GLOUCESTERSHIRE

(1) Brockworth, Phase 5, Perrybrook (SO 89432 17186): an excavation identified Roman ditches dating to the second to fourth century a.d. which formed part of a co-axial field system and were contemporary with several broad, shallow pits. Finds from these features suggest proximity to a settlement. A notable discovery was that of a probably dispersed hoard of 40 copper-alloy coins, including several mid- to late fourth-century issues. Three Roman inhumation graves were found; one was associated with finds of the mid-first to early second century a.d. and another with material of the mid-third to fourth century.Footnote ⁹⁴

(2) Cirencester, Waterloo carpark (SP 02644 02057): an evaluation recorded a wall and two limestone surfaces which corroborate the findings of a 1998 evaluation and suggest that at least one Roman building was present.Footnote ⁹⁵ Deposits associated with possible land reclamation during the later Roman period were also identified.Footnote ⁹⁶

(3) English Bicknor: a hoard of 1,486 radiates was found. The majority of the coins were recovered during a controlled excavation at the base of an earth and rubble boundary wall of unknown date. The date range of the coins extends from the joint reigns of Valerian and Gallienus (a.d. 253–60) through to those of Tetricus I and II (a.d. 271–74). Over half of the total number of radiates discovered fall within the Tetrarchic reign: 532 of Tetricus I (a.d. 271–74) and 257 of Tetricus II (a.d. 273–74). Deposited c. a.d. 272–74 or later.Footnote ⁹⁷

(4) Lydney, Archers Walk, B North Phase II (SO 64360 03614): an excavation identified at least two phases of Roman activity. The first phase included drainage and enclosure ditches, pits, post-holes and a possible crop-processing oven, all dating to the first to second century a.d. Later features included a late second- to fourth-century agricultural or industrial complex with associated domestic activity, including four masonry structures, a well, ditches and pits. A cremation grave also dates to this period. Evidence for terracing of the land prior to construction of the buildings was identified, and field-boundary ditches, trackways and isolated refuse pits were also present.Footnote ⁹⁸

(5) Stroud, Cotswold Canal, The Missing Mile (SO 78011 06061; SO 77347 07069): an archaeological evaluation of two parcels of land was undertaken. The northern parcel of land was evaluated in 2019 and the southern in 2020. The 2019 evaluation found evidence for mid- to late Roman settlement, concentrated on a gravel spur which formed an area of high ground in the north-eastern part of the site. The evidence took the form of ditches, pits and the possible remnants of a wall.Footnote ⁹⁹

SOUTH GLOUCESTERSHIRE

(1) Chipping Sodbury, Land west of Trinity Lane (ST 73041 82818): an evaluation identified a Roman rural settlement comprising several ditched enclosures, a possible droveway and various pits, post-holes and other features. The finds assemblage indicates an early Roman phase preceding a more substantial later Roman phase.Footnote ¹⁰⁰

BATH AND NORTH-EAST SOMERSET

(1) Midsomer Norton, land west of Silver Street (ST 66267 53146): an excavation identified a Late Iron Age to early Roman square enclosure that was reworked several times over the course of the first to fourth century a.d. The layout of the enclosure sequence and presence of domestic-related activity is suggestive of an enclosed farmstead. Possible structures were identified in the interior of the later enclosures, with the artefactual assemblage suggesting an intensification of domestic activity from the second century onwards. Evidence of crop-processing activities and possible metal-working was also identified.Footnote ¹⁰¹

DEVON

(1) Exeter, St Sidwell's Point/Exeter bus station (SX 92551 92877): an excavation uncovered the defences of a previously unknown enclosure 500 m north-east of the legionary fortress and early town. The defences comprised two parallel north-west–south-east-orientated ditches of substantial size and depth (fig. 19). The easternmost, outer ditch (Ditch B) had a pronounced V-shaped profile with very steep, straight sides and a narrow, rounded base. The inner ditch (Ditch A) was larger and deeper with an asymmetric ‘Punic’ profile and a steeper outer slope. The sides sloped to a sharply tapered base. Ditch A contained several fills suggestive of a bank on the internal south-western side. Mid-first- to second-century a.d. pottery was retrieved from several mid- and upper fills in Ditch A, along with part of what may be an early Roman copper-alloy armlet. A substantial pottery assemblage was recovered from Ditch B and is largely consistent with filling from the later first century a.d. onward. In both ditches, small amounts of later material accumulated in shallow depressions at the top of the features. Two third- or fourth-century land-boundary ditches to the east of, and truncating, the earlier defences (Ditches C and D) are likely to be broadly contemporary with this final infilling. The nature of the enclosure defined by the defences is uncertain, as is whether it was contemporary with the occupation of the legionary fortress (c. a.d. 55–75) or, as seems more likely from the dating evidence, should be associated with the early town. The defences are unlikely to be associated with the probable fort at Princesshay outside the north-eastern defences of the legionary fortress.Footnote ¹⁰²

FIG. 19.

St Sidwell's Point/Exeter bus station: site plan. (© Cotswold Archaeology)

(2) North Tawton, land south of The Barton (SX 6616 9997): extensive geophysical survey over 35 ha has offered a greater understanding of the Roman complex first mapped from the air as earthworks and cropmarks.Footnote ¹⁰³ This was subsequently extended by analysis of LiDAR data by Bryn Gethin, Hugh Toller and Chris Smart.Footnote ¹⁰⁴ The south-western corner of a postulated legionary marching camp has been located, allowing for the proposal that the total enclosed area covered c. 12.5 ha. To the south and bisected by the former Okehampton to Exeter rail line, a substantial rectilinear ditched enclosure, measuring 610 by 450 m and with an internal area of c. 27.5 ha, has been confirmed. The known Roman road (Margary 492a) that ran west from Exeter passes exactly midway through this enclosure.Footnote ¹⁰⁵ The geophysical survey shows evidence for extensive roadside buildings on either side of this thoroughfare and, importantly, a T-junction with an equivalent occupied road branching north from the centre point of the enclosure. Other occupied streets are evident, with at least two set out in a regular fashion from the principal roads. Work is ongoing, but this new evidence hints at significant complexity and the possibility that North Tawton developed into a substantial civilian settlement west of Exeter.Footnote ¹⁰⁶

(3) Sampford Courtenay (a) Rowden Moor (SX 658 999–SX 636 972): a combination of LiDAR data imagery, aerial photography, ground observations and excavation has provided evidence for a previously unrecorded length of 3.4 km of a Roman road heading in a south-westerly direction from the western bank of the river Taw at North Tawton (SX 658 999) and across Rowden Moor towards Okehampton.Footnote ¹⁰⁷ The road, the line of which is suspected to be of military origin, is thought to have been a continuation of the Roman road that ran from the legionary fortress at Exeter to a crossing of the river Taw (Margary 492a).Footnote ¹⁰⁸ An extrapolation of the road line from its last attested location at Trehill Farm (SX 636 972) suggests that it heads towards a crossing of the river Okement at Okehampton, probably within sight of the known Roman fort which lies just to the north-east of the modern town.Footnote ¹⁰⁹

(b) Falcadon Farm (SX 654 998): LiDAR data, aerial photography and geophysical investigations have revealed the presence of a suspected small Roman army marching camp (or a construction camp?). The camp is located about 200 m north-west of the Roman road described above at Rowden Moor, 300 m east of a ford on the river Taw and approximately 600 m east of the Roman military complex at North Tawton.Footnote ¹¹⁰ The north and south linears measure 158 m and the east and west 150 m; corners are rounded. No internal structures were discernible.Footnote ¹¹¹