3. Ripae et Litora

3. “Ripae et litora,” Supply and security on the riverine and coastal edges of the Roman Empire
Maandag 22 August 2022, Steigerzaal

Session Chairs: Wouter Dhaeze, Erik Graafstal, Tom Hazenberg & Jeroen van Zoolingen
Affiliation: City of Oudenburg, Belgium
Affiliation of co-organiser: city of Utrecht, the Netherlands
Affiliation of second co-organiser: Hazenberg Archeologie, the Netherlands
Affiliation of third co-organiser: city of The Hague, the Netherlands

Session Abstract: At its height the Roman empire’s edges stretched over 29,000 km, half of which bordered the sea. While most of the imperial shores needed little protection, the northern provinces, both on the Continent and in Britain, saw military investment along the coasts in various forms and contexts from the 1st to the early 5th century. A special feature of this seeming backwater was the transport link between the German Rhineland and Britain. This corridor was a vital piece of infrastructure throughout the Roman period. With both the German and British fleets involved in its operation, and many harbour and military sites along it known, we are singularly well informed about this lifeline of the northern Empire.

This session will explore the interconnections between supply and security on the exposed water frontiers of the northern provinces. While Roman land frontiers have been studied in detail, they cannot simply be seen as a blueprint for coastal systems. The latter are often merely seen as extensions of land frontiers in reaction to seaborne raiders, but their purpose, development, operation and tactics are in fact more complicated, and unique for each situation, depending i.a. on the local geography, sealanes and flows of traffic. Recent work on the British shores and the North Sea and Black Sea coasts suggests that their military dispositions can be concerned not just with coastal defence, patrol and transport security, but also with taxation, logistics or communication. River frontiers also deserve a closer look, as they are functional hybrids, merging frontier security and river logistics. In the Rhine delta, especially, the picture gets blurred as coastal and riverine infrastructures merge.

This session intends to explore this broad theme of supply and security under four headings, concentrating on the northern shores and river corridors. The organisers invite papers on the following subjects:

  • Hubs, harbours, ships
  • River transport, navigation and the maritime seascape
  • The range of activities of the German and British fleets
  • Coastal security systems: southern North Sea, Cumberland Coast and Litus Saxonicum
  • Comparative perspective from Danube and Black Sea region.
Time Presenter (s) Presentation
9.00 Welcome & Introduction
9.20 Sytze van Heteren Natural drivers of coastal behaviour along the North Sea
9.40 Alistair McCluskey Prisoners of ethno-geography: Transnational dynamics of warfare between Late Iron Age Ireland and the Roman frontier in Britain?
10.00 Wouter Dhaeze Sea frontiers along the Channel and the North Sea. Development, purpose and tactics


10.50 Jeroen van Zoolingen Defending dunes and marching along marshes: details of the Antonine coastal limes between Rhine and Meuse
11.10 Arjen Bosman The Roman bases at Velsen
11.30 Jane Harrison The Mystery of the Marsh: the western end of Hadrian’s Wall, UK
11.50 Philip Smither Shore-ing Up Britain


13.30 Kim Cohen Lower Rhine River palaeolandscape mapping: understanding meander dynamics below and along and flood deposition around and above Limes archaeological horizons
13.50 Ulrich Stockinger Go with the flow. Antunnacum/Andernach and the supply of material from its hinterland to the Lower German ripa
14.10 Thomas Becker River Main as route of military supply – new archaeologicial and geoarchaeological research
14.30 Norbert Hanel Bridge construction and maintenance work at the ripa Rheni and on the upper course of the rivers Moenus and Mosa as strategic construction measures in the 4th century AD – Results from dendrochronological and historical studies


15.20 Ian Longhurst Chesters Road Bridge
15.40 Ronald Visser Securing the transport of timber on water
16.10 Andrew Tibbs 1st C Scotland
16.30 Christoph Rummel When’s a fleet a fleet? Classes and legions on the water
16.50 Yardeni Vorst Operation Zwammerdam ships: Putting things together on rowing and steering on the Roman Rhine

Natural drivers of coastal behaviour along the North Sea
Sytze van Heteren, Geological Survey of the Netherlands

While rivers formed clear-cut natural borders to past empires, seas and their coasts acted as more diffuse margins. As a river migrated, the border typically migrated with it. As a coast migrated, borders were influenced indirectly; there was no direct enemy or rival on the other side. Some of the coastline changes during Roman times were man-made or self-inflicted, but others were the result of natural drivers such as sea-level rise, storm surges, and sedimentation and erosion patterns. Awareness of the possible roles of these natural changes is important when analyzing and framing the history of the limes. Reconstructing these changes remains a key challenge that requires archeologists and geologists to work together. The limes bordering the North Sea, on both the British and the continental side, and the Irish Sea traversed different types of coastal area. Soft-sediment coasts showed the dynamics that still mark them today, with shifting estuaries (and people’s fortunes), breaching and healing sand and gravel barriers, and fixed or highly mobile dunes. Together, they moved the coastline landward and seaward during periods of land loss and recovery, of change and stability. Challenges and opportunities resulted, with settlements and fortifications suffering during some times and benefiting at other times. Cliffs moved landward. As their edges were eaten away over the centuries, there was only one way to go for coastal communities. By putting incomplete geological and archeological records together, we are able to say more about coastal and border dynamics than when these records are considered separately. Thus far, this has been done more extensively for man-made or self-inflicted changes than for their natural counterparts.

Prisoners of ethno-geography: Transnational dynamics of warfare between Late Iron Age Ireland and the Roman frontier in Britain?
Alistair McCluskey, Independent Researcher

Analysis of contemporary and recent historic warfare has underscored the significance of transnational dynamics, such as diaspora, that interact with or contest the existing political geography of the conflict region in question. This paper will apply the transnational dynamics of conflict to three aspects of the Roman frontier of Central Britain to demonstrate the value of this framework in understanding: (1) the human terrain networks in late Iron Age Britain and Ireland; (2) their relationship with the physical geography of the region; and (3) their relationship with the Roman military strategic response. The geographic links between Late Iron Age Ireland and Roman Britain are seldom studied in relation to Roman frontiers, with the Irish Sea more often perceived as being a barrier, frequently polarizing research perspectives between either the Late Irish Iron Age or the Romano-British Iron Age. When contact across the Irish Sea has been considered, the inquiry has tended to reflect upon the presence of Roman material in Ireland in the first instance. Although Roman and Iron Age research fields are well developed, they struggle to reconcile with each other beyond the presence – or otherwise – of each other’s material culture on their respective sites. This paper will argue that the transnational and connected character of the North British Iron Age communities around the Irish Sea was exploited by some to pull Rome north as a strategic ally, while others used these connections to resist Rome’s advance and develop their own strategic advantages where possible. In this respect, new perspectives of the frontier emerge including a possible ‘west facing’ limes along the coast, rivers, Pennines and Southern Uplands to complement the north facing mural barriers, developed over a military campaign extended across some 80-90 years between AD 70-160.

Sea frontiers along the Channel and the North Sea. Development, purpose and tactics
Wouter Dhaeze, City of Oudenburg, Belgium

The harbours along the shores of the southern North Sea and the Channel were of the utmost importance for communication, army supply, and trade between the Continent and Britain and along their coasts. From the Early Empire on the military took measures to protect critical infrastructure along this exposed stretch of the northern frontier. The military dispositions were concerned not just with coastal defence, patrol and transport security, but also with taxation, logistics or communication. The position of the known forts, not installed at the mouth of the rivers or tidal inlets, but some kilometres more inland, and usually offering not a perfect lookout for raiders, suggest that communication and control over the supply chain prevailed upon coastal defence. This does not mean that, in the case of raids, the available infrastructure or troops could be used, but always most economically. The water frontiers along the North Sea and Channel were not simple copies of the land frontier systems. Their purpose, development, operation and tactics are more complicated and unique for each situation. Military investment along these shores was continuously adapted to changing sea lanes and flows of traffic, changing morphology of the coastal areas, internal agendas of emperors, usurpers or governors, and changing barbaric threats. A lot of questions remain on for example the statute of these shores in the Late Roman period and the tactics of the coastal defence. Were they considered provincial borders or military borders as well? How did the available personnel in the military installations operate? Besides a general overview of the sea frontiers of the North Sea and Channel, this lecture elaborates further on these questions.

Defending dunes and marching along marshes: details of the Antonine coastal limes between Rhine and Meuse
Jeroen van Zoolingen, Municipality of The Hague, The Netherlands

Much of the Roman coastline in The Netherlands has been lost to erosion by the North Sea. The The Hague region is one of only a few areas where some of this landscape remains. Now mostly covered by medieval dunes, at least two military sites have been excavated. The very first military activity in the area is known from the Ockenburgh site, where in the first half of the 2nd century AD a temporary camp arose. Not lang after, around 150 AD a praesidium was installed adjacent to it, soon followed by a vicus covering the features of the camp. The outpost was occupied by a cavalry unit with intervention, patrol, and communication as its primary tasks. It functioned in a system of multiple outposts between the rivers Rhine and Meuse, of which at a distance of some 7 km the Scheveningseweg site is best understood. Though no military structures could be identified, multiple militaria and tools indicate a military vicus. Further inland, the Roman military undertook large scale infrastructural works. Around the same time when the Ockenburgh outpost was build, a primary Roman road between the rivers Rhine and Meuse was built/refurbished and dedicated. Around 160 AD a harbour was installed at the Roman town near present-day Voorburg (Forum Hadriani), which lay along this Roman road and was reachable through a canal connecting the major estuaries. In the last 10 years many of these sites where (re-)studied and published. As a result, the area features some intriguing detail to understanding the development and functioning of a coastal defence and guard system during the 2nd century AD.

The Roman bases at Velsen
Arjen Bosman, Military Legacy

The last significant presentation at the Limescongress on discoveries in Roman Velsen dates from 1995. Since then a lot has been achieved, mainly by research of the final base of Velsen 2. Insight on the outlay of the several bases of both Velsen 1 and Velsen 2 and their surroundings has changed dramatically. Also the difference between the forts of Velsen 1 and Velsen 2 has become more apparent, by indepth study of all findgroups. The dates at which the Romans founded and left both bases have become more solid. The supply of both forts is better known as is the type of troops which occupied Velsen.

The Mystery of the Marsh: the western end of Hadrian’s Wall, UK
Jane Harrison, University of Newcastle, WallCAP

This paper re-examines the western end of the Hadrian’s Wall, between Luguvalium (Carlisle) and Maia (Bowness-on-Solway). This stretch of the Wall faces the Solway Estuary to the north, unlike the eastern end where Tyne river and its mouth lie to the south of the Wall, and the paper asks whether this landscape position determines the character of the Wall structures along the estuary. To investigate the question a number of more recent small-scale pieces of fieldwork are brought together with older investigations and new work by the WallCAP Project in the area. The investigation takes a landscape perspective in examining an area of Hadrian’s Wall that has seen less archaeological work than the central and eastern sections of the Wall. Overall, the landscape is low-lying and marshy, punctuated by hillocks of higher ground, which have always been the focus of settlement. How did this influence the development of the Wall constructions? There has been much speculation about what happens to the Wall structures between Milecastles 73 and 76, across the Burgh Marsh and Drumburgh Moss, where no evidence has been discovered for the curtain wall or indeed any aspects of the Wall. Did the landscape and security dictate a different approach here or has the Wall been lost to erosion? Does comparison with other estuarial and riverine stretches of the Limes in other parts of the Empire help answer these questions? In this part of the World Heritage Site, the Wall structures and forts are mostly not visible and there are no spectacular stretches of upstanding remains, so the paper examines how landscape archaeology approaches can advance understanding of this area of the Limes.

Shore-ing Up Britain
Philip Smither, University of Kent/English Heritage/Portable Antiquities Scheme

Since the 1922-1939 excavations at Richborough, research into the late Roman shore forts in Britain took off. However, not only did archaeological evidence come up against historical evidence, most notably in the form of the Notitia Dignitatum, but the archaeological interpretation is on ground as unstable as the Thanet Sands. Returning to the Richborough excavation archive and reassessing the published material on the shore forts in Britannia has revealed that much was either omitted or misinterpreted. We can now also go beyond a simple historical narrative of the shore forts and provide a context within their landscape. It has often been noted that the shore forts vary in construction and this evidence has been used to provide a sequence for the shore forts which is often at odds with the archaeological evidence and historical narratives. A reassessment of the evidence in light of new archaeological analysis shows that in almost all cases the forts were the brainchild of Carausius and Allectus. Furthermore, their use in the 4th century suggests a close link with the main settlements in the hinterland. As the focus of the shore forts has mostly been in isolation a view of the wider landscape shows clear interactions between the shore forts and its environs. A major problem with interpreting the shore forts is how they are presented in the Notitia of the early 5th century. This presentation often led to them being interpreted as a unified system. While this is likely the case to some degree, this falls under a reutilisation of the forts in the AD330-40s. A further reutilisation, at least at Richborough, likely occurs in the c.AD380s when the site layout is significantly altered. There is a clear need to re-examine the archives of the shore forts, of which Richborough is only the beginning.

Lower Rhine river palaeolandscape mapping: understanding meander dynamics below and along and flood deposition around and above Limes archaeological horizons
Kim Cohen, Utrecht University, Dept. Physical Geography

For a variety of reasons, the Holocene geological-geomorphological record of the Lower Rhine valley and deltaic river branches are intensively studied by geoscientists, paleo environmentalists and archaeologists alike. Palaeogeographical reconstructions, nowadays stored in digital map datasets, iteratively integrate that accumulated knowledge. For the Lower Rhine (and hence for the Limes that runs along it) they are very well developed, tightly age- and elevation-constrained. The contribution will cover channel dynamics: what older and younger meanders were present to settle on top of and along? What continued meander reorganization is seen during and after Roman times? What entire branches switched on and off? It will then cover flooding during and since Roman times, and the overbank deposits that this left.

Go with the flow. Antunnacum/Andernach and the supply of material from its hinterland to the Lower German ripa
Ulrich Stockinger, Universität Basel

Antunnacum/Andernach is located on a strategic position on banks of the Middle Rhine, at the north end of the Koblenz-Neuwied Basin and next to the resources and products of the Mayen region. Thus, it was an important part of the logistics to supply construction material (tuff, basalt) as well as millstones and pottery especially to the ripa downstream. While the site served as a garrison only in the Early Imperial period and in Late Antiquity, it retained its special importance to meet the demands for these goods throughout the Roman period and beyond. Starting from the first large-scale excavation within the Roman imperial civil settlement resp. the fortification of Late Antiquity conducted from 2008 to 2014, this paper presents an overview of the current state of research on Antunnacum. The questions addressed include the role of Antunnacum in the supply chain to the Roman troops and for the civilian trade to the ripa region, the organisation and development of these commercial operations and the influence and interactions between the military and civilian sphere.

River Main as route of military supply – new archaeologicial and geoarchaeological research
Thomas Becker, Landesamt für Denkmalpflege Hessen, hessenARCHÄOLOGIE, Außenstelle Darmstadt, Andreas Vött, Lea Obrocki, Johannes-Gutenberg-University Mainz, Institute for Geography, Geoarchaeology

The rivers Lahn, Main and Neckar are the largest within the Roman province Upper Germany, which connect the Rhine with the eastern parts of the province. Rivers Lahn and Main originate outside the Roman Empire and cross the Limes at different positions. The course of the Neckar concentrates on the provincial territory. The Main is the one of three rivers, which has the longest course through plain terrain outside the low mountain range (Mittelgebirge). The plain terrain on both sides of the river is an intensely settled area. The river and its stream network play a central role in supplying the military on the northern border section of the province. At the same time, it was used to transport raw materials from the border to the hinterland. As part of the conquest of the provincial area on the right bank of the River Rhine, it not only offers supply options, but also has a strategic function. In recent years, both archaeological and geoarchaeological research was carried out along the river, which supplemented and expanded this picture of the water course used and controlled. This provides information on the varying use of the river from the 1st to the 4th century AD.

Bridge construction and maintenance work at the ripa Rheni and on the upper course of the rivers Moenus and Mosa as strategic construction measures in the 4th century AD – Results from dendrochronological and historical studies
Norbert Hanel, Universität zu Köln, Archäologisches Institut, Thomas Frank, Universität zu Köln, Institut für Ur- und Frühgeschichte, Labor für Dendroarchäologie / Dendro-Archiv NRW

In recent years, dendrochronological studies have provided important new insights into the construction and repair of Roman bridges on the Rhine and on the upper reaches of the Meuse and the Main estuary. In combination with historically documented events, these results allow a new evaluation of the bridges in the overall strategic concept of late Roman military leadership in the course of the 4th century AD. This is because the joint examination of dendrochronological data and historically documented events means that dendrodata, which were previously incomprehensibly young and therefore subject to criticism, can now be casually incorporated into the historical development and complement the understanding of bridge construction and maintenance/repair as part of late Roman military strategy. According to current knowledge, the dendrochronologically dated bridge construction measures can be linked to historically handed down campaigns against the Franks:

    • with the campaigns of Emperor Constantinus I in 310/315 AD
    • with the construction programme that transcended provincial boundaries and the military campaigns under Valentinianus I around 367 AD
    • with the conflicts under Valentinianus II/Arbogastes with tribes on the right bank of the Rhine between 388–393 AD.

As a result, not only fortress construction programmes directly on the ripa Rheni, but also the restoration of infrastructure, especially bridges, both in the hinterland and directly on the Rhine border can be interpreted as military strategic measures.

Chesters Road Bridge
Ian Longhurst,

The Romans originally carried Hadrian’s Wall across the North Tyne at Chesters on a pedestrian bridge. Later the pedestrian bridge was demolished and replaced with a road bridge. Beyond its existence, nothing about this road bridge is correctly understood. M.J.T. Lewis rightly suggested that the archaeology at Chesters was of a river wharf. (“Roman navigation in northern England? A second look” Journal of the Railway and Canal Historical Society vol. 32 (1995) 417-422). The wharf has instead been misinterpreted as a grossly oversized bridge abutment.

    1. No one leaves a road bridge exactly on the military frontier. An outwork, most likely along the line of General Wade’s road, had to protect the bridge.
    2. The stone riverside quay had a wheel stop fitted to the edge to allow carts to be loaded directly from barges in the river.
    3. The road bridge serviced resupply convoys whenever the army invaded Scotland and was built for the move to the Antonine Wall. Routing supplies for the army in Scotland through Chesters avoided a 150 metre hill on Dere Street and made additional fodder resources available.
    4. Later, the quay was extended and a water course constructed at the back of the quay, to combat the silting up of the quay.
    5. In a probable late Roman phase, when the quay was entirely silted up, a timber quay structure was probably built in front of the stone quay, presumably for another invasion of Scotland.
    6. Sick and wounded soldiers were brought back on the empty supply carts, suggesting a hospital function at Chesters.
    7. The bridge was the probable head of navigation. Developing the Tyne for navigation as far as Chesters under Antoninus Pius opens a question on river navigation on the Neckar in his reign.

Securing the transport of timber on water
Ronald Visser, Saxion University of Applied Sciences

Wood is one of the most important resources, both in the present world as in the past. It is one of the most sustainable and strongest building materials. The use of wood is versatile and it is not surprising that many structures along the Limes were made of wood or timber. Wood was needed for nearly everything in the Roman world, not only the ships that travelled along the rivers, the carts that moved along the roads, the pegs to secure tents, but also tools and weapons were made (partly) of this material. Various attempts were made to calculate an estimate of the amount of wood needed in the Roman period. For the Lower Rhine region, some studies state that the local landscape could have provided sufficient (fuel) wood and timber to sustain the local population, others debate this conclusion. In addition, there is strong evidence that timber was also used that must have been transported over hundred(s) of kilometers. The question remains how this wood was transported. The most logical route would have been over water ways. This paper will present and discuss evidence for the transport of wood and timber over water. The use of, for example, rafts or ships will have had consequences for security along the rivers. What methods or means were used? And how has this impacted the security?

1st C Scotland
Andrew Tibbs, University of Durham

Roman fortifications, in 1st century Scotland, are almost always positioned in locations which enable them to exert control of an area, usually through visible command of valleys where the indigenous population move through; on the coast, the fortifications are usually located at river mouths, restricting access to those military installations located upriver. This suggests a relationship between securing new territories, controlling indigenous populations, and the use of the coastal and river networks by the military. However, there have been few studies which have considered the role and usage of Scotland’s coastal waters and rivers by the Roman military, during any of the various incursions into north Britain. But there is a growing body of both archaeological and circumstantial evidence which suggests that Roman ships were active in Scottish waters, and may have had a pivotal role in the invasion, occupation, and securing of north Britain by Imperial forces. This paper looks at the potential for the coastal and inland waterways to have played a significant role in the 1st century invasion of Scotland, including the potential interconnectivity between the fortifications located on the same river networks, the capacity for the coast and river networks to have been used as transport corridors, and the extent of control that the military may have had of these, and the impact of this on the indigenous population.

When’s a fleet a fleet? Classes and legions on the water
Christoph Rummel, Roemisch-Germanische Kommission of the DAI

In past Congresses of Frontier studies, Roman Fleets, their ranges of operations and tasks have repeatedly been discussed. Both in Britain and on the Rhine – but also on the Danube and in the Mediterranean – there is ample evidence that legions maintained naval branches. Building on discussions at Newcastle and Viminacium, this paper will collate and analyse data on Roman legions that were evidently involved in maritime activity from Britain, the Rhine and Danube frontier, as well as from other parts of the Empire. On this basis, it will attempt to reconstruct their role in the naval frontier infrastructure of Rome. A key question to be addressed is to what extent the operational spheres of legions and established fleets can even be defined and if so, whether they complemented each other, overlapped or whether legions took on the role of fleets in fields and regions where these did not operate. In this way, the paper seeks to contribute one key aspect towards a better understanding of communication, supply and security in the waterscapes at the edge of the Empire.

Operation Zwammerdam ships: Putting things together on rowing and steering on the Roman Rhine
Yardeni Vorst, Restoration team ‘Zwammerdam ships’

In 2018 a team of dedicated volunteers, carpenters, shipbuilders and archaeologists started to rebuild and reconstruct a river barge found nearly fifty years ago in the Dutch village of Zwammerdam. The ship was one of six Roman-period vessels found in the harbour area of a Rhine-based fortress. It was named ‘Zwammerdam 2’ in order of its discovery and the ship was conserved for future study and display. Since its excavation, the construction was studied on paper as well as in actual parts and onward from 2008 it was recorded digitally piece by piece. Fitting together the ship in 3D helped to solve certain questions and supported the reconstruction of a vessel that was incompletely preserved. However, some aspects of the construction remained unclear. Starting to rebuild the actual ship in 2018 was like putting theory into practice and it forced us to think again about each nail and joint from an interconnected perspective. Sharing our thoughts not just with archaeologists but with our team of historical shipbuilders and generally practical people, benefited the process greatly. We reexamined the evidence for a rowing setup in the bow of the ship and considered the implications for using it effectively, gaining new insights. We also designed an experiment to explore steering and the attachments of a steering oar behind a river barge of Zwammerdam type, based on the archaeological find of a unique Zwammerdam steering oar. In our presentation we will discuss the archaeological evidence found for rowing onboard Zwammerdam 2 and other river barges of similar type. We will also discuss what we think is missing and put out some ideas on steering these vessels, moving up and down the rivers in Roman times.