Carchemish, an outpost of the Hittite Empire and an Assyrian city, overlooks the Euphrates and is located today on the Syrian-Turkish border (Figure 1). Recently, with improved accessibility to the site in Syria coupled with the availability of high resolution satellite imagery, Carchemish has begun to reveal a wealth of new and exciting information. The layout of Carchemish is essentially tripartite (Figure 2): a high walled acropolis, a walled lower inner town and a roughly quadrangular outer town with additional ramparts enclosing an area approximately 95ha in extent. In 1911 the Berlin-Baghdad Railway was constructed through the site, and the Turkish-Syrian border later established along it. Thereafter, the excavated citadel and part of the outer town was situated within Turkey. The rest of the outer town, almost 40 per cent of the total intramural site, now lies in Syria. Under the sponsorship of the British Museum the site was excavated between 1911 and 1920 (Hogarth 1914; Woolley 1921; Woolley & Barnett 1952).
Since 2006, the Land of Carchemish Project, a collaboration between Durham University and the University of Edinburgh, has been conducting an archaeological survey of some 500km² on the west side of the Euphrates in Syria, south of Carchemish (Wilkinson et al. 2007). In 2009, the Directorate General of Antiquities and Museums in Syria granted permission for an intensive survey of the Carchemish outer town area as defined in Woolley's 1921 report, and, with support from the Global Heritage Fund, we initiated the Carchemish Outer Town Project. Our primary objective is to record the ancient site, to assess the impact of modern development and identify the archaeological features under imminent threat from the spread of the town of Jerablus, in order to assist in the site's long-term conservation and management.


In May 2009, our investigations within the outer town began with surface collection of two sample areas within fields lying between the outer town wall and the South Gate of the inner town, collecting copious Iron Age ceramics and other small finds (Figure 3). In July 2010, we re-assessed the 1920 site plan by C. Leonard Woolley and P.L.O. Guy. Careful reading of Woolley's narrative describing the outer town ramparts and excavations resulted in our distinguishing 48 points worth revisiting. Woolley's map was then geo-rectified and superimposed on a satellite image (Figure 2). We then visited the locations, recording the visible remains and according each point a waypoint, taking photographs and giving a new description.
According to Woolley, the outer town wall was, for the most part, '...flimsy and shallow, a single course of small stones...' (Woolley 1921: 55), a surprising feature for a city as important as Carchemish. Taking advantage of large-scale excavations made by local farmers in the fields bordering the ramparts to the west, we determined that the outer town ramparts were much more substantial than Woolley had surmised. They were constructed through a combination of layered soil, Euphrates river gravels and broken limestone (Figure 4). Near the West Gate, gravel extraction pits revealed an exposure of mud-brick walling at least 14m in length encased within the rampart, perhaps an earlier feature or a component in the construction (Figure 5). On the south-eastern side of the site, the town wall is less prominent than on the western side, but topographic profiles illustrate that it remains as a low feature to the south-east of the Jerablus football stadium, extending eastwards to a minor scarp revealed in a Pleistocene river terrace. The profiles show the outer town ramparts are a substantial feature, still approximately 50m wide and 1.5–2.0m high.

North–South and East–West cross profiles and transects to collect pottery samples, extending 250 and 300m in length across the outer town, established the approximate topography of the site, the depth of occupational mounding, artefact densities and periods of occupation. In conjunction with evidence from CORONA satellite images, these demonstrated the existence of a major feature we have termed the 'inner anomaly' (Figure 6). This topographic feature may have formed part of a previously unrecorded city wall or ramparts, perhaps subsequently shrouded by occupational debris. The highest concentration of Iron Age pottery (and arguably, occupation) was either on this feature or in its immediate vicinity.
Drawing upon an array of geospatial data, including early-twentieth-century maps, CORONA images taken in the late 1960s and high resolution satellite imagery from IKONOS, Quickbird and GeoEye, we were also able to map archaeological features inaccessible to us; for instance, previously unrecorded architectural remains just north of our 2009 collection areas, near the South Gate of the inner town in Turkey (Figure 7).

Artefacts collected during 2009 and 2010 indicate a dominance of ceramics contemporaneous with Iron Age 2 levels at other sites in the region, especially Tells Shiukh Fawqani (Bachelot & Fales 2005), Jurn Kabir (Eidem & Ackermann 1999), and Afis (Mazzoni 1992). Most parallel types appear in later eighth- and seventh-century contexts on these sites. The assemblages from the transect collections and along the outer wall indicate that the main phase of occupation was later than that suggested in the original excavation reports; our material suggests that it derives mainly from the time when the city was under Assyrian control, namely after their installation of a governor in 717 BC until the conquest of the city by the Babylonians in 605 BC. The surface ceramics imply that the Assyrians were responsible for the enlargement of the city, and that Carchemish was more significant in that period than has been previously assumed. Evidence of occupation in the periods preceding the Iron Age consists of a handful of sherds of possibly Middle Bronze Age date. The post-Iron Age ceramics are represented by small amounts of Hellenistic and Roman wares, making it difficult to determine the scale of the settlement in these periods.
We thank Dr Bassam Jamous, Director, and Dr Michel Al-Maqdissi, Director of Excavations, at the Directorate General of Antiquities and Museums, Damascus, for facilitating the Land of Carchemish Project as well as Drs Nadim Fakesh, Sa'ir Yarta, Yusuf Kanjo and Mr Mohammed Ali, of Aleppo Museum. Field and technical assistance was provided by Nikolaos Galiatsatos, Andrea Ricci, Dan Lawrence and Danny Donoghue. The Project is made possible through funding and support from Durham University, the University of Edinburgh, the Christian Albrecht University in Kiel, Germany, the Council for British Research in the Levant, the Arts and Humanities Research Council, the British Academy and the Global Heritage Fund.
*Author for correspondence