Monument record 8204/1 - Excavated Remains of Roman North Gate

<1> Earthworks Archaeological Services, 2018, Northgate Steps, City Walls, Chester: An Archaeological Watching Brief, R4271 (Client Report). SCH8626.

An archaeological watching brief was undertaken between June 2016 and April 2018 during the dismantling of the North Gate Steps, City Walls, Chester (centring on NGR SJ 40400 66668). Removal of the steps exposed masonry of the inner face of the City Wall and also structural remains relating to the defences of the Roman legionary fortress. In addition to a section of the fort's rampart (see 8201/9), two parallel masonry walls, each aligned north-south, were exposed. They are thought to represent the remains of the east and west walls of a tower representing the Roman north gate (porta decumana). It is suggested that, like the Roman east gate at Chester, the north gate could have consisted of two gate towers (the remains found representing the west tower), with two carriage ways imbetween separated by a spina. Additionally, a large semi-circular pit cut into the bedrock, and potentially contemporary with the rampart, may represent a posthole associated with an earlier, timber-built gate. If so, significantly, this is the first evidence of any timber gate structure found at Chester. The construction of the steps to the City Wall in the nineteenth century removed some elements of both the Roman rampart and masonry gate tower; in the case of the latter, this clearly resulted in the stepped appearance of the remains encountered.

The post hole was located approximately 2.5m from the east terminus of the exposed stretch of rampart material. It was sub-circular in shape, measuring up to 1.1m in diameter and cut to a depth of 0.55m into the surrounding bedrock. The sides of this feature were steep and the base flat. The posthole contained clear traces of grey clay-silt around the edge and a loose fill of crushed sandstone. A single, large block of sandstone had been deposited across the top of the fill, perhaps to seal and level off the feature. The location and scale of the large rock-cut posthole – of similar size to those of the east and west gateways of the extensively excavated legionary fortress at Inchtuthil, Scotland – suggests that this potentially belonged to a timber gate tower, which was later replaced by the masonry tower. Based on the Inchtuthil model, the post-hole could have belonged to a square, nine-post timber gate tower, likely contemporary with the earth rampart of the primary Roman defences.

The remains of the eastern wall of the masonry tower, which potentially replaced the timber structure, consisted of three in-situ courses of block-in-course masonry. The large blocks of red sandstone used in the construction of this wall measured between 0.3m and 0.45m thick, up to 0.9m wide and in excess of 0.6m long. The thickest blocks belonged to the foundation course and were set within a trench of corresponding depth cut into the sandstone bedrock. On the west side of the wall, the second course of masonry oversailed the foundation course and the edge of the trench by approximately 0.3m. Overall, the wall measured in excess of 2.0m wide, 1.1m in height from the underside of the foundation course, and extended northwards beneath the later post medieval City Wall. Damage was noted to the south faces of the blocks, the surviving foundation course having been cut back considerably. Indeed, south of the in-situ remains and where the foundation course should have extended beneath the kerb of Water Tower Street, the wall had been totally removed. This cutting back and removal of the remains resulted in the artificially stepped appearance of the wall, rising from south to north. A thin bedding of lime mortar was noted both on the upper surfaces of some of the blocks and on the exposed surface of the bedrock where the foundations had been removed.

The west wall of the tower lay 2.5m to the west of the east wall, the remains surviving at an even shallower depth below current ground level than those of the east wall. Five courses to the west wall survived in situ and this too contained blocks of squared sandstone similar in character to those encountered in the east wall. However, here the largest block observed, the foundation block, measured 2.3m wide and some of those in the upper courses measured up to 1.2m wide. The fourth and fifth courses were offset by 0.15m from the third course on the west side and by 0.30m on the east side. Thus the wall narrowed from 2.3m at the base to 1.6m in the uppermost courses; it stood to a height of 1.6m. As with the east wall, thin deposits of lime mortar were noted and, here too, damage had occurred to the faces of some of the blocks, resulting once again in the remains exhibiting an artificially stepped appearance from south to north.

Measured from the surviving outer face of each wall, the foundations as encountered measured approximately 6.7m across the external faces and 2.5m between the internal faces at their narrowest point. If the missing fourth and fifth courses of the east wall were offset like the lower courses of the west wall, then the overall dimensions of the structure across these upper courses could be reasonably estimated at 6.4m externally and 3.3m internally. The internal offset may have supported a raised timber floor at, or close to, the contemporary ground level, the undulating bedrock beneath having not been modified to serve this function.

Immediately to the west of the west masonry wall, the earlier fort rampart had been truncated by a trench infilled with red sand with small fragments of sandstone. This trench represents the point where the earth rampart of the primary defences was cut back to accommodate the north gate masonry tower. A trench filled with sandstone rubble, with larger, squared blocks to the north edge of the trench, was recorded towards the west end of the site. This could represent the remains of a Roman revetment wall. A single sherd of black-burnished ware dating to the mid second century was recovered from the lower rubble deposit, along with some Roman ceramic building material.

<2> Dodd L., 2018, Discoveries at the Northgate Steps, Chester, 2016: The North Gate of the Roman Fortress, Vol 88, 2018, p.55-64 (Article in Journal). SCH8630.

This published account of the archaeological work undertaken between 2016 and 2018, presents a further discussion of the remains encountered and their interpretation as the Roman North Gate. Evidence for the other gates at Chester, in particular the East Gate, is discussed. The proposed reconstruction of the two phases of the Gate (timber and masonry) is presented and discussed.

<3> Earthworks Archaeological Services, 2015, An Archaeological Assessment of Proposed Remedial Works at King Charles Tower Steps and North Gate Access Steps, City Walls, Chester, R3836 (Client Report). SCH7965.

An earlier phase of the monitoring and investigation work of the North Gate steps and City Wall occurred in 2015, when a trial pit was excavated at the east end of the steps structure. It appears that this earlier phase of work also idenitified masonry relating to the Roman defences and Roman phase north gate. The trench identified that the stone steps were founded on a single course of sandstone foundation, the trench for which had been cut into the rubble backfill of a potential subterranean feature. At the northern end of the trial pit, a masonry structure was encountered beneath the basal course of the City Wall. This structure and the rubble backfill are thought to relate to the medieval phase of the Northgate (see 12000/12 for further detail). Significantly, this masonry structure had been mortared to an earlier masonry structure employing large, squared blocks of sandstone, the size and character of these blocks are indicative of a Roman date.

<4> Earthworks Archaeological Services, 2020, Northgate Steps, City Walls, Chester: Further Archaeological Watching Brief, R4440 (Client Report). SCH8885.

Following on from the archaeological monitoring work undertaken between 2016 and 2018 (sources 1 and 2), a further watching brief occurred between September 2019 and April 2020, prior to the rebuilding of the Steps and City Wall.

Full excavation of the previously identified large rock-cut posthole demonstrated that this could have accommodated a squared timber post perhaps as large as 0.60m square, twice that of the standard 0.30m square post size considered standard for the east and west gateways of the extensively excavated legionary fortress at Inchtuthil, Scotland. One side of the posthole was sloped to allow insertion and placement of the timber post prior to erection, the postholes encountered at Inchtuthil were similarly prepared.

Also, to the west of the North Gate, the evidence encountered during the excavation of stone rubble core material beneath the City Wall has demonstrated that the upper deposit of rubble is of post-medieval date, below this the rubble core is Roman in date. Furthermore, the stratigraphic level at which the these deposits meet coincides with the level of the large and in situ blocks of masonry interpreted as belonging to the Roman revetment wall of the masonry gate tower.