The EW stone-lined drain
4. Threshold with step 5462
5. Wall 5442
6. Wall 5443
It is worth noting, however, that the sequence of surfaces and repairs in the courtyard of the 1961 Byzantine house reflects continuous use over a period of about 100 years from the 11th through Early 12th centuries according to the pottery dates.
The primary characteristic of the deposits in the courtyard is the superimposition of surfaces. On the basis of the sequence of floors, it can be said that the latest preserved surface in the courtyard was composed of flat lying roof tile fragments and cobbles excavated in 2 parts: surface 5300 is the segment N of the EW wall 5508 (61-G) and surface 5630 (lot 2008-70). Beneath the surface 5300=5630, was another tile-cobbled surface. The portion N of wall 5508 (61-G) was excavated as deposit 5324 while the portion South of the wall remains to be excavated next season when wall 5508 is dismantled. The tily surface 5324 transitioned gradually to a hard packed dirt surface excavated as deposit 5327. Beneath the tile and cobble surface 5324 was a series of shallow lenses and pebbly patches intended to level depressions in the courtyard surface coordinated between deposits 5679 and 5685 (5633; 5634; 5635; 5657; 5661; 5664; 5668; 5669; 5672; 5674; 5665 lotted as 2008-69).
In the NW corner of the courtyard where the robbing trenches 5549 and @@ intersect, a lens of dumped construction debris 5398 appears to coordinate with the leveling patches under the tile and cobble floor 5324. Under this lens of debris was the pebble surface 5564 (top elevation 84.84). It abutted wall 5463 and was cut through by the trench for robbing portions of that same wall (robbing trench fill 5545 and 5852; cut 5549). The pebble surface 5564 also preserved some cement (5560), which lay on it, presumably residue from the mixing of this material for a construction project.
Beneath the pebbly surface 4464 was another patched pebbly floor (patch 5565, pebbly surface 5566) and beneath it two more hard packed surfaces.. Because of the cut around well 61-9 (5864) and the cut of the foundation trench for wall 5543, the pebbly surfaces 5564 and 5566 and the dirt surfaces 5616 and 5618 are truncated and their relationship with the deposits on the E side of well 61-9 (5864) must be reconstructed.
Preserved under pedestal of the terracotta and stone-lined drain (5213) in the NE corner of the courtyard was another sequence of pebble surfaces that continue under the NS rubble foundation (5649 ), 5291, 5293, 5641, 5645 and 5656. These likely continued northward to meet the north wall of the courtyard 5741 (NB 230 wall 59) but were truncated by excavations in the 1960’s. Due to comparable elevations and sequence of layers, the surfaces W of the cut for well 61-9 (5864) are imagined to represent the same stratigraphic sequence as those W of the well.
The sequence of patched pebble surfaces in the NW and NE corners of the courtyard seem to be replicated in the area just N of wall the EW wall 5508. The sloping surface 5679, surface 5683 (lotted as 2008-67), pebble patch and surfaces 5726, 5730 and 5757. The foundation trench for the NS wall 5442 cut through a lens of fill 5728, overlying 5730 and abutting 5726. The NS wall 5442 extends from 1031.83 to 1033.56 N and 267.09 to 267.66 E with a top elevation of 84.95 and bottom elevation of 84.43. The random courses of roughly hewn limestone blocks and a spoliated marble molding abut the EW wall 5443 to the N, forming a corner. Like wall 5443, wall 5442 appears to be built on an earlier structure. The function of the walls 5442 and 5443 is not clear.
The pebble surface 5291(?), pebbly surface 5697 and packed dirt surface 5651are cut by an EW trench that extends from under wall 5649 up to the cut around well 61-9 (5864). This trench appears to mark the clearing out of an earlier EW drain with built stone walls (5863) that follows the same course as the later terracotta drain (5213). The fill of the drain-clearing trench, deposits 5288, 5289, 5646 and 5908, was lotted as 2008-58.
Deposit 5697, likely contemporary with the pebble surface 5656, is the resting surface of a limestone step associated with a threshold 5462 in wall 59-G (between wall 5463 and 5741). The surface 5697 was overlain by 5696, which continued underneath the blocks of the threshold 5462 (1034.93 to 1035.72 N and 268.68 to 270.19 E; top el. 85.08 and bottom el. threshold 84.87, step 84.63). The raising of the threshold and addition of the step likely marks the raising of the floor in the room N of the courtyard. The threshold 5462, removed on May 29th, and the coordinated surfaces (5697, 5656, 5651) appear to be contemporary.
Under the pebbly surfaces 5697 and 5656 was another uniform surface extending between the drain-clearing trench (cut 5673) and wall 59-G (wall 5741 and threshold 5865), (deposits 5737, 5742 and 5800). Into this surfaces a large pit filled with ash and construction debris was cut (cut context 5702: 1034.08 to 1035.59 N; 268.89 to 272.42 E, top el. 84.72 bottom el. 84.04). The differentiated fills of the pit, deposits 5699, 5704, 5705 and 5716, were lotted as 2008-59.
With the removal of the surface 5751, a series of lenses and patches (5690, 5691, 5692, 5694, 5709, 5748, 5756) and the martyr 5771, a rather continuous surface extending from the N wall of the staircase 5883 the line of the EW drain 5863 was revealed. The material produced from the excavation of this surface, coordinated between deposits 5773 and 5764, was lotted as 2008-66.
The excavation of the surface coordinated between 5773 and 5764 revealed an earlier pebbly surface 5803. Through the surface 5803 S of the drain and 5799 (and perhaps 5797), a shallow linear cut was made outside the N and S line of the built stone drain (cut context 5781, 1033.31 to 1034.10 N, 269.85 to 271.91 E, top el. 84.65, bottom el. 84.56). This cut appears to be intended for the laying of cover-slabs for the drain that were subsequently robbed out. The foundation trench for the N wall of the staircase 5783 (NB 230 wall 61-G) was also cut through surface 5803. The fill of the foundation trench 5795, lotted as 2008-65, was only partially excavated this season; it remains unexcavated where it continues under the rubble pier foundations 5782 and the pedestalled wall 5508. The build courses of limestone blocks and tile stringers that make up N wall of the staircase 5783, first exposed by Lattimore on Date (NB 230 p. wall 61-G) extend from 1030.08 to 1030.68 N and 269.77 to 271.52 E with a top elevation of 85.23 and bottom elevation of 84.39. The foundation trench fill 5795 was slightly over-dug, revealing an earlier EW wall on top of which wall 5783 is built. These blocks are also visible in the bottom of a rectangular pit E of the staircase (1029.00 to 1030.25 N and 270.60 to 271.25 E, top el.84.41, bottom el. 84.25) excavated as deposit 5518. The pit fill 5518 abutted the N wall of the staircase 5783 to the N, the NS wall 38-F (NB 229 p.188) to the E, the rubble packing of the staircase to the W, and the unexcavated layers into which the pit cut 5521 was made to the S. Excavation of the pit fill 5518, lotted as 2008-55, produced an iron adze (MF-2008-6), an iron ladle (MF-2008-3) and scalloped door plaque with keyhole and latch cuttings (MF- 2008-4).
The tops of the stones forming the walls of the drain 5863 (1032.83 to 1034.68 N and 269.35 and 273.60 E top el.84.60 bottom el.84.31) are flush with the surfaces 5823, 5836 and the unexcavated deposit beneath 5799. Because no cut was discernable outside the line of the stones that line the drain, it is imagined that the walls of the drain 5863 were built directly against the walls of trench cut for its construction. The drain 5863 appears to continue under wall 5649 into the eastern entrance corridor and the western extent is truncated by cut around well 61-9 (5864) with which the drain does not communicate. The fill of the drain was excavated as deposits 5592 and 5651.
The removal of the surface 5902, coordinate with the surfaces 5900 in the NW corner of the courtyard, revealed the foundation trench fill 5905 for the EW wall 5906 on top of which wall 5443 was built. Both wall 5906, which extends from 1032.91 to 1033.53 N and 265.36 to 267.01 E (top el. 84.61; bottom el. 84.23), and its foundation trench are truncated to the west by the robbing trench for the NS wall that form the W boundary of the courtyard (cut 5859). Wall 5906 appears to be part of a courtyard installation of unknown function, perhaps the foundations for a staircase. Another pebble and packed earth surface 5956, 5857 and 5851, picked out with patches 5817, 5819, 5833 and the shallow ash filled pit 5829, was revealed by the removal of the surfaces 5803 and 5815 south of the line of the drain. A patch of cement (deposit 5827) that is likely residue for cement mixing in the courtyard for a construction project in the house was also revealed. The surface in the NE part of the courtyard excavated as deposits 5900 and 5902 likely corresponds to the surfaces S of the line of the drain excavated as deposits 5803 and 5815, which in are contemporary with the superimposed surfaces 5823 and 5961 in the NE. The excavation of the surface 5823 produced a bronze ring with incised decoration (MF-2008-23).
The removal of the surface 5868 in the NW corner of the courtyard revealed the foundation trench fill 5873 running along the south face of wall 5741 (59-G). Wall 5741, also revealed by Lattimore (NB 230 p.178) and built of courses of roughly hewn blocks with tile stringers and half a spoliated column drum, extends from 1035.41 to 1036.52 N and 270.08 to 275.80 E (top el. 85.16, bottom el. 84.12). Excavation of the fill of the foundation trench 5873, revealed that wall 5741 incorporates and is built onto the earlier NS wall 38-G (NB 230 p.95), which forms the E wall of the room N of the courtyard, and with which wall 5741 forms a corner. The gap between wall 5463 and 5741 represents the doorway that allows for communication between the area of the courtyard and the room N of the courtyard. It is punctuated by a block with a doorjamb cutting, structure 5865. The surface 5868 is likely contemporary with the surface 5909 in the NE corner of the courtyard as the removal of this surface revealed the fill if the foundation trench that runs along the south face of wall 5463 and the trench for the robbing out of part of that wall (5549). This foundation trench remains unexcavated. Wall 5463 extends from 1034.95 to 1035.81 N and 265.58 to 268.81 E (top el. 85.16; bottom el. 84.15. The eastern portion of wall 5463 was revealed by Lattimore on May 12th, 1961 (NB 230 p.95) and the western portion was revealed by the removal of the fill of the robbing trench 5545 and 5852. The EW robbing trench 5549, filled by 5546 and 5852, truncates the earlier NS robbing trench 5859.
The surface 5909 in the NW corner of the courtyard, coordinated with the surfaces 5868 in the NE and the pebble surfaces south of the drain 5851, 5856 and 5857 (lotted as 2008-61) likely represent the first phase of the use of this area as a courtyard associated with the EW wall 5463 and 5471 and the threshold 5865. The removal of these surfaces mark the end of the pebbly surfaces appropriate to an outdoor area and reveal the foundation trench for the earliest wall associated with the floors above.

ROOM NORTH OF THE COURTYARD:
Directly N of the courtyard and in communication with it through two successive thresholds is a room bounded to the N by the EW wall 5562 (58-G NB 230), to the E by the NS wall 38-G (NB 230) to the S by the EW wall 5463 and 5741. The W boundary of the room is likely wall 5473 (71-G NB 230) in the Frankish period that is later largely robbed out (NB 235 p.19). It is possible that before the construction of wall 5473 (71-G) the room extended westward all the way to the NS wall 5724 (260.12 to 261.59 E), though a NS robbing trench 5830 (264.66 and 265.42 E) presents another candidate for the western boundary of the room. This season the room was excavated between the NS robbing trench 5830 and wall 38-G running W to E and wall 5562 (58-G) and the N face of 5463 running N to S, except a few deposit that communicate with the courtyard to the S via a threshold. Lattimore reports ceasing excavation in Room 12, which includes both the Byzantine courtyard and the room N of the courtyard at the strosis 84.765 (NB 235 p. 178). Excavation in this room began with the removal of what appeared to be a wheelbarrow ramp from the excavation in the 60’s abutting wall 5473 and wall 5562 (deposits 5459, 5461 and 5465).
The later of the two thresholds between wall 5463 and 4741 (5462), which overlaid the latest deposits in the room, was removed and subsequently revealed an earlier threshold (5865). A series of floors and leveling fills (5569, 5570, 5571, 5572, 5585 and 5587) were removed. The excavation of 5572 and 5585 revealed a small area of tile cobbling where wall 5473 (71-G) abuts wall 5562 (58-G). The tiles can be seen to continue westward underneath wall 5473 (71-G). Both deposits 5585 and 5587 were martyred during excavation. to prevent contamination from the scarp to the west. These martyrs remain to be excavated at a later date. The removal of the fill 5587 revealed a chunky layer of broken tile and marble as well as cobbles, because this deposit runs underneath the tile floor along wall 5562, excavation of this fill is delayed until wall 5473 (71-G) can be removed.
The removal of the fills and floors of higher elevation in the west part of the room was followed by the excavation of a trench along the N face of wall 5463 (fill 55573; cut 5574). The removal of the fill of this trench revealed second cut at a lower elevation, which appeared to be the foundation trench on the N side of wall 5463 (5910). The fill of the lower cut was excavated as deposit 5613. The level at which the cut was made and width of the trench are comparable with the foundation trench on the S side of wall 5741 (5874). However, the cut seems to end at 267.6 E and does not extend along the length of wall 5463. Further excavation of the deposits to the west is necessary to clarify the issue.
A series of small pits and trenches were excavated on both the N and S sides of Wall 5463 (5712, 5721, 5725, 5904, 5898, 5610, 5557). Their function is unknown. A larger pit filled with cobbles and boulders that abutted wall 38-G was also isolated and excavated (5583, 5584).
Two more lenses of fill were excavated in this room 5578 (lotted as 2008-60) and 5800. Resting on the fill 5800 was a block with three rectangular cuttings, possibly for a small tripod, it was removed as structure 5809 and placed on top of wall 5562 (58-F).
The threshold 5865 consists of a rectangular block with doorjamb cutting set in the gap between the walls 5471 and 5463. The block rests on a surface that slopes to ward the S into the area of the courtyard. The courtyard floor 5800, which bridged both spaces also assures communication between the lower courtyard surface and raised floor of the room N of the courtyard. This like explains the addition of a step in the later threshold (5462).




ROOM SW OF THE COURTYARD:
Daniel Geagan conducted excavations in this area (Agora SW-J) in the 1963 season (NB 254; 262). The room bounded by wall 50-J to the N, wall 51-J (structure 5360) to the E, wall 52-J to the S (structure 5285, 5411; 5866) and truncated by the foundations for Turkish wall 2-J. Geagan dicovered a built pithos (pithos #1) in the center of the room (NB262 p.35, 38-9). The elevation of the mouth was reported as 84.664. He excavated this room down from 84.91 to a floor level at 84.663 (Fill A; lot 1293; NB 262 p.36) and notes a foundation trench along the N face of wall 52-J (structure 5285, 5411, 5866) but does not excavate it. In the 2008 season, we removed the backfill in this room, revealing the floor described by Geagan.

FRANKISH BASEMENT WITH IRON OBJECTS:
Immediately south of the room SW of the courtyard of the 1961 Byzantine house is an adjacent room bounded to N by wall 52-J (structure 5285, 5411; 5866), to the E by wall 49-J (structure 5284), to the S by wall 23-J (structure 5216).
Geagan excavated this room in the 1963 season down to a floor at 84.377. He left a martyr along the NS wall 5284 (49-J) that wrapped along half the stretch of the EW wall 5411 and it is this martyr and floor that we excavated this season after removing the backfill from the area. A post-season photograph (63-18-25, vol.19 p.8) shows the extent of Geagan’s excavation of the room.
Excavation of the stratigraphic deposits in this area began with the removal of structure 5285 (1027.33 to 1027.77 N; 263.70 to 264.80 E; top el.85.18 bottom el. 85.03). Structure 5285 is the filling-in of a doorway that communicated with the room to the N. The removal of structure 5285 revealed a plaster floor that covered the threshold block and doorjamb cuttings of the earlier doorway (5415). The removal of this surface revealed a packed dirt floor (5416), which also covered the doorjamb cuttings. Beneath the floor 5416 was another floor 5429 that appears to be associated with the threshold (structure 5866). The structure 5866 consists of a threshold block set in a gap punched through the earlier EW wall 5411. On either side of the threshold block are rectangular cuttings for the insertion of wooden doorjambs. These are cut into blocks that belong to the earlier wall 5411.
Beneath the floor 5429 were a series of fills (5432, 5441, 5448, 5452, 5453). This sequence of leveling fills appears to correspond to Geagan’s fill G (lot 1289; coins 63-340, 63-340a; NB 262, p.56 #14 and 19 for notes; NB 254 p.182 for section). The removal of these fills revealed an ashy floor (5456). The removal of the ashy floor (5456) and two lenses of leveling fill beneath it (5458, 5460) revealed a layer of fill with many rooftile inclusions. The broken tiles were reserved during excavation to look for joining fragments, but the mixed tile types and lack of joins indicated that this lens (5472) was re-deposited destruction debris rather than a destruction layer. The tile layer 5477 appears to correspond to Geagan’s fill G2 (lot 1290; coin 63-351; NB 262, p.67 #3 for notes; NB 254 p.182 for section). The tile-filled layer was cut through by the foundation trench (cut 5471; fill 5466) for the NS wall 5484.
Wall 5484 (49-J), which extends from 1023.80 to 1027.20 N and 265.32 to 265.80 E (top 85.77; bottom 84.31), was uncovered by Geagan on May 29th (NB 262 p.55-55). Beneath the tily layer was another layer of fill (5477). The fill 5477 corresponds to Geagan’s fill H (lot 1291; NB 262, p.70 #7 for notes; NB 254 p.182 for section). The excavation of the fill 5477 produced a glass bead necklace (MF-2008- 18) and revealed a packed dirt floor 5891 (el. 84.35). This is the same floor (el. 84.377) that Geagan exposed in the SW corner of the room and ceased excavation in 1963 (NB 262 p70). An iron axe (MF-2008-9), iron spearhead (MF-2008-8), iron sickle (MF-2008- ), a handle and a nearly complete coarse mug (C-2008-7) were found in situ next to three articulated goat vertebrae in the NW corner of the room where the EW wall 5411 is truncated by the Turkish wall 2-J (5217). The objects were removed as context 5507. A shallow pit (5890) cut through the floor 5891. The fill was excavated as deposit 5889 and produced a base fragment of a zeuxippus bowl. The removal of the floor 5891 revealed a NS robbing trench, likely associated with a roman wall 62-J (NB 262 p.156 ff.). The fill of the robbing trench remains unexcavated. The bottom of the EW wall 5411 (52-J) has not yet been reached, nor has a foundation trench for this wall been isolated along its S face. Wall 5411 extends from 1027.28 to 1027.89 N and 262.64 to 265.88 E (top el. 85.73). It is abutted by wall 5284 with which it forms a corner, and truncated to the W by the Turkish wall 2-J (5217).
Wall 5216 (23-J) runs EW from 1023.24 to 1023.88 N and 263.00 to 265.60 E. It is built against wall 5284 and on to wall 32-F (NB229, p.181), which runs along the same orientation as 5216, and is truncated by the Turkish wall 2-J to the W. Geagan excavated the foundation trench for wall 5216 on the S side of the wall (lot 1408; NB 262, p.15 #15-16; p.26 #16 for notes; NB 264, p. 182 for section, disturbance V).
Because of the elevations of the fill 5477 and floor 5891, which both produces Frankish material, and the lack of a foundation trench on the N side of wall 5216, this space is interpreted as a basement.

ROMAN WALLS:

Late Roman
The walls whose foundation trenches we have excavated have been dated to the Late Roman Period. The foundation trench of 5218 was excavated in two discrete deposits (5221, 5235). The upper layer was dated by pottery and coinage to the 5th and 6th centuries and the lower one to the 4th. In any case, the deposits from this foundation trench show that Wall 5218 was constructed during the Late Roman Period, no later than the end of the sixth century C.E. This interpretation supplements the previous work done by Heidi Broome-Raines, who gave the trench for the E face of this same wall in the room N of room F,West a Late Roman date based on a coin deposited in it from the late 4th century (context 5060). This coin gives a secure terminus post quem for the trench and wall. The pottery and coin finds from the west side of the wall locate this trench and its wall within the fifth or sixth centuries. Geagan also reports excavating the foundation trench for wall 5218 (wall 22-J NB254, p.146 #16 for notes; p.197 for section; lot 1389), though it appears that he excavated on the portion S of wall 5215 (56-J=38-E).

Middle Roman
The Middle Roman period is represented in our area by the fills bounded by Walls 5216 (NB 254 23-J), 5218 (NB 254 22-J), 5215 (NB254 56-J=38-E), and 5217 (NB254 2-J) and starting securely at 85.47 El. (5248). None of the these wall abut or bond with one another, however walls 5218 and 5215 are separated by gap that appears to be a doorway emphasized with an orthostate and likely belong to the same phase of use. There were many contexts below 5248 that were all Middle Roman and seemed to be associated leveling fill (5248, 5266, 5277, 5311, and 5320). After taking only a thin layer off we started to come down on a large secondary deposit of tile (5311 lotted as 2008-46). This along with the contexts below it down to 85.02 El. (5320 lotted as 2008-47), where we stopped digging, turned up dates to 300 +/- 25 CE. This all seems to be part of a leveling fill.
A pit (Fill 5171, 5183 (lotted as 2008-44), 5351, 5350, 5331; cut 5175) containing material dating to the first half of the fourth century CE was filled along the north face of Wall 5215. The excavation of the pit fill 5331 (lotted as 2008-45) produced a plastic terracotta lamp (L-2008-1), a seated muse figurine fragment (MF-2008-24) and a fragment of a terracotta comic mask. Beneath the pit fill the foundation trench for wall 5215 was detected and excavated (fill 5352; 5254 cut). The E edge of the foundation trench for wall 5215 (cut 5354) is formed by an ashlar limestone block. This block is on the same orientation as Geagan’s NS wall 62-J (NB 262 p.197 for cross-section) exposed on the S side of wall 5215. A similar block is visible in the bottom of the foundation trench for the EW wall 5216. Both blocks are in alignment with an unexcavated NS robbing trench N of wall 5216.