Excavations by K. M. Kenyon in Jerusalem 1961-1967: Volume V Discoveries in Hellenistic to Ottoman Jerusalem Centenary volume: Kathleen M. Kenyon 1906-1978

By K. Prag

The fifth volume in the series of final reports on the work of the Joint Expedition to Jerusalem in the 1960s describes the discoveries made in six sites in the ancient city and places them in the archaeological and historical context of Jerusalem and the surrounding lands. Among the most debated issues are the extent of the occupation of the city during the Iron Age, the location of the southern defence line in Herodian and Roman times, and the date of the destruction of an Umayyad palatial structure. There is fresh information on the civic amenities of the southern half of the Byzantine city, and on the structure of the Ottoman city defences built under Sulaiman the Magnificent in the sixteenth century. Fine glazed pottery, both locally made and imported, and the wide range of materials reaching Jerusalem through trade and pilgrimage, reflect elite patronage and the high status of the holy city under Islamic rule.

• Language: English
• Publisher: Oxbow Books
• Release date: Aug 6, 2008
• ISBN: 9781782975144

Book preview

1. Introduction

Sites and choices

During the first (1961) season the excavations in Jerusalem already embraced seven sites, in locations ranging from the centre of the Old City to the extreme south of the Roman and Byzantine city. In 1962, when the Royal Ontario Museum joined the partnership, the number of areas increased, and in all 21 areas were listed alphabetically. Additional investigations outside the scope of the main sites were also undertaken during the project, such as the work on the late third millennium cemetery on the eastern slopes of the Mt of Olives and on the Byzantine tomb in the grounds of the American Colony Hotel (Prag 1995; Marshall and Prag 1995). Large area excavation was only undertaken in two places, Area A and Area L, though Site M and the combined Sites J and S were quite extensive, but not connected. Only in Site L was a more comprehensive architectural record obtained. Some areas were very small. In many ways the project was consistent with the Kenyon method at Jericho, where the complete multi-phase stratigraphic record of a multi-period site had priority over the broad area exposure of any single period. Maps locating the sites were provided in all the Kenyon publications, and on a larger scale by Tushingham (1985); part of the latter was re-published by kind permission in Eshel and Prag (1995). All these maps illustrate the locations in relation to the contours. For this volume, another of the maps in the archive has been redrawn, with the sites in relation to the early-twentieth century roads, paths, houses, boundaries and some of the major known ancient sites (Fig. 1). This map locates all the sites excavated in the vicinity of the south-east ridge, and reports on some of these sites are published in this volume.

Kenyon’s selection of sites for excavation was part problem-oriented, part pragmatic. For the latter, as Jerusalem is, and was, a densely inhabited city, the selection of sites was governed by what land was available to the Joint Expedition. It was not just a matter of the presence of roads and housing which limited access, but the distribution of sacred sites, cemeteries and valuable garden land in and around the modern city. There were also negotiations with owners to be surmounted, and costs to be considered. Large spaces without buildings were scarce, and inevitably areas such as the Armenian Garden (Site L), the area south of the Haram (Sites G, J and S) and the open space in the Muristan (Site C) drew Kenyon’s attention.

Concerning Kenyon’s choice of the sites described in this volume, Site V on the south-east ridge was one of the few unencumbered areas available for excavation, and it was hoped to find remains going back to the time of the original settlement (Kenyon 1966, 88). Site E was a rescue operation initiated during domestic construction work in 1961 (Kenyon 1962, 85–86; 1974, 48, 195, 201, 271). Here, as happened on other occasions, the builder/landowner reported the finding of archaeological features to Kenyon, and then willingly permitted access. Site E provided an opportunity to investigate the date of the earliest occupation on the western hill at a location on its eastern slope which was not far from the Old City. The purpose of the excavation at Site D was to investigate the earliest date of the occupation (and thus probable enclosure) of the east side of Mt Zion and of the central Tyropoeon Valley (Kenyon 1962, 85). The landowner, the Assumptionist Order, was very supportive of the project, which added to the long history of archaeological investigation on the property. Kenyon (1962, 84) described the rationale of working at Site B as linked to the question of the wall joining the southern ends of the eastern and western ridges. This investigation obviously also related to the question of the date of the first occupation of the western hill.

As to Sites G and J, the opportunity provided by an open area, with the potential to explore the possible Solomonic extension of the town northwards from the south-east ridge, had attracted Kenyon to the area south of the Haram as early as 1960, but permission was difficult as it was endowed land in the care of the Awqaf. In 1961 the Department of Antiquities of Jordan conducted excavations as a preliminary to the building of a school on the western part of the area. In 1962 permission for the work on Sites G and J was obtained by the Joint Expedition. However, there were problems because of the proximity of the excavations to the Haram which resulted in the closure of the excavations in 1963. Permission for work in Site S was obtained in 1965, with the aim of dating the foundation of the city wall at this point. Kenyon attributed the base of the city wall to Aelia Capitolina in the second century AD (Kenyon 1967a, 70; 1968, 98; 1974, 53, 223), which, she proposed, took advantage of a rock scarp which had been left by the quarry workers who respected the (hypothetical) line of the much older Solomonic city wall (cf. E. Mazar 1994, 65; B. Mazar in Mazar and Ben-Dov 1971, 27, fig. 15.1). The aim, to investigate the date of the city wall at this point, was combined with the opportunity for further investigation of the ancient city.

e9781782975144_i0002.jpg

Figure 1. Jerusalem: general site plan for the area south-east of the Old City. Based on revised editions of the Survey of Palestine 1937, originally drawn at 1:1250. Drawn by Mark Roughley.

The common research themes linking most of these areas were the date of the enclosure/occupation of the eastern slopes of the western hill and the Tyropoeon Valley; and the investigation of the ancient cities of the second and first millennia BC, and their defences. Inevitably not all sites produced evidence exclusively linked to these themes, and some produced no relevant evidence at all.

The order in which the site processing has been done for this volume was mainly a matter of choice by the writer. The unpublished sites could, to some extent, be grouped. From the map (Fig. 1) it can be seen that Site V is rather isolated, and in fact produced no information on the pre-Roman period. Sites B, D and E lie on the eastern slopes of the western hill and all were linked to the investigation of the date of the occupation of this area. It also seemed sensible to begin work on what appeared to be relatively uncomplicated, smaller sites. Sites G, J and S are also a spatially inter-related group, but the decision to process these sites was governed by the logistics growing out of the initial problems of locating the archive and materials from Sites G and J, not all of which were in Manchester. Once the task of locating the material was completed, and as a result considerable inroads had already been made on the recording and analysis, it made sense to continue and complete the work on these areas.

The order in which the sites are described in this volume is partly geographical, but mainly chronological in terms of the major results from each site. The progression is as near chronological, from Roman to Ottoman, as the varied nature of the material permits.

As to the remaining unpublished sites, Sites S.II, R and M in the Ophel area, and Sites N, K, W, O, F and X in the lower east side of the Tyropoeon Valley, many are linked by a common theme, the investigation of the enclosure of the city on the south-east ridge. These will be reported in Volume VI, with the remaining specialist reports, including the report on the glass currently being studied by Dr M. O’Hea. Site S.II was omitted from this volume, as its stratigraphic sequence and results were for the most part separate from those of Sites S.I, III–VI.

Terminology

Franken and Steiner (1990, 5–6) have already provided descriptions of the Kenyon terminology. Each of the sites was named alphabetically, apparently in the order each site was opened. A context designation in Site S (e.g. S.300.5) begins with an upper case S, followed by Arabic numerals indicating the ‘page number’ and the ‘level number’ (the context within each designated area). Many excavation reports now re-order the terminology to provide a consistent list of contexts from early to late, but the huge quantity of material from Jerusalem, and its complexity, precludes this undertaking. The records of Sites A and L each contained around 5000 contexts. Rather than introduce another layer of terminology, the original context names have been retained throughout the reports.

In each site, there may be more than one trench; the trenches were identified with Roman numerals (e.g. in Site S there were six trenches, S.I–VI). The trench number is not normally included in the context number, as the ‘page number’ is sufficient for identification within each site. ‘Page number’ refers to an allocated number in the site notebook, which defines a designated area within the trench (not the running page numbers of the site book itself); the ‘page number’ is purely an identity number and has little or no stratigraphic value, it is not a ‘locus’ number, as it may embrace several different stratigraphic phases, but should provide guidance to which area was dug within the trench. The ‘page number’ is not repeated within a site, regardless of how many trenches were excavated at that site, but the numbering begins with a ‘Page 1’ in each site. The number of trenches and contexts in Site A was so great that in the day-to-day running of Site A over seven seasons, in the pre-computer 1960s and with numerous site supervisors and site books, the allocation of ‘page numbers’ very occasionally got out of hand and ‘page numbers’ were doubled, so a new system of AA numbers was introduced in which the page numbering could again begin from Page 1. This was not a problem on other sites where batches of page numbers were allocated to each trench (e.g. pages 300–399 to Site S.IV). The ‘level number’ or context also does not necessarily represent the stratigraphic order, rather the order of excavation or recording; if ‘level/context 9’ underlies ‘level 8’ either the notes in the site book or the section drawing should show this; but ‘level 9’ may well lie in another part of the trench and be stratigraphically later than ‘level 8’. The full context mark on a sherd or object would normally be e.g. Jmp.S.300.5, for Jerusalem Site S, page 300, context 5; and the ‘page number’ (in this case S.300) will indicate this came from Site S.IV. This system (also in use at Jericho) was designed to be brief and efficient for identification and for marking finds, and flexible when the stratigraphy was subjected to later phasing analysis. To allow for occasional lapses in the correct copying of numbers, the site supervisor was also required to give a brief description of the context (e.g. ashy or rubble) on both the manilla tags (one inside and one outside the packaging) attached to all finds, a description which could be correlated with the description in the site notebook if necessary.

Registered objects have a registration number between 1 and 7828 enclosed in a box which was usually marked in ink on the object or on an attached card tag, or on the packaging, along with the context number; coins were registered separately as c1–c3793 (c = coin); a separate register was made of the items found in the Iron Age caves in Site A in 1967, and these items were registered C1–C1307 (C = Cave). Fortunately there was no overlap between the items in the two last systems.

Methodology

For the Kenyon sites, the archive contains the original pencil-drawn main sections on millimetre paper at a scale of 1:25 (mostly drawn by Kenyon in collaboration with the site supervisor at the end of the season) and the original plans and final inked versions made by the surveyors. The site books contain the supervisors’ records augmented by sketch sections and plans, usually measured and to a given scale. All these were worked through, with a context data base, site matrix and phasing developed for each site. The site matrices are usually too large and unnecessarily complicated for publication, but the contexts within each phase are listed in the reports. The plans and sections were redrawn in inked versions without correction for publication, some measured sketches in the site books were also re-drawn, and some new additional phase plans were drawn. The lettering was incorporated on the digitised version, with any final corrections. A draft history of the site was composed on the stratigraphic and architectural evidence. The pottery (still bagged by context, or retrieved from the Kenyon ceramic typologies) was then recorded by phase, the date of the ceramics of each phase separately assessed, and then with the registered finds, correlated with the stratigraphic record and any final adjustments made. A limited selection of the pottery was drawn, and most of the already drawn registered pottery re-integrated in the final pottery figures. Text reference is made to the presence of well-known and well-illustrated types, rather than publishing an extensive redundant corpus of ceramics. As all the registered finds had been drawn by a series of expedition draughtsmen in the 1960s, the style varies accordingly, but the original system of pottery drawing, with the section on the left, is retained. The report was then revised, and finally set in the context of other finds and records.

The fine quality of the plans and sections produced by the expedition is an aspect of Kenyon’s work which has been insufficiently emphasized, and although redrawn, the plans and sections in this volume do follow the originals unchanged. Where the phasing and interpretation of the stratigraphy varies from the section drawing, the variation is based on the evidence from the site supervisors’ notes and/or the ceramics and other finds from relevant contexts. Kenyon undoubtedly would have added soil definitions to the sections in the final report, as in the Jericho publication, but for reasons of clarity, this has not been done.

Layout

The illustrations have been integrated in the text for each chapter (excepting the colour illustrations which are set at the end of the volume). Every effort has been made to cross reference from the text to the illustrations. Working from the figures back to the text, the figure captions indicate the phasing within each site so that locating references and descriptions in the text from the figures should not be difficult.

Ceramics

The Kenyon ceramic typology occupies many boxes, covering as it does the occupation of Jerusalem from the late fourth millennium BC to the twentieth century AD; and is organized by shape and to a lesser extent by ware or period. It has been described by Franken (Franken and Steiner 1990, 61–66), along with the reasons for rejecting its use. Franken broke the typology down into chronological sections, and published a little Bronze Age material, but focussed particularly on the great bulk of material from Site A in order to produce an Iron Age II typology based on fabric and manufacture. The current volume is concerned principally with first century AD, Byzantine, Umayyad/Abbasid and Ottoman ceramics. As a decision was taken not to publish a large redundant corpus of pottery shapes, reference has been made to examples published elsewhere, in particular from the work of Clamer (1997), Magness (1993), Tushingham (1985) and Wightman (1989). The figures therefore do not reflect the complete typology of any period and should be used in conjunction with the discussion in the text. Ceramic typologies by period/site are retained in the archive.

2. Site V. The Byzantine re-structuring of the city

Introduction

Site V lies on the east side of the ridge on which the ancient city was located, c. 200 m. south of Site A and the Gihon Spring, and c. 150 m. north of the Birkat al-Hamra. It lies between the 674 and 675 m. contour levels (grid ref. 1724.1310) (Figs 1 and 2). Kenyon (1966, 88) excavated here as it was one of the few unencumbered areas available for excavation, and it was hoped to find remains going back to the time of the original settlement. Even in the course of one season she could conclude that nothing earlier than Roman and Byzantine quarries survived in the area, and there was no hope of finding any trace of early Jerusalem. Two small trenches (with a total area of 76 sq. m. and total volume of approximately 213 cu. m.) were excavated on a tree-planted terrace against the east face of the house of Hassan Khalil. This terrace is in the area above and to the west of the footpath which descends the length of the ridge to the Birkat al-Hamra. According to Kenyon’s brief description (1966, 88; 1974, 264) the results at Site V showed that intensive quarrying in the area dated to the second century AD. Ussishkin (1993, 342) also noted that the area was quarried in the late Roman period.

Site V is located in the area of the broad flat terraces on top of the ridge, where relatively few excavations have been conducted. Although the ground level drops rapidly to south and east, here it is supported by modern dry-stone terrace walls (Fig. 3). Some of the masonry in the terrace, directly below and to the right of the figure in the photograph, is clearly older. The context of the site is now better known than it was in the 1960s (for the relationship to the other excavations in the area, see Shiloh 1984, fig. 3). The site is located approximately 20 m. inside the line of the ‘inner’ Iron Age II city wall (Shiloh 1984, fig. 33). Shiloh’s re-investigation of the area outside this city wall revealed IA II housing and Hellenistic terracing (Shiloh 1984, fig. 9) which can now be seen to be within an ‘outer’ late IA II city wall (Reich and Shukron 2001, 79, 85). The boundary or terrace walls, dated by Shiloh to the Hellenistic period, are very similar in construction to the segments of older walls in the terrace in Figure 3 (cf. Shiloh 1984, pl. 11: 2, Strata 7–6, W152). The wall below the path could well have formed part of a series of such terrace walls, for the stratigraphy in the Site V trenches proves that the terracing was ancient.

The excavations by Weill in 1913–4 (Weill 1920; 1947; see also Kenyon 1974, pls 8, 9) in the area immediately to the east and south of Site V showed dramatic evidence for quarrying, but also for baths and cisterns of the Hellenistic and Herodian periods. Among the rock-cuttings were three unusual tunnel-like features, which Weill identified as the destroyed royal tombs of the House of David. Kenyon was of the opinion that the use of plaster in these ‘royal tombs’ indicated that they were cisterns, though of unusual shape; but the shape can be parallelled on a larger scale in the Hellenistic Struthion pools at the Antonia fortress. Ussishkin (1993, 298–300), following his study of the Iron Age tombs at Silwan, concluded that the quality of the stone carving in ‘Weill’s Royal Tombs’ made it unlikely that these were royal tombs. While it is not satisfactory to compare putative tenth and ninth century tombs with known tombs of the eighth and seventh centuries, the hypothesis that these cuttings were part of the royal cemetery in the tenth century BC is generally regarded as unlikely (e.g. Kenyon 1967b, 188; Avigad in Stern 1993, II. 710), but it has not been disproved.

Report on Site V.I

Trench V.I was opened in July 1965, and worked during that season only. The area of V.I was initially less than 12 × 4 metres (Fig. 4). The west end proved relatively shallow, but parts of the centre and east end of the trench were excavated to a much greater depth. The maximum depth reached was seven metres in the north-east corner of the trench without reaching bedrock, by which time the area excavated had shrunk to 3 × 2 m. and finally to 3 × 1.50 m.

Summary of six stratigraphic phases

Phase 6 (the earliest) had a rock-cut and plaster-lined cistern (which may be Herodian or earlier in date). The cistern was truncated during subsequent quarrying of the rock. Kenyon proposed that this quarry dated to the second century AD. Inside the cistern cut was a horizontal rubble and silt fill, attributed to Phase 5 (Byzantine). Phase 4 comprised later tip fills which sealed the cistern area, also fairly horizontal, with a rather poor fragment of a wall (Byzantine). Phase 3 contained the principal structural phase, with walls, pits and robber trenches (late Byzantine/early Islamic). Phase 2 contained two dubious wall fragments and pits (Early Islamic). Phase 1 consisted of modern garden fill.

Contexts, features and finds

Phase 6: no contexts

A cistern, of which only a tiny portion of the west edge was exposed near the centre of the trench (Fig. 5), was rock-cut and plaster-lined (Fig. 6). What appears to be an open, stone-built and plaster-lined channel runs towards it from the west. This channel was later used as the foundation for Walls A and C (Figs 7 and 8). There is no stratigraphic link between the features in the central area and the built, plastered channel/Wall A at the extreme west end of the trench, so it is not possible to define the relationship between the east end of the channel and Wall C. It could be suggested that the channel (at a higher level than the preserved fragment of cistern) belongs to Phase 6, and that the rock-cut steps,

e9781782975144_i0003.jpg

Figure 2. Detail from 1935 map, showing location of Site V in relation to modern features, including the path down to the Birkat al-Hamra, and Weill’s dump south of Site V. (Drawn by R.J.D and R.A.N).

Wall X and entrance pier (Fig. 9) also belong to Phase 6, when the channel ran through an arched opening in the partly plastered Wall X, which was partly cut from the bedrock and which was later blocked by the building of Wall C (Fig. 10).

The cistern appears to have gone out of use when cut by a quarry. If the plastered channel at the west end of the trench was contemporary with the cistern, then the deposits related to the channel may also belong to this phase (V.5.1–4, 9, 10), but they are here attributed to Phase 5.

e9781782975144_i0004.jpg

Figure 3. Site V: the figure stands on the path down to the Birkat al-Hamra with terrace walls behind and in front; Site V is on on the upper right on the terrace behind and above the figure, with the rock cuttings and dumps of Weill’s 1913–1914 excavations to the left; facing south. Neg. no. 65/213.

e9781782975144_i0005.jpg

Figure 4. Sites V.I and V.II: relation of the trenches and alignments of the walls in the main building phases in both areas.

e9781782975144_i0006.jpg

Figure 5. Site V.I: plan of the bedrock features and Walls A and X (22/9/1965).

There were no finds, as the phase is essentially defined by rock cuttings under later deposits.

Phase 5: contexts V.I: (5.1–4, 9,10); 7.13–38

Only a small area was excavated. The stratigraphically earliest contexts excavated at the east end of the trench, consisted of an almost level fill of silt, sand and clay deposits varied by earth and stones, but the sequence was not excavated to bedrock (Figs 11, 12). The excavator noted possible floor levels (surfaces? = V.7.20, 21) in this fill, and a level beneath containing corroded bronze lumps and charcoal fragments (V.7.25). The lowest levels excavated consisted of fine silt and broken plaster (V.7.30, 32 and 33).

Pottery: the sherds kept from 24 contexts are Byzantine, mainly of fifth to eighth century date. The forms are those defined by Magness (1993) and Hayes (1972). There are rouletted and flanged bowls including: a quantity of LRRW flanged and rouletted bowls (Late Roman C Form 3F, Hayes 1972, 336–338, fig. 69: 23); FBW Rouletted Bowl Forms 1 and 2A; FBW Bowls Form 2B (three examples in V.7.25, three sooty examples in V.7.27, cf. Fig. 146: 6; Magness 1993, 198 dates these from the mid-seventh century); Arched-Rim Basins Forms 1, 2, 2A and possibly a variant of Form 3 (Fig. 19: 1, perhaps cf. Wightman 1989, fig. 24: 10, seventh–eighth centuries?); Rilled-Rim Basins; Incurved-Rim Basin (Magness 1993, 210, eighth to tenth centuries); a storage jar with a long-fold on the outside of the neck (cf. Damascus Gate, Wightman 1989, pl. 49: 7, late Byzantine/Umayyad); FBW Juglet Form 2A, one with gouging (Fig. 19: 2; cf. Magness 1993, 240, Juglets Form 2A, mid-sixth to early eighth century AD); roughly made storage jars (cf. Site J, Fig. 150: 5); one fragment of pie-crust decoration; Cooking Pots Form 4A and an unusual type, perhaps close to Cooking Pots Form 4B (Fig. 19: 3); and two fragments of Magness Form 2 lamps. Although many of the types continue through to the eighth century and later, the deposits seem to belong to the sixth/early seventh century, and lack any of the early Islamic metallic or glazed wares. Very few earlier sherds were noted, and just one IA II ring-burnished body fragment. The limited quantity of material recovered from the contexts at the west end of the trench (V.5.2–4, 10) fits the same range.

e9781782975144_i0007.jpg

Figure 6. Site V.I: the scale lies in the truncated cistern cut in bedrock, with the plaster facing just visible to right of scale; later quarrying can be seen in the upper right; facing south. Neg. no. 65/201.

e9781782975144_i0008.jpg

Figure 7. Site V.I: plan of the upper features including Walls X, C and B (29/7/1965).

Coins: c3068, 3069, 3081, 3122, 3123, 3124, 3133, 3134, 3135, 3136, 3137, 3138. The latest coins possibly date to between the fourth and sixth centuries AD. Registered finds: 5685 Hellenistic jar stamp (Fig. 21: 1; Fig. 263); 5732 lid fragment (Fig. 19: 13; cf. Magness 1993, 248.3, Lids Form 2, sixth to mid-eighth century.); 5796 Byzantine marble bowl fragment (Fig. 21: 11, cf. reg. no. 5016, and cf. Tushingham 1985, fig. 74: 18); 5819 limestone fragment (Fig. 22: 3); 5951 iron nail; glass fragments: 6048, 6163, 6169, 6465.

Phase 4: contexts V.I: 2.28–29, 32, 37–41; 3.1–10; 4.2–5, 8; 7.3, 4, 7a, 12

Mainly tip and horizontal fill cut by later deposits. Wall F.

Wall F is only illustrated by a sketch in the site notebook. It appears to be older than Wall E and to be little more than a line of ten stones which curves slightly across the north-east corner of the trench, with one or two courses surviving. At most it appears to be a terrace wall (cf. Phase 4 in Site V.II). It may be indicated in Figure 12. Pottery: pottery from 20 contexts was kept; almost all the material is Byzantine, with a few residual fragments dating to the first–second century AD but nothing earlier. Amongst the residual material were two imbrex tiles with the stamp of the Tenth Legion, first to third centuries (Fig. 20: 1, 2). The same ceramic types as were noted in Phase 5 included FBW Bowls Form 1, Rouletted Bowls Form 1, Rilled-Rim Basins, Arched-Rim Basins Form 2A and Late Roman C Form 3; ribbed and rouletted sherds were well represented; there were no lamp fragments. There were however a number of fragments which are closer in both fabric and shape to types occurring in the earlier levels of Site J, cf. Figs 147: 4, 9, 16; 149: 13, 20. A small jar with string-cut base (reg. no. 5426) is very similar to those found in Site J (Fig. 149: 14–17). The latest material may date to the seventh/ eighth centuries.

e9781782975144_i0009.jpg

Figure 8. Site V.I: west end of trench, water channel/Wall A beneath scale; bedrock to left; facing south-west. Neg. no. 65/ 42.

Coins: c2724, 2768, 2802, 2877 all range from second century BC to the first century AD. Coin c2848 (which probably dates to the fifth/sixth centuries) may also have come from this phase, but the context number is uncertain.

Registered finds: 5077 glass; 5237 spindle whorl (Fig. 22: 6); 5320 glass; 5388 imbrex tile, reverse stamp LXF (Fig. 20: 2); 5426 small jar with string cut base (Fig. 19: 4); 5493 imbrex tile fragment, part of rectangular stamp, possibly LXF (Fig. 20: 1); 5546 iron nail.

A fragment of plaster from V.7.3 was kept: white wall plaster in two layers, with a rather polished surface and very few inclusions.

Phase 3: contexts: V.I: 1.18, 25–26, 28–40; 2.7, 10–12, 14–27, 30–31, 33–36, 42; 3.11; 4.1, 6, 7, 9–16; 5.5–8, 11; 6.1–22; 7.2, 5–11

Remnants of occupation/buildings = Walls A, C, D and E; removal of Wall C (= V.1.40); foundation trench of Wall E (= V.2.27); robber trench of Wall E (= V.2.11), numerous pits (Pit B = V.2.30–31, 33–36, 42 + 7.2; Pit C = V.2.21; Pit D = V.2.15–16 + 4.6; Pit E = V.3.11, 4.1, 6, 7); possible floors = V.2.20 +22, 2.26, 6.4. Walls and features: Wall A appears (from a sketch in the site notebook) to be a fairly substantial rubble wall, but is not well recorded; 11 courses of rubble masonry survived to a height of 2.10 m., with some quite large stones, ranging in size up to 30 cm. Much of this may be a fairly deep foundation (foundation trench(?) = V.1.7 +7a). Alternatively the sketch may be a conflation of bedrock + Wall C (see below). Spot height for bedrock = top of Wall A in this area = 673.34 m. (Fig. 5).

e9781782975144_i0010.jpg

Figure 9. Site V.I: west end of trench, showing the door jamb cut in bedrock at the north end of Wall X, and rock-cut steps leading down to it; facing north-east. Neg. no. 65/43.

Wall C is shown on the plan (Fig. 7) as a fairly substantial wall on a similar alignment to Wall A, running north-east/south-west with a second isolated fragment to the north-east. Perhaps only one course of Wall C survived, except in the blocking of the arch cut for passage of the Phase 6 channel in Wall X to the west (spot height for top of ‘Bedrock Wall X’ = 672.81 m.). There is no discussion in the site notebook of the relationship between the various elements discovered in the trench; the photographs (especially Fig. 10) show that Wall C blocks the Phase 6 water-channel (leading to the cistern in the centre of the trench?). The respective spot heights are: plastered channel at west end = 673.18 m.; the top of Wall C = 672.27 m.

Wall D is just a remnant shown on a sketch plan near the east end of the trench, with four stones and a few nearby pieces of decayed mudbrick (= V.2.22) which may be later than floor V.2.20. It might be interpreted as a return or end wall connected to Wall C (Fig. 4).

e9781782975144_i0011.jpg

Figure 10. Site V.I: rock-cut Wall X, Wall C in centre, and quarry steps in foreground around the scale; facing west. Neg. no. 65/44.

e9781782975144_i0012.jpg

Figure 11. Site V.I: south section, central area only.

Wall E just touches the south-east corner of the trench (Fig. 11); two rubble/foundation courses remain, with a foundation trench (= V.2.27) which cuts contexts V.2.29, 29a, 37 and 38. The alignment may be parallel to Wall C. The absolute height of the foundations of Wall E lie between 671.90 and 672.70 m.

If these five wall fragments do belong together, then a rectangular structure could be proposed. Pottery: the Phase 3 pottery included two IA II fragments (including a bowl from context V.5.6, with ring burnished red slip inside, and narrow bands of red/black and white paint outside, cf. Franken and Steiner 1990, 101, Small Bowls Class 2.2a); and a quantity of first century AD and early Byzantine residual material, but the pottery was predominantly sixth to eighth centuries in date, including many forms which first appeared in the sixth century. It included a quantity of LRRW, particularly Late Roman C Form 3 bowls, but otherwise the pottery was similar to that seen in Phases 5 and 4, with many plain incurved bowls in thick well-fired light wares which are closest to Rouletted Bowls Form 4 (Magness 1993, 192.7). The commonest forms were Rouletted Bowls Form 1 (nine examples), Form 3 (two examples), and Form 4 (one example); FBW Bowls Form 1B (one example), Form 2A (one example), Form 2B (five examples); only one fragment of FBW/metallic fabric with pare-burnishing was noted; Shelf-Rim Basins (five examples), Rilled-Rim Basins (one example); Arched-Rim Basin Form 1 (four examples) and Form 2A (five examples); Cooking Pot Form 4A (one example); Storage Jar Form 4A (one example), Form 4B (two examples); FBW Jug Form 1A (one example), Form 1B (two examples) including a shoulder fragment with incised band combing and gashes. A jug in cooking pot fabric was found in V.4.14 (Fig. 19: 9); Magness noted various forms of jugs and juglets produced in cooking pot ware (1993, Jugs Forms 2B, 4 and 5). Few lamp fragments were kept, just a small fragment of a Roman lamp, the base of a Magness Form 3 lamp and part of a Form 3A lamp (mid-sixth to early-eighth centuries). A few noteworthy fragments included a flask rim from V.5.7 with a straight cylindrical neck, a hole pushed through the wall of the vessel and the neck applied to it in the same rather rough way as spouts are applied to jugs in Site J (cf. an Umayyad flask from Amman, in Northedge 1992, fig. 132: 3), with poorly-made, twisted handles crudely applied at right angles to the neck and shoulder. Although the shape is the same as in Byzantine flasks (cf. Zohar, Harper 1995, 28, fig. 14: 104), the Site V.I fragment had no indentation on the inside of the neck, and the handles appeared to be straighter. It is not clear whether the neck and body were made separately on the Zohar examples, which otherwise seemed to be of similar fabric, and which dated to the sixth, possibly fifth century. Equally characteristic of the Phase 3 deposits were storage jars with long folded rims (Fig. 19: 10, six examples) which were also found in Site J (cf. Damascus Gate, Wightman 1989, fig. 49: 7, late Byzantine/Umayyad; cf. from Zohar, Harper 1995, fig. 12: 55). No pottery was kept from floors V.2.20 and 2.22, or from foundation trench 2.27. The pottery notes record ‘Byzantine’. The pottery from Pits B, C, D and E were recorded as Byzantine, with some sherds earlier than the sixth century.

e9781782975144_i0013.jpg

Figure 12. Site V.I: a) north section, part of east end only. b) east section.

The latest diagnostic fragments included a FBW Bowl Form 2A and five scattered examples of FBW Bowls Form 2B (all mid-seventh to tenth centuries, Magness 1993, 198) including one with pare-burnishing from V.6.5 (cf. Fig. 146: 6–7), two fragments of early Islamic Cream Ware, one of Type A2 from context V.2.10, and one of Type B2 from context V.2.7, which according to Walmsley (2001) did not appear before the second half of the eighth century. The latter context included a fragment of a yellow-glazed bowl (Fig. 19: 5) similiar to the early ninth-century glazed fabrics from Site J (cf. J.9.2, J.9.4 and Fig. 151: 2 from J.14.6). The Site J fragments tended to have black painted lines, but this fragment has a plain yellow decayed glaze inside and out, thinner on the exterior; and the fabric was pale, soft and frit-like but with calcite and basalt temper. These three probably ninth-century or later sherds came from the upper levels included in Phase 3, which were cut by Pit A of Phase 2 (see Fig. 12, north section); it is probable therefore that they postdated the rather destroyed remains of structures in Trench V.I. From the same levels came three flared flask necks with pronounced ridge below the rim, examples in V.1.34, V.2.10 and V.2.42 (Fig. 19: 8) characterized by thin, hard pink fabric with calcite temper, one with grey core, two with dark red slip outside and dripping over the inside. The rim on two examples tapers, on one example it thickens; the loop handle starts at the rim. They were similar to Magness 1993 Jugs Form 2A (second to fifth centuries). The nearest parallels were from Site J Locus J.1.6b, from Mafjar (Baramki 1944, fig. 15: 7) and Zohar (Harper 1995, fig. 14: 90). Coins: c2965, 2973, 2984, 3044, 3078, 3079, 3080, 3082, 3086, 3087, 3093, 3110, 3117, 3118. Five coins were illegible, three dated between the second century BC and the third century AD, and five were possibly fifth and sixth century AD in date.

e9781782975144_i0014.jpg

Figure 13. Site V.II: quarried blocks and bedrock; facing south. Neg. no. 65/203.

e9781782975144_i0015.jpg

Figure 14. Site V.II: north section.

e9781782975144_i0016.jpg

Figure 15. Site V.II: lower plan (18/8/1965).

e9781782975144_i0017.jpg

Figure 16. Site V.II: upper plan (6/8/1965).

e9781782975144_i0018.jpg

Figure 17. Site V.II: Walls C, A, B and D; facing north-west. Neg. no. 65/40.

e9781782975144_i0019.jpg

Figure 18. Site V.II: Walls A, B, D and C; facing east. Neg. no. 65/39.

Registered finds: 5010 limestone stopper (Fig. 22: 4); 5016 white marble bowl rim (Fig. 21: 12; cf. Tushingham 1985, 238, fig. 74: 19 Byzantine IA, c. AD 550–575); 5056 grey limestone disk base (Fig. 21: 10); 5067 flat dark limestone fragment (Fig. 21: 14); 5251 copper alloy pin fragment (Fig. 21: 2); 5362 glass; 5679 a large fragment of a Byzantine storage jar (Fig. 19: 12; cf. Magness Jugs Form 1B, mid-sixth to early eighth centuries); 5680 a large fragment of a Byzantine storage jar (Fig. 19: 11; Magness Form 4B(?), fifth to sixth centuries); 5733 Byzantine juglet fragment (Fig. 19: 6; probably Magness 1993, Jugs Form 6A, late third to early eighth centuries); 5734 Byzantine juglet fragment (Fig. 19: 7; probably also Magness Jugs Form 6A); 5915 iron ring (Fig. 21: 3); 5946 iron nail; glass fragments: 6028, 6045, 6049, 6051, 6054, 6062, 6143, 6149, 6158, 6168, 6173, 6174, 6185, 6195, 6199.

Note on contexts V.I. 6.16–19: a high percentage of the registered finds from Site V.I came from contexts 6.16–19, in all 19 objects. There was also a series of nine coins from V.6.19. This stratigraphic context was poor, consisting of soft earth and a hole, which seemed to fill an area against bedrock near the top fill of the cistern. The context could belong to the Phase 5 fill, and to the same context as the three coins found in V.7.13; the contexts might also reflect an occupation pocket in Phase 4, as some large fragments of registered pottery also derived from V.6.16, 6.17 and 6.18. Both coins and other finds from this group indicated a sixth century or slightly later date.

Figure 19. Pottery from Site V.I

1. Basin, possibly a variant of Magness Arched-Rim Basins Form 3, with thick everted rim, slightly incurved walls, and bands of straight and wavy incised lines; fabric: fine orange-brown, pale buff slip; temper: fine and medium calcite, red clay. Phase 5. V.I.7.15.

2. Juglet, handle missing, swollen funnel neck; gouged decoration; fabric: orange-pink with pink-cream surface (FBW); temper: rare large calcite. Phase 5. V.I.7.32.

3. Ribbed cooking pot with inverted and everted rim, short straight neck; fabric: red-brown, soft, brittle, sooty; temper: lot of quartz grains. Phase 5. V.I.7.31.

4. Small jar with string cut base, rim missing; fabric: drab, grey core; temper: some fine white; cf. Fig. 149: 14–17 from Site J, Phase 4 and later. Reg. no. 5426. Phase 4. V.3.4. (EMO).

5. Bowl with everted lip and decayed yellow glaze inside and out; fabric: soft, whitish, frit-like; temper: basalt and calcite. Cf. Fig. 151: 2 from J.14.6. Phase 3 (or 2?). V.I.2.7.

6. Juglet fragment, slightly swollen neck, high shoulder, straight walls tapering to base, slightly ribbed; fabric: fine drab ware firing red at core; temper: calcite; base and much of body missing. Reg. no. 5733. Phase 3. V.6.16. (LEE).

7. Juglet fragment, stump base, high shoulder and straight walls tapering to base; fabric: fine light orange-buff, buff slip; temper: calcite; upper part missing; base blackened by fire. Reg. no. 5734. Phase 3. V.6.16. (LEE).

8. Flask neck; fabric; brown with grey core, pink surface; temper: fine calcite, red clay, tiny black particles. Phase 3. V.I.2.42.

9. Jug; fabric: thin red-brown cooking pot ware; temper: red clay and calcite. Phase 3. V.I.4.14.

10. Jar with long folded exterior rim; fabric: red, with cream slip; temper: medium quartz, medium calcite, red pebbles. A very similar but smaller fragment in V.6.11 has a rim diameter of 11 cm., red fabric, with large angular calcite temper. Phase 3. V.I.5.5.

11. Large fragment of storage jar, short neck thickening to rim, ridge at base of neck; fabric: light red, cream slip; temper: calcite; base and part of sides missing. Reg. no. 5680. Phase 3. V.6.18. (RMS).

12. Large fragment of storage jar, everted triangular rim, loop handle, straight and wavy incised band combing on shoulder; fabric: light orange-buff, buff core; temper: calcite; one handle, rim and neck only. Reg. no. 5679. Phase 3. V.6.17. (RMS).

13. Lid fragment; fabric: pinky-buff, drab core; temper: medium to fine white; rim blackened by fire. Reg. no. 5732. Phase 5. V.I.7.26. (EMO).

14. Bowl with everted lip, carinated wall, flattened base with two incised rings; fabric: metallic with pare-burnishing, brown with grey core, firing red at the surface; temper: calcite and basalt, with some red clay. Interior is crudely painted with bands and dots of thick white slip/paint and bands and lines of black paint, but pattern is worn. Cf. Fig. 148: 9 and Pl. 3b from J.9.4. Reg. no. 5393. Phase 2. V.I.1.9a. (MM).

15. Jug or flask neck, grooved, with funnel-shaped neck; fabric: near metallic. Phase 2. V.I.1.12b.

16. Jar, with long folded exterior rim; fabric: thin, well fired, orange with buff core and cream slip; temper: some fine calcite, red clay and rare black particles. Phase 2. V.I.1.9.

e9781782975144_i0020.jpg

Phase 2: contexts: V.I: 1.5–17, 19–24; 2.4–6, 8–9, 13

Wall B, removal of Wall B (= V.1.22); numerous pits (V.2.4a; Pit A = V.2.5+6; 2.8); much gravel wash, earth and stone fill; there may be evidence for occupation (V.1.14) above a floor (V.1.19) and a fireplace (V.1.21) Notes on Wall B: Wall B seemed to be little more than five stones in rather uneven alignment, immediately north-east of Wall A in the centre of the trench (Fig. 4). Pottery: pottery from 18 contexts was kept; one IA sherd was noted but not kept; there was a little first century AD material, otherwise the pottery was very similar in range to that in Phase 3, and no later material was recorded. One body sherd of Eastern Sigillata A was kept, and some fragments of LRRW were recorded. The most commonly kept fragments were Rouletted Bowls Form 1(12 examples), but Form 2A and Form 3A were also present, FBW Bowls Forms 2B (cf. Magness 1993, 199.6) and 2D, Shelf-Rim Basins, Rilled-Rim Basins, Arched-Rim Basins Form 1 and Form 2A, a wishbone pan handle fragment from Casseroles Form 2 of sixthseventh century date (Magness 1993, 213.1–3), a possible Holemouth Jar Form 1A variant, FBW Jugs Form 1C and 2B (Fig. 19: 15; sixth–eighth century; cf. Harper 1995, fig. 12: 54), and Storage Jar Form 7. Three folded jar necks from V.1.9 and 9b were noted, usually in hard orange fabrics, some near metallic (Fig. 19: 16). A painted metallic ware bowl (Fig. 19: 14) should also indicate an eighth-century date. The pottery from the occupation, floor and fireplace was inevitably rather unhelpful, with three examples of Rouletted Bowls Form 1, but included the base of a FBW Bowl Form 2B from V.1.21 (mid-seventh to ninth/tenth centuries). A number of the pottery forms only appeared in the seventh century and continued in production for sometime thereafter. It seems likely that the phase should date at the earliest to the seventh century but certainly there was some later eighth-century occupation.

Tesserae were noted in seven secondary contexts, with large (2 cm.) tesserae being recorded twice; roof tiles were also noted twice.

Coins: none.

Registered finds: 5007 limestone pestle (Fig. 21: 8); 5393 metallic painted ware bowl (Fig. 19: 14 and Pl. 3b).

Phase 1: contexts: V.I: 1.1–4; 2.1–3

Modern topsoil/surface earth, supported by modern terrace walls to the east and south.

Pottery: late Byzantine pottery from just three contexts was retained, with no reference to recent material; just one brown glazed sherd is mentioned.

Coins: c2599 was Islamic, but not more closely dateable. Registered finds: none.

Conclusions

There is no stratigraphic link between the extreme west end of the trench and the centre (due to the height of the bedrock at the west end), and it is not clear therefore how the various structural elements may have been related. The cistern and quarrying of bedrock reflect at least two stages in Phase 6, probably originating in the first/second centuries AD and earlier but there is no in situ material to date this stage. It is possible that the rock-cut steps, Wall X (the roughly dressed bedrock pier and doorway with jamb, Fig. 9) and a built water-channel (base of Walls A and C) belong in this earliest phase. Large dressed stones formed the sides of a channel (Figs 8 and 10) which led through a hole cut in Wall X (plastered on the east face), and towards a cutting in the bedrock (at the top of the scale in Fig. 10). All Phase 5 deposits above the bedrock date to the late Byzantine period, probably to the sixth/early seventh centuries. The lowest levels were horizontal fill, and clearly reflect the existence of a retaining wall beyond the trench to the east, possibly retained within the cistern, or possibly the walls visible in the side of the modern path (see the coursed masonry below the feet of the figure, and at centre right in Figure 3). The level of the rock-cut steps and doorway are at approximately the same level as the possible Phase 5 floors/surfaces noted in V.7.20, 7.21. In Phase 4 the fill continued to be horizontal, and presumably represented the same levelling process during the seventh-eighth centuries. The separation of the Phase 3 and 4 levels is not certain, and thus the pottery phases should be lumped rather than split. Phase 3 provided the only identifiable structural phase, on top of the Phase 5 and 4 fill, but was represented solely by some miserable remnants. The plastered channel at the west end may have formed the base of Walls A and C (Fig. 10). Walls A, B and C were possibly constructed on the same alignment as some of the walls in Site V.II to the east, which lie at approximately the same height and are supported by the same levelled fill. At the end of Phase 3 these walls were extensively robbed and cut by pits. Phase 3 is not earlier than the sixth century, and may have continued in use to the eighth century, possibly later, but the Islamic Cream Ware and the yellow-glazed sherd may be intrusive from Phase 2. Phase 2, a level of minor walls, pits and poor occupation, dates at earliest to the seventh century, also has material which can be dated to the eighth century onwards, but with no other glazed pottery than the glazed fragment from V.2.7 noted above, which should properly belong in this phase and may indicate a ninth-century or later date (but see discussion of yellow glazed ware in Chapter 6, Site J). Pottery kept from Phase 1 contexts was not representative.

No evidence for Iron Age occupation was recovered, and only three IA II residual sherds were recorded from the trench.

Report on Site V.II

Trench V.II is located 5.30 m. east of V.I and slightly to the north, on the same terrace among olive trees (grid ref. 1724.1310; Figs 2 and 4). The east end abutted the terrace wall which bounded the garden above the adjacent path leading down to the Birkat al-Hamra. The area of V.II initially was 7 x 4 m., but shrank to c. 4 x 4 m. and at bedrock was roughly 2 x 2 m. A maximum depth of 7.40 m. was excavated to bedrock.

Summary of six stratigraphic phases

Phase 6 (the earliest, a fill with mainly first/second century AD and a few Byzantine sherds) included and overlay the quarried bedrock. Phase 5 consisted of gravel and stone tip (mainly first/third century AD with some Byzantine sherds) which may be retained by the old wall visible in the terrace wall to the east. Phase 4 contained two walls which may be terracing to stabilize the upper fill (sixth/seventh century AD). Phase 3 is probably fill associated with Phase 2 (sixth/eighth centuries AD). Phases 2a and 2b (sixth/eighth centuries AD) included a building with two phases of use (Fig. 4). Phase 1 is a thick stratum of earth fill on which the modern garden is set.

Contexts, features and finds

Phase 6: contexts V.II: 53.21, 22; 54.1–5.

The lowest levels in the trench contained rock chippings overlying the bedrock, which had quarried grooves and roughed out masonry blocks (Fig. 13); some very large rock slabs, probably unused quarry blocks, were doubtfully described as Wall G (Fig. 14). Very stoney debris overlay this material.

Pottery: pottery from only five contexts was kept. It included two IA II sherds; pottery of the first century BC/AD including cooking pots, a jar, a fusiform unguentarium base, and sharp, rounded and fine ribbed sherds, mainly in dolomite clay. From the stoney debris came a FBW Bowl Form 2A with very steeply curved wall, red slip and fine dolomite clay (mid-seventh to tenth century); and a flask handle, first century AD. Kenyon described these deposits as fill of the first century AD but the rare Byzantine sherds suggest the debris may have been brought to the site in a later period. Alternatively these later sherds contaminated the loose stoney fill either in Byzantine times, or during excavation. Why Kenyon attributed the quarrying in V.II to the first century, and in V.I to the second century, is uncertain.

Coins: c2938 (AD 59), 3155 (AD 69), 3294 (37–4 BC). Registered finds: none.

Phase 5: contexts: V.II: 53.5, 6, 8–17, 19, 20.

Three to four metres of stone and gravel tip overlay the bedrock; the lower tip slopes from the north-east, while the upper tip slopes from the north-west (Fig. 14). The tip-lines must be retained by an older terrace wall to the east (Fig. 3).

Pottery: pottery from 12 contexts was retained for study, but one context, V.53.14, comprising the great main fill, was exceptional in that many more sherds were kept, and these have been described as a separate group. In the other 11 contexts, the same remarks apply as to Phase 6; almost all sherds belonged to the first century AD. Among these sherds were a) a red painted, thin-walled body sherd (Magness 1994b, 43–44), b) three fine red slipped pieces, c) a rouletted bowl, d) a fragment from the body of a lamp, e) and a fragment from a fine fluted-wall deep cup with ridged band. The latter is unusual (Fig. 23: 9). John Hayes suggests (see below) it may date to late in the first/early second century AD or later. It can be compared to a vessel from Petra, which Khairy (1982, 277, fig. 4: 2, 4) compared to vessels of Augustan date (cf. Bennett 1973, fig. 1: 1, post-Augustan). The Petra cup has a local Nabatean fabric but the Site V fragment does not. A small jar with a related type of decoration, but without provenance, is displayed in the Pergamon Museum, Berlin, described as Terra Sigillata and dated to the third century AD.

There was also a small admixture of earlier and later material. As well as one IA II sherd, the Byzantine fragments included three Rilled-Rim Basins (Magness 1993, 203.5, late third to sixth century), two FBW jars, and one Storage Jar Form 3 (Magness 1993, 223).

Pottery from context V.II.53.14

A large but curious sample of pottery was saved from the deep deposit of gravel fill. The repertoire is very limited and predominantly of first century AD date. It consists of 324 selected sherds, mainly rims, of which 121 (37%) were cooking pots. The whole sample was made up as follows:

1) Spoon/ladle/handled cup

Reg. no. 5654: ladle fragment(?), first century AD (Fig. 23: 8; cf. Lapp 1961, 176, Type 52.9A, miscellaneous cup, from Qumran 1b, 50–31 BC; cf. de Vaux 1954, fig. 3: 15; cf. Clamer 1997, pl. 2: 14, ladle, first century AD).

2) Bowls

Only seven fragments of bowls were kept/included in the group: a) two examples of thin, small incurved rim bowls, b) a similar piece with an exterior fold to the rim, and c) three examples of thin small carinated bowls, all in a pale buff or light pink fabric, but otherwise reminiscent of small Nabatean bowls or cups (cf. Tushingham 1985, fig. 20: 24, 28, Early Jewish, late first century BC to early first century AD). d) A rim fragment of a jar or a deep bowl (RD. 9 cm.) in fabric similar to cooking pot ware was also found (Fig. 23: 2; cf. Clamer 1997, pl. 20: 20, early Byzantine). The rim has an exterior fold and an interior lid indentation and the form may be related to a shape described as a holemouth jar by Magness (1993, 232.3, Holemouth Jars Form 1A, second to fifth centuries AD).

e9781782975144_i0021.jpg

Figure 20. Registered objects from Site V. (Above)

1. Imbrex fragment, part of rectangular stamp LXF(?); fabric: light orange-buff, buff slip; temper: sparse white. Reg. no. 5493. Phase 4. V.I.3.6. (ASH).

2. Imbrex fragment, reverse stamp LXF; fabric: orange, with orange-buff surface. Reg. no. 5388. Phase 4. V.I.3.6. (RMS).

3. Stone weight, pinky-buff limestone, chipped; weight 7 g. Reg. no. 4945. Phase 1. V.II.50.2. (DAJ).

Figure 21. Registered objects from Site V. (Opposite)

1. Round stamp on ajar handle; fabric: light orange-buff, buff slip; temper: some fine white; lime encrusted. Reg. no. 5685. Phase 5. V.I.7.34. (MM).

2. Copper alloy pin or awl, L. 4.8 cm. Reg. no. 5251. Phase 3. V.1.31. Discarded.

3. Iron ring; D. 4.7 cm. Reg. no. 5915. Phase 3. V.6.19. Discarded.

4. Stone mould(?); smooth rectangular block of white limestone, one end rounded, other sides smooth; finely engraved design; 4 x 3 x 2 cm. Reg. no. 5798. Phase 5. V.II.53.14. (LEE).

5. Copper alloy object, heavily corroded. Reg. no. 5485. Phase 4. V.II.51.4. Discarded.

6. Iron nail, D. of head 4.5 cm.; corroded. Reg. no. 6325. Phase 5. V.II.53.14. (SHE).

7. Limestone vessel, unfinished. Reg. no. 5813. Phase 3. V.II.51.26. Discarded.

8. Limestone pestle, L. 5.5 cm. Reg. no. 5007. Phase 2. V.I.1.11. Discarded.

9. Limestone core or pestle(?); H. 6.0 cm. Reg. no. 5047. Phase 4. V.II.53.2. Discarded.

10. Grey limestone disk base with incised concentric circles, lathe marks(?). Reg. no. 5056. Phase 3. V.2.24. Discarded.

11. White marble bowl fragment, shelf rim, smoothly finished. Reg. no. 5796. Phase 5. V.I.7.35. (AIA).

12. White marble bowl rim, shelf rim, poor quality marble, well finished. Reg. no. 5016. Phase 3. V.1.18. (RMS).

13. Limestone core(?). Reg. no. 5818. Phase 5. V.II.53.14. Discarded.

14. Flat dark limestone platter(?) fragment, flat on one side. Reg. no. 5067. Phase 3. V.4.10. Discarded.

e9781782975144_i0022.jpg

3) Cooking pots

i) There were 103 examples of a standard first century AD type (cf. Fig. 78: 12). The fabric is thin, brittle, red, with little trace of ribbing; the rim is slightly thickened, often somewhat concave on the top edge; the neck is short and only slightly everted; the handles are short, rather flat in section, do not rise above the level of the rim, and are set neatly to the shoulder; there is occasionally a slight variation in the detail, such as a narrow rather thick handle (possibly a Byzantine type?), instead of the normal strap. The closest parallels are from the Damascus Gate (Wightman 1989, 28:14, 29:12, first century AD; Lapp 1961, 187, Type 71: N2 and P, AD 0–68). Why so many examples were kept in the archive is unclear, unless it is representative of the occurrence in the deposit. One example has a pair of parallel oblique lines thinly incised in the wet clay on the handle; and there is a vertical strap handle in the same fabric which comes from the body of a vessel, instead of the more usual attachment at the rim and shoulder. For the most part only the handle with small area of attached rim and shoulder was retained, and it is not always possible to be certain that some fragments do not belong to the variety described below.

ii) There were 13 examples of a generally slightly heavier cooking pot with a wider mouth and a marked carination of the shoulder (‘square-shoulder cooking pots’, cf. Fig. 78: 11). Two examples have a rather elaborate rim, and a raised ridge at the shoulder (Fig. 23: 6). The closest parallels are from Site L (Tushingham 1985, fig. 22: 28, 29; fig. 23: 5, and fig. 24: 17, 18 which are generally described as Late Jewish c. AD 40–70; see also Lapp 1961, 190, Type 72.2, from Qumran AD 50–68; Clamer 1997, fig. 4: 12, first century AD).

iii) one example of a large cooking pot with a longer, rather thicker handle, which may be Hellenistic.

iv) two examples of a shallow casserole, one with short strap handle from rim to body (cf. Lapp 1961, 190, type 72K from Qumran, AD 50–68).

v) two horizontal casserole handles, one with rim (cf. Magness 1993, 212.2, Casserole Form I, with range from third to ninth century; this type with uptwisted handle may well be of the fourth/fifth century AD).

4) Storage jars

i) The majority of storage jars had a plain neck, 50 examples, of which 31 had a collar, and 19 either lacked a collar or