top of page

Water Newton Roman Fort

The 2 hectare site of the Roman fort at Water Newton is on slightly elevated ground between Durobrivae to the south-east and Ermine Street's crossing point of the River Nene to the north.

NVAT undertook an excavation of the fort in 2012 in conjunction with Middle Nene Archaeological Group (MidNag) and with financial support from English Heritage. The excavation was led by Dr Stephen Upex. Until then the site was known only from aerial photographs. The primary objective was to provide dating evidence of the fort in order to inform debate regarding the advance of Roman troops following the Claudian invasion of AD43.


Aerial photo showing the three defensive ditches cut by the West Gate (right side) and the South gate (top)

The fort is of typical playing-card shape with three defensive ditches on its eastern, southern and western sides, which were the sides not immediately protected by the Nene or the Billing Brook. A partial geophysical survey showed the fort’s ditches, gateways and some internal features.

The Excavation

Trenches were located over the terminal ends of the defensive ditches of the south gate of the fort, and also over the west gate where the military ditches were suspected. It became apparent that layers that corresponded to the military phase of the site were overlain by considerable deposits that represented occupation in the period after the fort had been abandoned. Eventually the military ditches were revealed at both gateways. Each of the ditches was ‘V’ shaped in profile, with the inner ditch being the deepest at 2.1m.

As had been expected from geophysical survey and knowledge from similar forts it was apparent that walls would have been made from turves piled up at the front and back of the rampart and then filled with loose soil. Any structures would have been wooden. 

Finds were pitifully scant with insufficient pottery or other artefacts to give, with absolute confidence, a positive date for the fort’s occupation. However, sufficient material was accumulated to suggest that the fort was occupied at some point between AD 60-65; perhaps significantly in the period after the revolt of Boudicca in AD 60.


West Gate area from the west with a figure in each ditch - and the fort roadway to the right.


West gate area. The military ditches show on the left and right edges of the excavation box. Later 2nd - 4th century ditches can be seen running along and also cutting the later road line. (Photo - Ian Schreiber)

The West gate’s middle ditch with “V” shaped profile and  “ankle breaker” slot 


The military road surface of packed gravel with later limestone road surfaces above

Occupation and Dating

Sparse finds and absence of silting in the ditches led to the conclusion that the fort lasted for a very short period - maybe only months. At the end of its life, perhaps after the establishment of the adjacent Nene crossing, it had been de-commissioned by its partial and deliberate destruction.

The post-military history of the excavation site was far more complex than was expected. The former military road into the fort’s west gate was resurfaced several times over the next 300 years with imported limestone which was packed over the earlier military road surface of packed gravel. The wear on this re-surfaced road was considerable with ruts developing in some areas. This traffic may have been associated with increasing settlement to the northwest of the site, within the modern parishes of Stibbington and Wansford, where we know there were villas and manufacturing sites including potteries.

As well as through traffic there was evidence of a series of buildings. None of these was fully uncovered but the bases of large posts and internal and external flooring surfaces indicated that there were permanent buildings constructed on the site and probably lining the roadway which originally had led into the fort. These buildings dating to the 2-4th centuries could have been structures which developed within the suburban sprawl of the growing town of Durobrivae which lay some 400 m to the east, across the Billing Brook. The numbers of coins that were found, especially those recovered from the west gate, suggested that there may have been some form of market or series of shops on or close to the site.  

Road Realignment

An early north-south route ran close to the Longthorpe and Lynch Farm forts, 5km to the east of Water Newton. These forts have been dated to just after the Roman invasion of AD43. The meandering Nene valley here is wide and prone to flooding making it unreliable in winter as a crossing point. In the context of the Boudican revolt and of Roman expansion to the north a more secure river crossing would have been desirable.


The probable line of the early military road and its later realignment

This new crossing became the long-term route of Ermine Street.  What was probably a wooden bridge on stone abutments was constructed and, south of the river, Roman engineers elevated the road by 1.5 m on an artificially constructed causeway. The Water Newton fort may well have provided temporary accommodation for troops and engineers to survey, construct the new road and build the bridge over the Nene. Once this phase of road and bridge works was completed the troops moved away and the fort was abandoned.


Fort - brown rectangle. Yellow line - Ermine Street. Red line replacement route when the Roman bridge collapsed. Blue line - Great North Road

The collapse of the Roman Bridge across the Nene, which may have been a 6th or 7th century phenomenon, would have necessitated a new means of crossing the Nene. Traffic probably switched to a ford 5km to the north-west (at the site of the later medieval village of Wansford). Initially the route to the ford would have been the road through the old fort. At some later time, perhaps as late as the era of coaching along the Great North Road, a short cut was made which remains the route of the modern A1.  

Water Newton Fort and Durobrivae

Durobrivae had previously been regarded as a vicus settlement outside the neighbouring fort. The 2012 excavation suggests an alternative explanation for the town’s location and development.

The key feature is the river crossing and the route of Ermine street – not the fort which was of very temporary significance. The settlement initially developed close to the fort simply because this was the closest dry land to the bridge. It then extended to the south-east along the line of Ermine Street over the following centuries.

More Information

A more detailed report of the excavation and the points arising is provided in an article originally published by  the CAMUS project at Castor.

bottom of page