ESSEX

(1) Colchester, 60 Creffield Road (TL 98702 24712): the Colchester to Gosbecks Roman road was projected to run through the site, which is also within a Roman cemetery area, and the road was identified in 2019. Further work in 2020 has added detail to its layout and history. Phase 1 (first century) consisted of a 7 m-wide carriageway with a roadside ditch, a 2 m-wide footway and a second ditch on either side. In Phase 2, probably in the early second century, the carriageway was widened to 10 m and a metalled surface and two new roadside ditches were added. Further phases of metalling show that the carriageway was maintained and in use until the late fourth century. To the east were four first- to second-century pits. These are probably pyre-debris pits. Only one produced cremated human bone, but all four showed evidence of scorching around the edges/base and contained burnt animal bone and pottery sherds.Footnote ⁶⁹

(2) Colchester, Roman northern link road (east: TM 00330 27608; west: TM 00203 27442): this probable Roman link road appears to be a branch off the Coddenham to Colchester road (Margary 3c) heading to Colchester's northern rather than eastern gate. It would have provided an approach to Colchester using a different bridge over the river Colne. The road is clearest across High Woods Country Park (fig. 15), with prominent cuttings. These are so large that they could have been the original route of Margary 3c. Old roads on the OS first-edition maps line up with both sides of the High Woods length, connecting to the traditional route of Margary 3c and a long suggested Roman road heading north from Colchester's northern gate.Footnote ⁷⁰

FIG. 15.

DTM LiDAR plot (no vegetation/buildings) showing the prominent cuttings in High Woods Country Park indicating the course of a probable Roman road to the north of Colchester. (Base LiDAR data is © Crown Copyright 2020)

(3) Cressing, Tye Green (TL 7834 2037): excavation revealed the remains of a large Late Iron Age and Roman settlement built on a prominent ridge overlooking the Brain valley.Footnote ⁷¹ During the Roman period, the Iron Age settlement was expanded, with an enlargement of the main enclosure and an increase in activity to the east. There is evidence that, at some point during the first century a.d., the main enclosure was cleared and several roundhouses were burnt down, the waste from which was backfilled into the ring ditches. The newly enlarged enclosures, which were the main settlement focus during the first and second centuries, contained evidence for crop processing and storage, including the remains of an unusual granary constructed of 12 large posts (c. 0.8 m in diameter), with the charred post remains surviving in the base of several of the post-pits. The granary was located at the entrance to the enclosure and could have functioned as a storage place for grain taxes. On the western side of the original enclosure was a discrete area of intercutting gullies and pits, which truncated an infilled section of the enclosure ditch. These pits and gullies, which saw activity throughout the life of the settlement, contained assemblages currently interpreted as feasting deposits, including large amounts of animal bone and oyster shell, alongside numerous votive offerings. In total, this discrete area produced over 100 brooches of first-century b.c. to third-century a.d. date, Iron Age coins, dozens of Roman coins, hairpins, beads, finger-rings and a copper-alloy cockerel figurine. The gully features themselves could represent the remains of a shrine, which developed over two centuries, suggesting the area was held in special regard by its inhabitants well into the mid-Romano-British period.Footnote ⁷²

(4) Dovercourt, land west of Low Road (TM 23250 30180): excavation targeting an area of multi-period cropmarks revealed features of Roman (second- to fourth-century) date. Ditches defining a co-axial field system and a large rectangular enclosure laid out on high ground were found. There was no evidence of domestic occupation, but three corn-driers and fragments of millstones and quernstones show that cereal processing was carried out here. There was also a large watering hole with metalled base and evidence for animal management. The site may have been a stock-management and cereal-processing zone associated with either Little Oakley Roman villa or the postulated villa(s) at Dovercourt.Footnote ⁷³

(5) Harwich, a possible Roman road: a road north-east from Colchester was missed by Margary but recorded by FaircloughFootnote ⁷⁴ and is included on the 2016 OS Roman Britain Map. A branch off this road was previously suspected (TM 07962 29220), but only for a short distance. This branch route can be clearly traced with LiDAR as far as Wix Lodge (TM 14800 28504). The route coincides with Chequers Road, Little Bromley, before passing through Horsleycross Street (TM 12370 28649). After Wix Lodge clues are less frequent with possibly the most significant visible at TM 19185 29677. This would suggest that the road changed alignment slightly at Wix Lodge and headed for Harwich, approximated by the modern straight roads of Ramsey Road and Main Road (Old). At Harwich, the Harwich Stone Band rock outcrops. It is one of the few decorative stones native to Essex and was therefore a very important local resource. One of the Roman theatres in Colchester is believed to have used Harwich stone for its foundations. A Roman road to the Harwich area would therefore make perfect sense (fig. 16).Footnote ⁷⁵

FIG. 16.

LiDAR plot showing the course of a probable Roman road towards Harwich. The evidence is secure to Wix Lodge but intermittent to its east. (Base LiDAR data is © Crown Copyright 2020)

(6) Kelvedon, Monk's Farm (TL 8605 1932): excavation was undertaken across three areas (A–C). The most significant remains were concentrated in the eastern part of the development area, in Area A, and consisted of a complex of small conjoined rectilinear enclosures/paddocks.Footnote ⁷⁶ Discrete features were relatively scarce and no definite structures were identified, though the fills of the enclosure ditches produced relatively substantial assemblages of finds. The pottery assemblage is largely of second- to third-century date, and initial analysis suggests that industrial-type activity was being undertaken at the site, with relatively large quantities of slag and some briquetage coming from the enclosure ditches. The most notable feature, however, was a well/waterhole, which had been backfilled with extremely finds-rich deposits, including a very large assemblage of pottery and ceramic building material, probably deriving from a nearby, relatively affluent, settlement. These remains represent the first evidence for significant Roman activity to the north of Kelvedon Roman town (Canonium) and the contemporary London to Colchester road. Further analysis of the site should provide important insights into activity in the hinterland of the town.Footnote ⁷⁷

(7) Wixoe to Braintree, Roman road: a previously unknown Roman road (15 miles/24.1 km in length) was spotted in LiDAR imagery south of the Roman settlement of Wixoe (fig. 17). The road would have initially crossed the river Stour at or near the modern bridge before turning to the south. The first alignment headed up towards Moat Road (TL 7130 4120). Here the second alignment began and ran to TL 7190 4060, where another short alignment commenced. On this alignment there is fine surviving agger (TL 7306 3989) and a ‘captured’ footpath south-west of Ridgewell. Another slight turn onto the longest alignment occurs at TL 7320 3980. There is a prominent cutting on this line at TL 7590 3700, and the road would eventually have merged into the route of the A1017, approximately at the bridge site, now culverted, in the centre of Sible Hedingham (TL 7790 3460). From there to Braintree the modern road probably overlies the Roman line.Footnote ⁷⁸

FIG. 17.

Map of the route of the newly discovered Roman road from Wixoe to Sible Hedingham and ultimately, probably, to Braintree. Inset: detailed LiDAR plan of the roads around the Roman settlement at Wixoe. (Base mapping is derived from Ordnance Survey Opendata © Ordnance Survey; base LiDAR data is © Crown Copyright 2020)

(8) Wixoe to Radwinter, Roman road: a long-suspected Roman road has been confirmed for its full route (7.5 miles/12 km) by LiDAR (fig. 18). It takes a very direct route to Radwinter village where it meets up with the Great Chesterford to Great Dunmow Roman road. Leaving Wixoe over the river Stour bridge, a modern footpath represents the road to Hill Lane, which is ‘captured’ by the Roman line (TL 69396 42643). There are signs of the agger passing to the north of Steeple Bumpstead at TL 68154 41935 and TL 67298 41449. Near Boblow Hill, the Roman line is followed by a public footpath and parish boundary for around one mile (1.6 km). This is the length that was long assumed to be Roman but without a proven connection either to its east or west. Towards Radwinter the agger is evident at TL 61759 38082 and TL 60753 37430. The alignment of the road indicates that it formed a junction with the Chesterford to Dunmow Roman road (Margary 300) just to the north of the present-day crossroads in the village, close to the village school (TL 60662 37362).Footnote ⁷⁹

FIG. 18.

Oblique 3D LiDAR image of the Roman road from Wixoe to Radwinter. (Base LiDAR data is © Crown Copyright 2020)

NORFOLK

(1) Postwick, Broadland Gate infrastructure (TG 2907 0902): trial trenching ahead of the construction of access roads within the Broadland Gate development identified an area of Roman settlement. In particular, the complete footprint of a substantial three-bayed Roman building measuring 9 by 5 m was excavated and recorded. The pottery found in association with the building suggests a date for construction (or possibly disuse) of the mid-third century a.d. The rectangular building was aligned approximately NNW–SSE and comprised two parallel rows of four large post-holes set in the ground at regular intervals of c. 3 m. Morphologically, the post-holes were generally sub-oval shaped in plan and measured 0.6–1.0 m across and 0.35–0.65 m deep, with clearly defined post-pipes indicating that posts with diameters ranging between 0.3 and 0.5 m were used in its construction. Environmental evidence and the presence of an extra, smaller post-hole suggest that one of the posts was replaced, possibly after localised fire damage. Although much of the pottery recovered is typical of local manufacture, none appears to have come from the nearby second-century Postwick pottery kilns, and the pottery and activities at this site probably occurred approximately 100 years later, in the mid-third century a.d. The pottery assemblage is dominated by utilitarian domestic wares, some sooted, and indicates that settlement occurred in the immediate vicinity of the agricultural building. The few finewares indicate limited access to, or need for, fine Roman tablewares.

Of great interest is a large deposit of flax and gold-of-pleasure seeds in one of the structural post-holes; these were possibly grown and cultivated for their oils. Amphorae may have been used for the storage, and possibly sale, of oil produced on this site, and fragments were found in the buildings’ post-holes. Further remains of this Roman site have been uncovered in neighbouring plots 2a and 2b (see below). It is anticipated that further remains will be uncovered during the development of the neighbouring plots.Footnote ⁸⁰

(2) Postwick, plot 2a, Broadland Gate (TG 2897 0901): archaeological trial trenching took place prior to development of this plot within the larger Broadland Gate development. Previous work in the plot to the east and within the road lines indicated the presence of a Roman farmstead. The majority of the evidence relates to land partition during the early to mid-Roman period (mid-/late first to mid-second century a.d.), which is of interest as excavations on the adjacent (to the east) plot 2b and east of plot 2b identified activities associated with a farmstead during the mid- to late Roman period (mid-third century a.d.).Footnote ⁸¹

(3) Postwick, plot 2b, Broadland Gate (TG 2902 0906): archaeological trial trenching took place prior to development of this plot within the larger Broadland Gate development. Evidence relating to activities associated with a Roman farmstead, including pits, ditches and the remains of a second structure (the first recorded in the infrastructure excavation; see above), was recorded. These structures were mostly of mid-third century date, but had possibly continued in use from the mid-/late first to mid-second century a.d.Footnote ⁸²

(4) Thorpe St Andrew, plot 16a, The Oaks, Harvey Lane (TG 2542 0872): archaeological monitoring during the construction of a new house uncovered several early Romano-British settlement features, comprising ditches, pits and the bases of two oven-like features. One pit contained a dump of burnt material along with unabraded pottery sherds (including amphora) and large fragments of daub. Overall, the pottery, coinage and other finds provide a mid-first-century a.d. date for the activity here. The coin assemblage is particularly interesting as the majority of the coins, which were deposited within features, date to c. a.d. 41–50. It indicates a close military association with this relatively high-status site, which is also suggested by the date and range of Roman pottery vessels and hinted at by the presence of a discoidal horse-harness mount, a copper-alloy jug handle and a glass gaming counter. Previous archaeological investigations in the locality have recorded early Roman features dating to the first century a.d. in the form of a system of shallow gullies, pits, an oven and a possible grave which contained a pair of brooches.Footnote ⁸³

SUFFOLK

(1) Great Whelnetham, Fentons Farm, Stanningfield Road (TL 8760; WLG 038): archaeological excavation revealed three phases of Roman activity. The earliest phase is dated to the first to second century a.d. and consists of a large of number of quarry pits containing only minimal artefactual assemblages. The second phase is represented by a substantial boundary ditch running across the site on a broadly north-west–south-east alignment. The third phase, dated to the third to fourth century, consists of an inhumation cemetery containing at least 69 individuals. A number of these burials were decapitated, with evidence for execution, and other burials were interred in a prone position. Grave goods are limited but include bone combs and several small purple beads. The chronology of the burials is interesting, as some were deposited directly into the Phase 2 ditch. The site is positioned over a geological feature, consisting of a band of sandy material flanked by clayey deposits rich in gravel; it seems likely that the archaeology either targeted or was influenced by this feature.Footnote ⁸⁴

(2) Lavenham, land adjacent to Bears Lane (TL9148; LVM 121): archaeological evaluation revealed a cluster of pits and ditches associated with a kiln, also a grave and a surface layer focused on the central-southern area of the site. The Roman pottery assemblage includes a samian-ware dish with a maker's stamp. The animal bone includes butchered cattle, horse and deer bone. Other finds include copper-alloy coins, a pin, a strap end and a bronze or brass disc bearing the bust of an emperor, possibly part of a brooch or similar adornment. In the late nineteenth to early twentieth century a Roman tessellated pavement was reported a short distance to the north of the site, possibly in the grounds of Grove House, but the details remain unconfirmed and anecdotal. This new evidence from Bears Lane, including quantities of fragmented tegula and imbrex roof-tile, appears to add credence to the presence of a significant Roman building in the vicinity. The kiln may also relate to this structure, as it included large fragments of imbrex tile and an unusual half-box tile, which may have been backfilled from a nearby bath-house. Alternatively, this may indicate that the kiln was designed as a furnace to heat a hypocaust or bath-house (and that it was not designed to fire pottery/tile or serve as a corn-drier).Footnote ⁸⁵

(3) South Elmham St Mary, Flixton Park Quarry, Site 16 extension (TM 2985; SEY 038): excavations revealed two substantial Roman features. The largest feature was a square enclosure 10 m across and formed by a continuous metre-wide ditch, aligned with the cardinal compass points and with a 1.5 m-wide entrance placed centrally on its western length. Within the enclosure was a centrally placed pit, which contained no finds and was of an uncertain function. Also discovered was a well-preserved Roman pottery kiln, which was 1 m in diameter and 1 m deep with two large kidney-shaped internal pedestals and a well-fired archway. The empty nature of the kiln, devoid of pottery and kiln furniture, suggests it was deliberately closed down and cleaned out.Footnote ⁸⁶