The following is a summary of excavations of Frankish and Late Byzantine levels by the Green team in North of Nezi in the 2008 season. Our area lies directly N of the courtyard of the Byzantine house. The excavation area is defined in the north by a scarp created during 20th century excavations of the south Stoa (1046.50 N), in the south by EW running wall 5562 (NB 230 wall 58-G; 1039.10 N), in the E by NS running wall 5430 (NB230 wall 77; 276. 50 E), and in the W by a NS running scarp (260.00 E) created by earlier excavations (see NB 253 p. 141).

In the first session from April 7th to April 23th, Amit Shilo and Thanos Webb oversaw the excavation of an area defined to the north by 1044.50 N, to the south by 1038.75 N, to the east by 264.50 E, and to the west by 260.00 E. In the second session from May 15th to 23rd, Christina Kolb, Sarah Lima (for the first week), and Thanos Webb oversaw the excavation of an area defined in the north by 1044.50 N, in the south by 1039.00, in the east by 276.50 E, and in the west by 261.50 E. In the third session from May 26th to June 13th, Thanos Webb oversaw the excavation of an area defined in the north by 1046.50 N, in the south by 1038.75 N, in the east by 268.60 E, and in the west by 260.00 E. We worked under the supervision of excavation director Guy Sanders, assistant field director Alicia Carter. We excavated with Thanasis Sakellariou as pickman and Kostas Arberoris as shovelman and barrowman.
Our main objective at the start of the excavation season was to get a better understanding of the interface between our area and the Byzantine courtyard immediately to the south. An EW running wall (5562) was the southern limit of our area and the northern limit of the room N of the courtyard of the 1961 Byzantine house (the excavation of the courtyard was summarized by Jody Cundy and Megan Thomsen, as well as Sarah Lima). A secondary objective was to determine whether our area was interior or exterior space. In order to accomplish these objectives and maintain the open area excavation methodology, we sought to remove all the trench martyrs left by previous excavations. In practice this meant we were constantly moving to the top of next martyr in order to stratigraphically bring these later deposits down to the earlier phases exposed in the 1960’s. Interpretation of our area was often tricky. In some cases, using the 1960’s excavation notebooks proved helpful in explaining the presence of remaining martyrs and scarps. In most instances, however, we were unable to connect the information recorded in old notebooks with what we saw in the field. The 1961 notebooks that relate to our area are 230 and 235 and they were kept by Steven Lattimore. The 1963 notebook that relates to our area is 253 kept by Charles K. Williams. This summary will be presented in two sections beginning with the area excavated west of wall 5677 and followed by the area east of wall 5677.

Area W of wall 5677 (Agora SW-H, NB253, Room 2E, p.23ff.)
In this area we excavated this season a robbing trench, a sequence of floor levels with an associated hearth and dry storage pithos. This area is bounded to the E by wall 5677 (NB253 1-H), that is composed of random courses of rough-hewn blocks and tile with punctuated with rectangular orthostates. To the N is wall 5842 (NB253 2-H) which appears to abut wall 5677 to the E and continues westward outside of our excavation area. To the S is the EW wall 5562 (NB253 3-H and NB230 58-G) that is abutted by the NS wall 5677 and continues eastward to roughly 270 E, where it meets the NS wall 38-G. The N-S running martyr he created is the western most extent of our area (260.00 E).
C.K. Williams first excavated this area in 1963 (NB 253), revealing three floor levels and a storage pithos (Corinth photos v. 18, page 42; neg, 63.9.24). The half exposed pithos was still in situ when we started excavating. The edges of stepped floors left by Williams had eroded over the past 45 years into a slumped mound. Excavation this season began with the removal of the fill of a robbing trench in wall 5677 (fill 5181;cut 5195). It appears to be the robbing of only one block from this wall. Then a series of fills into which the robbing trench was cut were excavated on the W side of the wall (5219, 5223, 5225, 5227). The removal of these fills revealed floor 5229, which corresponds with Williams’ floor 1 (NB253 p.129). Floor 5229 appears to be associated with both wall 5677 and wall 5842. Floor 5229 and floor 1 both show signs of extensive burning (NB 253,p. 28). We reserved a sample for flotation from context 5229 to recover any potential carbonized botanicals. Beneath floor 5229 is another floor 5234. It is likely that context 5234 is the same as Williams floor #2 (NB 253 p.129). Floor 5234 also abuts both wall 5677 and 5842. The pithos that abuts wall 5842 appears to still be in use when floor 2 (5234) is laid down. The removal of floor 5234 revealed another floor 5841 through which a hearth pit (5838; 5839; 5849) was cut. There was another floor beneath 5841, deposit 5860 in which the pithos was sunk (cut 5851; 5850). Beneath this floor were several lenses of fill (5872, 5879, 5885). Deposit 5885 was dumped fill consisting of large cobbles and fragmented roof tiles. Within the deposit’s matrix in the southeastern corner where walls 5677 and 5678 meet, a cache of over 50 iron objects was discovered. The iron objects include: large nails, bent and broken iron strips, broken and folded over knives, and one object with a series of small holes which we are unable to determine its function. Most significant perhaps are the multiple examples of door-lock hardware. Cover plates with keyholes, square boxes which would hold the locking mechanism along with a keyhole, and hinges in various stages of completion. This marks the end of excavation in this area this season, although deposit 5885 remains unfinished. Further excavation in this area might reveal the foundation trenches for walls 5677, 5842 and 5562 and clarify the function of the installation 5888.

Area E of wall 5677

Like the area W of wall 5677, the area E of the wall is bounded to the S by the EW wall 5562. The boundary to the N is an excavation scarp at 1046.50 N, and to the E by the NS wall 5430 at 277.00 E. The area appears to belong to a building adjacent to the 1961 Byzantine house. Excavation in this area this season revealed several robbing trenches and large pits, a roof-tile-built pithos and several floors. The latest features in this area are two NS walls, or perhaps foundations for walls, that overlie the earlier contexts. Permission to dismantle wall 5430 and 5394 has been requested. We excavated two sequential floors that abutted wall 5677 (5172=5186, 5189=5202). They date to the Frankish period. Cutting these floors was a deep, Frankish pit (5169) that destroyed part of a Late Byzantine roof tile-pithos (Corinth Photographs v. 18: page 36, neg. 61. 27. 5.). The roof tile pithos (5504) was then backfilled in the Frankish period. A few modern sherds contaminated the first layer of fill (5165) and subsequent fill levels (5167 and 5496) were dated to the Frankish period. The last Frankish fill (5496) within the pithos contained three complete dog skeletons. Underlying the Frankish fills within the roof tile-built pithos was a final Late Byzantine fill (5505). This fill was not completely excavated because it served as internal support for roof tile-built pithos.
Another Frankish pit (5758) was discovered near the scarp of the South Stoa. Half of the pit had been removed when the South Stoa was excavated.
Frankish EW wall 5678 made of stone-rubble construction and has no bonding material. Since no foundation trench has been discovered, it is possible that 5678 is the rubble foundation for a wall rather than the wall itself. The construction of the 5678 is earlier than the digging of Frankish pit 5169. Also, 5678 cuts an earlier Byzantine robbing trenches 5714 and 5802.

Four Byzantine pits dating to the 12th were excavated (5629, 5595, 5776, 5772). The fill of the pit 5772 produced redeposited 11th century material and is dated by stratigraphy to the 12th century. Two of the pits 5629 and 5595 were large and similar in size being two meters in diameter and over a meter deep. Their fills were different; 5595’s fill (5567) was mostly dirt with only 15% inclusions. In comparison, 5629 had 50% inclusions. Whereas the fill of 5629 represents what we would expect to find in a refuse pit, pit 5595 appears to have a different use that remains obscure.
The pits 5772 and 5776 distinguish themselves on account of the majority of their fill was made up of ash. These are interpreted as pits for the disposal of burned material, whether from hearths, cooking or industrial operations.
A roof tile-built pithos (5504) was also discovered. It appears to have been built while Late Byzantine floors 5528 and 5525 were in use. Its 12th century fill (5505) was disturbed by the later digging of Frankish pit 5169.
Two Late Byzantine NS robbing trenches (5714 and 5802) were also discovered. 5714 is cut into the rubble and tile fill of 5712. It abuts NS Byzantine wall 5677 and is cut by EW Frankish wall 5678. At the base of 5714, two marble floor tiles were found in situ. The fill over these tiles dates to the Middle Byzatine period (late 10th/11th npd). These tiles appear to be from the same marble-tile floor as 5710. Also, it is possible that this marble tile floor appears in the scarp of the cut for the roof tile-built pithos (5504).
A large NS robbing trench (5802) cuts the floor revealed by 5531. It abuts wall 5562 in the south and is cut by pit 5629 in the north. This deep and long robbing has four large, cut blocks at its base and what appears to be a Roman cistern on its east-facing side.

The two of the walls in the area have not been phased (NS 5677, and EW 5562). Foundation trenches have yet to be found in any of the areas excavated thus far. As for walls NS 5394 and NS 5430, their dates are also contingent on the discovery of foundation trenches. At the moment, it appears that wall 5562 predates all the walls in our area.