Pictures from:

Handout from the 2007 season Open Day

Publication: (available online):
Lockyear, K., T. Sly and A. Popescu with contributions from Mihaela Ciausescu, Clive Orton, Jane Sidell and Robin Symonds (2006-2007). 'The Noviodunum Archaeological Project 2000-2004: results and conclusions from the pilot seasons.' Peuce, New Series, 3-4, pp. 121-158.

Summary Report of the 2000 and 2002 Seasons

The site

The site of Noviodunum is situated on a small hill on the southern edge of the Danube in the Dobrogea region of Romania (see Figure1). The Danube now forms the border with the Ukraine in this region, but in the past it has formed the border between the Roman and Byzantine Empires and barbaricum, and between Ottoman dominated Dobrogea and Russian dominated Bessarabia. Each has left its mark on the site with Roman, Byzantine, Ottoman and 20th century defences overlying each other at this key spot, the last easy crossing of the Danube before the multitude of channels and marshes of the Danube delta. During the Roman period, not only were various Roman army units based at the site, e.g. the Legio I Iovia Scythica, it was also the base of the Roman lower Danube fleet, the Classis Flavia Moesica, later known as the Classis Ripae Scythicae. Alongside the military installations was a large subsidiary 'civil' settlement, and an extensive cemetery. The fortress was paired with that at Aliobrix across the river, near the modern settlement of Orlovka in the Ukraine.The site is now a national archaeological reserve, and the subject of several projects including NAP.


Figure 1: The location of Noviodunum

The Project

NAP is directed by the author alongside Adrian Popescu (now of the Fitzwilliam Museum Cambridge, formerly of the Institutul de Arheologie "Vasile Pârvan'', Bucuresti and Timothy Sly of the Department of Archaeology, University of Southampton, and in close collaboration with Victor Henrich Baumann of the Institutul de Cercertari Eco-Muzeale Tulcea. Preliminary visits were made in 1998 and 1999, and two seasons of fieldwork have been undertaken in 2000 and 2002. Funding has been provided by the Institute of Archaeology and the Graduate School at UCL, and the Department. of Archaeology, University of Southampton.

NAP's long term aim is to examine the development of the site at Noviodunum, and to investigate the relationship between the site and its immediate hinterland, the lower Danube frontier zone and the wider Empire.
Previous work

Interest in the site dates back to the 19th century when Polonic and Tocilescu both visited the site and produced sketch plans. The earliest modern excavations date to the 1950s and took place on the shore of the Danube. Erosion of the river bank revealed a number of structures including the walls of the site, some bastions, a gate, Roman baths and a putative church (Barnea et al.1957,Barnea and Mitrea 1959). Lack of stratigraphy on the beach forced the excavators to rely on structural relationships and mortar samples to phase the remains. Traces of the structures on the beach can still sometimes be seen (Fig. 2). Further rescue work was undertaken on the shore in 1970--71 (Barnea and Barnea 1984), alongside limited excavation elsewhere on the site (Barnea 1977) which found, amongst other things, a statue of a lion.


Figure 2: One of the late Roman bastions on the shores of the Danube in 2000

To the south of the site, along the line of the roads, lie a number of Roman burial mounds some of which have been excavated over the last forty years, mainly under the direction of Gavrila Simion (Simion 1984, Simion 1994, Bujor and Simion 1961). Mound XXX was particularly rich containing, amongst other things, a complete marble statue of a woman (Fig.3) and the torso of a man, and a large marble sarcophagus containing a rich collection of finds, and the preserved hair of the deceased (Simion 1994).


Figure 3: Marble statue found in burial mound XXX by Gavrila Simion.

The Romanian expedition to the site has been concentrating for a number of years on the large late Roman tower on the southern side of the site (Fig.5, A). The tower (Fig.4) is a massive structure which dominates the site to the south. It is to be conserved and left open to the public.


Figure 4: The late Roman tower being excavated by Henrich Baumann. Photograph taken in June 2002.

Lumps, bumps and ditches: surveying the topography

Fig.5 presents the results of the topographic survey so far. The survey was conducted using Leica total stations (Fig. 6) and over 23,000 readings have been taken, largely at 5m intervals although extra readings were taken where necessary to accurately model the site. The largely open nature of the site is ideal for this type of survey (Fig. 7), although thistles, cows (Fig. 8), sheep-dogs and the heat - often over 40oC - make working on the site arduous.


Figure 5: Topographic survey of Noviodunum as at the end of the 2002 field season.
A: the 'Big Tower'; B: the Ottoman Fortress; C: the Roman Fortress; D: putative location of harbour;
E: outer 'defence' of civil settlement; F: valley to east of main site; G: modern quarry;
H: 'civil' settlement; J: burial mound/command and control point.


Figure 6: Surveying the site using a total station.


Figure 7: View across the site from the south.


Figure 8: Cows stop work! Vegetation does not!

When completed, the survey will form the framework within which the project can proceed. Even at its present stage, however, the survey raises a number of questions about the site and current interpretations of it. The last detailed survey of the site published was that by (Stefan 1973) derived from aerial photographs of the site and a visual inspection of it. He proposed that outside the main fortified promontory the civil settlement was defended by three lines of ditches and banks (Fig. 9, defences labelled I, II and III). The topographic survey, alongside on-site observations, brings this interpretation into question. For example, the outer defensive line III looks somewhat different on our survey (Fig. 5, E), no bank survives, and the large natural valley F makes its location seem rather unlikely. This is shown particularly clearly in Fig.10. The complicated pattern of channels near H (which continue into the unsurveyed area) also look very unlike defences. The true nature and origin of these features is one question currently being addressed by NAP, but it seems likely that they are in fact natural.


Figure 9: Map of the site published by Stefan:1973


Figure 10: A 3D image of the site derived from the survey data viewed from the east

Picking up the pieces: the fieldwalking survey

In order to begin to understand the development of the site, a field-walking survey was started in 2002. So far, only two fields have been walked systematically (Fig.5) in lines 5m apart and divided into 20m long sections. All pottery, ceramic building materials and other finds were collected, and the number of large stones was counted, in a band approximately 1m wide. Some 5,300 sherds of pottery and 92kg of brick and tile were collected, along with some fragments of glass bracelets, a couple of coins and some other artefacts. This material only took a day and a half to collect, but two weeks to wash and process!

The pottery and tile was analysed by Robin Symonds of the Museum of London Specialist Services. The pottery was divided into Roman wares (in the case of Noviodunum this is material up to the early 7th century) and Byzantine / post Byzantine wares (largely dating to the reoccupation of the site in the 11th and 12th centuries). The Roman material included large quantities of amphorae, large storage / transportation vessels which are a good indication of where goods were coming from. Most of these amphorae were relatively local Balkan types (Tomber and Williams 1986) Type A being the most common, although one was from Rhodes and some 'hollow foot' examples (Kapitön II) are thought to come from somewhere around the Aegean (Peacock and Williams 1986 Classes 9 and 47, respectively). As well as the amphorae, some fine wares were found including a fine red glossy ware (terra sigillata), some of which may be from Gaul, but most of it appears to be Eastern (Hayes 1972).

The analysis of the pottery from this site has only just begun, but will be a key part of the project, not only as dating evidence, but also in enabling us to investigate the site's relations with the wider world, and its hinterland, for example with the pottery production site at Telita - Valea Morilor (Baumann 1995).

"Electrocuting the grass'': the resistivity survey

Two resistivity surveys were conducted on the site (Figs.5, 12). The eastern survey was unsuccessful. The second western area was chosen because (Stefan 1973) had noted a feature on the aerial photograph that he interpreted as the western wall of the Roman fortress (Fig. 9, line 0). It was hoped that the survey might be able to detect and confirm the presence of this feature. The results, overlain on the contours from the topographic survey, are given in Fig.11. The dark area which runs diagonally across the surveyed area represents an area of high resistance, usually representing a wall or a layer of rubble. Close examination of the contour lines show a low bank in the same area, which went unnoticed on the site.

The resistivity plot does, however, suggest that the feature has a corner in the south-west of the plot, and it is also suspiciously parallel to the Ottoman fortress. Although we have confirmed the existence of this feature, its interpretation remains open. The area surveyed will be expanded in 2003 and it is hoped that the wider picture may make the feature clearer to interpret.


Figure 11: Resistivity survey results overlain on the topographic survey.


Figure 12: Resistivity survey underway.

Drilling for data: the auger survey

From previous work it was known that deposits on the site were, in some places at least, several metres deep. Additionally, Adrian Popescu suggested that the low lying area (Fig. 5, D) was one possible location for a harbour, now silted up. Observations in 2000 had also questioned the nature of the outer 'defences' (see above) and it was suggested that they might in fact be old water channels. It was decided, therefore, to take environmental samples using an auger. The survey was supervised by Ash Rennie (UCL) and the results examined by Ash and Jane Sidell (UCL / English Heritage). Nineteen samples were taken (Fig. 5), some up to 4.5m deep (Fig. 13).


Figure 13: Taking samples from 4.5m down!

In the area of the proposed harbour, these deposits were over 4m deep and surprisingly homogenous. They appear to be soil eroded from the higher parts of the site to the east and south. The sample taken at the lowest point of the `harbour' (Fig. 5, BH 18) did reveal some more clayey deposits at the very bottom which could represent riverine deposits A nearby borehole produced a small sherd of Roman pottery at a depth of four metres. Further evidence for erosion of the site is provided by the Byzantine cemetery (Fig. 5, south of G) which was undergoing rescue excavations in 2002. Here the burials were only 20 to 30cm below the surface. These findings obviously raise many questions about the site: when did this erosion take place? Why? How has it affected the surviving deposits? The site was under cultivation in the 1950s as shown in an aerial photograph published by (Stefan 1973).

In 2003 it is planned to excavate a 5m deep test pit by machine to recover environmental evidence which may help us to explain the changes to the immediate landscape, as well as sinking more auger holes.

The landscape around Noviodunum has obviously undergone significant changes, much of it relatively recently. The most obvious difference is the large scale reclamation of the Danube flood plain for agriculture leaving sites such as the fortress at Dinogetia surrounded by fields rather than water. In the longer term it is hoped the project will be able to address some of these issues.


In two short seasons of work NAP has perhaps posed more questions than it has answered. Some of the current ideas about the site appear to be in need of revision and the project is working towards being able to do this on the basis of detailed information. The work being undertaken provides a strong foundation on which the project can build, and within which the results of future surveys and excavations can be integrated.


BARNEA, I. and A.L. BARNEA (1984). ‘Sapaturile de salvare de la Noviodunum’, Peuce 9: 97–105, 503–518.

BARNEA, I. (1977). ‘Noi descoperiri la Noviodunum’, Peuce 6: 102–121.

BARNEA, I. and B. MITREA (1959). ‘Sapaturile de salvare de la Noviodunum (Isaccea)’, Materiale si Cercetari Arheologice 5: 461–473.

BARNEA, I., B. MITREA and N. AGHELESCU (1957). ‘Sapaturile de salvare de la Noviodunum’, Materiale si Cercetari Arheologice 4: 155–174.

BAUMANN, V. H. (1995). Asezari rurale antice în zona gurilor Dunarii. Tulcea: Institutul de Cercetari Eco-Muzeale.

BUJOR, E and G. SIMION (1961). ‘Sapaturile de salvare din cimitirul roman de la Isaccea (r. Tulcea, reg. Constanta)’, Materiale si Cercetari Arheologice 7: 391–99.

HAYES, J. W. (1972). Late Roman Pottery. London: British School at Rome.

PEACOCK, D. P. S. and D. F. WILLIAMS (1986). Amphorae and the Roman Economy: an introductory guide. London: Longman.

STEFAN, AL.-S. (1973). ‘Noviodunum. Studii de foto-interpretare arheologica’, Buletinul Monumentelor Istorice 42 (1): 3–14.

SIMION, G. (1994–1995). ‘Ensemble funéraire de la nécropole tumulaire de Noviodunum (Isaccea)’, Dacia, New Series: 38–39: 121–149.

SIMION, G. (1984). ‘Descoperiri noi în necopola de la Noviodunum. Raport preliminar’, Peuce 9: 75–96, 481–502.

TOMBER, R. and D. F. WILLIAMS (1986). ‘Late Roman amphorae in Britain’, Journal of Roman Pottery Studies 1: 42–54.

Bookmark and Share