Julius Caesar first invaded Britain on 23 August 55 BC. Within a month, he was gone, and although his army – fewer than 10,000 strong – did not campaign beyond east Kent, the invasion caused a sensation back home. By crossing the sea, Caesar had ventured beyond the world known to Romans – what they called the orbis terrarum – and he had brought Britain under the authority of Rome. In less than a year, he would return – and now, Andrew Fitzpatrick reports, the location of his landing point may have been revealed.
Caesar’s invasion of Britain was actually part of a much broader campaign: the nine-yearlong Battle for Gaul that was waged between 58-51 BC. Our fundamental source of information about this conflict is Caesar’s own written account. Roman generals had to send dispatches, called litterae, about their activities to the Roman Senate, and would also write reports, or commentarii. The book we know today as Caesar’s De Bello Gallico is based on these writings. In it, Caesar describes how he campaigned across modern France, Belgium, and west Switzerland, and how he twice crossed what the Romans regarded as major boundaries: the river Rhine, which he presented as the border between the Gauls and the Germans, and the sea that was believed to encircle the Roman world.
One of the reasons that Caesar gave for invading Britain in 55 BC was that the Britons had assisted the Gauls in their wars against Rome in 57 and 56 BC. When Caesar returned to these shores less than a year later in 54 BC, he had a different motivation, though. During the preceding winter Mandubracius, a prince of the British Iron Age tribe the Trinobantes, had fled to him in France and asked for his protection. In response, in 54 BC the Roman army campaigned through Kent, Greater London, Hertfordshire, and Essex before winning a decisive victory that ensured the safety of the Trinobantes. The resulting peace settlement that Caesar enforced, secured by the taking of hostages, required the tribes of south-east England to pay tribute to Rome. It was this accord that made south-east England part of the Roman Empire.
It was traditionally believed that in both 55 and 54 BC Caesar landed at Walmer, 10km north of Dover – a view energetically propounded in 1907 by Thomas Rice Holmes in his book Ancient Britain and the Invasions of Julius Caesar. Despite the widespread acceptance of this view, it has never been supported by archaeological evidence. Recent research efforts have set out to pin down the location of the launch sites more precisely.
NEW ROAD, NEW THEORY
In 2010, the Oxford-Wessex Archaeology joint venture completed a massive series of excavations in south-east Thanet, undertaken in advance of the East Kent Access Road 2 (see CA 266). Among the many important – and sometimes spectacular – finds was a large but unprepossessing ditch at Ebbsfleet. This feature was 5m wide and 2m deep, and the pottery found in it dated it to the 1st century BC. As the route of the new road cut across the ditch in two places, we were able to establish that it enclosed an area at least 500m long north– south. An Iron Age village had stood in the northern part of the site until the ditch was dug, at which time it seemed to have been abandoned. The purpose of the ditch was clear – it was a defence – but its date was a puzzle. Large defensive works were rarely constructed in southern England during the late Iron Age, and there was no evidence for buildings anywhere else in the defended area. What, then, was the enclosure for?
A clue could lie in the site’s location. Today, the Ebbsfleet enclosure is less than a kilometre from the sea, but 2,000 years ago it would have been much closer to the shore. Could it be that the ditch had formed part of the defences of one of Julius Caesar’s bases, and that at least one of his landing sites was further north than previously believed? It was clear that further evidence was needed. Thanks to a grant from the Leverhulme Trust, a new research project has been re-examining the evidence for Caesar’s invasions of Britain.
RETURN TO EBBSFLEET
Our investigations have included further work at Ebbsfleet, with geophysical surveys in 2015-2016 and excavations in 2016-2017. All the fieldwork has been undertaken by volunteers, coordinated by Kent County Council, working alongside professional archaeologists. The first phase of the research proved promising: geophysical survey identified the sites of two possible entrances to the enclosure. Both of them were examined in trial excavations. Due to the greater footfall at points of entry, finds are most likely to be discovered there.
The magnetometry survey also revealed the outline of the western part of the enclosure, though, as it has not been possible to explore the entire site, the full plan of the defences is not yet known – the eastern edge is likely to lie below reclaimed land that is now occupied by a new, and heavily landscaped, golf course. Even so, the defences seem likely to enclose an area of at least 20ha, and they appear to have been dug in individual sections, giving the plan a rather angular look. Wherever a section has been dug across the defence, it has proved to be a standard design – in each of the nine slots the same thing has been found: a ditch measuring 5m wide and 2m deep, with a broad, flat base, and walls that slope at 45º. The Iron Age village, or at least the site of it, also seems to have been deliberately included in the defensive circuit, which then ran along the ridge of the peninsula before enclosing Ebbsfleet Hill, a low rise to the south-west.
Evidence for the possible entrances proved less clear cut, though. Our excavations in 2016 ruled out one altogether – Trench 2 showed that there, the ditch was actually continuous – while in Trench 1 the other possible entrance was found to be covered by much deeper layers of rain-washed material than had been expected, up to 1m deep in places. This meant that the plan of the area could not be established, although traces of a possible trackway with a cobbled surface were found – something that we would explore further when we returned to the site.
WEAPONS AND BONES
In the autumn of 2017, the next phase of investigations confirmed that an entrance did indeed exist in Trench 1. As resources were limited, attention was concentrated on recovering its plan. Details began to emerge: the ditch terminals were squared and the causeway between them was less than 3m wide, while the ditch was a little deeper near the entrance and its slopes were steeper, their angles closer to 65º than 45º. No trace of a gate could be found – while it is possible that one stood outside the area of the excavation, it is equally possible that there was none. As for the track that ran through the entrance, it comprised only a single layer of beach pebbles (other than where it crossed the causeway between the ditches, at which point the metalling was thicker), directly on top of which we found a potin coin (a bronze coin made in a mould using a tin-rich alloy) which confirmed a late Iron Age date for the surface.
The track also yielded several iron objects – badly corroded and difficult to identify with confidence until conservation is complete, but it is clear that at least some weapons, including swords and spearheads, are present. It was not only artefacts that we recovered, either – bones from at least two people were also found on the surface of the track, mostly fragments of skull and intact thigh and arm bones. At least one of the femurs shows signs of a cut that was inflicted by a sharp blade.
It was the area of the site that had been ruled out as a possible entrance in 2016 that produced our most important find to-date, though. Near the bottom of the ditch we found a mix of mid-1st-century BC pottery, human and animal bone, and the tip of an iron weapon. When conserved, the object was revealed to be a Roman spear or pilum.
Further clues to a Roman link to the Ebbsfleet defences came from their shape: the flat-bottomed ditch is very similar in size and shape to the Roman siege works at Alésia, France, which in 52 BC was the site of the decisive – and very bloody – showdown in the Battle for Gaul. This style of ditch was used to defend large areas rather than a fort: the ditches of the irregularly shaped Roman camps at Alésia were V-shaped. Meanwhile, the dimensions and layout of the entrance at Ebbsfleet are also tantalisingly similar to the Roman fort at Hermeskeil, Germany, which is dated to 51 BC. Excavation of the western entrance there also failed to reveal any evidence for a gate.
CAESAR’S TOPOGRAPHIC CLUES
What of Julius Caesar’s account of the invasion? Unhelpfully, he did not say exactly where his army landed in 54 BC, simply stating that it was at the place that they had learnt the previous year was ‘the best place of disembarkation’. But he did provide three important clues about the site. The first is contained in his description of the crossing. So that the army could embark and disembark in daylight, the Roman fleet had set sail in the evening. Caesar’s plan was to sail northwards from France with the tide, while using the wind to help cross the channel westwards. But at around midnight the wind dropped and the tide carried the fleet too far north, so that at sunrise they saw Britain far behind them on the port side.
The second clue is in Caesar’s description of the landing itself. He says that the Britons had gathered to oppose the disembarking army, just as they had in 55 BC, but, taken aback by the size of the Roman fleet (numbering 800 ships), the Britons withdrew ‘and concealed themselves on the high ground’. As for the location of the landing site, Caesar gives us hints in two separate places, describing it as the best place to land and with a soft or ‘sandy, open shore.’ All of these clues point to Pegwell Bay.
Pegwell Bay is large, flat, and sandy, the biggest bay of this type on the east Kent coast. Its north end, at Ramsgate, is formed of chalk cliffs that rise up to higher ground, while the southern side is delineated by the mouth of the Wantsum Channel where the River Stour flows into the sea. Once a wide marshy area, the Wantsum Channel was systematically reclaimed in the medieval period, and today is an area of flat, low-lying land, into which the finger of higher ground that is home to the defended Ebbsfleet site projects. Today, the site is about is 1km from Pegwell Bay, but in the 13th century AD the sea was only 500m away (a medieval sea defence of this date survives), and it seems likely that in the 1st century BC the eastern side of the Ebbsfleet enclosure lay next to the sea.
The curvature of the earth means that a sailor in a small boat some way out to sea can only see high ground – the kind of land that Caesar must have been describing when he said that ‘at sunrise he sighted Britain left afar on the port side’. The land to the south of Thanet is low-lying, while to the north is the Thames Estuary and, beyond that, the low-lying coastline of Essex. The only high ground that would have been visible to Julius Caesar would have been the cliffs near Ramsgate.
Caesar’s description of an open shore is consistent with Pegwell Bay, and with 800 ships and an army of perhaps 20,000 men and at least 2,000 horses to land in a single day, it is the strongest candidate for ‘the best place of disembarkation.’ Moreover, the higher lands of Thanet are the only high ground in the area to which the Britons could withdraw, as Caesar reported. Taken together, the
A STORM BREAKS
The Roman fleet landed on 5 July 54 BC. Leaving a force equivalent to a legion to build the base and watch over the ships, Caesar records that at first light he marched the rest of the army to attack the Britons. Their intended victims took refuge in a hillfort, but this was soon overrun – almost certainly the fort of Bigberry, just west of Canterbury, which is the only hillfort known in Kent that could have been reached after an overnight march.
It was a resounding victory, but the Romans’ sense of triumph was not to last. No sooner was the battle won than Caesar received news that the fleet had been badly damaged in a storm. The galleys are likely to have been beached on arrival, but the transports were still riding at anchor when the ferocious weather caused their anchors to break loose, and the vessels collided with each other before they were cast onto the shore. Caesar writes that he returned to the coast immediately, where he decided to beach all the ships and protect them by building a single line of defence that connected the ships with the camp. The single ditch found at Ebbsfleet is consistent with it being this defence. It would have been a defence that, as at Alésia, was used to protect a large area – in this case the one occupied by the 800 ships that had been hauled up onto the shore.
The repair of the fleet was imperative because Caesar always intended to return to France at the end of the fighting season. The army in Britain had no supplies of food for the winter, and if it was unable to return to the Continent it could have been surrounded and starved into submission by the Britons. Nor were the legions left in France strong enough to prevent an uprising there by themselves. The army worked for ten days and nights to haul all the ships ashore and to build the defences. With this completed, and the base ‘strongly fortified’, Caesar left the original garrison and all the skilled workmen in the legions to repair the ships while he led his troops back out into Kent to restart the advance. The main army would only return to sail back to France once they had achieved victory and the peace settlement that bound Britain to the Roman Empire. Over 2,000 years later, archaeological traces of these historic events are now beginning to emerge from the ground.
This feature appeared in CA 337.