Archaeology in Edinburgh

As part of ongoing research on The Transition from Foraging to Farming in Southeast Europe, being undertaken in collaboration with Dr Ivana Radovanovi, Mr Bonsall spent nine days in Serbia from 28^th May to 7^th June 2000. Most of that time was spent at the Department of Archaeology in Belgrade where the collections from Lepenski Vir are housed, and two days were spent in Kladovo in northern Serbia where material from other Iron Gates sites is kept. During the visit, a total of 136 samples were collected for AMS radiocarbon and stable isotope analyses. The visit to Serbia was at the invitation of the Institute of Archaeology in Belgrade. The Institute generously provided hotel accommodation in Belgrade and Kladovo. Travel and additional subsistence expenses were met by a grant from the British Academy under the Joint Activities Scheme.

Isle of Ulva Project

Clive Bonsall

At intervals during the year further progress was made with post-excavation work for the Isle of Ulva project. Groups of undergraduate students from the Department undertook wet sieving and hand sorting of shell midden samples recovered during the 1999 excavation. The work was supervised by Catriona Pickard.

The Dana-Faynan-Ghuwayr Early Prehistory Project, Spring 2000 Season

Bill Finlayson

This project is a study of the early prehistoric settlement of the wadis Dana, Faynan, and Ghuwayr in southern Jordan directed by Dr Bill Finlayson (Director of the Council for British Research in the Levant) and Prof. Steven Mithen (University of Reading). The project involves survey, test-pitting, trial excavations and palaeoenvironmental research. Much of the focus of the project has been around the mouth of the Ghuwayr, where a Pre Pottery Neolithic A site has been located. Field survey on the terraces of the Faynan has identified extensive Middle Palaeolithic flint scatters.

In spring 2000 the survey area was extended to the uplands in the area between Wadi Hamra and Wadi Al Bustan, lying North of El Khureiba and South of Dana village. The principal objective of the season was to investigate the exploitation of the flint resources in this area at the top of the Wadi Ghuwayr system. The survey comprised three main elements, firstly recording the nature of the flint sources, secondly assessing the artefactual material, and thirdly collecting artefacts from sites defined by the assessment. The main interest in the survey area was based on the premise that this was the most likely source area for flint during the prehistoric occupation of Wadis Faynan and Ghuwayr. The apparently abundant flint in those wadis is now mostly too badly rolled to be useable, although this situation may not have applied in the past, as many cores have been found with rolled and battered cortex.

The eastern and western slopes of the hills in the survey area were very distinct in character, with the eastern slope comprising long relatively shallow slopes (c 45 degree angle), and the western side marked by a much steeper slopes. None of the flint seams outcropped on the eastern slopes, while their presence on the western slopes provided much of the structural form of this side, including a number of overhangs above cliffs. Survey commenced with walking a number of transects. Initially this involved two people walking either side of a tape, each counting flints by square meter. A series of these transects were undertaken, demonstrating that the density of surface finds was generally very low. Transect survey was followed by collection from scatters that had been identified within the transects. In some areas gridded survey was conducted without picking up artefacts, and then collecting from 4 x 4m squares.

The first reason for the survey concerned the assessment of sedentism in the PPNA. We needed to establish whether the type of occupation near the lithic source material was slight, relating only to quarrying and initial knapping, or whether there was evidence for more substantial settlement. The other important aspect was to assist the interpretation of the lithic assemblages from the Wadi Faynan area (including those of PPNB date and later). This would be enormously assisted by understanding the nature of the source materials and how they were exploited.

The most immediately striking observation of the fieldwork is the significant quantities of Palaeolithic material recovered. This is despite the relatively steep topography, indeed most of the material shows very little evidence of any transportation. The hills all appear to suffer from erosion, visible both in the number of gullies present and in the amounts of apparent slopewash at the base of the hills. There is some direct evidence that in places this erosion post-dates these Palaeolithic knapping events, where the material appears to be incorporated in a thick band of fine silts. Some of the best concentrations of Palaeolithic material are currently in clearly eroding situations where the visibility of the material is directly related to their exposure. The overlying soil is shallow. It is possible that recent years of drought are causing the current erosion as the organic binding of the soil surface is lost, but again it is difficult to explain how the Palaeolithic scatters have remained intact until now.

There were Middle Palaeolithic assemblages in all survey areas. They contained Levallois points, scrapers, notches, denticulates, and burins. Flakes were often dihedral or facetted, and many showed signs of discoidal previous removals. Many assemblages also had a large blade element. This may represent a transitional period from Middle Palaeolithic to Upper Palaeolithic, although there have also been suggestions that the Southern Jordanian facies of the Middle Palaeolithic contains many blades, possibly as a by-product of point production.

Four assemblages also contained bifaces, mainly of a moderate size, with straight edges, perhaps suggestive of a Late Acheulian presence. Later periods were very scantily represented, with only two assemblages showing signs of undiagnostic later small flake technology. One of the assemblages contained a Chalcolithic/EB retouched tool.

On the basis of our current knowledge it is difficult to clarify chronology or how these compare with other Levantine sites. There have been many finds of Middle Palaeolithic scatters throughout Jordan in recent years. Two of the more thoroughly investigated sites in southern Jordan are within the Wadi Hasa and the Ras en Naqb surveys. These sites have given some picture of Middle Palaeolithic lithic assemblages in the area, although their chronology and how these sites fit in with other Levantine sites remains a matter of debate. At ‘Ain Difla in Wadi Hasa, the high proportion of blades is seen as typical of Tabun-D type assemblages. At the Hisma sites, however, this is seen as a by-product of the production of broad based Levallois points. These assemblages are seen as Tabun-B type. ‘Ain Difla is dated early for Tabun D – 90-180kya. The Hisma sites are more recent – 48-69kya.

Our sites clearly have similarities with both these areas in the significant proportions of blades present and the presence of Levallois points. Like the Hasa site, narrow, elongated Levallois points are present in the assemblages, but overall length: width ratios remain to be analysed. Also like the Hasa site, a variety of core types is present, and at some sites platform cores outnumber bifacial cores. It is interesting that some of the Hisma material is interpreted as transitional or very early Upper Palaeolithic on the basis of numerous blades with little platform preparation.

Lower Palaeolithic sites have been identified throughout Jordan, especially Late Acheulian ones. In Southern Jordan, Middle Acheulian has been identified at Mashari’a 4 near Dhra, and possibly Jerf al Darawish near Tafileh. The later Acheulian is better represented, with sites in numerous locations in the Hasa (MacDonald et al 1983), Ras en Naqb, Fjaje, near Shobak and Wadi Faynan as well as scattered earlier finds. The Fjaje site is situated close to the current survey area, overlooking the Wadi el-Bustan, and comprises an extensive spread running over several kilometres, but is confined to a narrow strip no more than 250m wide at the top of the escarpment of the highland plateau at c 1200m asl. Although broadly attributed to the Late Acheulian, there are clearly some Middle Palaeolithic and probably Upper Palaeolithic elements present.

Our sites show more affinities with the Later Acheulian. The bifaces are mainly 90-120mm in length, and ovate, cordiform, or amygdaloid in shape. Most have straight cutting edges, are worked around the entire piece and have oval or pointed tips rather than cleaver edges. They are usually symmetrical, and many pieces have fine flaking around the edge. There are some exceptions in the more massive pieces, sometimes with cortical butts or more crudely retouched. The debitage includes small, flat ‘biface thinning flakes’ with radial removals on the dorsal face. This is similar to the situation at Fjaje, but we do not believe we can attribute all of our material confidently to a single period.

Despite the apparent taphonomic problems associated with the Palaeolithic material, the presence of so many discrete sites, represents a find of major significance. They show a variety of interesting features such as the number and range of hand axe types, the type of blade technology, the associated use of Levallois technique, and the potential for locating apparently in situ material. Coupled with evidence being accumulated in the Faynan area the project will now be in a position to consider the nature of Middle Palaeolithic occupation over the whole wadi system. There appear to be open-air sites with the potential of in situ material in both areas, raising the possibility of future invasive fieldwork. It is of interest that in the upper area there may be a Middle to Upper Palaeolithic, or an early Upper Palaeolithic, presence which to-date is missing from the lower area. The sites also raise the question of date attribution to different assemblages, with some material being classifiable into Late Acheulian, Levallois Mousterian Tabun B or D, and possibly early Upper Palaeolithic. As surface, or near surface scatters it is possible that all periods may be represented, but, unlike the enormously extensive spread at Fjaje, we seem to have relatively small, discrete sites, making long time sequences less likely. This appears to conflict with the fact that all sites with bifaces have a Levallois element.

The absence of any easily identifiable material relating to the later occupations down in Wadi Ghuwayr and Faynan is at first sight disappointing. It does however appear to provide some evidence in support of the hypothesis that PPNA exploitation of these resources was based on brief visits rather than extended stays. Whether this absence indicates that exploitation of these flints was restricted to collecting and very crude initial knapping, not immediately identifiable, or whether any such material may be buried under slopewash cannot at present be proved. Two strands of evidence will be used to investigate this further. Firstly, the more detailed examination of the non Palaeolithic material may indicate whether the samples collected may relate to the periods in question. Secondly, a consideration of the taphonomy of the Palaeolithic material may help to assess the possibility that later periods may have been removed by erosive processes, or buried under the subsequent slopewash. A detailed geomorphological study of the area is clearly required. Walks by project members over the last two seasons have demonstrated that access between Faynan and the flint sources is relatively easy. A return trip can be achieved in a day. Good quality flint is therefore readily available in large quantities. In addition, given the changes in environment between Faynan and the upper reaches of the wadis, such expeditions could easily be fitted into subsistence activities exploiting different environmental niches. PPNA sites do appear to be hard to locate however and the absolute absence of another site in the nearby plateau area cannot be dismissed out of hand. At present the scarcity of PPNA sites continues to cause problems in understanding settlement patterns, population levels, or indeed how the subsistence economy of this crucial period operated.

We are grateful to the support and assistance of the Department of Antiquities of Jordan, without whose permission the project would not have taken place, and team members, especially Anne Pirie of Durham University.

Scotland’s First Settlers Project

K Hardy & CR Wickham-Jones

Introduction

Scotland’s First Settlers was set up in 1998 as a regional study of the Mesolithic around the Inner Sound, on the Atlantic seaboard, western Scotland. Given the importance of the sea in the Mesolithic, both as a resource and for transport, the project has taken for its focus the seascape defined by Skye and the mainland: an area with a large coastline incorporating many islands.

The initial aims of the project were to identify new Mesolithic sites within the study area, which would be followed by selected excavations. Prior to SFS, three Mesolithic sites were known in the area, An Corran, in north east Skye, Redpoint in Torridan and Shieldaig, at the north end of the Applecross peninsula. Work at An Corran in the early 1990’s had highlighted the potential for survival of shell midden material in the area, something which was previously thought to be rare in Scotland. Scotland’s First Settlers was particularly interested in looking at shell middens in order to examine their internal composition, for information on lifestyle and environment and to examine the relationship between midden and non midden sites.

In 1999 a short test season demonstrated the great archaeological potential of the area. Survey work revealed 33 new sites and four of these were test pitted, two of which turned out to be Mesolithic. The newsletter from this first season may be found on the Internet (http://www.pabay.org).

Survey and Test Pitting

Field work in 2000 was concentrated in Applecross, but survey also took place at Staffin, and on the islands of Scalpay and Pabay. The coastal survey team of Martin Wildgoose and George Kozikowski, from Skye, recorded a further 66 sites and rockshelters. This brings the total to 104, comprising 74 caves and rockshelters, 21 open lithic scatters, and 9 open shell middens, many of which have visible archaeological remains.

39 of the new sites on the Applecross peninsula and on the Crowlin islands were test pitted in order to assess preservation and date. Finds from the test pits are not only Mesolithic. Many of the rockshelters were in use in recent times and these contain relatively modern remains. Others contain evidence of earlier activity such as a fragment of Norse comb.

Limited shovel pitting was also undertaken, and it produced two new open air sites with stone tools in the vicinity of Applecross bay.

In 1999 the rockshelter site of Sand, just to the north of Applecross, was identified as a particularly promising site with a deep stratified shell midden and stone tools. Analysis over the winter of 1999 confirmed that it was, indeed, Mesolithic (see table of dates) and so it was selected for detailed excavation in 2000. The aims of the excavation were to open and assess an area of midden and to study the surrounding area, a large grassy slope in front of the rockshelter. For this reason, two trenches were opened, across and down the site, each measuring 26m long by 2m wide.

The shell midden lies at the top of the terrace just outside the rockshelter. It lies only a few centimetres below the surface turf and extends for approximately 4x5m. The midden is made up mainly of dry limpet shells, but closer inspection reveals other shellfish, and some fish bones as well as animal and bird bones. There are also tools of bone, stone and antler, together with the waste from tool manufacture.

The unconsolidated nature of the midden and the absence of any interruptions or stabilisation layers suggest that it accumulated over a short, possibly continuous, space of time. At the time of occupation some 8000 years ago sea levels were higher and there would have been a brackish salt marsh some 30m from the site. The abundance of shells in the midden shows that the mesolithic inhabitants of Sand had chosen their spot well for there were plentiful local food resources even if larger animals were scarce in the harsh winter conditions.

Much of the midden material has still to be processed, but from the sample that has been sorted there is evidence for a variety of artefacts including: many bone limpet scoops; a varied lithic assemblage including microliths and the debris from the manufacture of stone tools; fine shell beads; a fragment of antler, probably from an unfinished harpoon; a scallop shell from which a rectangular plaque has been cut; and a boar’s tusk.

Post excavation analysis will start in earnest in 2001 and generous support has been promised by Historic Scotland, though other funds are still being sought. The most important task to be undertaken at this early stage is the sorting of the samples of midden material that were brought back from the excavations. We had hoped to use local people with support from Ross and Cromarty Enterprise to work on this as excavation progressed, but in the event this could not be organised and so a lorry-load of material, carefully bagged into old fertiliser sacks and meticulously documented, was sent back to Infirmary Street.

Preliminary work from the summer of 2000 suggests that the first visitors to the rockshelter were working antler and stone to make tools. Shellfish collection, mainly of local limpets, was very important, however, and gradually the large midden pile built up over the early remains. The shellfish seem to have been cooked, for fragments of stone “pot-boilers” were abundant in the midden, together with the bevel ended bone tools that were used for extracting and processing the flesh. At the same time, the knapping of stone tools continued and items of jewelry and worked shell indicate that there was also time for other pursuits. It is clear that activities at Sand were not restricted to obtaining and processing food alone. Further material has been sent for radiocarbon dating and the results of this are awaited with interest.

Fieldwork at Sand has also included the collection of a variety palaeo-environmental information for it will be of vital importance to build a picture of the world in which the early inhabitants were living. Dynamic environmental change throughout this period is well-documented elsewhere in Scotland so that the information available from SFS will help to increase our knowledge of the early post-glacial in Scotland as a whole. Geomorphological survey by Mike Cressey, of CFA, has identified 4 relict Holocene shorelines around Sand. These provide evidence of the changes in sea level that have taken place through the ages. Preliminary analysis of the shape of a sample of prehistoric shells by Ruby Cerón-Carrasco indicates that violent wave action and high storm frequency may have been prevalent at the time of collection. Collaboration with Professor Kevin Edwards at Aberdeen University will shed light on vegetational history, both on and off-site, and the project is working with Dr Robert Sheil of Newcastle University to look at the history of soil development and early land use.

The inhabitants of Sand were part of a Mesolithic network that operated across the Inner Sound and further afield. They got stone for their tools from Rum (30km to the S, a source of bloodstone), and Staffin on Skye (10km to the W, a source of baked mudstone and siliceous chalcedony). In addition they also used local stones: cherts; quartz; and agates. Analysis of these stones can help us to identify the patterns of contact, communication and movement in this part of W Scotland.

Several local residents worked with us during the dig and since 1999 we have been able to extend our relationship with the local community through work with the primary school and local lectures. A very successful Open Day was held, and attracted visitors from all over the north of Scotland who came to meet the archaeologists, view work taking place on site and study some of the finds. Lectures have also been given further afield, and more events are planned for 2001.

The Society of Antiquaries of Scotland, The Society of Antiquaries of London, The Russell Trust, The Prehistoric Society, The Applecross Estates Trust, Historic Scotland, The Munro Fund, University of Edinburgh, Ross and Cromarty Enterprise Leader II, Private donations

The Centre for Field Archaeology; Applecross Estates; DERA BUTEC and the Works Services department SERCO, Kyle of Lochalsh; RTB Applecross; the crew of LENIE from the RMAS; RJ Macleod Ltd; Highland Council Roads Department; Q Banting; Howie Firth; Lynne Kelly; Lorna Lumsden; and Mike Summers.

Janice Adamson Steven Birch; Richard Bourne; Kirsty Cameron; Mike Chase; Donna Clark; Mike Cressey; Sheila Duthie; Ivy Hancock; Kevin Hicks; George Kozikowiski; Hazel MacFarlane; Fergus MacLeod; Murdo Nicholson; John Patrick; Kath Small; Ian Suddaby; Kirsten Thompson; John Tutt; John Fulcher; Alison MacLeod; Kati Kohler and Bandi; Ann Wakeling; Martin Wildgoose.

Camas Daraich was discovered in the up-cast from a newly constructed track at the southern tip of the Sleat peninsula, Skye, in November 1999. Initial inspection revealed the presence of a large lithic scatter comprising artefacts made predominantly of Rum Bloodstone and including narrow blade microliths. The site was identified as potentially of great interest due to its location on the 20m raised beach which dates to the Late-glacial period, and to the presence in the assemblage of possibly early lithic types: tanged points. Historic Scotland funded initial work to cover the exposed sections and prevent further damage, and a small assessment excavation was carried out in May 2000 with the objectives of assessing the size, preservation levels, artefact types, and date of the main site.

Preliminary work by Martin Wildgoose and Steven Birch in the winter 1999-2000 resulted in a total of 2647 flaked lithics, collected from the spoil alongside the track throughout. Another 261 came from three subsidiary sites.

Excavation work in 2000 concentrated on one trench (Tr1) across the new track, together with four test pits. Three soil pits were also dug. The preservation of a greasy black occupation layer in the vicinity of Tr1 was confirmed. This layer sits directly on the raised beach, evidence from the test pits indicated that it covered a discrete, limited, area, and it contained artificial features: at least one scoop and a possible hearth.

An assemblage of 2013 pieces of flaked stone was recovered from the excavation, together with some coarse stone and pumice. The presence of charcoal and carbonised hazelnut shell was also confirmed.

The flaked stone assemblage incorporates several large pieces (including a scraper and obliquely blunted points) that are typologically early and are not common in Scotland. In addition there are narrow blade microliths that would conventionally be more appropriate to the post-glacial settlement of Scotland. The stratigraphy suggests that the assemblage may be divided into two: the narrow blade lithics lie in the ploughsoil, overlying the assemblage of larger pieces which seems to be associated with the dark occupation layer on top of the raised beach. It is notable that the material in the ploughsoil is scattered across the whole field with a wider distribution than that of the black greasy layer and its associated lithic material. The lithics are made mainly of Rum Bloodstone, with some flint and other chalcedonies. A few pieces are made of a white siliceous material that is not easily paralleled from other lithic sites in the area. As these include most of the larger artefacts and this material is particularly associated with the occupation layer, its origin and nature clearly merits further attention and this will be the subject of study in 2001.

There are three pieces of pumice. One came from the pre-excavation collection, and the other two from the occupation deposit. None are obviously worked, but all have dating potential from tephra studies.

Four samples from features within the black occupation layer were submitted for radio carbon dating. These dates place Camas Daraich at the start of the recorded human settlement of Scotland.

Preliminary palaeo-environmental work by Mike Cressey has combined on-site work with previous research by Selby (1997) to show that two marine transgressions are recorded within the vicinity of the site. These events did not impact on the Camas Daraich site. The age of the main raised beach is probably Devensian (70,000-10,000 BP) with the lower storm beach representing the Late Glacial Shoreline. A lower shoreline consisting of steepening ramps of beach boulders backed by a low terrace is in all probability attributed to the Main Postglacial Shoreline which at 10m OD (estimated) marks the culmination of the Main Postglacial Transgression at about 6000-6500 BP. The position of the Camas Daraich site on the high raised platform is probably due to the shelter afforded by the flanking cliffs on the east and western side. The fact that the in-filled basin examined by Selby (1997) was formerly a freshwater lagoon may have added to the attraction for Mesolithic settlement, though access to sweet water would not have been a problem given the amount of streams and small rivers in the area.

The site at Camas Daraich is of considerable interest, as the presence of potentially early broad blade lithic industries in the Scottish Mesolithic has long been problematical. Camas Daraich provides a date for both broad blade material and a possible dated association for tanged points and it provides one site where it should be possible to examine the relationship between a broad blade industry and a more common narrow blade industry. Further excavation is not feasible at present, but it is hoped to be able to preserve the site for possible future research. Work in 2001 will concentrate on preparing the information for publication and is to be grant aided by Historic Scotland.

We would like to thank the Clan Donald Centre, Armadale, and Roger Winkleman for allowing access to the croft. Steven Birch found the site and he and Martin Wildgoose undertook the pre-excavation work. Historic Scotland are to be thanked for providing funds to get the project off the ground at short notice and for support throughout the 2000 field season. The Centre for Field Archaeology helped to support the fieldwork. Finally, thanks to George Kozikowski, Mike Cressey, Andrea, and to Murdo Nicholson who all worked with us.

Selby K., 1997, Late Devensian and Holocene Relative Sea Level Changes on the Isle of Skye, Scotland. Unpublished PhD University of Coventry, August 1997.

The evolution of ancient technology and society in later prehistoric Europe: The example of the Iron Age salt workings at the Dürrnberg-bei-Hallein, Ld. Salzburg, Austria

Excavations (left) in July 2000 showing the section through the La Tene spoil heap at
the saltmine entrance. Wooden haft (right) of La Tene mining pick in saltmine shaft.

2000 saw the final year of a three-year Large Grant from the Australian Research Council supporting continuing work at this key site for the later prehistory of Central and Western Europe, the other main co-partners being Dr Thomas Stöllner (Vorgeschichtliches Seminar, Philipps Universität, Marburg), overall director of the project, and Dr G. C. Morgan (School of Archaeological Studies, University of Leicester).

Following on previous above-ground work on the site of what dendro-chronological dating has shown was an early Medieval and not a prehistoric Mundloch or mine entrance, exploratory excavation just over the German border indicated tantalising hints of an extensive La Tène settlement although without — so far — any attendant structural remains, the team turned its attention to another area where geophysical surveying had indicated the possibility of a considerable dump of material thrown out from another adit. Partial excavation of this area indicated that here indeed was a major re-deposition of early La Tène material.

Below ground one of the most interesting small finds in addition to the wealth of organic material — textiles (including silk), tools, wooden lighting tapers, shoes (some of children), paleo-faeces (providing fascinating information or diet and parisitology) — was part of an iron brooch of Middle La Tène type, the first evidence for ancient workings in the area extending beyond the early La Tène phase.

The project, being undertaken together with Dr M. Ruth Megaw under a commission from the Clarendon Press, Oxford and funded by the Australian Research Council and Flinders University, Adelaide, completed nearly a decade of collecting which has required research extending from Bulgaria to Britain. The aim of the project is not so much to attempt to provide a new edition of Paul Jacobsthal’s Early Celtic Art (Oxford, 1944) as to present an overview of material discovered since Jacobsthal completed his own fieldwork in 1935 and to discuss changing approaches to this material.

A fitting end to the building up of a database comprising a catalogue raisonné modelled on that originally included in The art of the European Iron Age: A study of the elusive image (Bath and New York, 1970) has been a series of annual visits to Wiesbaden where unprecedented access has been granted to the extraordinary wealth of material recovered from the latest series of excavations below the hill-top settlement of the Glauberg bei Glauburg-Glauberg, Wetteraukreis. The life-size of a warrior found beside a large barrow covering two rich warrior graves, with his very La Tène ‘leaf-crown’, shield and sword and his very Mediterranean composite cuirass literally stands for the cultural cross-currents out of which grew what one may continue to call ‘early Celtic art’.

The dual aims of this season’s work from March to June 2000 were to study material in preparation for final publication of the funerary record and to resolve outstanding stratigraphic problems on site by limited excavation.

In order to provide a framework for the publication programme, to discuss progress and how best to disseminate information, we held a Project workshop in Edinburgh for all major contributors. Bolger, Campbell, Cerón-Carrasco, Lunt, Mavros, Parris, Peltenburg, Philip, Prag and Watt took part. During the day’s meeting, contributors provided progress reports, common concerns about the database were raised, we agreed on the establishment of a web page and we finalised a revised schedule for submission of first drafts.

At our excavation house in Jerablus, all tomb objects were assembled by tomb groups, final pottery recording was undertaken, 80% of drawing and photography was completed and databases for this volume were refined and completed. Analysts are now receiving the results of this preparation and the pro-forma Catalogue publication chapter which will need their input. The balance of human bone, 14C samples and palaeobotanical samples were successfully exported and are now being dealt with by specialists.

Clarification of stratigraphic issues entailed excavation in deposits of two periods, 1A, the Local Late Chalcolithic, and 2B, the fort and its extra-mural occupation belonging to the Early Bronze Age. For terminology and location of areas, see Peltenburg et al. 2000.

During the 1999 season, it became evident that the Uruk Period occupation in Area III rested on Local Late Chalcolithic material of unknown depth. This season we showed that the latter consisted of shallow remains of activities situated directly on the natural silts of the Euphrates floodplain. They were examined in a 70 m2 area, but they were disturbed by Uruk pits, mainly below Building 2185. Ephemeral remains include postholes, mudbrick walls, hearths and surfaces. It was not possible to define a coherent architectural unit, but we may nonetheless conclude that the site was founded in this precarious, riverside locale before Uruk pottery was used in the settlement. The valuable ceramic sequence will be studied in a future post-excavation season.

Outstanding issues were to attempt to date and characterise the initial occupation of the fort in Area IV, to clarify the development of the fortification system in Area I and to establish the chronological status of the Lower Sector lying to the south of the fort in Area II.

In order to evaluate reasons why a fort was imposed on the earlier Early Bronze Age settlement, we required an exposure of the initial occupation of the fort and investigation of its relationship with the underlying occupation in Area IV. This probe would also provide much needed evidence concerning the chronology of the first phases of the fort. To accomplish these goals, a 30 m2 sounding was effected in the base of our 1999 excavation between the fort wall and the south limit of excavation.

It seems that what we had hitherto referred to as the fort wall (999) in Area IV was but an extension founded on material typical of the glacis thrown up against the fort walls in Areas I and III. On this basis, we may infer that the fort extension took place soon after the primary occupation when the whole interior was artificially raised and the free-standing circuit walls were strengthened by a broad glacis.

The earlier fort walls, 3040 and 3152, are contiguous and on the same alignment as 999. Glacis material abutted the northern faces, hence the fort on this part of the site had been strengthened by an additional wall or buttress before the glacis was erected against it. While this evidence represents a significant addition to our understanding of the history of the fort, the time that it took to unravel these deposits meant that we were unfortunately unable to reach the initial occupation as planned.

Here we hoped to refine our understanding of three previously identified, linked architectural elements: fort, extension and annex. In the course of this work, it became clear that the narrow entrance through the annex wall probably led to an internal route which extends beyond the limit of excavation. So, we do not know how or if the route may have led to entrances through the extension and fort walls which lay further inside.

Removal of a baulk at the southern end of the extra-mural passageways in the Lower Sector yielded a Period 3 inhumation burial, the first secure deposit recovered from the Iron Age. Amongst limited grave goods was a bronze omphalos bowl similar to those found at Deve Hòyòk. Several EBA graves were also cut into Passage 990, the latest of the sequence. This area, therefore, became a burial ground after it had ceased to function as a route into the fort and presumably after phase 1 of high status Tomb 302. There were no burials in Tomb 302 after phase 1, so these satellite graves were clustered around a monument for the ancestors. In total, eight graves were recovered from Area II this season.

Attempts were made to establish if the southern terminals of the passageways were attached to an outer wall that enclosed this low extra-fort zone. Erosion had severely damaged deposits here and no associated wall was located. One fairly intact mudbrick wall stratigraphically below the terminals had contemporary deposits with exclusive Uruk sherdage. This is approximately at the same absolute height as Uruk deposits below the South Terrace, hence Uruk period occupation may exist over an extensive and fairly accessible area to the south of the main tell.

E. Peltenburg, E. Eastaugh, M. Hewson, A. Jackson, A. McCarthy and Tom Rymer, Jerablus Tahtani, Syria, 1998-9: Preliminary Report, Levant 32, 2000, 53-75

Further fieldwork, developing the programme underway since 1995, was conducted in association with Dr Pierre-Yves Milcent, Université de Toulouse 2, and Dr Olivier Buchsenschutz, Ecole normale supérieure, Paris, around Bourges in central France during August and September. Since the later nineteenth century, evidence has been accumulating to suggest that the site, Avaricum of the Bituriges in the first century BC, was also a significant node in the settlement pattern of west-central Europe at the Hallstatt/La Tène transition. Initially, this evidence consisted principally of objects from funerary contexts, supplemented by stray finds, some from the wetland adjacent to the rivers Auron and Yèvre which define the promontory on which Avaricum was later centred. Over a dozen Etruscan metal items have been recovered from the vicinity. During 2000, it proved possible to examine two sites relevant to the status of the site in the fifth century BC, with a team drawn from the participating universities and the Service archéologique municipal of the ville de Bourges, and with financial support primarily from the Mairie de Bourges and the Service régional de l'Archéologie.

Les Carrières à Bachon was a major barrow, edged by a substantial ditch, some 40m in overall diameter, located on a slight eminence within a limestone plateau above and to the east of the river Auron, some 4 km S of the town centre. Located by trial trenching undertaken by AFAN, this barrow lay on the margins of the Lazenay sector, where a number of excavations have been conducted over recent years by both the Service archéologique municipal and Buchsenschutz and the writer; the report of these is now in final draft. Total excavation at the tumulus revealed the surviving mound to contain a complex architecture: substantial annular deposits of limestone, cut turfs, unfired clay blocks, gravel, and brilliant white limestone rubble extracted from the surrounding ditch made up the mound, perhaps originally up to 7 m high. The ditch itself displayed evidence of recuts. Whilst successive phases are entailed in the construction and maintenance of this elaborate monument, all could belong in the fifth century BC. Unfortunately, the primary burial within a small cylindrical pit cut into the subjacent limestone had been wholly robbed, probably shortly after deposition. It may have held an imported metal vessel and a cremation. The contents were certainly sufficiently important to attract robbers, who fortunately only marginally disturbed the adjacent burial of an inhumed 7 or 8 year old child, placed in a wooden coffin within a rectangular pit filled with limestone rubble. The gravegoods accompanying this youngster comprised: a copper alloy torc and piece of toilet equipment; two brooches and an elaborate belt-fitment of iron; and a gold ear-ring. Some items have associated leather or cloth. Subsequently, a massive cist was built into the clay-brick mound. Although robbed in modern times, fragments of an adult inhumation and of a bronze torc survived. A funerary pyre erected on the old ground surface, with the cremated remains of an adult, was sealed by the clay brick addition to the mound. Iron wagon-box fittings in considerable quantity suggest the burning of the upper part of a vehicle here. A gold pin, decorated with a ram’s head, was recovered undamaged within this pyre debris - the first substantial La Tène A goldwork from Berry.

Some 3 km east of the historic centre, successive campaigns at Port Sec Nord have revealed traces of semi-subterranean early La Tène workshops within a partially-disused army supply base: another peace dividend. This site had been initially examined by AFAN archaeologists; we were lucky to get the opportunity to carry out further excavations at very short notice in September, and the speed with which the Service régional de l’Archéologie arranged an excavation permit is acknowledged. These excavations were co-directed by Ms Laurence Augier of the Service archéologique municipal. The artefact-rich fills of these workshops indicate that settlement occurred in the vicinity, but extensive topsoiling failed to provide confirmation of the survival of other structures of Iron Age (or indeed any pre-Recent) date in the vicinity. Distributed unevenly over 300m on a gentle slope above an affluent of the Yèvre, the secondary fills of these near-square structures of c. 5m2 provided evidence for both copper alloy jewellery manufacture (unfinished brooches) and bone-working. Associated with this was domestic debris including locally-made wheel-finished pottery, fragments of quernstones, spindle-whorls, and animal bone as well as Mediterranean imports including sherds of Attic pottery (including pedestals), Massaliote amphorae, and ‘céramique à pâte claire’.

Both sites extend our knowledge of the periphery of Bourges in the fifth century BC, and highlight the special characteristics of its apparent wealth. The workshops from Port Sec Nord are paralleled south of the town centre at St Martin des Champs, again associated with fifth century imports including Attic red-figure ware. A third series of apparently-similar features has now come to light in the autumn of 2000 on the site of the former military hospital, again on the periphery of the assumed core of the ‘princely seat’ which may have occupied the high ground around the cathedral SE of the river confluence. Bourges seems to be unique in having multiple series of workshops (known individually elsewhere, e.g. Eberdingen-Hochdorf (Baden-Wurttemburg) and Bragny (Saône), in close proximity to its apparent core. The discovery of the richly-accompanied child and of the pyre with its wagon parts add new dimensions to the funerary rites of the site’s fifth century élite, even without Etruscan metalwork.

Both these sites will be developed during the course of 2001, when it is hoped than an opportunity will arise to examine contemporary settlement evidence within the core of the town of Bourges. A site known to have been vacant at least since medieval times has been identified by M Jacques Troadec, the municipal archaeologist, close to where apparently-elaborate late Hallstatt structural evidence was recorded in salvage work half a generation ago. Negotiations for access, a permit to excavate and research funding are under way. Even if successful, there can however be no guarantee of the survival of protohistoric deposits at this location.

A preliminary account of the 2000 season, outlining the results of both the above excavations, has been submitted to Antiquity and is scheduled to appear in March (Augier et al, forthcoming).

Augier, L., Buchsenschutz, O., Froquet, H., Milcent, P.-Y. and Ralston, I. in press “The 5th century BC at Bourges, Berry, France: new discoveries”, Antiquity 75, 2001.

This small enclosed site, set in woodland in the south of the departement of Indre, western Berry, has long been recognised as likely to be one of the late Iron Age enclosed sites attributable to the Bituriges cubi (e.g. Ralston & Buchsenschutz 1975), for example on account of its possession of a clearly-marked Zangentor. In 1999, as part of the European Union-supported project collectif de recherches on Berry, Dr Buchsenschutz had an opportunity to sink a trial trench into the rampart: the discovery of long iron spikes demonstrated that this small fort had been protected by a wall of murus gallicus type. Le Châtellier at les Pornins thus becomes the sixth confirmed example of this style of defence in the territory of the Bituriges cubi. As well as occurring on major sites e.g. Les Tours at Levroux and Argentomagus, this elaborate fortification type thus – as was suggested earlier by a record of the finds of nails in the rampart at Meunet-Planches, also in Indre – appears too on much smaller enclosed hill-forts in Berry.

In the late autumn, the opportunity was taken by the writer to plan les Pornins by EDM survey. This exercise confirmed the hill-fort to consist of an univallate enclosure sitting in relatively level terrain above the course of the ruisseau de la Claise on its SW side. The rampart is fronted by a ditch, from which it is intermittently separated by a narrow berm. Parts of its circuit are relatively rectilinear, perhaps betokening the timber framework within this defence. As noted previously, there is a single, well-marked inturned entrance, with all other gaps in the rampart circuit appearing wholly superficial. The interior, now in woodland, is archaeologically featureless in ground-perspective survey.

There is restricted evidence of pre-Recent damage, consisting of localised quarry scoops in the front face of the rampart, presumably to win facing stones for re-use in buildings elsewhere. It is hoped that there will be an opportunity for further excavation at this or another of the small murus gallicus defended forts of Berry, to provide evidence to compare and contrast with that provided elsewhere by major oppida enclosed by fortifications of this type.

Ralston, I. & Buchsenschutz, O. 1975 “Late pre-Roman Iron Age forts in Berry”, Antiquity 49, 8 - 18.

Finlaggan is an archaeological site of national importance, being for a period the residence of the Lords of the Isles during the 14th and 15th centuries. It was their administrative centre, place of assembly and site of their Council meetings. The site consists of two islands in Loch Finlaggan, Eilaean Mor and Eilean na Comhairle.

Since 1990 archaeological investigations by a team led by Dr David Caldwell of the National Museums of Scotland have revealed many mediaeval and earlier structures on both islands. An underwater reconnaissance by the Scottish Trust for Underwater Archaeology in 1994 and a series of test-pits by the National Museums of Scotland in 1997 recorded submerged midden deposits off the northern edge of Eilean na Comhairle adjacent to Eilean Mor.

Excavation of a section through these deposits has since been achieved, by a team from The Edinburgh University, by the construction of a dam and the removal of loch water: this allowed an extensive programme of wet sieving, flotation and sampling for the retrieval of palaeoenvironmental remains to be completed.

A midden recorded to a depth of 1.85m was exposed which had been sealed by the accumulation of a series of alluvial deposits of sand and gravel and the subsequent collapse of stonework from Eilean na Comhairle. All midden deposits which were excavated or revealed in section appeared to be in situ. A number of mediaeval artefacts were recovered during the excavation including sherds of pottery, a decorated copper alloy band, a small silver brooch and what has provisionally been identified as a pair of dog collars with a large number of brass fittings, such as those which appeared to have adorned hunting dogs of the Mediaeval period. The collars consisted of two metal alloy chains containing clasps decorated with ring and dot and a pattern of punch marks. Both terminals of the chains had similar opposing zoomorphic decoration. None of the artefacts so far examined from the upper deposits of the midden contradicts a mediaeval date of between the 12th and 14th centuries.

Three periods of occupation have been indicated by radiocarbon dating; Iron Age, Early Historic and Mediaeval. Two distinct episodes of midden accumulation were identified separated by a layer of sand and large stones. Both middens were extremely rich in organic remains containing considerable quantities of animal bone including those of cattle, wild pig, red deer, sheep and fish. The remains of plants, seeds, molluscs and insects were also recovered and a number of pieces of leather and wood, including round wood of birch, shavings, offcuts and several worked pieces. One large burnt wooden stake was recorded in situ in the lower layers. The stratigraphic succession was similar to that recorded in test pits by the National Museums of Scotland suggesting a consistency in the formation of both middens which is unusual for the nature of this type of deposit. The greater quantities of wood apparent in the lower midden including withies and what appeared to be collapsed wattle may have been part of a structure behind which some of the midden could have accumulated.

Subsequent analysis of the organic remains as part of a palaeoenvironmental programme of research has produced a series of insect assemblages, which it is hoped will reveal information concerning resource use and the living conditions relating to human occupation of the island. Each sample that has been processed has contained insect remains, predominantly those of Coleoptera (beetles), with a minimum of 20 different species in each sub-sample analysed. Some contexts are particularly rich and contain upwards of 40 species and over 70 individuals within each sub-sample. Preservation in most cases is excellent, though many body parts were fragmentary. The richest contexts in terms of arthropod remains include those from the Early Historic and Mediaeval periods. .

The overall impression is, not surprisingly for such a site type, one of decay. This varies from indications of a sour, wet foul habitat such as dung to a sweeter drier one, such as decaying hay or straw through the presence of a number of mould feeding species.

Differing environments or activities are indicated within contemporary phases across the area of the site. Differences are also evident between chronological horizons, providing the potential for inferences to be made concerning changing site use. In relation to this a few puparia of Diptera (True flies) have been noted from the lower mediaeval layers, providing a breeding population for analysis. This may indicate the presence of faecal material or carrion, both a favoured pabulum of many species within this Order. A number of wood-boring beetles are also present in certain contexts and may originate from some of the structural wood remains which were encountered during excavation.

A particularly significant discovery in an Iron Age context is the remains of a number of flea heads. Although their identification is tentative, their host appears to be human, so their presence has implications concerning the living conditions and hygiene of the site's occupants and/or site use within this period.

A more detailed interpretation of the insect remains will require further identification to genus and species level.

This report is a summary account of the results of Phase1 of an archaeological survey undertaken between 1998 and 2000 at Duddingston Loch in Edinburgh. The survey was designed to record, identify and assess any archaeological remains within the loch and its immediate surrounding areas using a combination of geophysical survey, fieldwalking and underwater survey. The content of Phase 2 currently being carried out is also highlighted.

An investigation of existing archaeological and palaeoenvironmental sources regarding Duddingston Loch and its immediate environs was undertaken. A number of primary and secondary sources, maps and illustrations have been consulted and an examination of the records held by the National Monuments Record (NMRS) concerning the loch completed. A print out of these records will be incorporated into the final report.

In the past a number of artefacts have been recovered from the loch covering a time period of at least 5000 years. Information on these artefacts was obtained from the NMRS entries, particularly on their circumstances and contexts of discovery with a view to comparison and correlation with any new discoveries. Relevant first edition Ordnance Survey map coverage has been examined and other readily available 18th and 19th century cartographic sources held at the National Library of Scotland Map Library in Edinburgh have been examined for evidence of former activity and structures within the study area. A number of oblique aerial photographic records held by the NMRS and vertical stereo photographic coverage held by the Royal Commission on Ancient and Historical Monuments of Scotland (RCAHMS) were studied to identify any archaeological features within the loch and the survey area and for comparison with existing conditions at the study site. There has also been liason with a number of other interested parties namely Historic Scotland, Scottish Wildlife Trust and Scottish Natural Heritage before and during this survey.

A number of areas were identified by fieldwalking and ground survey for future coring and land-based resistivity, magnetometry and ground radar.

A bathymetric survey using an echo-sounder to assess the uniformity of depth across the loch and a survey of the loch bed using sidescan sonar to produce a topographic map and identify the presence of any features on the surface of the loch bed, was undertaken by Dr Robert Duck of the Tay Estuary Research Centre, Dundee. A number of anomalies were indentified on the loch bed on the south side in the middle third of the loch; these tied in well with evidence from a number of written sources of the recovery or presence of archaeological remains, (Dick, 1780), (Wilson, 1863) and (Baird, 1898). These will require further investigation using geophysical and underwater survey by a team of divers. In addition, a possible wave-cut ledge was identified on the northern side of the loch to the east of the breakwater. This suggests that at some point in the past the loch may have been smaller than its present dimensions.

A sub-surface survey of the loch bed using ground radar was initiated by the Department of Civil Engineering, The University of Edinburgh with a view to identifying individual features on the surface and within the loch bed. The results suggested the presence of a feature within the loch bed near to the surface proximate to the position of the anomalies recorded by sidescan sonar. An exact correlation of the positions has not been possible at this point. However, the consistency of one area within the loch producing results of this nature augurs well. A large feature, 40 to 50 metres in length and 5 metres wide, has been identified by resisitivity survey adjacent to the loch edge and awaits further prospection by magnetometer.

An underwater survey by divers to explore the loch with a view to gauging the depth of the loch sediments was initiated. The sediments appeared to be of considerable (1 metre or more), but variable depth.

Further geophysical survey is planned for the Summer of 2001 including a magnetometer survey of the loch bed and further use of ground radar in accessible loch areas. The possiblity of including another geophysical technique - chirp sonar is being considered. This technique would potentially provide further information regarding any sub-surface features.

The above loch survey will run concurrently with a geophysical survey of the immediate surrounding land and infilled areas of the loch. Resistivity survey ( using a resistivity meter, RM 15), magnetometry (using a FM 18 magnetometer ) and a conductivity survey (using an EM 31 conductivity meter), if available, will provide information regarding any potentially detectable sub-surface archaeological remains in these areas.

It is hoped that the positions of all forms of survey used in this study will be integrated using one GPS system.

As Scheduled Monument Consent has been kindly granted by Historic Scotland coring will be carried out where possible in a number of areas to acquire palaeoenvironmental samples. The survey will provide a series of cores from the loch bed and surrounding areas which will allow analysis of changing oxygen isotope levels (by Professor Kevin Edwards, University of Sheffield) and acquisition of a pollen and stratigraphic sequence for a reconstruction of the late glacial and holocene vegetational history at both the local and regional levels. Some samples from the cores may also provide the basis for undergraduate dissertations within the Department of Archaeology, The University of Edinburgh.

A detailed underwater survey of the loch bed will be undertaken by a team of divers after the geophysical survey has been completed and the results analysed. This will allow particular areas of interest revealed by the survey to be examined in detail.

Analysis, interpretation of results and reports by individual specialists will follow the completion of the survey. These will be integrated in a final report which will be produced in due course with the results being offered to the Society of Antiquaries of Scotland for publication.

Archiving - A complete archive of results and interpretation will be deposited in the NMRS and reproducible elements in the form of a report will be supplied to Historic Scotland and the Park Library.

The potential strength of the survey lies in its multi-disciplinary approach and the integrated nature of the different survey techniques, particularly the remote sensing and underwater elements. the data obtained so far relating to the loch are promising, revealing a series of anomalies on the loch bed which require further investigation. If the experimental elements of the survey prove to be of value, this combination of different techniques (concentrated underwater survey in conjunction with appropriate remote sensing), could in the future be applied to similar loch environments.

I would like to thank Ian Morrison of the Department of Archaeology, The University of Edinburgh for his assistance and guidance regarding the health and safety elements of the underwater survey and Ian Armit, then Inspector of Ancient Monuments at Historic Scotland for his help and advice regarding the study area. Also to express my gratitude to the Scottish Wildlife Trust, Scottish Natural heritage and the local warden, Colin Mclean at the Bawsinch Reserve, for access to the loch and its surrounding areas. Lastly my thanks to Ian Mack, elder at Duddingston Kirk for the anecdotal evidence he supplied regarding the loch.

La Porte du Rebout at Mont Beauvray, Saine-ey-Loire and Nievre, France: Reconstruction

The writer’s involvement in this project during the year was limited to two brief visits from Bourges to deal with ‘snagging’ following the completion of the full-scale reconstruction of the murus gallicus flanked inturned entrance of the Porte du Rebout (the excavation of which has been published in full: Buchsenschutz et al, 1999), the largest entrance in the principal, inner rampart around the oppidum and certainly a candidate for the most extensive gateway in any of the late Iron Age oppida of continental Europe. This project was undertaken in association with M Paul Bernoud, Architecte en chef des monuments historiques, Dijon, and MM Vincent Guichard and Jacky Gorlier of the Centre archéologique européen du mont Beuvray.

The first reconstruction of the northern bastion, which had been rapidly built some years ago and which had become progressively dilapidated, was demolished totally prior to the erection of the new work; the southern bastion, fronted by its original rock-cut ditch, was rebuilt for the first time as part of this project. Only one phase of the complex sequence of rebuildings identified archaeologically is represented by the in situ reconstruction. The ensemble, through which the public enters the site uphill from the major museum at the Col du Rebout, now offers a good impression of the monumental character of the late Iron Age architecture on the site and literally acts as a gateway to the field investigations which are still underway there.

In light of the very partial evidence for structural arrangements furnished by the excavations of the fairway between the two inturns, nothing has been built in the entrance-way itself (through which the tarred access road to the site still passes). It is undoubtedly the case, however, that the Porte du Rebout was never furnished with the kind of elaborate gatehouse that is postulated for distinctly narrow entrances, such as the eastern gate at Manching, or gate D at Zavist.

Buchsenschutz, O, Guillaumet, J.-P. and Ralston, I. 1999 Les remparts de Bibracte: recherches récentes sur la Porte du Rebout et le tracé des fortifications. 316pp. Glux-en-Glenne: Collection Bibracte, 3.