Contextual Analysis of the Use of Space at Two Near Eastern Bronze Age Sites

Part 3: The Excavations at Kilise Tepe (1994-98)

J. N. Postgate

Contents

Location of the Site
The Archaeological Strata
History of the Project

Early Bronze Age
Middle Bronze Age (Level IV)
Late Bronze Age (Level III)
The Stele Building (Level IIa-c)
The Stele Building (Level IId)
The Eastern Building
The NE Area
Level IIe
Middle and Later Iron Age: Levels IIf-h

Other Soundings

Soundings on the E Side of the Mound

The Byzantine Period

Bibliography

Context information for samples taken at KT 1994-1997




Location of the site

The mound of Kilise Tepe sits on a natural conglomerate bluff which projects from the terrace flanking the left side of the valley formed by the Kurtsuyu river, close to its junction with the river Göksu. Especially with the accumulation of up to 13m of archaeological strata, this hill provided a dominating position at the south-east end of the Mut basin. It controlled the main route going from the central Anatolian plateau south-eastwards down the left side of the Göksu valley, but also overlooked access to the crossing of the Göksu which took a second route onto the south-western side of the valley or further west via Gülnar (or 1st millennium B.C. Meydancik) and down to the Mediterranean near Aydincik (Kelenderis).

Although the climate of the Mut basin is hot and dry (July and August temperatures in the 40's and rain very unusual for 4 months in the summer), cereal crops and olives are grown on the surviving terraces and the gentle slopes, both unirrigated. Irrigated crops, which include vegetables, fruit trees, vines and stands of poplar, are very successful. There are small springs scattered across the landscape, including one significant source at the foot of the Kilise Tepe mound itself, but the principal supplies are from the Kurtsuyu, which feeds a triangular area above its junction with the Göksu, and a copious spring at Pinarbas some 5 km above the right bank of the river from which a considerable area of orchards and gardens is watered.

The Göksu itself has cut too deep in its bed to serve for gravity-flow irrigation, and the thickets and sand-banks along its course are mostly free of human interference. Away from cultivation the landscape is sparsely vegetated, and the steep scarps left by erosion are mostly bare white limestone dotted with occasional bushes. The commonest livestock in the area today are goats, though there are some sheep, and cows are kept for milk; wild pigs are quite often encountered marauding crops.

The Archaeological Strata

The earliest occupation of the site, as found in our H20 deep sounding, seems to have been in the Early Bronze Age, provisionally between 3000 and 2600 BC (date range left deliberately broad). Thereafter it looks as though it was continuously occupied into the early 2nd millennium, that is into the Middle Bronze Age (Level IV). Probably there was no break of occupation between the Middle Bronze Age and the Late Bronze Age (Level III). Level II, the earliest phases of which are perhaps still technically Late Bronze Age, represents a new architectural layout, but again there was probably no break in occupation at this stage. Level IId is almost certainly of 12th century date, while Level IIf is late 8th or early 7th century B.C. Whether there was a break between these two, and if so where in the sequence, remains uncertain.

We have some ceramics but no architectural strata belonging to later Iron Age phases and to the Hellenistic age (4th-3rd centuries B.C.). Finally, almost certainly after a period of desertion, the mound witnessed the construction of a large Byzantine church, and there are up to 2m of Byzantine occupation both on the mound itself and at some places round its flanks. The original building may belong to the 4th or 5th century, and coins suggest some occupation as late as the 11th century A.D.

History of the Project

The first academic notice of the site was given by J. A. Mellaart who visited it in the course of a survey of the Göksu valley in the 1950s. The location of the site, under the name of Maltepe, was described and sherds from it published by him. In 1965 D. H. French also collected more sherds from the surface in the course of a re-examination of the Göksu valley sites. During the 1980s a scheme was drawn up for damming the river Göksu at Kayraktepe, about 15 km above Silifke, as a result of which the mound would be flooded. In 1993 the British Institute of Archaeology at Ankara decided to request a permit for rescue excavation at the site and this was granted in time for a first season of work to begin at the site in July 1994, under the direction of J. N. Postgate. In the course of this season it was established that the correct name of the site is Kilise Tepe (Maltepe proper being the other side of the Göksu). There followed four further seasons (1995-98), after which it has been decided to call a halt; one factor in this decision was the revision of the projected height of the Kayraktepe barrage which means that Kilise Tepe is no longer under threat of flooding.

In response to the status of the project as a rescue excavation, there was no requirement on the part of the Turkish Directorate to carry out an area survey before starting excavation, and attention was immediately focussed on the tepe itself. The site was cleared, excavated and recorded with a grid of 10m squares; in the course of excavation these squares were often subdivided into four 5x5m squares labelled a (for NW), b (NE), c (SW) and d (SE). During the first two seasons the surface of the mound was entirely cleared of loose sherds, stones and other items, which provided us with a considerable range of ceramic wares and a preview of what might be present under the surface. The prospect which had initially attracted the Institute to the site was the opportunity to investigate a Late Bronze Age settlement. It rapidly became apparent that these levels lay beneath an overburden of at least 3m of archaeological deposits. In consequence, most of our work was concentrated at the north-western end of the tepe. Here the steep slope of the mound gave access to strata of different periods and has enabled us to establish a cultural sequence.

In the text which follows we describe the character of the different levels in order to provide a general setting for the units from which samples were chosen for analysis. After each section the relevant units are listed, arranged broadly chronologically, but for ease of reference the units are listed again in numerical order at the end, with the column headings.

Early Bronze Age

The earliest levels were investigated in square H20. Here a large robber tunnel had been dug into the side of the mound, and in the interests of speed we enlarged and squared off this hole to provide a cross-section through the lowest strata. These were founded on the natural conglomerate bed-rock at about 13m below the summit of the mound. Seven building levels were clearly distinguishable (lettered from e down to k), in some cases with more than one construction phase as well. Although in Levels g, h and j in particular the walls were well constructed and finished with good-quality clay plaster, horizontal exposure was in most phases no more than 5x5m, and does not allow us to reconstruct the plan of any of the buildings involved. The walls and rooms are not unexpectedly big, and their other characteristics are consistent with domestic houses. Two of the levels had suffered destruction by fire, Level j, with a number of pots and other items caught up in the destruction, and Level g, which did not have in situ artefacts. After level f it seems that the area may have been an open space, to judge from the absence of architecture and the number of storage pits. Artefacts recovered are principally spindle-whorls, flint tools, and of course pottery. The ceramics, which have been studied by Dr. Dorit Symington, have parallels in the Early Bronze Age repertoires of other Anatolian sites, notably Tarsus, and on this basis Levels k, j, and i are provisionally assigned to Early Bronze II (2800-2600), and Levels h, g, f and e to Early Bronze III (2600-2000). It should be stressed that these are provisional assignations and that the dates are very approximate. We are awaiting the results of C-14 analysis on some samples from these levels.

UnitPur.Lev. TypeSq. DescriptionSample
5429a5k4H20cW half, construction surface below 5428 (//1891)97/33
1858a5j4H20cRm 83, destr. debris96/53
1871a5j4H20cRm 83, destr.debris97/01
5412a5j4H20cW half, destr. debris N of W225 (//5417)97/19
1852a5i4H20cE half, packing below 185196/30
1853a5i4H20cE half, packing assoc. with W792, below 185296/39
1855a5i7H20cP96/9696/49
1838a5g/h4H20cPacking below Phase G floors96/16
1836a5g4H20cRm 62, fill below occ. surface below 1831, 183596/10
1812b5f6H20cRm 55, fill below 181095/31
5310a5f45H20dRm 52, fill below 530797/29
5328a5f48H20dRm 53, filling of cut for W23897/16
5335a5f45H20dRm 51, fill below 533197/34
1810a5e/f4H20cE strip, packing below 180795/27
1822a5e7H20cP95/65 fill95/38
5321a5e7H20dP97/5497/12
1288a5e7H20dP97/7, below 1289 

Middle Bronze Age (Level IV)

Two Levels of Middle Bronze Age occupation were excavated in squares H19 and H20, over an approximate area of 50 square metres. Both had suffered extensive burning, but the earlier (Level IVa) is better preserved. On the east side there was probably an open space; here from among the ashy debris several baked clay loom-weights were recovered. A room 6.20 x 4.40m to the south-west had an impressive white-plastered central hearth, with two pottery jars built into a bench on the west side and 5 or 6 unbaked clay containers occupying benches in the north-east corner of the room. It is impossible to say for certain whether, and how, this space was roofed, but in the course of the destruction 16 courses of brick from the upper part of its east wall had fallen across the room and we can be sure that it stood originally at least 1.88m high. A step led down from a doorway in the north-west end of the room into a second room, most of which had been dug away by the villagers' tunnel. Here there had been a small fireplace in the south-western corner, and more mud-brick benches against the south-east and north-east walls, with more storage vessels.

Level IVb was laid with little levelling more or less directly on top of the destruction debris of IVa. We did not identify any architecture, and the whole area seems to have been open space. Various jars had been lying on the surface at the time of the fire, and at the west side there were two fire installations. The pottery from this level and from IVb includes grey-ware beaked jugs and other types with parallels at Tarsus which have been assigned to the Middle Bronze Age, say 2000-1600 B.C.

UnitPur.Lev. TypeSq. DescriptionSample
1258a4a4H20dRm 43, destr debris below 125395/01, 04, 10, 13
4273a4a4H19a/bRm 42 SE, collapsed masonry etc (//4263-4)96/58, 59
4278a4a6H19a/bRm 42, burnt occ. surface96/57
4261a4a/b4H19a/bRm 42, packing below 4255 (//4257)96/35
4249b4b6H19a/bBurnt debris on IVb surface below 424896/18, 20
4251a4b4H19a/bBurnt occ. deposits on IVb surface below 424996/21

Late Bronze Age (Level III)

Level III occupies a stratum about 2m deep with at least 5 architectural phases. Owing to the slope of the mound, very little survived of Level III in squares H19 and H20, and the stratum is known from a strip measuring about 20m from northeast to southwest and 10m from northwest to southeast, stretching from J20a in the northeast to H19b in the southwest, with the majority of the work in I20 and I19.

The earliest three phases of this level (IIIa-c) are attested in H19b, directly above the Level IVb destruction material, and are represented by similar stone foundations and relatively clean associated packing. Phase IIId was a more substantial rebuild, and here in the SE corner of the square several courses of very clean yellow mud-brick survived on the stone foundations, with yellow clay plaster on the faces of the wall. This phase was also reached in I19 and I20, and is the earliest to give a coherent (though incomplete) building plan, with two rooms and an external courtyard area. The NW limit of the building is provided by a single wall running for at least 15m from NE to SW, more or less in conformity with the modern contour of the mound; owing to slope of the mound, only the lowest foundation stones survived. To its south, on the W side of I19, was a space slightly more than 5m in width with mud-brick walls each side (Rm 32). Its size points to an open space or courtyard, although its E wall had received several coatings of fine quality clay plaster, usually typical of interior walls. The space was filled with clean brick from the collapse or destruction of the walls, and the floor was not reached here.

In the next-door room to the E (Rm 30), which measured 5.50m (N-S) x 2.50m (E-W), at least two occupation phases were recognized, with their floor surfaces 0.25m apart. The walls were plastered, though not as elaborately as in Rm 32, and in the later phases two large storage jars stood close to the east wall. Parts of their bases remained in situ, other sherds lay round about, and these were mixed in with a granular clayey coating which had been applied to the outside of the jar, formed over a framework of reed. Both here, and in the IIId deposits excavated in H19b, it was clear that there had been a gradual accumulation of occupation surfaces over the lifetime of the building.

To the east of Rm 30 there seems to have been an open space in Phase IIId, bounded on the west by the east wall of the room and on the north by the northern wall; limits to the south and west were not reached. The final surface here showed extensive signs of burning, and lying on it were tumbled lumps of mud-brick masonry. At the north side, individual mud-bricks could be seen tilted as they slumped off the stone foundation, making it clear that the material filling the courtyard was indeed debris from the destruction of the building itself, rather than constructional materials brought in from elsewhere. Mixed in with this debris was a large quantity of potsherds, which provide us with our best sample of the Late Bronze Age ceramics at the site, including red-burnished ware and pieces of libation arm. Similar material came especially from the stratified Level IIIa-c layers in H19b.

Within the courtyard, about 4m from its west wall, was a large fire installation (FI 97/10). Although some stones from its base remained in place, not enough survived to allow us to reconstruct the superstructure, but it occupied a space some 4.50 x 2.50m. This was covered with layers of fine ash, principally black and light grey, but in places there were small unshaped logs, completely carbonized, which lay as though they had fallen from a flimsy shelter. These were sampled for dendrochronological dating by the Carolyn and Malcom Wiener laboratory under Prof. P. I. Kuniholm, and one piece with 192 rings suggests a date of 1380 B.C.

After this episode it seems that the building was reconstructed, although this phase (IIIe) is virtually only represented in the courtyard area. The west wall of the courtyard was moved slightly to the east, and the stones of its foundation rest in part on the IIId destruction material. Lying over the courtyard floor of this phase is a band of burnt debris which was instantly recognizable because it was rich in carbonized food remains, in particular figs which had been dried and stored on strings. The large fire installation was also covered by this surface and debris; further over to the east was a smaller hearth belonging to IIIe, but otherwise there were no features of any note.

UnitPur.Lev. TypeSq. DescriptionSample
4243a3a4H19aPacking below W703 and W74496/17
4246a3b5H19bPacking E of W74596/11
4231a3c4H19bPacking below 423096/03
4209a3c/d5H19bRm 32, occ. surface below 420595/28
1376b3d4J20aE ctyd. Destr. debris below 137295/53
4205a3d4H19bPacking assoc. with W737, W73895/32
1372c3e4J20aE ctyd. Burnt occ. deposits (//1388) below 137195/44
1378b3e6J20aFI 95/9 fill95/52
1381b3e4J20aBurnt occ. deposit round FI95/995/51
1388a3e5J20cE ctyd. Burnt occ. deposits below 138797/17
4017b3e4I20b/dE ctyd., packing below 4016 (//1426)95/34
4018b3e5I20b/dE ctyd., occ. surface below 401795/36
5519a3e4I19b/aRm 30 packing below 550997/27

The Stele Building (Level IIa-c)

Above the Level III structures the architectural layout of the area and the artefactual repertoire changed radically. The area in I19 and 20 previously occupied by Rooms 30-32 became an open courtyard, while to its east an approximately square building was erected with a solid stone foundation, its walls aligned quite differently. This we refer to as the "Stele Building" from the stone found in its Phase IIc courtyard. As for its date, it seems clear from evidence to be mentioned below that phase IId belongs to the early 12th century, and hence that it was founded in the 13th century at the latest.

Phases IIa-b:The earlier phases, IIa (which is the initial foundation), and IIb (the period of occupation between the first occupation and the IIc destruction), were only excavated in limited soundings beneath the IIc floors within the building, and in the western courtyard.

UnitPur.Lev. TypeSq. DescriptionSample
5526a?2-35I19aOcc. deposit round P97/70 (//5504)97/31
5501b2a7I19bP97/69, fill of NW part97/10
2884a2a/b5I19bOcc. surface W of W823, below 288097/09
1370c2a-b4J20aPacking of cut round NW corner of building95/37
4540d2b9K19aRm 2, sounding below IIc floor97/26
4551d2b9K19aRm 3, additional sounding to SW of 453597/30
1696a2b7J19bRm 3 FI 96/18 central hearth96/54
3963b2b8J19dRm 7. Sounding below IIc floor97/25
3965a2b5J19dRm 7, sounding below 396397/28
4004b2b7I20dP94/5 cut by P95/28 and P95/34 (//1421)95/24
4008c2b7I20dP95/34 fill95/25
2871a2b4I19bW ctyd., packing below 286797/06
2883a2b7I19bP97/41, below 287397/08
4534d2b/c9K19aRm 4 sounding below IIc floor in SW corner97/15, 23
4535d2b/c9K19aRm 3 sounding below IIc floor in NE corner97/23
2848a2c7I19dW. courtyard: P97/6 lower fill below 285897/03

Phase IIc: Our description of the building necessarily concentrates on Phase IIc, for which the plan and evidence for its use were preserved by a severe conflagration. The external dimensions of the IIc building are approximately 18m NE-SW and 14m NW-SE. It was mainly constructed on the walls of the earlier phases, and in several parts of the building one or two courses of stone were laid on the IIb mud-brick, followed by timbers laid longitudinally, upon which fresh mud-brick walls were constructed. In the case of W122 along the W side of Room 3, a stub of IIb mud-brick was actually left standing and re-plastered, so that the IIc foundation stones were at a level higher than the contemporary floor.

Room 3: At the heart of the building is a courtyard (Room 3) with a number of unusual features. There is a circular hearth at its centre, resting on a plastered rectangular area (some 2.3x1.8m) slightly raised above the floor and protected on the west by a brick partition. A single row of bricks was constructed diagonally across the NE corner of the room, and there are indications that this may have been an altar: like the hearth, it was present already in Phase IIb (and probably at the initial IIa foundation), and the IIb version of this feature preserved a small area where the clay plaster of the face curled over onto the top, giving it a flat upper surface and a height of around 80cm. From the space behind the bricks we recovered a curious collection of humble items, including clay spindle whorls, shells, and sheep/goat astragali.

At the SW corner of the court were three broad steps leading up into the space called Room 10, but in the SE , under thick deposits of burnt destruction debris, was a large slab of sandstone which had been cracked into over 70 fragments by the heat of the fire. It seems likely to have stood upright as a stele. It is 77cm high, 49cm wide and 16-17.5cm thick; it may have been smoothed on at least one face, but it has not been worked. There had been a design in thin red-painted lines on each face, but even on the better preserved (and flatter) face it is very hard to discern, and the significance of the design remains obscure. Nearby within the destruction material was a two-handled dish, associated with carbonized lentils and the skeleton of a snake. One hesitates to attribute any significance to this. Other architectural features of this space, including the two square sockets at the centre of the NE and SW walls, and the post-holes along the NW wall, remain obscure, and it is unclear whether we should assume there was a roof over it.

UnitPur.Lev. TypeSq. DescriptionSample
1686a2c4J19bRm 3. Destr. debris below 164296/45
1688a2c5J19bRm 3. Burnt occ. deposit below 168696/56
1560a2c4K19aRm 3. Destr. debris in SE corner (//1554)96/60
1348b2c4J20dRms 2-3, destr. debris of W122, below 134695/29
1353a2c4J20dRms 2-3, destr. debris below 135095/43

Rooms 1 and 2: At the NW corner Room 3 led through into the W end of Room 2, across some kind of low step or threshold. This room, measuring 5.5 x 2.5m, had no distinguishing features but was also filled with burned destruction debris to a depth of at least 0.60m. At the W end of Room 2 we presume there must have been a doorway through to Room 1, but at this point the slope of the mound and subsequent pits and other disturbance had virtually removed the masonry of the wall down to floor level. Room 1 itself had also been affected by the fire; its clay floor had been burnt black and hardened to the extent that when we first encountered it we thought it was the base of an oven. It also sloped very markedly down to the north, which probably reflects subsidence beneath the NW corner of the building. A plastered bench some 0.60m deep and 0.30 to 0.50m high was built against all three walls at the SW end of the Room. On the plastered face of the eastern bench, looking west into the room, there was a design in red paint, unfortunately as impenetrable as that on the stele, and above this on the bench top there was the emplacement for a pottery jar. Since the remaining rooms of the building are demonstrably utilitarian, one is tempted to conclude that Rooms 1 and 2 accommodated the more formal activities of the building, whether these were of a public or private, secular or religious nature. On the other hand, it will be suggested later that Rm 1 may in fact be an entrance room.

UnitPur.Lev. TypeSq. DescriptionSample
1356a2c3J20d/J19bRm 1. Bench along E side95/45;
1325b2c5J20dRm 1. Burnt occ. floor below 132495/11;

The Storerooms: Behind Room 3 to the east was a storage space measuring 6.75 x 2.20m called Room 4. There was the emplacement for a substantial door post to the south of the doorway and we may assume that the room could be closed off and locked. In the SW corner of the room a large jar (K19:294) remained in situ, and had contained a quantity of burnt cereal. Two other jars, associated with carbonized plant-foods stood in the NW corner and against the E wall, and the positions of another jar and a basket were also identifiable. The brickwork of the S wall of Rm 4 (W622) was bonded with the E wall of Room 3 and its own E wall, and may well date back to Phase IIb. Room 5 was not originally incorporated into the plan of the building, and indeed its access is only from the open space to the S (Room 6). It seems to have been an ad hoc creation, with its E wall supplied by the East Building, the N and W by the rear of the Stele Building, and its S wall by a rough wall composed in part of un-plastered pillars made by stacking square bricks on top of one another. The result was a space some 2.5m N-S and 3.3m W-E, which was also filled with extensive burned destruction debris. The sherds of, and emplacements for, five storage jars were identified along the N and W walls, accompanied by large quantities of carbonized barley and lentils.

UnitPur.Lev. TypeSq. DescriptionSample
4521a2b/c5K19aRm 4, floor round storage jar in SW corner97/18
1561b2c5K19bRm 4. Occ. surface below 1555 (//1995) 
1566a2c4K19a/bRm 4. Destr. debris, below 1551/297/02

A pair of storage or work rooms was also built along the S side of the building. Here Room 8 (4.2m. N-S and 3.3m E-W) retained the bases of four storage jars, of which the two against the N wall had contained wheat. A doorway in the N end of W620 certainly led through into Room 7. No storage jars were found here, although there was a deposit of carbonized olive stones close to the E wall. More enigmatic was a hoard of 92 sheep and goat astragali pushed into a small hole in the ground beneath the plaster floor against the S wall; with them were 12 small copper studs or rivets, perhaps from a perished container. Otherwise the floor surface of Room 7 was clear, but in among the destruction debris there was a quantity of flat stones, mostly fragmentary as a result of the heat, which must somehow have been a component of the superstructure.

Modern practice in the next-door village of Ksla is to insert a layer of flat stones along the top of mud-brick walls to act as a level emplacement for roof beams. However, since the room measures about 5.0m N-S and 5.5m E-W it is almost certainly too large to have been completely roofed, and the floor surface was plastered and showed no signs of posts or pillars to support a partial shelter. In the NW corner of Room 7 a circular storage pit had been dug. This was sealed by a IId wall, but can hardly have been in use in IIc since it is just inside the doorway leading into Rm 8. In the fill of the pit (P97/49) was a pottery jar (J19:320) and lying on its base there was a group of burnt bone implements which look like needles or crochet hooks for making netting or other coarse fabrics (J19:322). Finally, to the SE, the space called Room 6 is probably not a room, but an open space giving access to the East Building via Room 24. A large pit (P98/72) at the centre of this space was currently in use at the time of the IIc destruction, and showed signs of having been used and reused during its lifetime, suggesting that it needs to be considered one component in the activities in and around the building. Large stones and individual mud-bricks placed against the outside of the E wall of Room 7 also hint at some kind of activity in this space, but give no clues as to what kind.

UnitPur.Lev. TypeSq. DescriptionSample
4510a2c4K19cRm 5, destr. debris, below 450797/11, 21
4511a2c6K19cRm 6, carbonized matter in occ. deposit97/20
3929a2c4J19dRm 7, burnt occ. deposit97/05
3927a2c4J19cRms 7 and 8, destr. debris97/04, 07

Rooms 9 and 10: Throughout the rest of the Stele Building the IIc floors were readily identifiable, if only because it was on them that the burnt destruction debris was lying. Here neither floors nor debris were convincingly present, and in view of the staircase leading up to the W end of Room 10 from Room 3, we are forced to conclude that the IIc floor level here stood higher and was removed by later building activity (specifically, no doubt, the construction of the Phase IId building). It therefore becomes very difficult to interpret the function of these rooms. This difficulty is enhanced by features in the E end of Room 10. There is a curving arrangement of mud-bricks which respects the N and E walls of the room and would seem to have been in position in the IIc phase.

West of this is a large pit (P97/73) which was probably standing open at the time of the IIc destruction, since it was packed full of the characteristic burnt destruction debris. Above both the bricks and the fill of the pit were finely accumulating clay layers, heated to orange, which we would normally have associated with a fire installation. It seems necessary to assume that these belong to some IId activity below the regular floor level of that phase. There was some slight but rather equivocal evidence for an entrance from Room 6 at the SE corner of the room, though how this would have worked if the Room 10 floor was higher than +99.41 and the Room 6 ground surface around +98.60, is far from clear. My own suspicion is that a staircase (probably wooden but perhaps starting in brick) led up eastwards from the W end of Room 10 and hence that the bricks and pit at its E end were in the "cupboard under the stairs". Naturally, the possibility of a staircase raises the question of whether there was an upper storey, and whether or not Room 3 was roofed. These are questions which cannot be definitely answered.

Access: The floor level in Room 10 was above +99.41 and therefore as much as 0.8m above the Room 3 floor. Hence the need for the stairway. It was also as much as 1m higher than the floor levels in Rooms 7 and 8 to the S, and this poses the question of access. No comparable stairway was present; some enigmatic features in the S face of the mud-brick of W615 might hint at a wooden construction contrived in the NW corner of Room 8.

An even more crucial question remains unanswered, the location of the main entrance to the building. The layout strongly suggested that Room 9 might have been an entrance hall, leading in to Room 10 at the same, higher, level. However careful excavation along the exterior of the west wall of the building in 1998 failed to reveal any sign of an entrance here, or of any of the features which might have been associated with an entrance. This leaves Room 1 as the most likely candidate for an entrance hall: the slope of the tepe means that at the NW corner of the building the mud-brick walls have been completely eroded, leaving the highest course of the IIa stone foundations exposed. A doorway through the mud-brick superstructure could easily have been accommodated in either the N or the N end of the W wall of the room, and all trace of this would have disappeared.

The Stele Building (Level IId)

After the IIc destruction the Stele Building was rebuilt on much the same lines, but only fragments of the IId phase survived the attentions of the Iron Age and especially the Byzantine builders. The IId floors and walls were best preserved in the southern range of the building where both the IIc and the IId rooms were at a lower level; indeed, at one stage we had assumed the IId architecture in this area belonged to the IIc phase. Even here the IId walls and floors had been massively destroyed by later activities, including a recent robber pit in J19d. There remained disconnected pieces of plastered room walls and benches, underlying a layer of burnt destruction debris. On one surviving patch of floor, in the SE corner of Rm 8, there were sherds from several Mycenaean vessels, including deep bowls and cups attributable to LHIIIC, in the first half of the 12th century B.C. The only other well-provenanced Mycenaean pieces came from a pit sunk into the SE corner of the IIc Room 3: this pit was full of fresh destruction debris, and a IId stone foundation ran over its eastern side, so that it seems certain to belong to an intermediate phase after the IIc destruction and before the construction of the IId version of the building.

In contrast to the Stele Building, to its NE and SE there does not appear to have been an attempt to rebuild the walls in the triangular space occupied by Rooms 15 and 11, or over the Eastern Building. In both areas the IId phase was marked by a heavy layer of ash but no structures. On the other side the western courtyard was re-plastered and remained open as before, and IId destruction material was also found in this area.

UnitPur.Lev. TypeSq. DescriptionSample
1690b2c/d7J19b/dP96/102, below W71896/50
3943a2c/d8J19dRm 10, burnt surfaces below 391797/14
3967c2c/d7J19dP97/73, below 3957 
1540a2d4K19bOcc. layer? E of W770, below 153596/37
1988b2d4K20cRm 2, occ. surface, below 198096/31
1994b2d7K20dRm 4, P96/121 upper fill96/55
2825a2d5I19c/dW ctyd., ashy occ. deposit, below 282396/52

The Eastern Building

The E wall of Room 5 is not part of the Stele Building but is supplied by the W wall of Room 20 in the neighbouring structure, which is called the Eastern Building. This was also built of mud-brick on stone foundations, and was destroyed at the same time as phase IIc. Room 20 itself was paved with irregularly placed and unshaped stone slabs, over which lay a dense layer of black ashy soil. Above this the space was filled to a depth of about 1m with the calcined and distorted mud bricks tumbled from the superstructure. Caught up in this destruction were several pottery vessels, including pilgrim flasks, a trefoil-mouthed jug with clay stopper and at least two heavy pithos-like storage jars. Room 20 has all the characteristics of an open space. To the east it gave access to Rm 22, where a large pit in the floor seems to have been open at the time of the destruction, and through it to Rm 21, which was principally distinguished by the layer of bluish-grey ash directly overlying the floor.

Rooms 20, 21 and 22 appear to form the W end of a building of unknown extent, and were probably reached through an open court (Rm 24), over which destruction debris still lay but in reduced quantities. In general the floors in the Eastern Building are rather lower than the IIc floors in the Stele Building, and this serves to strengthen the suspicion that we may be looking at a basement area, with the more formal parts of the building on a higher storey.

The NE Area

Work in 1994-1995 exposed Level II occupation in the space NE of the Stele Building, in K20. In IIc this proved to be a triangular domestic area bounded on the NW by the solid stone foundations of a wall (W713) belonging to a building now entirely disappeared, and on the NE by a less substantial mud-brick wall which is on a roughly similar alignment and probably forms the SW limit of another building (for the IIc plan see Sonuçlar XVIII p. 453). Apart from a number of fire installations, including a range of three small hearths against the NW wall of Rm 15 (shown on plan as Rm 10), and a similar arrangement in Rm 12 the other side of the wall, there is little else to note in this space. In 1996 a small sounding was sunk below the IIc floor in this area to recover stratified IIb material. This revealed another oven in the NW corner, and another wall running parallel to W713, with cleanish floor deposits between them. Not enough was cleared to allow a confident reconstruction of the layout of the buildings at this date.

UnitPur.Lev. TypeSq. DescriptionSample
1935b2b5K20a/cRm 15 occ. levels below 1931 
1965a2b5K20aRm 15, occ. surface below 195896/13
1971av2b5K20a/cRm 15, occ. surface below 196596/19
1975a2b5K20a/cRm 15, occ. surface below 197296/22
1931a2c5K20a/cRm 15, ashy surface (//1142, 1948)95/30
1938b2e/d4K20b/dRubbly packing sealing W713, below 192495/35
1909a1/23K20dFI95/295/22
1942a2c/d4K20b/dRubbly destruction material N of W71395/39
1944b2c4K20b/dRm 15, destr. debris on floor below 194295/42
1945a2c4K20a/bRm 11, destr. debris on floor below 194295/40

Level IIe

Because the IIc walls and destruction debris were left standing up to 1m high when the IId building was erected above them, but much less debris was deposited in the western courtyard, and no doubt elsewhere, the IId Stele Building stood higher than the neighbouring ground surface. Hence it was itself almost completely cleared away by subsequent building works, some later levels did survive to its west and south. Thus in the western courtyard (I19) another courtyard floor was laid above the destruction debris lying on the IId floor; while to the south in J/K18 the IId walls and floors were cut away and replaced, at more or less the same level as the IId floors further north, by an open space with benches round it and a hearth, which all has to be assigned to a later phase, called like the third courtyard floor surface IIe. The indications in both areas are that in the IIe phase that the existing use and architecture of the area was respected, and we therefore tend to see it as a continuation of the IIa-d occupation sequence.

The only other area where significant IIe occupation was found is in K20. Here there was, as already mentioned, a destruction level attributable to the end of IId, and above this there was another occupation surface immediately below the Byzantine remains of Level I. We initially assumed this was an earlier Byzantine phase, but more careful examination of the pottery revealed no later material, and some distinctive Early Iron Age pottery including some large fragments of individual vessels which had been lying on the occupation surface. There was little in the way of associated architecture, but a pit, in use at the same time as the floor itself, contained many large pieces of at least 7 massive storage jars or pithoi. All this suggests that the area was in use for domestic purposes, much as it had been in IIc, and the need for large amounts of food storage is also reminiscent of the IIc building, and tends to support the supposition that this IIe phase was a continuation of IIa-d.

UnitPur.Lev. TypeSq. DescriptionSample
2640a2d/e4H18bPacking and burned debris E of W78196/46, 47
2818a2d/e4I19c/dW ctyd., destr. debris below 281496/42
2823a2d/e4I19cW ctyd., packing below 282096/51
4344a2e5K18aE side, burnt occ. deposit, below 434397/24
4363a2e5K18aN half, burnt occ. surface/ deposit below 436197/32

Middle and Later Iron Age: Levels IIf-h

Like IIe, evidence for the later Iron Age occupation of the NW corner of the tepe is confined to the space surrounding the Stele Building itself. To the east, between the massive Byzantine foundations and the IId destruction phase, there was an accumulation of fill without architecture, punctuated by many pits, including some which were exceptionally large and deep. This area must have been an open space after Phase IIe. In K/J18 to the south of the Stele Building there does appear to have been a building; in its later phases this was represented by a room with mud-brick walls and a number of mud-brick benches constructed along the inside of the walls, in some cases retaining the emplacements for and occasionally the bases of large pottery storage jars. To the west and north this room was delimited by stone wall foundations still standing up to 0.45m high, although built very sloppily with loose and ill-fitting stones (perhaps because within a foundation trench). We assume that these two walls are the eastern end of a building of which the majority was in H18 to the west; we did not excavate the NE quadrant of this square (H18b) and in the NW quadrant this level was virtually all removed by Byzantine developers.

As in earlier phases, I19 was largely open space in IIf, but in the SE corner of the square, built up against the stones of the NW room wall just mentioned, was a narrow kiln (FI 96/15). Its lower part which was sunk below ground was lined with medium sized stones, and what remained of the superstructure seems to have been clay lined, with traces of internal arches. Five metres away to the west, also bonded into a stone wall foundation but running SW into H18a, was a second similar kiln (FI 96/14). Here virtually nothing of the superstructure survived, but the fuel chamber was packed with loose ash, and a very high concentration of potsherds. These proved to come from a homogeneous group of painted and unpainted jars, plates and other forms, which belong to the wares known from Cyprus as White Painted IV and Plain White IV, and generally dated to the end of the 8th and beginning of the 7th century. This is not the place to consider whether these pieces were locally produced or imported, but they undoubtedly attest to relations with the Mediterranean at this date, and provide a useful fixed chronological point in our sequence.

These two kilns, and the associated walls, represent the end of our Phase IIf, and the last recognizable stratum before the Byzantine in the NW corner of the tepe. There are however a considerable number of pits in the I19 area, some probably storage pits, which have been assigned to Phases IIg and IIh. On the eastern side of the Stele Building, there does not seem to be any architecture after the IId destruction, and among the pits in this area is one, in K19d, which was cut down to the level of the floor of the IIc phase in Room 20 through all the destruction debris in Hellenistic times. This included among the fill an unmistakable piece of Megarian bowl and part of the bronze facing from a Greek hoplite shield with guilloche decoration.

UnitPur.Lev. TypeSq. DescriptionSample
4001c2b-e7I20dP94/1 S part (//1403), cut by P94/595/15
4003c2b-e7I19b/I20dP95/12 fill95/23
2859c2e/f7I19aP97/12 
1799a2e/f4I19c/dBurnt occ. debris below 179096/29
2391a2f14J18aS half, packing below 239097/22
2629b2f7H18bP96/78, below 262896/33
2636a2f7H18b/dP96/86 lower fill, below 263196/48
2370b2f4J18aPacking under mud brick features97/13
2802b2f4I19dFI96/15, light ashy fill below 177796/34
2817a2f6I19cFI96/14, ashy fill at base 96/44
1708b2f-h7I19bP95/1 fill, N part95/06
1781b2g7I19dP96/4896/23
1764a2h6I19dLower fill of FI96/5 
1743c1/2h6I19dFI96/1 ashy contents96/01

Other Soundings

In the first three seasons there was some excavation on other parts of the site. In the first season three 5x5m soundings were placed diagonally across the south half of the mound, in K14, N12 and Q10. N12 and Q10 both exposed Byzantine architecture overlying Hellenistic levels, and were taken no further down. K14 was subsequently incorporated into a second operation, a 50x5m trench bisecting the mound from W to E, about 10m S of the church, as it transpired. The upper levels in this trench were uniformly Byzantine overlying Hellenistic, but although adequately constructed stone wall or wall foundations were present from each period, there were no well-laid floor surfaces or plastered wall-faces, and the rooms were filled with loose and largely barren earth which did not invite any expansion of this operation. One sounding was sunk in K14 through Iron Age to the top of Bronze Age deposits, a floor level at 97m, but this was also unrewarding for stratigraphic purposes, largely because of two deep pits one above the other..

UnitPur.Lev. TypeSq. DescriptionSample
3037b29K14aS side, fill below 3027 
3078a27K14aP96/33, cut by P96/796/14
3081b27K14aP96/38 sealed be P96/3496/15
3083b27K14aP96/4596/24
3087a25K14aAshy occ. deposits in SW corner below 308596/26
3093b25K14aOcc. debris in cut below 309296/32
3096a27K14aP96/98 into 3803, below 309396/38
3067c2?7K14aP96/7 cut by P96/696/07
3359a1/25J14bOcc. debris in N below 335495/33
3031a17K14bP95/22, E half fill95/26
3047a17K14aP95/22 lower fill95/46

In I14 at the W end of the trench, more or less on the western edge of the mound, the results were of more interest. Here a sketchy Byzantine building level overlay the western part of a gigantic ditch running roughly north-south. 4.50m wide and 3m in depth. It was filled with a tumble of large stones interleaved with ashy layers rich in animal bone. The evidence of a handful of sherds from the base of the ditch suggests, though it does not prove, that it belonged to a period before the Hellenistic wares found elsewhere on the mound and after the latest Iron Age material we have recognized, perhaps the around the 6th century BC. The purpose of the ditch is obscure: it looks as though it continued to the north and south of the 5m stretch we excavated, and it is big enough and steep enough to have functioned as a defensive feature; but the value of a ditch running parallel to, and about 7m back from the steep western edge of the mound is far from obvious.

The ditch cut through two well defined architectural levels. In the upper level there was domestic architecture which had suffered destruction by fire and was correspondingly well preserved, with mud-brick walls on stone foundations. An open courtyard was flanked on the west by two rooms, that on the NW housing an oven. Below this in Level III, under a thick layer of packing, were solidly constructed stone wall foundations; parts of three rooms and a probable courtyard (Rm 94) were excavated, perhaps belonging to different buildings. The eastern room (Rm 91) was mostly removed by the cutting of the large ditch in I14b, but in its NW corner two large storage jars set into the floor survived in situ. The other rooms had clean surfaces and few distinguishing features.

Levels II and III in I14 probably belong within the time-span covered by Levels II and III at the NW corner and are therefore broadly Iron Age and Bronze Age respectively, but in neither case is there a sequence of phases so that they cannot coincide fully with the NW Levels, and it remains uncertain exactly where each belongs within the longer sequence.

UnitPur.Lev. TypeSq. DescriptionSample
3403b14I14a/bWall collapse, below 3401-295/02
3407a1/2k5I14aAshy areas below 340695/07
3409b1/2k5I14aAshy area below 340795/09
3410a16I14aFI95/495/18
3412a17I14aP95/26 fill, below 340695/12
3419a15I14bBurnt occ. layer below 340195/17
3432a25I14a/bOcc. surface W of ditch (S half) (//3466) 
3446b27I14aP95/78 fill, below 344595/41
3455b27I14aP95/81 fill, below 344895/47
3458a24I14a/bDestr. debris below 345795/50
3459a25I14a/bBurnt occ. surface below 345895/49
3469a27I14bP96/9, below 346696/05
3474b25I14/bN half, burnt deposit below 3472 (//3458-9)96/09
3478b27I14aP96/29, below 347596/27
3482a26I14aFI96/6, internal ashy fill96/12
3491b27I14a/bP96/58, below 347596/28
3495b2/37I14aP96/65, below 348996/25
3708a35I14Rm 91, occ. surface below 370296/40, 41

Soundings on the E Side of the Mound

In the first two seasons (1994-5) some excavation was carried out on the north-east side of the mound, principally in squares Q19 and R18. The main discovery here was a thick stone defensive wall running along the crest of the eastern slope; this must have belonged to the 1st millennium B.C., but its more precise date is impossible to establish. In Q19 within the line of the wall, and partly sealed by it there were mud-brick walls and rooms of Iron Age date, cut in one place by a rubbish pit from which a good sample of Hellenistic sherds was recovered. Outside the fortification wall to its east, in R18, we excavated an open area characterized, like the courtyard area in the NW sector, by numerous intersecting pits, whose layers of white phytoliths clearly marked them as storage pits. These were sealed by the wall, and belong to the Early and Middle Iron Age. We did not penetrate into layers contemporary with Level III in the NW sector, although the surface sherd material in this area makes it clear that this period must be present lower down. This operation, like K14, was unsuccessful in that the frequency of pits and lack of well-defined architecture prevented us from demarcating a clear stratigraphic sequence, but it is of interest in demonstrating that at least the Early Iron and Bronze Age occupation had extended eastwards beyond the upper edge of the modern site.

UnitPur.Lev. TypeSq. DescriptionSample
2402a27R18aP95/5 fill95/16
2463a25R18cOccupation deposit below 246295/48

The Byzantine Period

Placed transversely across the tepe, a little way to the NW of its centre are the foundations of a substantial Byzantine church. The original structure measured some 27x 16.5m, with stone foundations sunk to a depth of at least 1.30m, lime mortared and capped by a course of well-shaped flat slabs extending the full 0.90m width of the foundation. The original building had an apse at the eastern end, and a central nave separated from the N and S aisles by columns resting on the stylobates. Subsequently (and the date of these changes, as well as the first building, remains uncertain), the apse was taken down and rebuilt and the main structure replaced by a much smaller rectangular church in which four of the columns were reused. In a final phase, the apse itself was blocked off by a straight eastern wall, and probably destroyed.

Fragments of carved decorative capitals were found during the clearance of the surface soil from above these foundations, and in a few places small pieces of glass mosaic tesserae, and there is no doubt that the original church will have been a handsome building, dominating the south end of the Mut basin. There are no textual references which might give a clue to the site's identity or name during these centuries, in which fine Byzantine ecclesiastical buildings were also erected at Alahan and at Ala Kilise, just above Zeyne and some 10km from Kilise Tepe. Also in the church area of the mound there were an unusually high number of drain and tile fragments lying on the surface.

Elsewhere on the summit of the tepe the highest strata were invariably also Byzantine. 5m soundings in N, Q and K encountered walls of this date, usually above a Hellenistic stratum, but not enough of their plans was recovered to allow us to reconstruct building plans. At the NW end of the tepe there was somewhat wider exposure. At the NW angle of the mound a well-constructed stone wall was found to run exactly N-S, an orientation which suggests that it belongs to the same settlement design as the church itself, and was perhaps delimiting an enclosed area on its north side. Just to its east Byzantine wall-foundations lay just below the surface in I19. They were very scrappy remnants, but they were set in lime mortar and may therefore have belonged to a substantial structure now fallen away from the top of the slope. East of this again was an open space measuring more than 10m in each direction, filled with thick deposits of ash alternating with clayey layers, to a depth of about half a metre. The layers were remarkably horizontal, giving no indication of any one direction from which the debris might have been tipped, and it is therefore hard to say with which building they might have been associated. The E side of this ash-space was supplied by a very substantial structure, with stone foundations almost precisely 1m wide, and sunk into the underlying (principally Middle/Late Iron Age) deposits. The major outer walls form a gateway 6.25m wide; to the north and east of this are other less wide stone foundations, which were not sunk so deep but adopt the same alignment and are probably contemporary. Unfortunately the occupation surfaces which went with these walls, if they survived at all, fell within the 20cm of plough soil at the surface.

In I/J/K19/20, where there were no Byzantine walls, there were numerous pits of Byzantine date, recognizable by their carefully applied clay plaster lining and bell shape. Mostly these were dug from above the modern surface of the tepe, but in a few cases the pits were sealed by later Byzantine deposits, including the deep ash-lines in J19. There seems little doubt that these were designed as food storage pits, and one pit in J19 still retained carbonized emmer.

UnitPur.Lev. TypeSq. DescriptionSample
1602b15J19b/dStriated ash layers below 1601 (//2306)95/03, 08
1604a17J19bP95/14 fill95/05
1609b17J19b/dP95/15 upper fill95/14
1610b17J19b/dP95/15 lower fill95/19
1626c17J19dP95/33 fill95/20, 21
1644b15J19cAshy striation below 164596/04
1653b17J19aP96/1196/06
1656b1?7J19bP96/1996/36, 43
2607a16H18dPatch of ashy debris below 2606 

In S18a, a Byzantine building had been terraced back into the eastern flank of the mound; apart from a single large pottery storage jar in the corner of one room, there was nothing to indicate the function of the building.

Bibliography

Baker, H. D. et al. 1995, 'Kilise Tepe 1994', Anatolian Studies, vol. 45, pp. 139-191.

Postgate, J. N. 1996a, 'Kilise Tepe 1994: A summary of the principal results', XVII Kazi Sonuçlari Toplantisi I (Ankara), pp. 419-431.

Postgate, J. N. 1996b, 'Kilise Tepe 1995: A summary of the principal results', XVIII Kazi Sonuçlari Toplantisi I (Ankara), pp. 441-456.

Postgate, J. N. 1998a, 'Between the plateau and the sea: Kilise Tepe 1994-97', in Ancient Anatolia: 50 years' work by the British Insitute of Archaeology at Ankara, ed. R. J. Matthews (London, BIAA), pp. 127-141.

Postgate, J. N. 1998b, 'Kilise Tepe 1996: A summary of the principal results', XIX Kazi Sonuçlari Toplantisi I (Ankara), pp. 209-226.






Context information for samples taken at KT 1994-1997

ctyd. = courtyard     destr. = destroyed     occ. = occupation

Purity : a   believed pure
bslight, or slight risk of, contamination
cdefinitely and significantly contaminated
dno stratigraphic effort made

Class (see Preface, Appendix I, for more detail):
1   Natural deposits
2   Tip, midden, rubbish, dump
3   Structure
4   Constructional materials
5   Occupation sequence
6   In situ deposits
7   Pit fills
8   Miscellaneous
9   Mixed

1. Units from Squares in the NW of Kilise Tepe

UnitPur.Lev. TypeSq. DescriptionSample
1258a4a4H20dRm 43, destr debris below 125395/01, 04, 10, 13
1288a5e7H20dP97/7, below 1289 
1325b2c5J20dRm 1. Burnt occ. floor below 132495/11;
1348b2c4J20dRms 2-3, destr. debris of W122, below 134695/29
1353a2c4J20dRms 2-3, destr. debris below 135095/43
1356a2c3J20d/J19bRm 1. Bench along E side95/45;
1370c2a-b4J20aPacking of cut round NW corner of building95/37
1372c3e4J20aE ctyd. Burnt occ. deposits (//1388) below 137195/44
1376b3d4J20aE ctyd. Destr. debris below 137295/53
1378b3e6J20aFI 95/9 fill95/52
1381b3e4J20aBurnt occ. deposit round FI95/995/51
1388a3e5J20cE ctyd. Burnt occ. deposits below 138797/17
1540a2d4K19bOcc. layer? E of W770, below 153596/37
1560a2c4K19aRm 3. Destr. debris in SE corner (//1554)96/60
1561b2c5K19bRm 4. Occ. surface below 1555 (//1995) 
1566a2c4K19a/bRm 4. Destr. debris, below 1551/297/02
1602b15J19b/dStriated ash layers below 1601 (//2306)95/03, 08
1604a17J19bP95/14 fill95/05
1609b17J19b/dP95/15 upper fill95/14
1610b17J19b/dP95/15 lower fill95/19
1626c17J19dP95/33 fill95/20, 21
1644b15J19cAshy striation below 164596/04
1653b17J19aP96/1196/06
1656b1?7J19bP96/1996/36, 43
1686a2c4J19bRm 3. Destr. debris below 164296/45
1688a2c5J19bRm 3. Burnt occ. deposit below 168696/56
1690b2c/d7J19b/dP96/102, below W71896/50
1696a2b7J19bRm 3 FI 96/18 central hearth96/54
1708b2f-h7I19bP95/1 fill, N part95/06
1743c1/2h6I19dFI96/1 ashy contents96/01
1764a2h6I19dLower fill of FI96/5 
1781b2g7I19dP96/4896/23
1799a2e/f4I19c/dBurnt occ. debris below 179096/29
1810a5e/f4H20cE strip, packing below 180795/27
1812b5f6H20cRm 55, fill below 181095/31
1822a5e7H20cP95/65 fill95/38
1836a5g4H20cRm 62, fill below occ. surface below 1831, 183596/10
1838a5g/h4H20cPacking below Phase G floors96/16
1852a5i4H20cE half, packing below 185196/30
1853a5i4H20cE half, packing assoc. with W792, below 185296/39
1855a5i7H20cP96/9696/49
1858a5j4H20cRm 83, destr. debris96/53
1871a5j4H20cRm 83, destr.debris97/01
1909a1/23K20dFI95/295/22
1931a2c5K20a/cRm 15, ashy surface (//1142, 1948)95/30
1935b2b5K20a/cRm 15 occ. levels below 1931 
1938b2e/d4K20b/dRubbly packing sealing W713, below 192495/35
1942a2c/d4K20b/dRubbly destruction material N of W71395/39
1944b2c4K20b/dRm 15, destr. debris on floor below 194295/42
1945a2c4K20a/bRm 11, destr. debris on floor below 194295/40
1965a2b5K20aRm 15, occ. surface below 195896/13
1971av2b5K20a/cRm 15, occ. surface below 196596/19
1975a2b5K20a/cRm 15, occ. surface below 197296/22
1988b2d4K20cRm 2, occ. surface, below 198096/31
1994b2d7K20dRm 4, P96/121 upper fill96/55
2370b2f4J18aPacking under mud brick features97/13
2391a2f14J18aS half, packing below 239097/22
2607a16H18dPatch of ashy debris below 2606 
2629b2f7H18bP96/78, below 262896/33
2636a2f7H18b/dP96/86 lower fill, below 263196/48
2640a2d/e4H18bPacking and burned debris E of W78196/46, 47
2802b2f4I19dFI96/15, light ashy fill below 177796/34
2817a2f6I19cFI96/14, ashy fill at base 96/44
2818a2d/e4I19c/dW ctyd., destr. debris below 281496/42
2823a2d/e4I19cW ctyd., packing below 282096/51
2825a2d5I19c/dW ctyd., ashy occ. deposit, below 282396/52
2848a2c7I19dW. courtyard: P97/6 lower fill below 285897/03
2859c2e/f7I19aP97/12 
2871a2b4I19bW ctyd., packing below 286797/06
2883a2b7I19bP97/41, below 287397/08
2884a2a/b5I19bOcc. surface W of W823, below 288097/09
3927a2c4J19cRms 7 and 8, destr. debris97/04, 07
3929a2c4J19dRm 7, burnt occ. deposit97/05
3943a2c/d8J19dRm 10, burnt surfaces below 391797/14
3963b2b8J19dRm 7. Sounding below IIc floor97/25
3965a2b5J19dRm 7, sounding below 396397/28
3967c2c/d7J19dP97/73, below 3957 
4001c2b-e7I20dP94/1 S part (//1403), cut by P94/595/15
4003c2b-e7I19b/I20dP95/12 fill95/23
4004b2b7I20dP94/5 cut by P95/28 and P95/34 (//1421)95/24
4008c2b7I20dP95/34 fill95/25
4017b3e4I20b/dE ctyd., packing below 4016 (//1426)95/34
4018b3e5I20b/dE ctyd., occ. surface below 401795/36
4205a3d4H19bPacking assoc. with W737, W73895/32
4209a3c/d5H19bRm 32, occ. surface below 420595/28
4231a3c4H19bPacking below 423096/03
4243a3a4H19aPacking below W703 and W74496/17
4246a3b5H19bPacking E of W74596/11
4249b4b6H19a/bBurnt debris on IVb surface below 424896/18, 20
4251a4b4H19a/bBurnt occ. deposits on IVb surface below 424996/21
4261a4a/b4H19a/bRm 42, packing below 4255 (//4257)96/35
4273a4a4H19a/bRm 42 SE, collapsed masonry etc (//4263-4)96/58, 59
4278a4a6H19a/bRm 42, burnt occ. surface96/57
4344a2e5K18aE side, burnt occ. deposit, below 434397/24
4363a2e5K18aN half, burnt occ. surface/ deposit below 436197/32
4510a2c4K19cRm 5, destr. debris, below 450797/11, 21
4511a2c6K19cRm 6, carbonized matter in occ. deposit97/20
4521a2b/c5K19aRm 4, floor round storage jar in SW corner97/18
4534d2b/c9K19aRm 4 sounding below IIc floor in SW corner97/15, 23
4535d2b/c9K19aRm 3 sounding below IIc floor in NE corner97/23
4540d2b9K19aRm 2, sounding below IIc floor97/26
4551d2b9K19aRm 3, additional sounding to SW of 453597/30
5310a5f45H20dRm 52, fill below 530797/29
5321a5e7H20dP97/5497/12
5328a5f48H20dRm 53, filling of cut for W23897/16
5335a5f45H20dRm 51, fill below 533197/34
5412a5j4H20cW half, destr. debris N of W225 (//5417)97/19
5429a5k4H20cW half, construction surface below 5428 (//1891)97/33
5501b2a7I19bP97/69, fill of NW part97/10
5519a3e4I19b/aRm 30 packing below 550997/27
5526a?2-35I19aOcc. deposit round P97/70 (//5504)97/31

2. Units from Squares not in the NW Sector

UnitPur.Lev. TypeSq. DescriptionSample
2402a27R18aP95/5 fill95/16
2463a25R18cOccupation deposit below 246295/48
3031a17K14bP95/22, E half fill95/26
3037b29K14aS side, fill below 3027 
3047a17K14aP95/22 lower fill95/46
3067c2?7K14aP96/7 cut by P96/696/07
3078a27K14aP96/33, cut by P96/796/14
3081b27K14aP96/38 sealed be P96/3496/15
3083b27K14aP96/4596/24
3087a25K14aAshy occ. deposits in SW corner below 308596/26
3093b25K14aOcc. debris in cut below 309296/32
3096a27K14aP96/98 into 3803, below 309396/38
3359a1/25J14bOcc. debris in N below 335495/33
3403b14I14a/bWall collapse, below 3401-295/02
3407a1/2k5I14aAshy areas below 340695/07
3409b1/2k5I14aAshy area below 340795/09
3410a16I14aFI95/495/18
3412a17I14aP95/26 fill, below 340695/12
3419a15I14bBurnt occ. layer below 340195/17
3432a25I14a/bOcc. surface W of ditch (S half) (//3466) 
3446b27I14aP95/78 fill, below 344595/41
3455b27I14aP95/81 fill, below 344895/47
3458a24I14a/bDestr. debris below 345795/50
3459a25I14a/bBurnt occ. surface below 345895/49
3469a27I14bP96/9, below 346696/05
3474b25I14/bN half, burnt deposit below 3472 (//3458-9)96/09
3478b27I14aP96/29, below 347596/27
3482a26I14aFI96/6, internal ashy fill96/12
3491b27I14a/bP96/58, below 347596/28
3495b2/37I14aP96/65, below 348996/25
3708a35I14Rm 91, occ. surface below 370296/40, 41