Archaeology of Roman Rochester
Archaeologst Alan Ward introduces the Iron Age and Roman archaeology of Rochester. Featured in The Hidden Treasures, Fresh Expressions Project Archaeology Report, Keevill Heritage 2021.
Location and geology
Rochester was established at a crossing point over the River Medway (Figures 1 and 2; Plate2). Thiscrossing point was in use from at least the Late Iron Age, but was, probably, at that date by ferry ratherthan a ford. Even with a lower water table the depth, tides and currents would have been too greatto cross by foot or animal. Supposedly the nearest ford was four miles further to the south at LowerHalling, although one at Upnor has been suggested (Thornhill 1976, 126).
The Geological Survey of Great Britainshows that theCathedralsiteoccupiesbedrock of theLewesNodularChalkFormation.1 No superficial (drift) geology is noted, butin and around Rochester to thesouth of the High Street, the chalk bedrock is usually covered either by Brickearth, sand or sandy gravel,or‘brash’. The latter material is a buff-coloured eroded mix of sand, grit, chalk and flints,forming a solifluction deposit.2 Alluvial clay predominates to the north of the High Street within thearea of the Roman town walls, but because of its depth (c2.5m below ground near the town wall) it is rarely seen.
Brickearth was the only geological deposit seen in most of the excavations, but sandy gravel and flints in sandy clay were observed under this in the new cloister garth soakaway. Somewhere between the south cloister walkway and the road to its south known as The Precincts, there is a geological break between the Brickearth and sandy gravel/sand deposits. The latter are over 1m thick at the west end of the road where it meets St Margaret’s Street (Ward 2004a) and presumably directly overlie the solid chalk bedrock. The King’s Orchard Ditch in which the King’s School Assembly Hall was constructed c 1965 had been dug through gravel (Harrison and Flight 1968, 74-75; Flight and Harrison 1986, 15). The break in the geology takes place more or less along an east-west line just to the south of, or under, Garth House (built on the site of the monastic refectory), with the sandy gravel over solid chalk to the north and brickearth in the cloister to the south. A deposit of gravel of unknown thickness lies between them in a conjectural buried valley. This would have been created by water, ice or both, perhaps cutting through the brickearth after that was formed c 60,000 years ago or earlier. Chalk bedrock was seen by James T Irvine below the south wall of the cathedral in 1874 (Figure 4).
The High Street is orientated from west to east. It might be expected that the cathedral would be laid out parallel to the High Street (or that a more accurate east-west alignment might have been used). Instead it is parallel to the Roman town wall at a distance of c 40m).
This report is not the place to enter into a detailed description or discussion of the overall history or archaeology of Rochester, but some background knowledge of a settlement is always advisable prior to embarking on a study of one specific structure or theme within an area, whether that area be rural or urban. Deep archaeological stratigraphy extending back to the Late Iron Age exists in Rochester, but there are considerable problems in understanding that stratigraphy. This is due to a number of factors. These include the restricted nature of many (but not all) of the excavations, excavators’ often inadequate comprehension of the urban context within which individual sites were set, provision of limited or no location plans or section drawings, and a common failure to complete and publish. The following sections summarise current knowledge of the site’s development on this basis.
The town
Iron Age: Despite claims implying the presence of a tribal centre or even an oppidum at Rochester (Cunliffe 1982, 43; Philp 2014, 111), little has been found that can justify the claim (Figure 7). Ten fragments of a coin mould and some potin coins (Chaplain 1962; 1969: Harrison 1991) and, apparently, a gold stater (Philp 2014, 106) were found on or adjacent 50-54 High Street during excavations in 1961-2. Around 1200 sherds of Iron Age pottery were found on the same site, but unfortunately this material does not appear to have been the subject of specialist analysis. The site also contained the town’s only definite Iron Age feature: a 1m-wide and 1m-deep ditch exposed for a length of c.6.50m (Harrison 1991, 42, figure 1). Philp (2014, 113 and figure 1) shows another ditch surviving for a length of c 2m at right angles to the south.
Ward (2004a, 23) suggested that a large ditch at St Margaret’s Mews (between Boley Hill and Love Lane) that was in use in the medieval period was of Iron Age origin. Excavation apparently produced no Iron Age pottery (Dulley 1960, 197-8). This and the sheer size of the feature makes a medieval origin is much more likely.
There is little doubt that an Iron Age settlement existed at Rochester, but its extent and date-range are unknown. No buildings have yet been found, despite implications to the contrary (Chaplain 1962, l-li) no buildings have been seen (Harrison 1991, 45: Philp 2014, 116). Proto-urban status cannot be assumed on this basis.
Hardly surprisingly we have more knowledge of Roman Rochester (Figure 7), but it is still not possible to elucidate much of its plan or historical urban development. Its Roman name, Durobrivae, is known (Burnham and Wacher 1990, 76; Brooks 1994, 3); the name means ‘the fort by the bridges’ (Rivett and Smith 1981, 346). That meaning alone creates problems. Where were the fort and bridges, and given that only bridge is known historically, why the plural? The town is first mentioned in the early 3rd -century Antonine Itinerary (Rivett and Smith 1981, 346; Ward 1997a, 199), so duro could derive from its earthen defences dating to the last quarter of the 2 nd century or the early 3 rd -century masonry town wall. If that is accepted, however, what was the name of the settlement up to c AD 200? That may never be known. The ‘bridges’ element is easier to explain: it probably refers to the separate sections or spans between the piers of the bridge, literally a ‘bridge’ between two points which, when looked at as a whole, become ‘bridges’ (Rivet and Smith 1981, 349; Brooks 1994, 3).
Some writers (Harrison 1968, 76; Harrison and Williams 1979, 21) suggest that the town wall dates to the early third 3 rd rather than later, but the date is not secure (Detsicas 1987, 57), and the circuit may not be all of one period. At least one tower was present, at the north-west angle (Payne 1905, lxvi), which may have been a later addition (Ward 2011). The finding of the South Gate was mentioned by Payne (1895, 6) in his article on Roman Rochester, but apparently he did not see it himself. It was uncovered again in 1998 and was regarded as being contemporary with the earthwork defences (Ward 2002). The cobbles marking its position on Boley Hill represent only part of its length: the wall would probably have extended southwards for at least another 1.5m but were cut back, more probably in the Norman than in the Roman period.
The modern High Street more or less follows the Roman east-west route through the town. At East Gate3 it is probably on the same line, but it moves slightly to the north once it nears the bridge and, presumably, a West Gate. The suggestion that it was a dual carriageway for part of its route (Chaplain 1962, l-li: Philp 2014, 116) is incorrect. Buildings were present on both sides of this supposed road widening (Ward 2017). There is no reason to believe that a dual carriageway preceded any buildings later built upon its supposed course. This widening would form a gravelled open area between buildings and the road leading southwards from Watling Street (Figure 7).
The main north-south route is usually depicted as being on the line of the Boley Hill roadway (Figure 2) and the 19th -century King’s Head Lane and Northgate (correctly Pump Lane). No Roman road was found in Northgate when service trenches were dug in 2009, however, and a Roman building exists between Chertsey's Gate and the King’s Head Hotel (Ward in prep a). The north-south Roman street has to be that shown on the 1717 Bridge Warden’s Map of Rochester to the west of the King’s Head, not as now to the east. The latter route was constructed some time between 1717-72 (Figures 8 and 9). What is known in medieval documents as Cheldergate Lane almost certainly continued the line north across Watling Street to the Roman North Gate, called the Great Gate in the Anglo-Saxon period (Ward 1949; Ward 2005; 2011a; 2017). Cheldergate Lane was replaced probably c 1225 by Northgate, formerly Pump Lane.
Cemeteries are known to south and east of the town defences. The former, on Boley Hill, appears to have been largely for cremations, but one lead coffin is also known to have come from the area (Wheatley 1927, 159-164). Boley Hill has seen much disturbance from buildings and landscaping. No doubt some burials still remain in situ, but locating them will be difficult and probably a matter of chance. The cemetery outside the East Gate is also poorly represented. The area where inhumations are known was used for Brickearth quarrying (probably to make the bricks for Restoration House) in the late 16th century (Ward and Anderson 2003). A burial to the south of the East Gate observed in 1905 may also be part of this inhumation cemetery (Payne 1909, xc). Two burials are known from inside the town wall (Harrison 1981, 101; Cool 1981, 125-131; Gollop 2009). The presence of family burial plots seems likely, for the Roman law in relation to burial outside Roman towns does not seem to have been systematically applied to the smaller urban settlements (Esmonde Cleary 2000, 129). A burial c 3.4m below ground was also found in the basement of the former cathedral registry office on College Green, opposite the south-west corner of the cathedral (Payne 1902, lix-lx). This burial may be medieval, but its depth perhaps hints at a Roman date (although it does not fit easily with either of the two cemeteries just mentioned).
As far as buildings are concerned one probable and eleven definite masonry buildings have been observed to date. Another three possible or conjectured masonry structures may also exist and no doubt others remain to be found. Four timber buildings, one on dwarf walls, are also known. It is very likely that many more timber buildings would have existed (see Appendix 1). The known masonry and timber structures mostly lie to the west of the Cathedral and Northgate; none have been found in the north-east quadrant of the town. The buildings are only dateable to the Roman period as a whole, and no specific functions can be identified. Some at least would be domestic properties. No civic buildings can be identified at the moment.
Possible Roman structures had been identified at the Cathedral before HTFE. In 1898 Hope suggested that masonry found below the nave south wall was an apsed Anglo-Saxon church (Figure 10). This structure had been found and recorded by Irvine in 1876 during underpinning work (Figure 4). Irvine suggested that the remains were Roman in his notes. They are shown on Irvine’s 1876 sketch plan, along with a short length of straight wall to the west, passing below the nave wall. 4 Despite what Hope shows, the latter was not seen internally. Furthermore no definite walls of this building were seen in an excavation to the south of the nave in 1937 (Figure 11; Cobb 1938). There does not appear to be much (if any) evidence for Hope’s ‘Anglo-Saxon’ apse: the building is at least as likely to be of Roman date. Livett observed another Roman structure (possibly part of the same building) inside the nave while underpinning the west door. This is not shown on his plan, but is stated in his text (Livett 1889, 266-7). Externally, and cut by the foundations of the Anglo-Saxon church, Livett found the corner of a Roman wall and its foundation (Figures 12 and 13). There was c 0.9m of soil between this masonry and the Anglo-Saxon church: it is therefore most likely that the lower one was Roman. The levels of the masonry found by Irvine and Livett are more or less the same and the materials observed in their construction are typical Roman. Another Roman building may exist within the Sextry (or Deanery) Gate garden. In 1973-4 a mortar floor was found 5ft (c.1.55m) below the modern ground surface (Hayes 1974), at about 7.60m aOD: Roman floor levels along the High Street are found at c 7.15m aOD. There has to be an element of doubt in regard the date of the floor found in the Sextry Gate garden trench.
Explore the studies, conservation and excavations revealing the sites 1,400 year history.