The first trench measured 3.5m long x 1m wide and extended north from the wall of the church. It was excavated to a depth of 1m, then hand-augered a further 0.32m
down to the top of the natural substrate. The top of the church wall’s concrete foundations were found 0.55m below the current ground level, filling the only cut feature in the trench: the foundation cut for the church wall.

The vertical cut followed the line of the wall but protruded out at right angles from it along the trench 1.5m to the north. It was filled at its base by concrete, then by a dark sandy silt containing mid-2nd–3rd century Roman pottery. Its base had been made upon an earlier, loose rubble layer consisting of <50mm thick local ‘Lower Lias’ mud/lime-stone slabs.

The general stratigraphic sequence comprised mostly soft, dark sandy silts with diffuse horizons between contexts, and light rooting throughout. Except for the rubble layer, very few stone inclusions were encountered in any of these, highly bioturbated, dark earths.

Below the rubble layer were 2nd century and earlier layers. Pottery sherds found here included Parisian ware and grey wares, along with animal bones (cattle and deer) and a small fragment of tile. The overlying rubble layer was relatively loose, well-sorted and found in the southern 0.8m of the trench. It was poorly consolidated with little soil matrix and no datable remains.

Stratigraphically above the rubble were four successive accumulations of dark earth, each measuring 0.15-0.30m thick with very diffuse horizons between contexts. Small groups of late 3rd and 4th century pottery were recovered from each layer. This included mostly grey wares but also sherds from samian vessels, Dales ware, Swanpool wares and other colour-coated fine wares possibly produced in Lincoln.

A range of species was represented in the recovered animal bone assemblage: cattle, deer, pig sheep and dog/fox were all present. Several pieces displayed signs of butchery; cut and chop marks on some, with three long bone fragments exhibiting evidence of crafting. One bone/antler pin was recovered; the spherical head a common style adopted from the 2nd century into the post-Roman period.

The second trench measured 5m long x 1m wide and was aligned northeast to southwest, positioned over the location of a proposed septic tank. It was excavated to a depth of 0.8m at its northeast end – where excavation came down onto the top of a loose stone surface – and 1m at the southwest end. The trench was hand-augered at both ends to record the full stratigraphic sequence down to the top of the natural substrate, 1.4m below the current ground level. A small, 0.5m wide sondage was excavated against the stone surface to fully characterise the remains below.

The earliest archaeological layers found were from the auger cores taken at either end of the trench. They identified at least two stratified context units overlying the natural substrate. Both remain undated but presumed Roman based on the datable 3rd century pottery found stratigraphically above in dark earth layers. This included mostly grey wares, with colour-coated fine wares, Derbyshire ware and decorated samian ware all present. A small quantity of chop-marked animal bone (cattle) was also recovered.

The small area of stone flooring comprised a single thin (0.04m thick) layer of flat mud/lime-stones, overlying a layer of gravels. This had evidently been a levelling layer upon which the floor had been laid, although extensive root damage had deconsolidated what had originally been a compacted layer. A group of AD150-300 pottery was recovered from the stoney layer. It contained sherds from a Mancetter-Hartshill mortarium, an amphora, colour-coated beakers, and samian. Coarse wares were also represented, including a grey wares, Black Burnished ware and Derbyshire ware.

Cut and chop-marked animal bone (cattle, sheep and deer) was also present, as was a fragment of Roman tile (imbrex).

A thin layer of crumbling burnt dark grey mud/lime-stones was recorded in the trench section. It was stratigraphically contemporary with the floor surface but without the gravel underlay. Evidence for historic tree rooting was found to have been very destructive to buried structural remains.

Stratigraphically above the stone floor were two layers of dark earth, each roughly 0.30m thick with a very diffuse horizon between them. Groups of 3rd and 4th century pottery were recovered from each layer, along with several fragments of Roman brick, tile (imbrex and tegula) and a coarse ceramic tessera. The pottery included sherds from a paint decorated Mancetter-Hartshill hammerhead mortarium, colour-coated beakers, grey ware bowls and jars, Black Burnished ware, Derbyshire ware and a tazza possibly from the Swanpool industry in Lincoln.

The third trench measured 5m long x 1m wide and was aligned northeast to southwest, positioned over the proposed car parking area. It was excavated to a depth of 0.8m at its southwest end where a 1m wide sondage was dug and then augered to establish the full stratigraphic sequence down to the top of the natural substrate. The trench was hand-excavated to a depth of 0.5m across much of the rest of the trench. Augering at that depth confirmed a common stratigraphic sequence across the area. Sand and gravel substrate was found 1.6m below the current ground level.

Artefacts were not found during augering but small quantities of abraded finds were recovered during hand-excavation. Finds from the third trench had evidently been exposed to differing depositional conditions in comparison to those from the first two trenches.

Both the earliest and latest (topsoil) datable layers yielded similar groups of mid- to late 2nd century samian, fine colour-coated and grey ware pottery. Late Roman dark colour-coated fine ware was found below the topsoil in the upper dark earth layer. Butchered animal bone (cattle, sheep, pig and deer) was also found in low quantities.

The fourth and last trench measured 5m long x 1m wide and was aligned northeast to southwest, positioned between the Fosse Road and the churchboundary wall just south of the pedestrian gate. It was excavated to a depth of 1m at its southeast end and 0.45m next to the boundary wall. Two cut archaeological features were recorded: a foundation trench for a red brick wall, and the cut of a 3m wide ditch outside the boundary wall. Both features were originally constructed in the late 19th century likely around the time the church was built.

The boundary ditch and wall foundation trench were cut into a firm sandy gravel levelling layer that sloped down from the southeast. This overlay an earlier made ground layer that was lying above sand and gravel interpretated as natural substrate. There was no indication that any remains encountered in this trench predate the late 19th century.

The overlying stratigraphic sequence comprised ground layers associated with the raising of the road or filling of the ditch. Indeed, likely a combination of the two – in the later 20th century. At the northwest end of the trench the boundary ditch was filled by dark soil and rubble deposits that also yielded 20th century plastics and ceramics. Excavation was then ceased at this level because development will not impact beyond this depth.

Chris Casswell, 2024, Church of St Stephen, Brough, Nottinghamshire: Archaeological Evaluation (Unpublished document). SNT5872.